Download Medicines concordance: Draft full guidline for consultation

DRAFT FOR CONSULTATION
1
2
3
4
Medicines concordance and adherence:
involving adults and carers in decisions
about prescribed medicines
5
6
7
Full guideline
8
9
National Collaborating Centre for Primary Care
10
Draft for consultation, July 2008
11
12
When commenting please use page number and line number
13
Medicines concordance: full guideline DRAFT (July 2008) page 1 of 373
DRAFT FOR CONSULTATION
1
2
3
Table of contents
4
Recommendations ......................................................................................... 11
5
Interventions to increase shared decision-making about medicines (chapter 4)
6
....................................................................................................................... 11
7
Information regarding medicines for patients and practitioners when patients
8
are discharged from inpatient care (Chapter 6).............................................. 15
9
Patients’ experience of medicine -taking (chapter 5) ..................................... 16
10
Assessment of adherence (chapter 7) ........................................................ 17
11
Interventions to increase adherence to prescribed medication (chapter 8) 18
12
Reviewing medicines (chapter 9) ............................................................... 19
13
Key Priorities for Implementation ..................................................................... 9
1
Introduction ............................................................................................. 21
14
1.1
Aim of the guideline ......................................................................... 22
15
1.2
How the guideline is set out ............................................................. 22
16
1.3
Scope .............................................................................................. 23
17
1.4
Responsibility and support for guideline development ..................... 25
18
19
20
21
22
23
The National Collaborating Centre for Primary Care (NCC-PC)
25
1.4.2
The development team ............................................................. 26
1.4.3
The Guideline Development Group (GDG) ............................... 27
1.4.4
Guideline Development Group meetings .................................. 30
1.5
Care Pathway .................................................................................. 30
24
1.6
Research recommendations ............................................................ 30
25
1.7
Acknowledgements.......................................................................... 34
26
1.8
Glossary .......................................................................................... 34
27
1.4.1
2
Methods .................................................................................................. 41
28
2.1
Introduction ...................................................................................... 41
29
2.2
Developing key clinical questions (KCQs) ....................................... 41
30
2.3
Literature search strategy ................................................................ 42
31
2.4
Identifying the evidence ................................................................... 45
32
2.5
Critical appraisal of the evidence ..................................................... 45
33
34
2.5.1
Choice of outcomes .................................................................. 45
2.6
Health Economics methods ............................................................. 46
35
36
2.6.1
2.6.2
Health Economic evidence review methodology ...................... 47
Cost-effectiveness modelling methods ..................................... 48
Medicines concordance: full guideline DRAFT (July 2008) page 2 of 373
DRAFT FOR CONSULTATION
1
2.7
Forming recommendations .............................................................. 48
2
2.8
Areas without evidence and consensus methodology ..................... 48
3
2.9
Consultation..................................................................................... 48
4
2.10
Relationships between the guideline and other national guidance .. 49
5
6
7
8
3
2.10.1 National Service Frameworks ................................................... 49
2.10.2 Related NICE Guidance ........................................................... 49
2.10.3 Other national guidance ........................................................... 49
Principles and concepts used in the development of the guideline ......... 51
9
3.1
Patients’ rights to be involved in decisions about medicines ........... 51
10
3.2
What is meant by involving patients in decisions about medicines? 52
11
3.3
What are we trying to achieve in involving patients in decisions about
12
medicines? ................................................................................................. 55
13
3.4
Roles and responsibilities of patients and health care professionals
56
14
15
3.5
Understanding the influences on medicine taking by patients ......... 57
16
3.6
Terminology and structure of guideline ............................................ 60
17
3.7
Shared decision-making about medicines ....................................... 62
18
4
Interventions to increase shared decision-making about medicines ....... 65
19
4.1
Recommendations *listed at the start of the guideline according to
20
chapter ....................................................................................................... 65
21
4.2
Introduction ...................................................................................... 65
22
4.3
Process of shared decision-making ................................................. 66
23
4.4
Methods ........................................................................................... 69
24
4.5
Key Clinical Question: Is it possible to increase patient involvement
25
in decisions about medicines?.................................................................... 71
26
27
28
4.5.1
Methods of the evidence review ............................................... 77
4.5.2
Evidence review ....................................................................... 78
4.6
Key Clinical Question: How can practitioners elicit patients’
29
preferences for involvement in decisions about medicines?....................... 89
30
31
32
4.6.1
Methods of the evidence review ............................................... 92
4.6.2
Evidence review ....................................................................... 92
4.7
Key Clinical Question: What tools are available to help elicit patients’
33
beliefs about medicines? ............................................................................ 97
34
35
36
4.7.1
Methods of the evidence review ............................................... 99
4.7.2
Evidence review ....................................................................... 99
4.8
Key Clinical Question: What tools are available to help elicit patients’
37
information needs? ................................................................................... 103
Medicines concordance: full guideline DRAFT (July 2008) page 3 of 373
DRAFT FOR CONSULTATION
1
2
3
4.8.1
Method of the evidence review ............................................... 105
4.8.2
Evidence review ..................................................................... 105
4.9
Key Clinical Question: How can information about medicines be
4
provided for patients in order to enhance SDM in regard to medicines? .. 113
5
6
4.9.1
Evidence review ..................................................................... 114
4.10 Key Clinical Question: What information about medicines should be
7
provided for patients in order to enhance SDM in regards to medicine? .. 117
8
9
4.10.1 Evidence review ..................................................................... 118
4.11 Key clinical Question: Mental capacity narrative ............................ 128
10
11
12
4.11.1 **Roles and responsibilities of patients and health care
professionals......................................................................................... 128
4.12 Key Clinical Question: What information about shared decision
13
making and adherence should be recorded in patients’ notes? ................ 130
14
4.13
15
in reaching an informed decision? How effective are these tools? ........... 131
16
17
18
4.13.1 Methods of evidence review ................................................... 134
4.13.2 Evidence review ..................................................................... 134
4.14 Key Clinical Question: How can a practitioner elicit whether a patient
19
agrees with the prescription recommended by the practitioner? .............. 142
20
21
22
4.14.1 Methods of the evidence review ............................................. 143
4.14.2 Evidence review ..................................................................... 143
4.15 Key Clinical Question: What aspects of consultation style increase
23
patient involvement in decision-making? .................................................. 145
24
25
26
4.15.1 Methods of the evidence review ............................................. 147
4.15.2 Evidence review ..................................................................... 148
4.16 Key Clinical Question: Do interventions to increase patient
27
involvement increase length of the consultation? ..................................... 154
28
29
30
4.16.1 Methods of the evidence review ............................................. 158
4.16.2 Evidence review ..................................................................... 159
Patients’ experience of medicine-taking ............................................... 169
5
Key Clinical Question: What tools are available to support the patient
31
5.1
Recommendations *listed at the start of the guideline according to
32
chapter ..................................................................................................... 169
33
5.2
Introduction .................................................................................... 169
34
5.3
Key Clinical Question: what are the barriers and facilitators for
35
individuals in medicine-taking ................................................................... 169
36
37
38
39
40
5.3.1
Methods of the evidence review ............................................. 173
5.3.2
Evidence review ..................................................................... 174
5.3.3
The evaluation of medicines and the difficulties encountered by
people in evaluating medicines ............................................................. 175
5.3.4
Medicines and identity ............................................................ 177
Medicines concordance: full guideline DRAFT (July 2008) page 4 of 373
DRAFT FOR CONSULTATION
1
2
5.3.5
Ways people take their medication ......................................... 178
5.4
Update of qualitative evidence synthesis - Pound 2005 synthesis 180
3
4
5
5.4.1
Methodology of update ........................................................... 180
5.4.2
Evidence review ..................................................................... 181
5.5
Systematic reviews of barriers and facilitators for individuals in
6
medicine taking ........................................................................................ 190
7
8
9
5.5.1
Methods of the update ............................................................ 190
5.5.2
Evidence review ..................................................................... 190
Information regarding medicines for patients and practitioners when
10
6
patients are discharged from inpatient care ................................................. 193
11
6.1
12
chapter ..................................................................................................... 193
13
6.2
Introduction .................................................................................... 193
14
6.3
Key clinical question: What information regarding medicines should
15
be provided for patients and practitioners when patients are discharged
16
from secondary care ................................................................................. 194
17
18
19
6.3.1
Methods of the evidence review ............................................. 195
6.3.2
Evidence review ..................................................................... 195
Assessment of adherence .................................................................... 197
7
Recommendations *listed at the start of the guideline according to
20
7.1
21
chapter ..................................................................................................... 197
22
7.2
Introduction .................................................................................... 197
23
7.3
Key Clinical Question: What are the advantages and disadvantages
24
of self-report in assessing patient’s adherence? ...................................... 198
25
26
27
7.3.1
Methods of the evidence review ............................................. 203
7.3.2
Evidence review ..................................................................... 203
Interventions to increase adherence to prescribed medication ............. 218
8
Recommendations *listed at the start of the guideline according to
28
8.1
29
chapter ..................................................................................................... 218
30
8.2
Introduction .................................................................................... 218
31
8.3
Methods ......................................................................................... 219
32
8.4
Evidence to recommendations : difficulties in interpreting studies on
33
interventions to improve adherence ......................................................... 221
34
8.5
35
adherence? .............................................................................................. 224
36
37
Recommendations *listed at the start of the guideline according to
Key Clinical Question: Does change in dosing regime affect
8.5.1
8.5.2
Methods of the evidence review ............................................. 225
Evidence review ..................................................................... 226
Medicines concordance: full guideline DRAFT (July 2008) page 5 of 373
DRAFT FOR CONSULTATION
1
8.6
2
adherence to prescribed medication ........................................................ 233
3
4
5
8.6.1
Methods of the evidence review ............................................. 235
8.6.2
Evidence review ..................................................................... 235
8.7
Key Clinical Question: Does drug formulation and/or packaging affect
6
adherence? .............................................................................................. 239
7
8
9
8.7.1
Methods of the evidence review ............................................. 241
8.7.2
Evidence review ..................................................................... 241
8.8
Key Clinical Question: Is there any evidence on interventions that
10
aim to minimize side-effects in order to increase adherence? .................. 247
11
12
13
8.8.1
Methods of the evidence review ............................................. 249
8.8.2
Evidence review ..................................................................... 249
8.9
Key Clinical Question: How does the way the information is
14
presented (e.g. pictorial vs. written) affects adherence? .......................... 257
15
16
17
8.9.1
Methods of the evidence review ............................................. 259
8.9.2
Evidence review ..................................................................... 259
8.10 Key Clinical Question: Do specific forms of therapy (e.g. CBT) affect
18
adherence? .............................................................................................. 263
19
20
21
8.10.1 Methods of the evidence review ............................................. 267
8.10.2 Evidence review ..................................................................... 268
8.11 Key Clinical Question: Would a contractual agreement between HCP
22
and patient affect adherence? .................................................................. 278
23
24
25
26
8.11.1 Methods of the evidence review ............................................. 279
8.11.2 Definitions of contracts ........................................................... 279
8.11.3 Evidence review ..................................................................... 280
8.12 Key Clinical Question: Do reminders (and what types of reminders,
27
text messaging etc) increase adherence to prescribed medication? ........ 282
28
29
30
8.12.1 Methods of the evidence review ............................................. 285
8.12.2 Evidence review ..................................................................... 285
8.13 Key Clinical Question : Does being involved in self-monitoring (e.g. of
31
own blood pressure) increase adherence to prescribed medication? ....... 293
32
33
34
8.13.1 Methods of the evidence review ............................................. 297
8.13.2 Evidence review ..................................................................... 298
8.14 Key Clinical Question : Does the use of dosette boxes increase
35
adherence to prescribed medication? ...................................................... 307
36
37
38
8.14.1 Methods of the evidence review ............................................. 307
8.14.2 Evidence review ..................................................................... 307
Reviewing medicines ............................................................................ 308
9
Key Clinical Question : Effect of prescription charges/costs on
39
9.1
Recommendations *listed at the start of the guideline according to
40
chapter ..................................................................................................... 308
Medicines concordance: full guideline DRAFT (July 2008) page 6 of 373
DRAFT FOR CONSULTATION
1
9.2
Introduction .................................................................................... 308
2
9.3
Key Clinical Question: does medication review increase shared
3
decision–making or adherence?............................................................... 310
4
5
6
9.3.1
Methods of the evidence review ............................................. 314
9.3.2
Evidence review ..................................................................... 315
10
Health Economics and Medicines Concordance ............................... 324
7
10.1
Introduction .................................................................................... 324
8
9
10
10.1.1 How terminology of compliance, concordance and adherence is
used in HE literature ............................................................................. 324
10.2 Health Economics methods ........................................................... 325
11
12
13
10.2.1 Health Economic evidence review methodology .................... 326
10.2.2 Cost-effectiveness modelling methods ................................... 327
10.3 Shared decision making and medicine taking from a health economic
14
perspective ............................................................................................... 327
15
16
17
10.3.1 Patient involvement and shared decision making ................... 327
10.3.2 Medicine-taking ...................................................................... 330
10.4 Cost effectiveness literature review of Interventions to increase
18
adherence ................................................................................................ 332
19
20
21
22
23
24
25
10.4.1 Methods of the review ............................................................ 332
10.4.2 Interventions to increase adherence: cost effectiveness
systematic reviews narrative ................................................................. 333
10.4.3 Summary of findings and methodological issues highlighted by
systematic reviews ................................................................................ 333
10.4.4 Conclusion .............................................................................. 338
10.5 Update of the systematic review by Elliott ..................................... 338
26
27
28
10.5.1 Methods of the review ............................................................ 338
10.5.2 Update review narrative .......................................................... 339
10.6 Discussion ..................................................................................... 345
29
30
31
32
33
34
35
10.6.1 Theoretical discussion of review findings, caveats and
opportunities for future research ........................................................... 345
10.6.2 Recommendations for minimum reporting requirements ........ 347
10.6.3 Recommendations for future research ................................... 349
Please note the appendices are available as separate files.
Medicines concordance: full guideline DRAFT (July 2008) page 7 of 373
DRAFT FOR CONSULTATION
1
2
3
Preface
*to be added to final version
4
Medicines concordance: full guideline DRAFT (July 2008) page 8 of 373
DRAFT FOR CONSULTATION
1
2
Key Priorities for Implementation
3
 Offer all patients the opportunity to be involved in making decisions about
4
prescribed medicines. Establish what level of involvement the patient would
5
like. (1.1.1)
6
 Ask if the patient has any specific concerns about their medicines,
7
whenever you prescribe, dispense or review medication. Address these
8
concerns. (1.1.8)
9
10
11
 Offer patients information about medicines before the medicines are
prescribed. (1.1.10)
 Actively discuss information on medicines with the patient rather than just
12
presenting it. The discussion should take into account the patient’s
13
understanding and beliefs about the diagnosis and treatment. (1.1.12)
14
 The information offered to patients about medicines should include:
15
 what the medicine is
16
 how it works
17
 how the medicine affects their condition (that is, its benefits)
18
 potential side effects
19
 any important instructions on how medicine should be taken
20
 what to do if they miss a dose
21
 if the medicine should be continued following the initial
22
prescription.(1.1.13)
23
24
 Be aware that a shared decision may mean an agreement not to prescribe
25
a medication or for the patient to stop taking a medication. If in the
26
healthcare professional’s view this may have an adverse effect, then this
27
must be recorded. (1.1.20)
Medicines concordance: full guideline DRAFT (July 2008) page 9 of 373
DRAFT FOR CONSULTATION
1
 Accept the patient’s right to decide not to take a medicine, even when you
2
do not agree with the decision, as long as the patient has capacity to
3
consent. (1.1.22)
4
 Record the discussion of decisions about medicine and medicine-taking in
5
patients’ records where there are concerns about medicines. This may
6
include:
7
 patients’ beliefs and concerns about medicines
8
 information given to the patient
9
 potential problems with adherence
 plans for review of medication. (1.1.25)
10
11
12
 Non-adherence can be assessed by asking the patient if they have missed
13
any medications recently. Make it easier for the patient to report non-
14
adherence by:
15
 asking the question in a way that does not apportion blame
16
 explaining why you are asking the question
17
 using a specific time period such as ‘in the last week’
18
 asking about specific medicine-taking behaviours. (1.2.3)
19
20
 Tailor any intervention to increase adherence to the specific difficulties with
21
adherence experienced by a patient. Be aware that adherence can be
22
improved but no specific intervention can be recommended for all patients.
23
(1.3.2)
24
Medicines concordance: full guideline DRAFT (July 2008) page 10 of 373
DRAFT FOR CONSULTATION
1
2
Recommendations
3
Interventions to increase shared decision-making
4
about medicines (chapter 4)
5
Establishing patients’ wish for involvement
6
1.1.1 Offer all patients the opportunity to be involved in making decisions
7
about prescribed medicines. Establish what level of involvement the patient
8
would like.
9
1.1.2 Establish the most effective way of communicating with each patient
10
and where necessary consider ways of making information accessible and
11
understandable (for example, via pictures, symbols, large print and different
12
languages).
13
1.1.3 Avoid making any unwarranted assumptions about patient preferences
14
about treatment. Talk to the patient to find out their preferences about
15
treatment and note any non-verbal clues that may indicate you need to
16
explore the patient’s perspective further.
17
1.1.4 Tailor your consultation style to the needs of individual patients to
18
ensure that all patients can be involved in decision-making in the way they
19
wish.
20
21
How to explore patients’ beliefs
22
1.1.5 Encourage patients to ask questions about their condition and
23
treatment.
24
1.1.6 Ask patients open-ended questions because these are more likely to
25
uncover patients concerns.
26
1.1.7 Ask about the patient’s beliefs about medicines both before starting
27
new treatments and periodically during medication review.
Medicines concordance: full guideline DRAFT (July 2008) page 11 of 373
DRAFT FOR CONSULTATION
1
1.1.8 Ask if the patient has any specific concerns about their medicines,
2
whenever you prescribe, dispense or review medication. Address these
3
concerns.
4
5
How to provide information for patients
6
1.1.9 Explain the goals of medical treatment to patients, openly discussing
7
the pros and cons of proposed medication. The discussion should be at the
8
level expected by the patient.
9
1.1.10 Offer patients information about medicines before the medicines are
10
prescribed.
11
1.1.11 Check patients have any information they wish about medicines before
12
medicines are dispensed.
13
1.1.12 Actively discuss information on medicines with the patient rather than
14
just presenting it. The discussion should take into account the patient’s
15
understanding and beliefs about the diagnosis and treatment.
16
1.1.13 The information offered to patients about medicines should include:
17
• what the medicine is
18
• how it works
19
• how the medicine affects their condition (that is, its benefits)
20
• potential side effects
21
• any important instructions on how medicine should be taken
22
• what to do if they miss a dose
23
• if the medicine should be continued following the initial
24
prescription.
Medicines concordance: full guideline DRAFT (July 2008) page 12 of 373
DRAFT FOR CONSULTATION
1
1.1.14 To help patients make decisions about medicines, offer them relevant
2
information which is easy to understand and free from jargon.
3
1.1.15 Suggest where patients might find more information and support after
4
the consultation: for example, by providing written information or directing
5
them to other resources or expert patient groups (for example,
6
www.patient.co.uk, www.easyhealth.org.uk).
7
1.1.16 Be careful not to make assumptions about a patient’s ability to
8
understand the information provided. Check with the patient that they have
9
understood the information.
10
1.1.17 Information for patients should be structured and when possible
11
tailored to the needs of the individual patient.
12
13
How to support the patient
14
1.1.18 Clarify the patient’s goals from treatment.
15
1.1.19 Healthcare professionals have a duty to help patients to make
16
decisions about their treatment which are informed by an understanding of the
17
likely benefits and risks rather than by patients’ beliefs.
18
1.1.20 Be aware that a shared decision may mean an agreement not to
19
prescribe a medication or for the patient to stop taking a medication. If in the
20
healthcare professional’s view this may have an adverse effect, then this must
21
be recorded.
22
1.1.21 Be aware that patients may have different views from healthcare
23
professionals regarding balance of risks, benefits and side effects of
24
medicines.
25
1.1.22 Accept the patient’s right to decide not to take a medicine, even when
26
you do not agree with the decision, as long as the patient has capacity to
27
consent.
Medicines concordance: full guideline DRAFT (July 2008) page 13 of 373
DRAFT FOR CONSULTATION
1
1.1.23 Assess capacity to make an informed decision using the principles in
2
the Mental Capacity Act (2005). To lack capacity patients must (a) have
3
disturbance or malfunction of mind and (b) demonstrate lack of capacity to:
4
• understand the information relevant to the decision
5
• retain information for long enough to use it in the decision
6
• use or weigh information
7
• communicate the decision.
8
1.1.24 Encourage and support patients, families and caregivers to keep and
9
maintain an accurate list of all medications, including prescription and non-
10
prescription medications and herbal and nutritional supplements, and any
11
allergic or adverse reactions to medication.
12
1.1.25 Record the discussion of decisions about medicine and medicine-
13
taking in patients’ records where there are concerns about medicines. This
14
may include:
15
• patients’ beliefs and concerns about medicines
16
• information given to the patient
17
• potential problems with adherence
18
• plans for review of medication.
19
1.1.26 Consider that good communication is necessary for increasing patient
20
involvement and that there are methods for improving your skills for increasing
21
patient involvement in decisions about their care.
22
1.1.27 Be aware that simple interventions to increase patient involvement do
23
not necessarily increase the overall length of consultation. Any extra time
24
spent in a single consultation may be justified by benefits, particularly over the
25
course of a long-term condition.
Medicines concordance: full guideline DRAFT (July 2008) page 14 of 373
DRAFT FOR CONSULTATION
1
Information regarding medicines for patients and
2
practitioners when patients are discharged from
3
inpatient care (Chapter 6)
4
1.1.28 Before discharge from hospital, offer patients information about their
5
medicines. This information should include:
6
• what the medicine is
7
• why it is necessary
8
• how the medicine affects their condition
9
• its benefits and harms
10
• potential side effects
11
• any important instructions on how the medicine should be taken
12
• how to get a further supply
13
• likely duration of treatment
14
• what to do in case of adverse effects
15
• any special considerations (for example, drug interactions,
16
storage).
17
1.1.29 On discharge from hospital, give all patients a report containing:
18
• diagnosis
19
• a list of all medications the patient should be taking
20
• clear identification of new medications initiated during the
21
hospital stay
22
• clear identification of medications stopped during the hospital
23
stay with reasons
Medicines concordance: full guideline DRAFT (July 2008) page 15 of 373
DRAFT FOR CONSULTATION
1
• clear information on which medications should be continued
2
after discharge and for how long
3
• known adverse reactions and allergies
4
• any potential problems with adherence and any actions taken
5
(for example, a compliance aid or reminder system).
6
1.1.30 When a patient is transferred between services, the next provider of
7
care should be given a report with the following information:
8
• a written report of diagnosis
9
• a list of all medications that the patient should be taking
10
• clear identification of new medications initiated during the
11
hospital stay
12
• clear identification of medications stopped during the hospital
13
stay with reasons
14
• clear information on which medications should be continued
15
after discharge
16
• known adverse reactions and allergies
17
• any potential problems with adherence and actions taken (for
18
example, use of dosette box).
19
20
Patients’ experience of medicine-taking (chapter 5)
21
1.1.31 Be aware that patients’ beliefs about medicines, in particular patients’
22
concerns about medicines and patients’ belief in their personal need for the
23
medicines, affect how and whether they take their medicine.
Medicines concordance: full guideline DRAFT (July 2008) page 16 of 373
DRAFT FOR CONSULTATION
1
1.1.32 Elicit and address patients’ specific concerns about taking their
2
medicine. These may include concern about becoming dependent on
3
medicines and the side effects and adverse effects of medicines.
4
1.1.33 Discuss with the patient why they need the treatment. Adopt a common
5
sense approach that addresses the need for the treatment (the disease or
6
condition) and the solution (the treatment).
7
1.1.34 Be aware that patients may wish to minimise how much medication
8
they take.
9
1.1.35 Be aware that patients may wish to discuss:
10
• what will happen if they do not take medicine suggested by their
11
doctor
12
• non-pharmacological alternatives to medicines
13
• how to reduce and stop medication they may have been taking
14
for a long time
15
• how to fit the medicine into their daily routine
16
• how to make a choice between medicines if they believe they
17
are taking too many medicines.
18
1.1.36 Be aware that patients evaluate prescribed medicines using their own
19
subjective or objective indicators. These include stopping and starting the
20
medicine, altering the dose of the medicine and checking how the medicine
21
affects their symptoms.
22
23
Assessment of adherence (chapter 7)
24
1.2.1 Recognise that non-adherence is common, and most people are non-
25
adherent sometimes.
Medicines concordance: full guideline DRAFT (July 2008) page 17 of 373
DRAFT FOR CONSULTATION
1
1.2.2 Routinely assess adherence in a non-judgemental way as an integral
2
part of medicine prescribing, dispensing and review.
3
1.2.3 Non-adherence can be assessed by asking the patient if they have
4
missed any medications recently. Make it easier for the patient to report non-
5
adherence by:
6
• asking the question in a way that does not apportion blame
7
• explaining why you are asking the question
8
• using a specific time period such as ‘in the last week’
9
• asking about specific medicine-taking behaviours.
10
1.2.4 Consider using records of prescription re-ordering and pharmacy refill
11
records to alert prescribers and dispensers to non-adherence.
12
1.2.5 When a healthcare professional identifies issues around adherence this
13
should be recorded and communicated within the healthcare team to optimise
14
patient involvement and care.
15
16
Interventions to increase adherence to prescribed medication
17
(chapter 8)
18
1.3.1 Discuss with the patient who is not taking their medicines whether this
19
is because of beliefs and concerns about medicines or because of practical
20
problems.
21
1.3.2 Tailor any intervention to increase adherence to the specific difficulties
22
with adherence experienced by a patient. Be aware that adherence can be
23
improved but no specific intervention can be recommended for all patients.
24
1.3.3 Find out what form of support the patient would prefer to increase their
25
adherence to medicines. Together, you and your patient should consider
26
options for support.
Medicines concordance: full guideline DRAFT (July 2008) page 18 of 373
DRAFT FOR CONSULTATION
1
1.3.4 Consider whether to suggest that patients record their medicine-taking
2
in a diary to aid adherence.
3
1.3.5 Be aware that encouraging patients to monitor their condition may
4
increase patients’ adherence to prescribed medication.
5
1.3.6 Simplify the dosing regimen if this is a problem for the patient.
6
1.3.7 Use special packaging for the medicine when the standard packaging
7
is a problem for the patient.
8
1.3.8 Side effects can be a problem for some patients and in this situation:
9
• discuss how the patient would like to deal with side effects
10
• discuss benefits, side effects and long-term effects with patients
11
to allow patients to make an informed choice
12
• consider adjusting medication dosage
13
• consider switching to an alternative that has a different risk of
14
side effects
15
• consider what other strategies other than adjusting dosage or
16
type of medication might be used (for example, timing of
17
medicines).
18
1.3.9 Reminder systems are likely to be helpful to some patients. The
19
method of delivery should meet individual needs.
20
1.3.10 Ask patients if the costs of prescriptions are a problem for them. If cost
21
of prescription is a problem, patients may wish to know which medicines are
22
most important.
23
Reviewing medicines (chapter 9)
24
1.4.1 Offer repeat information and review to patients when necessary,
25
especially when treating long-term conditions with multiple medications.
Medicines concordance: full guideline DRAFT (July 2008) page 19 of 373
DRAFT FOR CONSULTATION
1
1.4.2 Review practitioner and patients’ beliefs about medicines at intervals
2
agreed with the patient because these may change over time.
3
1.4.3 Patients should have the decision to prescribe medicines reviewed at
4
regular intervals according to patient choice and patient need.
5
1.4.4 A review of medicines should include an enquiry into medicine
6
adherence and where non-adherence is identified, possible causes should be
7
clarified and an agreement made with the patient about any appropriate
8
action. Any plan should include a date for a follow-up review.
9
1.4.5 Healthcare professionals involved in medication review should inform
10
the prescribing doctor of the review and its outcome, particularly when the
11
review involves adherence issues and further review is necessary.
Medicines concordance: full guideline DRAFT (July 2008) page 20 of 373
DRAFT FOR CONSULTATION
1
2
1
Introduction
3
The prescription of drugs is now a central part of the delivery of medical care.
4
The total drugs budget for the NHS in 2006/07 was approximately £10.6billion.
5
Of this around £7.6 billion was spent in primary care. Expenditure on primary
6
care drugs has increased by over 60% in the last decade or an average of
7
4.8% in real terms each year1.
8
Reviews across different disease areas report that between 30-50% of
9
patients do not take or use the medicines prescribed for them. The patient
10
was in the past thought to be the ‘source of the problem of compliance’ but it
11
is now acknowledged that medicine - taking is a complex behaviour and
12
influencing this involves not only patients but providers of health care and
13
health care systems (World Health Organization, 2003). The term adherence
14
is now used to describe the extent to which patients’ medicine taking
15
behaviour matches agreed recommendations from the prescriber. This
16
understanding presumes an agreement between patient and prescriber.
17
Attention is therefore required to the decision-making process between
18
prescribers and patients.
19
The estimated costs of unused or unwanted medicines in the NHS have been
20
estimated to exceed £100 million annually2 The WHO report notes that while
21
interventions to promote adherence are not generally seen to provide savings
22
in economic studies from an institutional perspective, cost savings are
23
demonstrated from a societal view point due to improvements in patient
24
quality of life, indirect costs avoided and effect on productivity. As with all
25
NICE guidelines a cost impact report will be developed for this guideline once
26
recommendations have been finalised and this will be published with the final
27
version of the guideline.
28
While objective health and cost impacts of patients’ behaviour in relation to
29
medicines is important, the right of patients to make decisions in regard to
1
2
http://www.official-documents.gov.uk/document/cm73/7341/7341.pdf
http://www.official-documents.gov.uk/document/cm73/7341/7341.pdf
Medicines concordance: full guideline DRAFT (July 2008) page 21 of 373
DRAFT FOR CONSULTATION
1
their own health is accepted in modern practice. The approach taken by this
2
guideline is that patients have a right to be involved in decisions about
3
medicines to the extent that they wish and it is the role of health professionals
4
to facilitate and support patients in their involvement in decision-making and to
5
support patients in taking medicine if the decision has been to prescribe.
6
Adherence to medicines taking should be based on shared decision making
7
between the patient and the practitioner.
8
9
We have not made separate recommendations for carers and families. The
10
principle relationship is between patient and health care professional and the
11
patient has a right to decide who should be involved in their care. With patient
12
consent carers should have access to same levels of information and support.
13
There are an increasing number of healthcare professionals now involved in
14
prescribing of medicines, dispensing and reviewing of medicines. It is not
15
within the remit of a guideline to recommend which health care professional
16
carries out these roles. Our recommendations apply to all healthcare
17
professionals who do carry out these roles but professionals also need to
18
work within legal and professional codes
19
20
1.1
Aim of the guideline
21
Clinical guidelines are defined as ‘systematically developed statements to
22
assist practitioner and patient decisions about appropriate healthcare for
23
specific clinical circumstances.
24
This guideline gives recommendations to clinicians and others on how to
25
involve adults and carers in decisions about prescribed medication.
26
1.2
27
The recommendations for all the topics in each chapter are listed at the start
28
of the chapter. Both the evidence statements and narratives of the research
29
studies on which our recommendations are based are found within each topic
How the guideline is set out
Medicines concordance: full guideline DRAFT (July 2008) page 22 of 373
DRAFT FOR CONSULTATION
1
section. The evidence statements precede the narrative for each topic. Also
2
included in each chapter is a brief explanation of why the GDG made the
3
specific recommendations. The evidence tables with details of the research
4
studies that describe the studies reviewed are found in Appendix C.
5
1.3
6
The guideline was developed in accordance with a scope given by the
7
National Institute for Health and Clinical Excellence (NICE, ‘the Institute’). The
8
scope set the remit of the guideline and specified those aspects of the
9
identification and management of medicines concordance and adherence to
Scope
10
be included and excluded. The scope was published in April 2007 and is
11
reproduced here in Appendix A.
12
Whom the guideline is intended for
13
This guideline is of relevance to those who work in or use the National Health
14
Service (NHS) in England and Wales:
15
Population
16
Groups that will be covered
17
a) Adults, including those with comorbidities, learning disabilities or language
18
and/or cultural differences.
19
Groups that will not be covered
20
Children and young people. However, the guideline recommendations may be
21
considered for a child or young person who is deemed competent to express
22
a view on their prescription.
23
Healthcare setting
24
All consultations with healthcare professionals in any NHS setting that relate
25
to the initiation or review of prescribed medication.
26
Areas that will be covered
Medicines concordance: full guideline DRAFT (July 2008) page 23 of 373
DRAFT FOR CONSULTATION
1
a) Shared decision-making about medicines and medicine taking as reported
2
by the patient or carer. The guideline will focus on the barriers (such as
3
communication difficulties, cultural issues, low health literacy and physical
4
limitations), facilitators (including structural or procedural factors), beliefs and
5
health behaviours that influence decision-making and adherence.
6
b) Shared decision-making about medicines and medicine taking as reported
7
by the healthcare professional. The guideline will focus on the barriers (such
8
as communication difficulties, cultural issues and time), facilitators (including
9
structural or procedural factors), beliefs and health behaviours that influence
10
decision-making and adherence.
11
c) The effectiveness and cost effectiveness of interventions to facilitate the
12
process of shared decision-making about medicines (looking at time of
13
intervention – before, during, or after the consultation with the healthcare
14
professional; and mode of delivery). The target of the intervention may be the
15
patient, the carer, the prescriber, any healthcare professional providing
16
ongoing support or a combination of these.
17
d) The effectiveness and cost effectiveness of interventions to promote
18
adherence in medicine taking (looking at time of intervention – before, during,
19
or after the consultation with the healthcare professional; and mode of
20
delivery). The target of the intervention may be the patient, the carer, the
21
prescriber, the dispenser or any other healthcare professional providing
22
ongoing support or a combination of these.
23
e) The evidence on single or multiple medications as it relates to issues
24
around decision-making and adherence.
25
f) The skills and competencies required by prescribers to involve patient in
26
decisions regarding prescribed medicines.
27
Areas outside the remit of the guideline
28
The administration of medicines will not be covered. Administration is defined
29
as giving a medicine by introduction into the body (for example, orally or by
Medicines concordance: full guideline DRAFT (July 2008) page 24 of 373
DRAFT FOR CONSULTATION
1
injection), or by external application (for example application of an
2
impregnated dressing).
3
1.4
4
The NCC-PC is a partnership of primary care professional associations and
5
was formed as a collaborating centre to develop guidelines under contract to
6
NICE. It is entirely funded by NICE. The NCC-PC is contracted to develop four
7
guidelines at any one time, although there is some overlap at start and finish.
8
Unlike many of the other centres which focus on a particular clinical area, the
9
NCC-PC has a broad range of topics relevant to primary care. However, it
Responsibility and support for guideline development
10
does not develop guidelines exclusively for primary care. Each guideline may,
11
depending on the scope, provide guidance to other health sectors in addition
12
to primary care.
13
The Royal College of General Practitioners (RCGP) acts as the host
14
organisation. The Royal Pharmaceutical Society and the Community
15
Practitioners and Health Visitors’ Association are partner members with
16
representation from other professional and lay bodies on the Board. The
17
RCGP holds the contract with the Institute for the NCC-PC.
18
1.4.1
19
PC)
20
The NCC-PC is a partnership of primary care professional associations and
21
was formed as a collaborating centre to develop guidelines under contract to
22
NICE. It is entirely funded by NICE. The NCC-PC is contracted to develop five
23
guidelines at any one time, although there is some overlap at start and finish.
24
Unlike many of the other centres which focus on a particular clinical area, the
25
NCC-PC has a broad range of topics relevant to primary care. However, it
26
does not develop guidelines exclusively for primary care. Each guideline may,
27
depending on the scope, provide guidance to other health sectors in addition
28
to primary care.
29
The Royal College of General Practitioners (RCGP) acts as the host
30
organisation. The Royal Pharmaceutical Society and the Community
31
Practitioners and Health Visitors’ Association are partner members with
The National Collaborating Centre for Primary Care (NCC-
Medicines concordance: full guideline DRAFT (July 2008) page 25 of 373
DRAFT FOR CONSULTATION
1
representation from other professional and lay bodies on the Board. The
2
RCGP holds the contract with the Institute for the NCC-PC.
3
The development team
4
1.4.2
5
The development team had the responsibility for this guideline throughout its
6
development. They were responsible for preparing information for the
7
Guideline Development Group (GDG), for drafting the guideline and for
8
responding to consultation comments. The development team working on this
9
guideline consisted of the:
10
 Guideline lead
11
who is a senior member of the NCC-PC team who has overall
12
responsibility for the guideline
13
 Information scientist
14
who searched the bibliographic databases for evidence to
15
answer the questions posed by the GDG
16
 Reviewer (Health Services Research Fellow)
17
with knowledge of the field, who appraised the literature and
18
abstracted and distilled the relevant evidence for the GDG
19
 Health economist
20
who reviewed the economic evidence and assisted the GDG in
21
considering cost effectiveness
22
 Project manager
23
who was responsible for organising and planning the
24
development, for meetings and minutes and for liaising with the
25
Institute and external bodies
26
 Chair
27
who was responsible for chairing and facilitating the working of
28
the GDG meetings
29
The members of the development team attended the GDG meetings and
30
participated in them. The development team also met regularly with the Chair
Medicines concordance: full guideline DRAFT (July 2008) page 26 of 373
DRAFT FOR CONSULTATION
1
of the GDG during the development of the guideline to review progress and
2
plan work.
3
Other guidelines normally have a clinical advisor who is someone with an
4
academic understanding of the research in the area and its practical
5
implications to the service, who advised the development team on searches
6
and the interpretation of the literature. Due to the conceptual nature of the
7
guideline topic and the different academic stances on explaining such
8
behaviour the development team chose not to have a formal clinical advisor.
9
1.4.3
The Guideline Development Group (GDG)
10
A Chair was chosen for the group and his primary role was to facilitate and
11
chair the GDG meetings.
12
The GDG consisted of diverse multidisciplinary group with an interest and/or
13
expertise in medicines concordance. The Chair, a general practitioner with
14
special interest in epilepsy identified by the NCC-PC, oversaw the work of the
15
group.
16
Nominations for group members were invited from various stakeholder
17
organisations, selected to ensure appropriate combination of members
18
including healthcare professionals and patient representatives.
19
Each nominee was expected to act as an individual expert in their own right
20
and not as a representative of their parent organisation, although they were
21
encouraged to keep their nominating organisation informed of the process.
22
Group membership can be found in the preface to the guidance.
23
In accordance with guidance from NICE, all GDG members’ interests were
24
recorded on a standard declaration form that covered consultancies, fee-paid
25
work, share-holdings, fellowships, and support from the healthcare industry.
26
Details of these can be seen in Appendix E.
27
The names of GDG members appear listed below.
28
Full GDG members
29
 Dr Henry Smithson (Chair)
Medicines concordance: full guideline DRAFT (July 2008) page 27 of 373
DRAFT FOR CONSULTATION
1
GP Escrick Research Practice York, Senior Clinical University
2
Teacher Academic Unit of Primary Medical Care Sheffield
3
 Professor Rob Horne
4
Professor of Behavioural Medicine, Head of Department of Practice
5
and Policy, The School of Pharmacy, University of London, London
6
 Dr John Benson
7
Senior Lecturer in General Practice, General Practice and Primary
8
Care Research unit, Department of Public Health and Primary
9
Care, University of Cambridge.
10
 Mr Shaun Johnson
11
Patient representative, Melton Mowbray
12
 Mrs Alison Bowser
13
Independent patient advocate, patient representative for Medicines
14
and Healthcare Products Regulatory Authority, Cornwall and Isles
15
of Scilly PCT, Royal College of GPs
16
 Dr Mahendra Patel
17
Lecturer/Research Fellow, Institute of Pharmaceutical Innovation,
18
Bradford
19
 Mr Stephen Hemingway
20
Senior Lecturer in Mental Health, University of Huddersfield
21
 Mrs Bunis Packham
22
Nurse Consultant-Thrombosis and Anticoagulation, Barnet
23
Hospital, Barnet
24
 Mr Jim Blair
25
President of the Forum on Intellectual Disability at the Royal
26
Society of Medicine and Committee Member of the Royal College
27
of Nursing's National Learning Disability Forum
28
 Professor Peter Crome
29
Professor of Geriatric Medicine, Keele University, Keele,
30
Staffordshire and Hon. Consultant Geriatrician, North Staffordshire
31
Combined Healthcare NHS Trust, Stoke-on-Trent.
32
 Dr Peter Haddad
Medicines concordance: full guideline DRAFT (July 2008) page 28 of 373
DRAFT FOR CONSULTATION
1
Consultant Psychiatrist, Cromwell House, Greater Manchester
2
West Mental Health NHS Foundation Trust Honorary Senior
3
Lecturer in Psychiatry, University of Manchester
4
 Dr Jonathan Steel
5
Chairman GP Network, Royal College of Physicians, London.
6
General Practitioner, Uley, Gloucestershire
7
 Dr Sean Kelly
8
Consultant Physician and Gastrolenterologist, York Hospital, York.
9
Honorary senior lecturer Hull York Medical School.
10
 Dr Wendy Clyne
11
Assistant Director, Medicines Partnership Programme, NPC Plus,
12
Keele University Science and Business Park, Keele
13
14
15
16
17
Members of the GDG from the NCC-PC were:
 Dr. Norma O’Flynn
Guideline Lead and Clinical Director, NCC-PC
 Ms Elizabeth Shaw
18
Guideline Lead and Deputy Chief Executive, NCC-PC (until
19
February 2007)
20
 Ms Vanessa Nunes
21
Senior Health Services Research Fellow/Project Manager, NCC-
22
PC
23
 Ms Julie Neilson
24
Health Services Research Fellow, NCCPC
25
 Ms Stefanie Kuntze
26
Health Economist, NCCPC
27
 Mr. Gary Britton
28
Health Services Research Fellow, NCCPC (until October 2007)
29
 Ms Marian Cotterell
30
Information Scientist, NCC-PC
31
Medicines concordance: full guideline DRAFT (July 2008) page 29 of 373
DRAFT FOR CONSULTATION
1
Observers
 Ms Sarah Willett
2
3
Commissioning Manager, National Institute for Health and
4
Clinical Excellence (from December 2007)
 Ms Colette Marshall
5
6
Commissioning Manager, National Institute for Health and
7
Clinical Excellence (until August 2007)
 Dr. Francoise Cluzeau
8
Technical advisor, National Institute for Health and Clinical
9
Excellence
10
Guideline Development Group meetings
11
1.4.4
12
The GDG met on 10 occasions (with one two day GDG meeting), at
13
approximately 2 monthly intervals over a period of 11 months and 6 weekly
14
intervals over a period of 6 months to review the evidence identified by the
15
project team, to comment on its quality and completeness and to develop
16
recommendations for clinical practice based on the available evidence. The
17
final recommendations were agreed by the full GDG.
18
1.5
19
**SEE NICE version
20
1.6
21
The Guideline Development Group has made the following recommendations
22
for research, based on its review of evidence, to improve NICE guidance and
23
patient care in the future. The Guideline Development Group’s full set of
24
research recommendations is detailed in the full guideline.
25
The GDG noted the generally poor quality of research in the area of
26
medicines adherence and the potential clinical and economic gains that would
27
accrue from achieving and implementing a shared decision about medicines.
28
The GDG believe that there is an urgent need to provide specific adherence
29
funding streams to support structured programmes of research particularly
30
where the health gains from medicines adherence are likely to be high. The
Care Pathway
Research recommendations
Medicines concordance: full guideline DRAFT (July 2008) page 30 of 373
DRAFT FOR CONSULTATION
1
central theme underpinning this guideline is that adherence to medicines
2
taking is an individual behaviour that should be based on shared decision
3
making principally between the patient and the practitioner. The key research
4
agenda is therefore one of behaviour-change for practitioners and patients.
5
However, medicines carry the potential for harm as well as benefit and there
6
are also questions about what is good-prescribing and good medicine-taking.
7
There are therefore two agendas: an empirical agenda to address the
8
question of how adherence might be improved and a normative agenda to tell
9
us what the right thing to do is.
10
The research recommendations from this guideline are for research
11
programmes which are described below under the themes of A: Shared
12
decision making and the consultation; B: Barriers and interventions to
13
adherence and C: Groups for special consideration
14
15
A. Shared decision making and the consultation
16
Research questions:
17
a.
18
making, partnership and other models of consultation?
19
b.
20
practices that approximate to joint decision-making look like? What are the
21
strengths and weaknesses of such practices seen from the vantage point of
22
various stakeholders (e.g. prescribers, patients, funders)?
23
c.
24
that could re-distribute accountability more from prescribers to patients (e.g.
25
waivers)? What are the legal, policy, practical and psychological implications
26
of trying to share accountability differently?
27
d.
28
and emotional challenges of prescribing consultations designed to promote
29
informed choice and adherence to medication?
What are prescribers’ and patients’ attitudes to shared decision
What joint decision-making processes are possible? What do real world
Are there practical mechanisms in place (or available to put in place)
How can clinicians and patients be supported to deal with the cognitive
Medicines concordance: full guideline DRAFT (July 2008) page 31 of 373
DRAFT FOR CONSULTATION
1
e.
How can we facilitate the honest disclosure of medication-taking
2
behaviours within prescribing-related consultations and medication use
3
reviews? How can we equip health practitioners to respond appropriately and
4
effectively?
5
f.
6
risk of nonadherence or who are a priority for medication-review and
7
adherence support and how can we provide it.
8
Why this is important
9
The principles of shared decision-making have largely been developed from
How can we enable new and existing prescribers to identify patients at
10
theoretical and conceptual models. The competencies listed for shared
11
decision- making consist of a number of different skills and patients have
12
shown that they may value different aspects of shared decision making. While
13
the right of patients to be involved in decision-making in regard to their own
14
health care is accepted, the practice of shared decision making may result in
15
practitioners and patients playing different roles than they have to date in
16
health care consultations. This may have implications for responsibility and
17
accountability. Information asymmetries also need to be addressed and this
18
may require structural changes to health services and their delivery. Patient
19
related outcomes need to be included. There is a new and growing agenda
20
relating to non-medical prescribers (pharmacists, nurses etc.) This is a key
21
context issue and there are a range of questions relating to patient
22
perspectives on new prescribers and to new prescribers’ perceptions and
23
skills. The effects of new prescribers on patient adherence to medication
24
should be included in any research agendas designed to evaluate new
25
prescribers.
26
27
B. Barriers and interventions to adherence
28
Research Questions
29
a) What are the most clinical and cost effective methods for addressing
30
cognitive influences (such as beliefs and concerns about medication) and
Medicines concordance: full guideline DRAFT (July 2008) page 32 of 373
DRAFT FOR CONSULTATION
1
capacity (such as memory, manual dexterity, changes in routine) that result in
2
reduced adherence?
3
Why this is important
4
Few interventions have been systematically developed, using appropriate
5
theoretical models, nor have they have been modelled and piloted with
6
assessment of process variables as well as outcomes (as recommended in
7
the MRC framework for complex interventions to effect behaviour change).
8
Consequently it is difficult to tell why some interventions work and others do
9
not. Interventions should be developed using an appropriate theoretical
10
framework with a phased approach to testing that includes assessment of
11
process (i.e. the things that are targeted for change) as well as outcomes and
12
a need for an individual approach. Research should explore both clinical and
13
economic benefits and include patient outcomes.
14
C. Groups for special consideration
15
a.
16
prescriptions and adherence to prescribed medication?
17
b.
18
(children, young adults, elderly people) influence adherence and what are the
19
implications for interventions?
20
c.
21
multiple pathologies (and their informal carers) and what interventions are
22
required?
23
Why this is important
24
The value an individual places on sharing decisions with their practitioner has
25
been found to differ by groups such as the elderly and severity of condition.
26
Research into the factors and impact on adherence could inform clinicians
27
and shape clinical care. Multi-morbidity is common and occurs at all ages but
28
increases with age. Patients with multi-morbidity are often prescribed large
29
amounts of medication. Research is required to assess particular barriers
30
experienced by people with multi-morbidity and to investigate what type of
What are the effects of social disadvantage and ethnicity on accessing
How do the perceptions and life circumstances of different age groups
What are the particular barriers to medicines use for people with
Medicines concordance: full guideline DRAFT (July 2008) page 33 of 373
DRAFT FOR CONSULTATION
1
support is required to allow people to make decisions about medicines and to
2
adhere to those decisions.
3
4
5
6
1.7
Acknowledgements
7
**TO BE ADDED for final version
8
1.8
Glossary
Adherence
Adherence – ‘the extent to which the patient’s behaviour
matches agreed recommendations from the prescriber’.
Adherence emphasises the need for agreement and that
the patient is free to decide whether or not to adhere to
the doctor’s recommendation..
Compliance
Compliance – ‘the extent to which the patient’s behaviour
matches the prescribers recommendations’.
Concordance
Concordance – this is a recent term whose meaning has
changed. It was initially applied to the consultation
process in which doctor and patient agree therapeutic
decisions that incorporate their respective views, but now
includes patient support in medicine taking as well as
prescribing communication. Concordance reflects social
values but does not address medicine-taking and may not
lead to improved adherence
Medicines concordance: full guideline DRAFT (July 2008) page 34 of 373
DRAFT FOR CONSULTATION
Cost-benefit
A type of economic evaluation where both costs and
analysis
benefits of healthcare treatment are measured in the
same monetary units. If benefits exceed costs, the
evaluation would recommend providing the treatment.
Cost-
A type of economic evaluation where various health
consequences
outcomes are reported in addition to cost for each
analysis
intervention, but there is no overall measure of health
gain.
Cost-
An economic study design in which consequences of
effectiveness
different interventions are measured using a single
analysis
outcome, usually in ‘natural’ units (for example, life-years
gained, deaths avoided, heart attacks avoided, cases
detected). Alternative interventions are then compared in
terms of cost per unit of effectiveness.
Cost-
An explicit mathematical framework, which is used to
effectiveness
represent clinical decision problems and incorporate
model
evidence from a variety of sources in order to estimate the
costs and health outcomes. See also Markov model.
Cost-
An economic evaluation that finds the least costly
minimisation
alternative therapy after the proposed interventions has
analysis
been demonstrated to be no worse than its main
comparator(s) in terms of effectiveness and toxicity.
Cost-utility
A form of cost-effectiveness analysis in which the units of
analysis
effectiveness are quality-adjusted life-years (QALYs).
Decision analysis A systematic way of reaching decisions, based on
evidence from research. This evidence is translated into
probabilities, and then into diagrams or decision trees
which direct the clinician through a succession of possible
scenarios, actions and outcomes.
Medicines concordance: full guideline DRAFT (July 2008) page 35 of 373
DRAFT FOR CONSULTATION
Decision problem A clear specification of the interventions, patient
populations and outcome measures and perspective
adopted in an evaluation, with an explicit justification,
relating these to the decision which the analysis is to
inform.
Discounting
Costs and benefits incurred today have a higher value
than costs and benefits occurring in the future.
Discounting health benefits reflects individual preference
for benefits to be experienced in the present rather than
the future. Discounting costs reflects individual preference
for costs to be experienced in the future rather than the
present. For NICE economic evaluations, health outcomes
will be discounted at 3.5% and costs at 3.5% per annum,
following the recommendations of the UK Treasury.
Dispensing
Professional trained in dispensing medicine, generally a
professional
pharmacist or a general practitioner in a dispensing
practice
Dominance
An intervention is said to be dominant if there is an
alternative intervention that is both less costly and more
effective. See also extended dominance.
Dosette box
A type of compliance aid. Other terms used are NOMAD,
MANRAX and monitored dose system.
Economic
Comparative analysis of alternative health strategies
evaluation
(interventions or programmes) in terms of both their costs
and consequences.
Medicines concordance: full guideline DRAFT (July 2008) page 36 of 373
DRAFT FOR CONSULTATION
Extended
An intervention is extendedly dominated when it can be
dominance
dominated by a combination of two alternative
interventions (i.e. if x% of the population are treated with
intervention A, and y% are treated with intervention C, the
overall result will be an intervention strategy that is both
cheaper and more effective than intervention B). See also
dominance.
Extrapolation
In data analysis, predicting the value of a parameter
outside the range of observed values.
Forgiveness
The ability of a drug to sustain its pharmacological action
after a dose has been missed
GDG
Guideline development group who developed the
guideline
Health care
Any health care professional- specialists, general
professional
practitioner, nurse prescribers who are involved in the
(HCP)
prescribing of medicines or in the discussion with patients
about those medicines.
Health
The study of the allocation of scarce resources among
economics
alternative healthcare treatments. Health economists are
concerned with both increasing the average level of health
in the population and improving the distribution of
healthcare resources.
Health-related
A combination of an individual’s physical, mental and
quality of life
social well-being; not merely the absence of disease.
Informed
Informed adherence refers to an outcome of informed
adherence
choice in decision to take medicines and supported
adherence
Medicines concordance: full guideline DRAFT (July 2008) page 37 of 373
DRAFT FOR CONSULTATION
Incremental Cost
Is the difference in costs between two interventions being
effectiveness
compared divided by the difference in effect of the two
ratio
interventions. For instance if A and B are being compared
Cost of A – costs of B divided by effects of A- effects of B.
ICER
Incremental cost effectiveness ratio – this is the difference
between the mean costs in the population of interest
divided by the difference in the mean outcomes in the
population of interest
Life-year
A measure of health outcome that shows the number of
years of remaining life expectancy.
Life-years gained
Average years of life gained per person as a result o fan
intervention.
Markov model
A modelling technique used when a greater number of
health states needs to be considered. They are
particularly useful for disease in which events can occur
repeatedly over time.
Medicines
The term medicines is used in the guideline to apply to
drug treatments that patients may take orally or selfadminister such as creams to the skin and drops.
Medication
A face to face meeting between a professional and a
review
patient to discuss the patients medicines and medicinetaking behaviour
Opportunity cost
.The opportunity cost of investing in a healthcare
intervention is the other healthcare programmes that are
displaced by its introduction. This may be best measured
by the health benefits that could have been achieved had
the money been spent on the next best alternative
healthcare intervention.
Medicines concordance: full guideline DRAFT (July 2008) page 38 of 373
DRAFT FOR CONSULTATION
Persistence
The length of time from initiation to discontinuation of
therapy. Persistence is measured in units of time.
Perspective
(or viewpoint): This determines which costs to include. For
NICE evaluations the perspective is from the NHS and
includes costs to the NHS and Personal Social Services.
Costs to other public bodies and to patients and carers
may be considered as an additional factor.
Probabilistic
Probability distributions are assigned to the uncertain
sensitivity
parameters and are incorporated into evaluation models
analysis
based on decision analytical techniques (for example,
Monte Carlo simulation).
Quality adjusted
An index of survival that is adjusted to account for the
life-years
person’s quality of life during this time. QALYs have the
(QALYS)
advantage of incorporating changes in both quantity
(longevity/mortality) and quality (morbidity, psychological,
functional, social and other factors) of life. Used to
measure benefits in cost-utility analysis, QALYS are
calculated by estimating the number of years of life gained
from a treatment and weighting each year with a qualityof-life score between zero and one.
Shared Decision
.Shared-decision making (SDM) is described as a model
Making (SDM)
of decision making where information exchange is a two
way process in the consultation and both deliberation and
decision are made by both health care professional and
patient.
Specialist
One who has expertise in a particular field of medicine by
virtue of additional training and experience.
Medicines concordance: full guideline DRAFT (July 2008) page 39 of 373
DRAFT FOR CONSULTATION
Time horizon
The time span used in the NICE appraisal that reflects the
period over which the main differences between
interventions in health effects and use of healthcare
resources are expected to be experienced, and taking into
account the limitations of supportive evidence.
Unit-dose
Unit-dose packaging is the packaging of a single dose in a
packaging
nonreusable container.
Utility
This concept is applied in health care to mean the
individual's valuation of their state of well-being deriving
from the use of health care interventions. In brief, utility is
a measure of the preference for, or desirability of, a
specific level of health status or specific health outcome.
Willingness to
WTP refers to the amount that a decision maker is willing
pay (WTP)
to pay for an additional unit of outcome (e.g. an additional
QALY). If the WTP is higher than the ICER, the
intervention is cost effective. If not, the intervention is not
cost effective.
1
2
Medicines concordance: full guideline DRAFT (July 2008) page 40 of 373
DRAFT FOR CONSULTATION
1
2
Methods
2
2.1
Introduction
3
This chapter sets out in detail the methods used to generate the
4
recommendations for clinical practice that are presented in the subsequent
5
chapters of this guideline. The methods are in accordance with those set out
6
by the Institute in ‘The guidelines manual’. April 2006. London: National
7
Institute for Health and Clinical Excellence. Available from:
8
www.nice.org.uk/guidelinesmanual. The Guideline Development Process – an
9
overview for stakeholders, the public and the NHS describes how
10
organisations can become involved in the development of a guideline.
11
2.2
12
A series of key questions created from the scope was the first step in the
13
development of the guideline. The key questions formed the starting point for
14
the subsequent evidence reviews and facilitated the development of
15
recommendations by the GDG.
16
The key questions were developed by the project team with the guidance from
17
the GDG. Where possible, the questions were refined into specific research
18
questions by the project teams to aid literature searching, appraisal and
19
synthesis. However, due to the generic nature of the guideline, full PICO
20
parameters were not applicable to the developed research questions. The full
21
list of key questions is shown in appendix B.
22
A full literature search and critical appraisal could not be undertaken for all key
23
questions due to the time and resource limitations within the guideline
24
development process. The GDG and project teams therefore agreed
25
appropriate review parameters (inclusion and exclusion criteria) for each topic
26
area in accordance with those outlined in the scope.
Developing key clinical questions (KCQs)
27
Medicines concordance: full guideline DRAFT (July 2008) page 41 of 373
DRAFT FOR CONSULTATION
1
2.3
Literature search strategy
2
An initial scoping search for published guidelines, systematic reviews,
3
economic evaluations and ongoing research was carried out on the following
4
databases or websites: National Library for Health (NLH) Guidelines Finder,
5
National Guidelines Clearinghouse, Scottish Intercollegiate Guidelines
6
Network (SIGN), Guidelines International Network (GIN), Canadian Medical
7
Association (CMA) Infobase (Canadian guidelines), National Health and
8
Medical Research Council (NHMRC) Clinical Practice Guidelines (Australian
9
Guidelines), New Zealand Guidelines Group, BMJ Clinical Evidence,
10
Cochrane Database of Systematic Reviews (CDSR), Database of Abstracts of
11
Reviews of Effects (DARE) and Heath Technology Assessment Database
12
(HTA), NHS Economic Evaluations Database (NHSEED) National Research
13
Register and Current Controlled Trials.
14
The aim of the evidence review was to identify the most relevant, published
15
evidence in relation to the key clinical questions generated by the GDG. Due
16
to time constraints, full systematic reviews were not undertaken. However, the
17
evidence reviews were undertaken using systematic, transparent approaches.
18
The following bibliographic databases were searched from their inception to
19
the latest date available: Cochrane Database of Systematic Reviews (CDSR),
20
Database of Abstracts of Reviews of Effects (DARE), Health Technology
21
Database (HTA), MEDLINE, EMBASE, CINAHL, AMED (Allied and
22
Complementary Medicine Database), CENTRAL (Cochrane Controlled Trials
23
Register). When appropriate to the question PsycINFO was also searched.
24
The search strategies were developed in MEDLINE and then adapted for
25
searching in other bibliographic databases. Systematic reviews and
26
randomized controlled trials were searched for using methodological search
27
filters designed to limit searches to these study designs. These were devised
28
by the Centre for Reviews and Dissemination and the Cochrane
29
Collaboration. The economic literature was identified by conducting searches
30
in NHS Economic Evaluations Database (NHSEED) and in MEDLINE and
31
EMBASE using an economics search strategy developed by ScHARR at the
32
University of Sheffield.
Medicines concordance: full guideline DRAFT (July 2008) page 42 of 373
DRAFT FOR CONSULTATION
1
Databases of the results of the searches for each question or topic area were
2
created using the bibliographic management software Reference Manager.
3
The search strategies for all questions or topic areas developed for the
4
Medline database are detailed in appendix B. Details of all literature searches
5
for the evidence reviews are available from the NCC-PC. Further references
6
were also suggested by the GDG.
7
In line with the Scope, literature searches were undertaken to produce
8
evidence reviews on each of the following key topics:
9
10
 Interventions to enhance adherence in the context of prescribed
medication
11
 Shared decision-making in the context of prescribed medication
12
 Barriers to adherence and shared decision-making in the context
13
of prescribed medication.
14
Literature searches for the guideline were undertaken for the designated topic.
15
It was decided that individual literature searches for each clinical question
16
would result in a considerable amount of duplicated work, as the retrieved
17
evidence would potentially overlap from question to question.
18
Other more focused literature searches were undertaken as appropriate for
19
some of the key clinical questions. These were indicated by the GDG as key
20
clinical questions that due to their importance and possible impact on clinical
21
practice would require a focused literature search to ensure that no important
22
study would be missed out. These were:
23
24
25
26
27
28
29
30
 What tools are available to help elicit patients information needs
about medicines?
 . What tools are available to help elicit patients beliefs about
medicines?
 How can a practitioner detect whether a patient
agrees/disagrees with recommendation to take medicines?
 How can practitioners elicit patient’s preferences for involvement
in decisions about medicines?
Medicines concordance: full guideline DRAFT (July 2008) page 43 of 373
DRAFT FOR CONSULTATION
 Do interventions to increase patient involvement increase length
1
of the consultation
2
3
 Does change in dosing regime affect adherence
4
 Does drug formulation/packaging affect adherence?
5
 What is the effect of prescription charges/costs on adherence to
prescribed medication
6
7
 How can practitioners assess adherence?
8
 Do medication reviews increase adherence to prescribed
medication?
9
10
11
12
The specific search strategy for each topic area varied and was agreed with
13
the project team (with input from the GDG as necessary). The review
14
parameters were agreed with the GDG and aimed to provide the best
15
available evidence. For further details on the methodology and
16
inclusion/exclusion criteria please see individual evidence reviews.
17
The literature on barriers to shared decision-making and medicine taking,
18
shared decision-making and adherence to medication is not well indexed,
19
therefore, despite the comprehensive and detailed searching, some trials that
20
met our criteria may have been missed.
21
Updated searches were conducted for references published during the course
22
of the guidance development and final update searches were carried out in
23
May and June 2008. Only those studies where recommendations needed
24
substantial revisions were added in detail.
25
In line with NICE Equality scheme additional searches of the literature were
26
undertaken to ensure that general searches had located all evidence relevant
27
to vulnerable groups in the United Kingdom3.
28
3
www.nice.org.uk/aboutnice/howwework/NICEEqualityScheme.jsp
Medicines concordance: full guideline DRAFT (July 2008) page 44 of 373
DRAFT FOR CONSULTATION
1
2.4
Identifying the evidence
2
After the search of titles and abstracts was undertaken, full papers were
3
obtained if they appeared to address the key clinical question. The highest
4
level of evidence was sought. However, other types of quantitative evidence,
5
qualitative evidence and expert formal consensus results were used when
6
randomised control trials were not available. Only English language papers
7
were reviewed. Following a critical review of the full text paper, articles not
8
relevant to the subject in question were excluded. Studies that did not report
9
on relevant outcomes were also excluded.
10
The reasons for rejecting any paper ordered were recorded and details can be
11
seen in Appendix D.
12
2.5
13
From the papers retrieved, the Senior Health Services Research Fellow
14
(SHSRF) and the Health Service Research Fellow (HSRF) synthesized the
15
evidence for each question or questions into a narrative summary. These form
16
the basis of this guideline. Each study was critically appraised using the
17
Institute’s criteria for quality assessment and the information extracted for
18
included studies is given in Appendix C. The content and delivery of
19
interventions was poorly defined in many studies and it was difficult to decide
20
which studies should be included or excluded. The GDG advised on which
21
studies to include and exclude in these circumstances. Background papers,
22
for example those used to describe the concepts used in the guideline, were
23
referenced but not extracted.
24
2.5.1
25
26
When agreeing key clinical questions the GDG discussed the choice of
27
outcomes for each search. A variety of outcomes are currently found in
28
studies on shared decision-making but the outcomes primarily looked at were
29
patient preferences, identification of beliefs and patient agreement to the
30
decision.. Any additional information on factors which may have influenced the
31
study results and had an impact on the wider implementation of an
Critical appraisal of the evidence
Choice of outcomes
Medicines concordance: full guideline DRAFT (July 2008) page 45 of 373
DRAFT FOR CONSULTATION
1
intervention, such as participants’ age, ethnicity or social status; dropout rates
2
and payments or rewards given to participants, were recorded in the evidence
3
tables considered by the GDG. The primary outcome measure for all the
4
evidence reviews on interventions to increase adherence was adherence.
5
Adherence levels was the outcome also for studies examining medication
6
review.
7
8
2.6
Health Economics methods
9
Economic evaluation provides a formal comparison of benefits and harms as
10
well as the costs of alternative health programmes. It helps to identify,
11
measure, value and compare costs and consequences of alternative
12
treatment options. These outcomes are usually synthesised in cost
13
effectiveness (CEA) or cost utility analysis (CUA), which reflect the principle of
14
opportunity costs. For example, if a particular treatment strategy were found to
15
yield little health gain relative to the resources used, then it could be
16
advantageous to re-deploy resources to other activities that yield greater
17
health gain.
18
19
To assess the cost-effectiveness of interventions to increase adherence
20
(Interventions to increase adherence), we conducted a comprehensive
21
systematic review of the economic literature relating to medicines and
22
nonadherence.
23
24
In accordance with the NICE social value judgement the primary criteria
25
applied for an intervention to be considered cost effective were either:
26
27
a) The intervention dominated other relevant strategies (that is it is both less
28
costly in terms of resource use and more clinically effective compared with the
29
other relevant alternative strategies); or
30
31
b) The intervention cost less than ￡20,000 per quality-adjusted life-year
32
(QALY) gained compared with the next best strategy (or usual care)
Medicines concordance: full guideline DRAFT (July 2008) page 46 of 373
DRAFT FOR CONSULTATION
1
Health Economic evidence review methodology
2
2.6.1
3
The following information sources were searched:
4
•
Medline (Ovid) (1966-June 2006)
5
•
Embase (1980-June 2006)
6
•
NHS Economic Evaluations Database (NHS EED)
7
•
PsycINFO
8
•
Cumulative Index to Nursing and Allied Health Literature (CINAHL)
9
10
The electronic search strategies were developed in Medline and adapted for
11
use with the other information databases. The clinical search strategy was
12
supplemented with economic search terms. Titles and abstracts retrieved
13
were subjected to an inclusion/exclusion criterion and relevant papers were
14
ordered. No criteria for study design were imposed a priori. In this way the
15
searches were not constrained to randomised controlled trials (RCTs)
16
containing formal economic evaluations. Papers included were:
17
•
Full/partial economic evaluations
18
•
Considered patients over 16 years of age
19
•
Written in English, and reported health economic information that could
20
be generalised to UK.
21
22
The full papers were critically appraised by a health economist using a
23
standard validated checklist. A general descriptive overview of the studies,
24
their quality, and conclusions was presented and summarised in the form of a
25
narrative review.
26
Each study was categorized as one of the following: cost-effectiveness
27
analysis or cost utility analysis (i.e. cost-effectiveness analysis with
28
effectiveness measured in terms of QALYs or life year gained). Some studies
29
were categorised as ‘cost consequences analysis’ or ‘cost minimisation
30
analysis’. These studies did not provide an overall measure of health gain or
31
attempt to synthesise costs and benefits. Such studies were considered as
32
partial economic evaluations.
Medicines concordance: full guideline DRAFT (July 2008) page 47 of 373
DRAFT FOR CONSULTATION
Cost-effectiveness modelling methods
1
2.6.2
2
De novo modelling was considered for aspects of medicine taking. However,
3
due to heterogeneity in the population covered by this guideline this was not
4
possible. This is discussed in more detail in Chapter 10.
5
2.7
6
In preparation for each meeting, the narrative and extractions for the
7
questions being discussed were made available to the GDG one week before
8
the scheduled GDG meeting. These documents were available on a closed
9
intranet site and sent by post to those members who requested it.
Forming recommendations
10
GDG members were expected to have read the narratives and extractions
11
before attending each meeting. The GDG discussed the evidence at the
12
meeting and agreed evidence statements and recommendations. Any
13
changes were made to the electronic version of the text on a laptop and
14
projected onto a screen until the GDG were satisfied with these.
15
All work from the meetings was posted on the closed intranet site following the
16
meeting as a matter of record and for referral by the GDG members.
17
2.8
18
The table of clinical questions in Appendix B indicates which questions were
19
searched.
20
In cases where evidence was sparse, the GDG derived the recommendations
21
via informal consensus methods, using extrapolated evidence where
22
appropriate. All details of how the recommendations were derived can be
23
seen in the ‘Evidence to recommendations’ section of each of the chapters.
24
2.9
25
The guideline has been developed in accordance with the Institute’s guideline
26
development process. This has included allowing registered stakeholders the
27
opportunity to comment on the scope of the guideline and the draft of the full
28
and short form guideline. In addition, the draft was reviewed by an
29
independent Guideline Review Panel (GRP) established by the Institute.
Areas without evidence and consensus methodology
Consultation
Medicines concordance: full guideline DRAFT (July 2008) page 48 of 373
DRAFT FOR CONSULTATION
1
The comments made by the stakeholders, peer reviewers and the GRP were
2
collated and presented for consideration by the GDG. All comments were
3
considered systematically by the GDG and the development team recorded
4
the agreed responses.
5
6
7
2.10
Relationships between the guideline and other national
guidance
8
2.10.1 National Service Frameworks
9
The National Service Framework for Older People (2001) makes specific
10
recommendations for medication review in older people.
11
2.10.2 Related NICE Guidance
12
This guideline differs from most NICE guidelines in that it is not condition
13
specific but makes recommendations on how to involve patients in decisions
14
about medicines. This guidance should be used in conjunction with condition
15
specific NICE guidance which makes recommendations on what treatments
16
are clinically and cost effective. NICE and the National Patient Safety Agency
17
(NPSA) have recently produced joint guidance on medicines reconciliation
18
when adult patients are admitted to hospital (www.NICE.org.uk/PSG001).
19
2.10.3 Other national guidance
20
In formulating recommendations consideration was given to
21
22
23
24

Report of the Committee on Safety of Medicines Working Group
on Patient Information’ (2005)
 General Medical Council document, Consent: doctors and
patients making decision together
25
http://www.gmc-uk.org/news/articles/Consent_guidance.pdf
26
 Mental Capacity Act 2005
27
http://www.opsi.gov.uk/ACTS/acts2005/pdf/ukpga_20050009_en
28
.pdf
29
 Disability Discrimination Act 1995
Medicines concordance: full guideline DRAFT (July 2008) page 49 of 373
DRAFT FOR CONSULTATION
1
http://www.opsi.gov.uk/acts/acts1995/plain/ukpga_19950050_en_1
2
3
Through review of published guidance, personal contact and commenting on
4
guideline scope, endeavours were made to ensure that boundaries between
5
guidance were clear and advice was consistent.
6
Medicines concordance: full guideline DRAFT (July 2008) page 50 of 373
DRAFT FOR CONSULTATION
1
3
Principles and concepts used in the
development of the guideline
2
3
Clinical guidelines generally provide guidance on the management of clinical
4
conditions. To inform the recommendations evidence is sought regarding the
5
benefits and harms, as well as cost of treatments. This guideline seeks to
6
inform patient involvement in decisions about medicines across clinical areas
7
and as such is more interested in patient and health care professional
8
behaviour than evidence for individual treatments. The development of the
9
guideline required the Guideline Development Group (GDG) to engage with
10
topics more usually found in psychological and sociological literature as well
11
as philosophical and legal issues such as rights and duties of patients and
12
professionals. The GDG discussed these issues at length to develop a
13
working consensus which then allowed them to examine the literature and
14
develop recommendations. These discussions are included in the relevant
15
sections of the guideline where they informed recommendations. The GDG
16
also wished to see the principles they used to develop the guideline brought
17
together in one chapter. This chapter brings together those discussions from
18
different parts of the guideline. Some of this content is therefore repeated in
19
different chapters and this is indicated where appropriate by the use of double
20
asterisks at start and end of sections. **In the final version of the guideline we
21
will use hyperlinks to link these sections**
22
3.1
23
Patients’ rights to be involved in decisions about
medicines
24
The prescribing of medicines to patients has become a central part of the
25
delivery of modern medical care. Studies and commentaries on medicine-
26
taking by patients have often emphasized the objective health and cost
27
impacts incurred when patients do not take medicine as prescribed (World
28
Health Organization, 2003). There is an often unstated and perhaps
29
unrecognised assumption that patients should take medicines as suggested
30
by medical professionals (Pound, P., Britten, N., Morgan, M. et al , 2005).
31
While objective health and cost impacts of patients’ behaviour in relation to
32
medicines is important, the right of patients to make decisions in regard to
Medicines concordance: full guideline DRAFT (July 2008) page 51 of 373
DRAFT FOR CONSULTATION
1
their own health is accepted in modern practice. The approach taken by this
2
guideline is that patients have a right to be involved in decisions about
3
medicines to the extent that they wish and it is the role of health professionals
4
to facilitate and support patients in their involvement in decision-making and to
5
support patients in taking medicine if the decision has been to prescribe.
6
It is particularly important for people who are known to frequently experience
7
inequalities in health to have their right recognized to be effectively engaged
8
in decision-making e.g. those with learning disabilities, mental health
9
problems and people of black and ethnic minority ethnic origin. These
10
individuals must be afforded equal opportunities for healthier outcomes
11
through the effective provision of appropriate access and support.
12
Practitioners must be aware of their legal duties and responsibilities to make
13
‘reasonable adjustments’ in line with the Disability Discrimination Act (2005).
14
15
16
3.2
What is meant by involving patients in decisions about
medicines?
17
Early analysis of consultations between medical doctors and patients
18
indicated that consultations were primarily led by the health care professional.
19
Bain (1976) (Bain, DJG, 1976) a general practitioner, tape-recorded his own
20
consultations to examine what actually happened in consultations and found
21
that he talked at least as much as the patients did. Tuckett and colleagues
22
(1985) (Tuckett, D, Boutlon, M, Olson, C et al , 1985) found that doctors
23
frequently inhibited patients from asking questions and did little to encourage
24
patients to present their own view. Historically medical professionals have had
25
the dominant role in making treatment decisions. Charles and colleagues
26
(1999) (Charles, C., Gafni, A., and Whelan, T., 1999) outline a number of
27
assumptions on which this dominant role was based: a single best treatment
28
existed and physicians would be well versed in current clinical thinking;
29
physicians would apply this knowledge consistently to all patients and were in
30
best position to evaluate treatment decisions and tradeoffs; and because of
31
their professional concern for the welfare of their patients, physicians had a
32
legitimate interest in each treatment decision. Significant asymmetry between
Medicines concordance: full guideline DRAFT (July 2008) page 52 of 373
DRAFT FOR CONSULTATION
1
professionals and patients also existed in terms of education, income and
2
status. The assumptions underlying the dominant role of the professional have
3
been increasingly challenged as both medicine and society changed (Charles,
4
C., Gafni, A., and Whelan, T., 1997). More treatments for conditions have
5
become available with complex risk – benefit analyses required. It was
6
recognised that it is the patient who has to live with the consequences of
7
these decisions and might be in a better position than the professional to
8
evaluate and weigh these. Medical practice has also shifted away from acute
9
care towards both chronic long term care and preventative care which
10
requires a long-term commitment to medicines taking and may require
11
frequent monitoring of medications and illness. The principles of informed
12
consent and informed choice in relation to treatment decisions are now
13
enshrined in law and there is an inherent implication in these that the patient
14
is making a decision in relation to their own healthcare. Previous asymmetries
15
between health professionals and patients in terms of education and access
16
to information have also lessened.
17
The concepts of shared-decision making and patient-centredness are part of
18
the response to the need to recognise the role of the patient in medical
19
encounters and decisions. In the literature shared-decision making (SDM) is
20
described as one model of decision making. In this model information
21
exchange is a two way process in the consultation and both deliberation and
22
decision are made by both health care professional and patient. This is in
23
contrast to a ‘paternalistic’ model where information is given to the patient and
24
deliberation and decision are made by the health care professional or an
25
‘informed’ model where information is given to the patient and the patient
26
makes the deliberation and decision (Charles, C., Gafni, A., and Whelan, T.,
27
1999). Intermediate models are also recognised where decision may be led by
28
the professional or handed to the professional following full sharing of
29
information between both parties. Patient-centredness is an approach to the
30
patient which encompasses the sharing of power and responsibility. Mead and
31
Bower (2000) (Mead, N and Power, P, 2000) have described patient-
32
centredness as having 5 dimensions (1) adopting the biopsychosocial
33
perspective (this means using biological, psychological and social
Medicines concordance: full guideline DRAFT (July 2008) page 53 of 373
DRAFT FOR CONSULTATION
1
understandings of disease and illness experience); (2) understanding the
2
patient as person; (3) sharing power and responsibility between the doctor
3
and patient; (4) building a therapeutic alliance and (5) understanding the
4
doctor as a person. Despite the interest in these concepts and an agreement
5
that there is a moral value inherent in them, it is accepted by many working in
6
the area that the operationalisation of these concepts is still evolving (Elwyn,
7
G., Edwards, A., Mowle, S. et al , 2001), (Charles, C., Gafni, A., and Whelan,
8
T., 1997), (Duggan, P. S., Geller, G., Cooper, L. A. et al , 2006). The
9
difficulties relating to how to enact shared decision making include
10
overcoming asymmetry in knowledge and experience between patients and
11
health care professional, the difficulty in recognising a shared decision and
12
what this concept means in terms of responsibility of the clinician.
13
Current evidence exploring health care professionals views indicate that they
14
perceive difficulties in implementing SDM. A recent systematic review which
15
examined qualitative and quantitative research on health care professionals’
16
views about implementing SDM identified several perceived barriers (Gravel
17
2006) (Gravel, Karine., Legare, F., and Graham, I. D., 2006). The studies
18
included in the systematic review were primarily qualitative and the vast
19
majority of the practitioners were medical practitioners. The three most
20
commonly perceived barriers were time constraints, lack of applicability to
21
patient characteristics and lack of applicability to the clinical situation.
22
Thompson (2007) (Thompson, A. G., 2007) describes the current literature as
23
being primarily derived from the perspective of professionals – policy makers,
24
academics or medical professionals and not from the perspective of patients.
25
Using interviews and focus groups he describes patients wish for involvement
26
to be dynamic with demand varying according to the need for health care,
27
personal characteristics of the patient and the patient- professional
28
relationship. Wish for involvement is higher in this model when illness is
29
chronic, trust in the professional is low and the patient is active. Demand in
30
lower when illness is acute and serious, the patient passive and trust in the
31
professional is high.
Medicines concordance: full guideline DRAFT (July 2008) page 54 of 373
DRAFT FOR CONSULTATION
1
Surveys of patients indicate that when asked patients as a group do ask for
2
increased involvement in healthcare decisions. The Picker Institute (2007)
3
(Richards, C and Coulter, A, 2007) published a report examining patient
4
experience of the NHS between 2002 and 2006. Using the results of 26
5
national patient surveys they report patients’ information needs not being met
6
and patients wanting more involvement in health care decisions. In regard to
7
medicines, patients reported wanting more involvement in medication choices
8
than they currently receive. The surveys indicate that professionals were
9
giving less information about side effects over the time of the surveys and
10
patients felt that fewer decisions were shared decisions in 2006 than in 2004
11
and 2005.
12
The literature on decision-making first evolved in the context of life-threatening
13
diseases such as cancer (Charles, C., Gafni, A., and Whelan, T., 1997) and
14
included decisions for example as to whether or not to have surgery (Murray,
15
E., Charles, C., and Gafni, A., 2006).These were often one off decisions with
16
major consequences. In the case of medicines the initial involvement in the
17
decision to prescribe a medicine is necessarily followed by ongoing, often
18
daily decisions by patients to continue taking the medicine prescribed.
19
Involving patients in decisions to take medicines concerns not just the
20
decisions made within a consultation but also attention to the ongoing
21
decisions that patients about their medicines following the consultation with a
22
health care professional.
23
3.3
24
What are we trying to achieve in involving patients in
decisions about medicines?
25
The outcome that we wish to see from patient involvement is an informed
26
choice by the patient in regard to their use of medicines. The term informed
27
adherence has been used to describe an outcome of informed choice and
28
supported adherence (Horne, R and Weinman, J., 2004). In this
29
understanding achieving a shared agreement is limited if the patient is then
30
not supported to implement their intentions to take the medication as
31
recommended. Similarly, stipulating unconditional and unquestioning
32
adherence to prescribers’ instructions as our goal is, in most cases, not
Medicines concordance: full guideline DRAFT (July 2008) page 55 of 373
DRAFT FOR CONSULTATION
1
justified if the patient has not made an informed choice about taking the
2
medication.
3
In most cases the patient is free to decide whether to take the treatment or
4
not. However, the healthcare practitioner has a responsibility to help ensure
5
that the choice is an informed one. Informed patient choice, rather than
6
‘compliance’ is the desired outcome of the discussion. If the patient decides to
7
accept the prescription, then the aim is to facilitate appropriate adherence to
8
the agreed recommendations for how it should be taken to maximise its
9
efficacy and safety for the individual and optimise benefits and reduce risk.
10
Facilitating informed choice involves more than the provision of information.
11
Informing should be an active process, which involves more than simply
12
presenting the evidence. It also entails eliciting the patient’s beliefs and
13
identifying whether pre-existing beliefs might act as a barrier to an accurate
14
interpretation of the evidence. If the interpretation of information is influenced
15
by misconceptions about the illness and treatment, then the patient’s choice
16
may not be ‘informed’.
17
18
3.4
Roles and responsibilities of patients and health care
professionals
19
20
Concern has been expressed by practitioners that sharing decisions with
21
patients may conflict with their duty of care to patients or their legal or ethical
22
obligations (Stevenson, F. A., 2003). While the UK General Medical Council
23
(GMC) (2001)4 considered one of the key duties of a doctor to ‘respect the
24
rights of patients to be fully involved in decisions about their care’ for many
25
clinicians there is a legitimate area of concern or indeed of conflict between
26
respect for autonomy of the patient and the duty of beneficence when a
27
clinician feels uncomfortable about a patient’s wishes. The GMC (2008)5 has
28
recently updated guidance for doctors about patient consent which makes
29
explicit the right of competent patients to make decisions about their own
4
5
http://www.gmc-uk.org/publications/annual_reports/review_archive/report2002.pdf
http://www.gmc-uk.org/publications/business_plans/Business_Plan_2008.pdf
Medicines concordance: full guideline DRAFT (July 2008) page 56 of 373
DRAFT FOR CONSULTATION
1
healthcare. The guidance emphasises the need for doctors to maximize
2
opportunities and patients abilities to make decisions for themselves and that
3
doctors must respect competent patients’ decisions even if they do not agree
4
with them. Doctors do not have to provide a treatment to patients which they
5
believe is not in the patients’ interest but must under such circumstances
6
explain their reasons and other options, including a second opinion to the
7
patient. It remains important however for health care professionals to do their
8
utmost to ensure that patients’ choices are informed choices as outlined
9
above.
10
The Mental Capacity Act (2005) 6governs the making of decisions for people
11
who lack capacity in England and Wales. Under the Act doctors are advised
12
that they must work on the presumption that every adult patient has the
13
capacity to make decisions about their care, and to decide whether to agree
14
to, or refuse, an examination, investigation or treatment. A patient is regarded
15
as lacking capacity once it is clear that, having been given all appropriate help
16
and support, they cannot understand, retain, use or weigh up the information
17
needed to make that decision, or communicate their wishes7.
18
Doctors are advised that assumptions must not be made that a patient lacks
19
capacity to make a decision solely because of their age, disability,
20
appearance, behaviour, medical condition (including mental illness), their
21
beliefs, their apparent inability to communicate, or the fact that they make a
22
decision that health professionals disagree with8.
23
3.5
Understanding the influences on medicine taking by
patients
24
25
If health care professionals are to facilitate patient involvement in decisions
26
about medicines it is helpful and necessary to understand how patients
27
approach the taking of medicines, in particular the ongoing appraisal of
28
medicines that continues after a consultation. Investigation into why patients
6
7
http://www.opsi.gov.uk/ACTS/acts2005/pdf/ukpga_20050009_en.pdf
http://www.gmc-uk.org/publications/business_plans/Business_Plan_2008.pdf
8
http://www.gmc-uk.org/publications/business_plans/Business_Plan_2008.pdf
Medicines concordance: full guideline DRAFT (July 2008) page 57 of 373
DRAFT FOR CONSULTATION
1
do not take medicines as prescribed indicates that the decision to take
2
medicines and the continuing taking of medicines should be considered as a
3
complex behaviour. Fig 1 indicates a diagrammatic representation of current
4
evidence regarding factors and influences on medicine taking by patients. As
5
the figure shows there are a number of influences on patients. Many of these
6
factors interact and the arrows indicate the dynamic nature of the process.
7
Internal factors represent the beliefs and experiences of the patient. These
8
include the patient’s beliefs about their symptoms or disease, their beliefs
9
about medicines in general and their own reaction to medicines. These will
10
influence their intention to take a medicine as suggested by a health care
11
professional. Even when patients intend to take a medicine they can
12
experience difficulty in doing so because of practical problems such as
13
packaging or complexity of regime or they may forget to take medicines.
14
Patients conduct their own appraisal of the medicine they are taking and
15
evaluate its effects against their own expectations of what the medicine will
16
achieve. This feeds into their beliefs about their medicines which in turn
17
influences their intention to take or not to take the prescribed medicines.
18
External factors are those that feed into the patient’s internal appraisal
19
process. These include communication with family and friends and the
20
communication they have with their health care professionals. Information
21
about medicines will be available from multiple sources including
22
documentation patients gets with their medicines, from the pharmacist or
23
dispenser or from any other health care professional the patients comes into
24
contact with. Patients may get information from other patients who have taken
25
the same medicine. Patients may be influenced by the attitude in society to a
26
particular medicine or medicines in general and information may be received
27
from media sources.
28
The research evidence indicates that patients’ decisions about medicines are
29
made within the patients’ own frames of reference and make sense within the
30
patients’ understanding. Patients however often do not disclose to the health
31
professional any reluctance to take medicines or disagreement with the
32
doctor’s recommendation of a prescription or their non-taking of medicines.
Medicines concordance: full guideline DRAFT (July 2008) page 58 of 373
DRAFT FOR CONSULTATION
1
The onus is on the health professional to elicit and explore patients’ beliefs
2
and experiences and facilitate the patient making an informed choice about
3
whether or not to take a prescribed medicine.
4
When patients do not take medication as prescribed they may therefore be
5
doing this on an intentional basis i.e. they have made their own appraisal and
6
have decided to take the medicine in their own way or even not at all. This
7
may be done with full information about medicines, illness and consequences
8
of taking or not taking recommended medication.
INTERNAL FACTORS
APPRAISAL:
Assessment and interpretation of outcomes relative to expectations: - Did it work? Is it worth continuing?
Beliefs about
prescribed
medicine(s)
Perceptions of
illness
Symptom
interpretations
Background beliefs
about pharmaceuticals
as a class of treatment
INTENTION
TO TAKE
ADHERENCE/
NONADHERENCE
MEDICATION
Treatment preferences
PRACTICAL ABILITIEScapacity and resources
Perceptions of personal sensitivity to
medicines
Barriers include forgetting
and regimen complexity
EXTERNAL FACTORS eg
-Information
-Communication (Health care professionals, friends etc.)
-Media
-Cultural influences
cultural factors
healthcare policy
social support
9
10
Figure 1……………………………….Horne, R. Concordance, Adherence and
11
Compliance in Medicine-Taking. Report for the National Co-ordinating Centre
12
for NHS Service and Delivery Organisation R& D (NCCSDO) (2005), pp 139.
13
In routine clinical terms the factors included in figure 1 that are barriers to
14
medicine-taking can be considered as either practical barriers or perceptual
15
barriers. Perceptual barriers are ways in which individual patients think about
Medicines concordance: full guideline DRAFT (July 2008) page 59 of 373
DRAFT FOR CONSULTATION
1
their illness or condition and treatments both in general and specifically.
2
Practical barriers are those such as cost, memory or dexterity that affect
3
individuals ability to use the medication recommended to them.
4
3.6
5
The terminology used in the area of medicine-taking highlights the changing
6
understanding of medicine taking behaviour and the changing relationships
7
between health care professionals and patients. The terms compliance,
8
adherence and concordance are now often used interchangeably and
9
inappropriately to describe medicine taking by patients. The approach taken
10
by this guideline is to use terminology as recommended in the Report for the
11
National Co-ordinating Centre for NHS Service Delivery and Organisation R &
12
D on Concordance, Adherence and Compliance in Medicine Taking (2005)
13
(NCCSDO) (Horne, R, Weinman, J., Barber, N. et al , 2005).
14
Compliance has been the most commonly used term in relation to medicine-
15
taking and can be defined as ‘the extent to which the patients’ behaviour
16
matches the prescriber’s recommendations’ (Haynes, T. B., Taylor, D. W., and
17
Sackett, D. L., 1979). The term has been criticised as it is suggested it carries
18
an implicit assumption that it is the prescriber’s role to decide on the correct
19
medication and the patient has a passive role which is to take the medication
20
as he/she has been instructed. Non-compliance indicates a failure to follow
21
instructions and can be used as a pejorative term indicating a patient who is
22
unwilling to do as instructed by a prescriber who knows what the patient
23
needs.
24
Adherence is defined as ‘the extent to which the patient’s behaviour matches
25
agreed recommendations from the prescriber’. Adherence shifts the balance
26
between doctor and patient to suggest there should be agreement between
27
doctor and patient about the prescriber’s recommendation.
28
Concordance is less easily defined and its meaning has changed. It was
29
initially used to describe the consultation process by which agreement about
30
therapeutic decisions was reached by doctors and patients. It presumed an
31
exploration of patients’ views and their incorporation into the decision made.
Terminology and structure of guideline
Medicines concordance: full guideline DRAFT (July 2008) page 60 of 373
DRAFT FOR CONSULTATION
1
The term therefore addresses consultation processes and does not include
2
any aspects of medicine taking as do compliance and adherence. Some uses
3
of the term has included both communication and support for patients in
4
medicine taking. While the term concordance has been useful in drawing
5
attention to the need to engage patients in decisions there is currently no
6
agreed way in which one can judge whether a consultation has been
7
concordant. Concordance does reflect current social values where patients
8
are seen as active participants in their own healthcare but does not address
9
medicine-taking and may or may not lead to improved adherence.
10
We have chosen to discuss the consultation process regarding medicines
11
separately from actual medicine taking. We use the term ‘shared decision-
12
making about medicines’ to refer to the healthcare professional–patient/carer
13
consultation and the term adherence to describe patients’ medicine taking
14
behaviour. The guideline has looked separately at evidence about
15
interventions to increase patient involvement in the decision to take medicine
16
within the consultation and at evidence about interventions to support patients
17
in taking of medicines. While this division is useful when examining the
18
literature and making recommendations, the dynamic nature of medicine
19
taking must not be forgotten. Patients’ perceptions and beliefs will change
20
over time and the experience of taking a medicine will also influence patients’
21
intentions to continue taking that medicine and others prescribed. Fig 1
22
provides a representation of the patient’s pathway. The patient comes to the
23
consultation with their own beliefs and experiences (see section 5). In the
24
meeting with the health professional the patient’s complaint is assessed and a
25
prescription may be recommended by the health professional. Within the
26
consultation the decision as to whether or not the patient leaves with a
27
prescription may be led completely by the doctor or negotiated between
28
doctor and patient.. If the patient leaves the consultation with a prescription
29
they may or may not take the prescription to be dispensed, and even if the
30
medicine is dispensed they may or may not take the medicine.
31
Medicines concordance: full guideline DRAFT (July 2008) page 61 of 373
DRAFT FOR CONSULTATION
P-HP
encounter
Before
consultation
P
HP
P-D encounter
Medicine-taking
Takes medicine
P
P
D
P
Medicine
prescribed but
not dispensed
P-Patient
Medicine
dispensed
but not taken
HP Healthcare
Professional
D- Dispensing
professional
Arrows to indicate how patients attitudes are influenced by previous
experiences of medicines and medicine taking
1
2
Fig.1 Simplified representation of patient pathway
3
3.7
4
Models of shared decision-making for use in clinical practice have been
5
developed. Towle (1997) (Towle, A., 1997) suggested steps for shared
6
decision making and these have been adapted by Elwyn and colleagues
7
(1999) (Elwyn, G., Edwards, A., Dwyn, R et al , 1999) following exploratory
8
research with general practitioner registrars. They suggest that population
9
surveys cannot predict preference of individual patients for involvement and
10
patients’ preferences for involvement may vary according to clinical situation
11
so the involvement of patients in decision-making has to be explicitly
12
addressed. Neither can patients consider their role in actual making of a
Shared decision-making about medicines
Medicines concordance: full guideline DRAFT (July 2008) page 62 of 373
DRAFT FOR CONSULTATION
1
decision until they have information about their options and the risks and
2
benefits of those options. The following stages are suggested by Elwyn and
3
colleagues to involve patients in healthcare decisions and by implication this
4
also describes the competencies required by practitioners to involve patients
5
(Elwyn 1999) (Elwyn, G., Edwards, A., Dwyn, R et al , 1999).
6

Implicit/explicit involvement of patients in decision-making process
7

Explore ideas/fears and expectations of problem and treatments
8

Portrayal of options
9

Identify preferred format and provide information
10

Checking process: understanding of information and reactions
11

Acceptance of process and role in decision-making
12

Make, discuss, or defer decisions
13

Arrange follow up
14
While these models have been developed when considering a variety of
15
decisions we have used this model to provide an outline structure for our
16
discussion and recommendations about sharing decisions about medicines in
17
consultations. While a treatment can never be understood without reference to
18
the underlying disease, illness or symptom it is beyond the scope of this
19
guideline to make recommendations about general communication and about
20
how diagnosis and prognosis should be explained to patients. These do
21
overlap with our recommendations on communication about medicines but we
22
have not examined these areas explicitly.
23
Although we have used the term SDM in this guideline and have used
24
literature relating to SDM the understanding of the Guideline Development
25
Group is that this is a process and that we are addressing the ‘sharing’ of
26
decisions rather than defining what a shared decision is. Edwards and Elwyn
27
(2006) (Edwards, A. and Elwyn, G., 2006) have suggested that it is the
Medicines concordance: full guideline DRAFT (July 2008) page 63 of 373
DRAFT FOR CONSULTATION
1
process of involving patients in decisions that delivers benefits for patients
2
rather than attaching importance to defining who makes the decision. Given
3
the difficulties in defining precisely what a shared decision is we cannot
4
advocate that the outcome from this process has to be a ‘shared – decision’.
5
For many patients this may be the preferred outcome, other patients will
6
prefer to give the decision to the professionals and others to make their own
7
‘informed’ decision. Cox (2007) (Cox, K., Britten, N., Hooper, R. et al , 2007)
8
in a study about prescribing of medicines in general practice, found that 39%
9
of patients wanted the G.P to share the decision, 28% wanting the G.P to be
10
main decision-maker, 17% wanted the GP to be the only decision-maker, 14%
11
preferred that the patient be the main decision-maker and 2% the only
12
decision-maker. Given the evidence that patient involvement in choices about
13
medicine is lower than desired by patients and that doctors are not good at
14
recognising which patients want involvement our recommendations aim to
15
make the process of involvement more explicit and to increase overall patient
16
involvement.
17
Medicines concordance: full guideline DRAFT (July 2008) page 64 of 373
DRAFT FOR CONSULTATION
1
2
4
making about medicines
3
4
Interventions to increase shared decision-
4.1
Recommendations *listed at the start of the guideline
according to chapter
5
6
4.2
Introduction
7
Shared decision making can be broadly defined ‘as a decision-making
8
process jointly shared by patients and their health care provider’ (Legare
9
2008) (F Légaré, S Ratté D Stacey J Kryworuchko K Gravel L Turcot ID
10
Graham, 2008). As discussed in chapter 3 the concept of shared decision-
11
making evolved as a development from a predominantly paternalistic model of
12
professional and patient interactions. Makoul and Clayman (2006) (Makoul,
13
Gregory and Clayman, Marla L., 2006) found thait the most commonly
14
occurring themes in discussion of shared decision making were patient values
15
and preferences, options, partnership and patient participation, with 17 other
16
concepts also given considerable weight. In a review of communication about
17
medicines Cox (2004) (Cox, F, Stevenson, N, Britten, N et al , 2004) sets out
18
how patients and professionals need to have two way discussions in which
19
they exchange information and views about medicines.
20
** This section is a repeat of section 3.4 from chapter 3 **
21
**The outcome that we wish to see from patient involvement is an informed
22
choice by the patient in regard to their use of medicines. The term informed
23
adherence has been used to describe an outcome of informed choice and
24
supported adherence (Horne, R and Weinman, J., 2004). In this
25
understanding achieving a shared agreement is limited if the patient is then
26
not supported to implement their intentions to take the medication as
27
recommended. Similarly, stipulating unconditional and unquestioning
28
adherence to prescribers’ instructions as our goal is, in most cases, not
29
justified if the patient has not made an informed choice about taking the
30
medication.
Medicines concordance: full guideline DRAFT (July 2008) page 65 of 373
DRAFT FOR CONSULTATION
1
In most cases the patient is free to decide whether to take the treatment or
2
not. However, the healthcare practitioner has a responsibility to help ensure
3
that the choice is an informed one. Informed patient choice, rather than
4
‘compliance’ is the desired outcome of the discussion. If the patient decides to
5
accept the prescription, then the aim is to facilitate appropriate adherence to
6
the agreed recommendations for how it should be taken to maximise its
7
efficacy and safety for the individual and optimise benefits and reduce risk.
8
Facilitating informed choice involves more than the provision of information.
9
Informing should be an active process, which involves more than simply
10
presenting the evidence. It also entails eliciting the patient’s beliefs and
11
identifying whether pre-existing beliefs might act as a barrier to an accurate
12
interpretation of the evidence. If the interpretation of information is influenced
13
by misconceptions about the illness and treatment, then the patient’s choice
14
may not be ‘informed’.**
15
This chapter outlines research evidence, the deliberations of the Guideline
16
Development Group and their recommendations in relation to patient
17
involvement in decision-making about medicines. The chapter is organised
18
around the key clinical questions in this area. These explored the process of
19
patient involvement . Involving patients in decision about medicines involves
20
attention to the process of patient involvement but is also informed by what is
21
known about patient views and experiences of medicines and medicine-
22
taking. Research studies have explored patients’ views about medicines and
23
although this information is not a substitute for discussion with individual
24
patients, it can sensitise health care professionals to how patients’ approach
25
medicine taking. The evidence about patient experience is presented
26
separately in chapter 5 for clarity but there is inevitable overlap between
27
recommendations developed from these separate evidence reviews.
28
4.3
29
** This section is a repeat of section 3.7 from chapter 3 **
30
**Models of shared decision-making for use in clinical practice have been
31
developed. Towle (1997) (Towle, A., 1997) suggested steps for shared
Process of shared decision-making
Medicines concordance: full guideline DRAFT (July 2008) page 66 of 373
DRAFT FOR CONSULTATION
1
decision making and these have been adapted by Elwyn and colleagues
2
(1999) (Elwyn, G., Edwards, A., Dwyn, R et al , 1999) following exploratory
3
research with general practitioner registrars. They suggest that population
4
surveys cannot predict preference of individual patients for involvement and
5
patients’ preferences for involvement may vary according to clinical situation
6
so the involvement of patients in decision-making has to be explicitly
7
addressed. Neither can patients consider their role in actual making of a
8
decision until they have information about their options and the risks and
9
benefits of those options. The following stages are suggested by Elwyn and
10
colleagues to involve patients in healthcare decisions and by implication this
11
also describes the competencies required by practitioners to involve patients
12
(Elwyn 1999) (Elwyn, G., Edwards, A., Dwyn, R et al , 1999).
13
14
 Implicit/explicit involvement of patients in decision-making
process
15
 Explore ideas/fears and expectations of problem and treatments
16
 Portrayal of options
17
 Identify preferred format and provide information
18
 Checking process: understanding of information and reactions
19
 Acceptance of process and role in decision-making
20
 Make, discuss, or defer decisions
21
 Arrange follow up
22
While these models have been developed when considering a variety of
23
decisions we have used this model to provide an outline structure for our
24
discussion and recommendations about sharing decisions about medicines in
25
consultations. While a treatment can never be understood without reference to
26
the underlying disease, illness or symptom it is beyond the scope of this
27
guideline to make recommendations about general communication and about
28
how diagnosis and prognosis should be explained to patients. These do
Medicines concordance: full guideline DRAFT (July 2008) page 67 of 373
DRAFT FOR CONSULTATION
1
overlap with our recommendations on communication about medicines but we
2
have not examined these areas explicitly.
3
Although we have used the term SDM in this guideline and have used
4
literature relating to SDM the understanding of the Guideline Development
5
Group is that this is a process and that we are addressing the ‘sharing’ of
6
decisions rather than defining what a shared decision is. Edwards and Elwyn
7
(2006) (Edwards, A. and Elwyn, G., 2006) have suggested that it is the
8
process of involving patients in decisions that delivers benefits for patients
9
rather than attaching importance to defining who makes the decision. Given
10
the difficulties in defining precisely what a shared decision is we cannot
11
advocate that the outcome from this process has to be a ‘shared – decision’.
12
For many patients this may be the preferred outcome, other patients will
13
prefer to give the decision to the professionals and others to make their own
14
‘informed’ decision. Cox (2007) (Cox, K., Britten, N., Hooper, R. et al , 2007)
15
in a study about prescribing of medicines in general practice, found that 39%
16
of patients wanted the G.P to share the decision, 28% wanting the G.P to be
17
main decision-maker, 17% wanted the GP to be the only decision-maker, 14%
18
preferred that the patient be the main decision-maker and 2% the only
19
decision-maker. Given the evidence that patient involvement in choices about
20
medicine is lower than desired by patients and that doctors are not good at
21
recognising which patients want involvement our recommendations aim to
22
make the process of involvement more explicit and to increase overall patient
23
involvement.**
24
The GDG were first interested in whether interventions can improve patient
25
involvement in decision-making. The evidence in this area comes from a
26
literature which examines overlapping areas of patient communication,
27
patient-centeredness and shared decision-making. The outcomes from these
28
studies include a variety of patient –orientated outcomes such as knowledge
29
and satisfaction. A scale for measuring patient involvement, the OPTION
30
scale, has been developed and validated in the UK (Elwyn, G., Edwards, A.,
31
Wensing, M. et al , 2003) (Elwyn, G., Edwards, A., Hutchings, A. et al , 2005).
32
The scale has not been used widely and recent comparison of this scale with
Medicines concordance: full guideline DRAFT (July 2008) page 68 of 373
DRAFT FOR CONSULTATION
1
an instrument developed in the US called the Informed Decision Making
2
instrument showed poor overall agreement although good agreement on
3
elements that appeared conceptually similar (Weiss 2008) (Horne, R and
4
Weinman, J., 2004).
5
When exploring the process of shared decision-making on more detail we
6
used the stages and competences outlined by Elwyn (1999) (Elwyn, G.,
7
Edwards, A., Dwyn, R et al , 1999) to structure the evidence.
8
For the purposes of this review the steps of the process of most interest to the
9
GDG were :
10
1. the elicitation of patients preferences for involvement in decision –
making,
11
12
2. .the exploration of patient beliefs,
13
3. the identification of preferred format and provision of information,
14
4. .support for the patient in decision-making
15
5. .the elicitation of patient agreement
16
4.4
Methods
17
Searches were conducted to gather the most relevant evidence relating to the
18
key clinical questions on shared decision-making. Due to time constraints,
19
exhaustive systematic reviews were not conducted (see Cochrane Review
20
Methods). Initially an overall search of relevant Randomised Controlled Trials
21
(RCTs) and Systematic Reviews was conducted using shared decision-
22
making terms. The articles retrieved were then separated under the relevant
23
key clinical questions. However, this did not answer all the Key Clinical
24
Questions succinctly so further searches were done using lower grades of
25
study design.
26
After the key terms were searched we generated a list of abstracts which were
27
sifted to find relevant articles. Full articles of those deemed to have relevance
28
to the question were obtained. These were then further assessed as to their
Medicines concordance: full guideline DRAFT (July 2008) page 69 of 373
DRAFT FOR CONSULTATION
1
match with our inclusion and exclusion criteria for the question and either
2
included or excluded. We were specifically looking for those studies regarding
3
medicine-taking, which ruled out a lot of studies from the oncology and
4
surgical fields.
5
We extracted the evidence from each included article for study quality, then
6
brought together the results into an evidence review for each key clinical
7
question. The evidence reviews were structured into explanatory narratives for
8
each article. These were then combined to provide evidence statements of the
9
overall evidence. It proved difficult to separate out the content of some of the
10
interventions contained in systematic reviews e.g. there is overlap in
11
interventions which explore improving communication between patient and
12
practitioner and those exploring how information should be presented and
13
discussed between practitioner and patient. We have therefore included some
14
studies in more than one section and have indicated this where appropriate.
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Medicines concordance: full guideline DRAFT (July 2008) page 70 of 373
DRAFT FOR CONSULTATION
1
2
4.5
Key Clinical Question: Is it possible to increase patient
involvement in decisions about medicines?
3
Related References
Evidence Statements (summary
Evidence into
of evidence)
recommendations
Patient-centred
communication in
the consultation
Rao (2007) (Rao,
1. One systematic review of
Jaya K., 2007)
interventions to improve
The GDG were interested
communication behaviours of
in what evidence was
physicians and patients found that
available to indicate that
the interventions studied improved
patient involvement in
the patient-centred communication
decisions about
behaviours of physicians and
medicines can be
residents significantly. There were
increased. The evidence
modest effects for the interventions
in this area is complex
to improve patients’ communication
and the GDG requested
behaviours.
further searches to look
at process of patient
Rao (2007) (Rao,
2. In one systematic review there
involvement in more
Jaya K., 2007)
was a mix of significant and non
detail. The GDG were
significant results, depending on
convinced however that
the communication behaviour
the systematic reviews
studied – e.g. 5/7 found significant
indicated that practitioner
changes in patients’
skills could be improved
communication patterns. All of
and that these
these included skills practice as
improvements as well as
part of the intervention, the other 2
some patient
studies were low intensity;
interventions did result in
Rao (2007) (Rao,
3. One systematic review
increased patient
Medicines concordance: full guideline DRAFT (July 2008) page 71 of 373
DRAFT FOR CONSULTATION
Jaya K., 2007)
investigated intensity of
involvement. The GDG
interventions and found that higher
considered it important
intervention intensity was clearly
that practitioners were
related to improvements in
aware that skills could be
physician communications,
improved through further
whereas this was less pronounced
training.
for patient communication.
Lewin (2001) (Lewin,
4. One systematic review found
S. A., Skea, Z. C.,
that some interventions to promote
Entwistle, V. et al ,
patient-centred care in
2001)
consultations may lead to
significant increases in the patient
centredness of consultation
processes. They concluded that
there is limited and mixed evidence
on the effects on health care
behaviours or health status
Patient
involvement in the
consultation
Rao (2007) (Rao,
5. Three systematic reviews found
Jaya K., 2007);
mixed results as to whether or not
(Harrington (2004)
interventions increased patient
(Harrington-Jane,
involvement in the consultation
Noble Lorraine,
(described below). One RCT study
2004); Kinnersley
found an increase in patient
(2007) (Kinnersley,
participation.
P., Edwards, A.,
Hood, K. et al ,
2007); Loh (2007)
(Loh, Andreas,
Medicines concordance: full guideline DRAFT (July 2008) page 72 of 373
DRAFT FOR CONSULTATION
Simon, Daniela,
Wills, Celia E. et al ,
2007)
Harrington (2004)
6. In one systematic review, 10
(Harrington-Jane,
RCTs found a significant increase
Noble Lorraine,
in participation, while 5 RCTs found
2004)
a non-significant increase.
Kinnersley (2007)
7. In one systematic review patient
(Kinnersley, P.,
participation was increased in 8 out
Edwards, A., Hood,
of 14 RCT studies and no effect
K. et al , 2007)
found in 5 RCT studies.
Wetzels (2007)
8. One systematic review found
(Wetzels, R.,
limited evidence of interventions
Harmsen, M., van,
aimed specifically at improving
Weel C. et al , 2007)
older patients’ involvement in
consultations, they found only three
RCTs.
Wetzels (2007)
9. Those RCT studies that did
(Wetzels, R.,
investigate older patients’
Harmsen, M., van,
involvement found that the
Weel C. et al , 2007)
interventions resulted in the
patients asking more and different
questions, and so they may
become more active in
consultations due to pre-visit
preparation.
Medicines concordance: full guideline DRAFT (July 2008) page 73 of 373
DRAFT FOR CONSULTATION
Type of behaviour
evoked
Harrington (2004)
10. Question-asking was found
(Harrington-Jane,
equal in significant and non-
Noble Lorraine,
significant trends in one systematic
2004);
review (Harrington 2004). In
another systematic review
Kinnersley (2007)
(Kinnersley, P.,
Edwards, A., Hood,
K. et al , 2007);
Wetzels (2007)
(Wetzels, R.,
Harmsen, M., van,
Weel C. et al , 2007)
(Kinnersley 2007) 6 out of 17
studies found significant increases
and the other 11 studies found no
significant effects. Another
systematic review (Wetzels 2007)
found one study where more asked
questions and one study where few
prepared questions.
Harrington (2004)
11. In one systematic review there
(Harrington-Jane,
was a significant increase in
Noble Lorraine,
clarifying issues.
2004)
Harrington (2004)
12. One systematic review found a
(Harrington-Jane,
significant increase in patients’
Noble Lorraine,
perception of control over health
2004)
and for an active role in health
care, recall of information,
adherence to recommendations,
attendance and clinical outcomes.
Harrington (2004)
13. One systematic review
(Harrington-Jane,
(Harrington 2004) found a
Noble Lorraine,
significant increase in satisfaction
Medicines concordance: full guideline DRAFT (July 2008) page 74 of 373
DRAFT FOR CONSULTATION
2004); Kinnersley
for only two RCTs however there
(2007) (Kinnersley,
was high satisfaction found overall.
P., Edwards, A.,
In one systematic review
Hood, K. et al ,
(Kinnersley 2007) 14 out of 23
2007); Loh (2007)
studies showed no change and 5
(Loh, Andreas,
had increased satisfaction. In
Simon, Daniela,
another RCT (Loh 2007) Patient
Wills, Celia E. et al ,
satisfaction was significantly higher
2007).
in the intervention 29.8 (2.7) than
the control group 27.0 (3.6),
p=0.14.
Type of
intervention
Rao (2007) (Rao,
14. In four systematic reviews most
Jaya K., 2007);
of the interventions were written,
Harrington (2004)
e.g. booklet/checklists, some were
(Harrington-Jane,
videotapes or face to face
Noble Lorraine,
coaching.
2004); Kinnersley
(2007) (Kinnersley,
P., Edwards, A.,
Hood, K. et al ,
2007); Wetzels
(2007) (Wetzels, R.,
Harmsen, M., van,
Weel C. et al , 2007)
Harrington (2004)
15. One systematic review found
(Harrington-Jane,
that face-to-face and video
Noble Lorraine,
interventions showed more
2004); Kinnersley
significant results than written
(2007) (Kinnersley,
interventions (Harrington 2007).
Medicines concordance: full guideline DRAFT (July 2008) page 75 of 373
DRAFT FOR CONSULTATION
P., Edwards, A.,
Another systematic review found a
Hood, K. et al ,
small to moderate significant
2007)
increase for writing alone and
coaching.
Wetzels (2005)
16. One RCT of a consultation
(Wetzels, R.,
leaflet (including open and pre-
Harmsen, M., van,
structured questions) to improve
Weel C. et al , 2007)
involvement of older patients,
showed an increase in satisfaction
but no effect of the leaflet on
involvement, enablement or
satisfaction. However there was a
significant result of reporting more
psychological symptoms to their
GPs.
Little (2004) (Little,
17. One RCT found that a general
P., Dorward, M.,
leaflet which asked patients to list
Warner, G. et al ,
issues they wanted to raise and
2004)
knowledge that the doctors
expected questions led to a
significant increase in satisfaction
but no other aspects. The leaflet
was significantly more effective
when consultations were short. The
leaflet increased the number of
investigations by the doctor.
Wilkinson (2002)
18. One RCT of an education
(Wilkinson, C. R. and guidebook to encourage and
Williams, M., 2002)
support participation found no
significant differences in the
effectiveness of their primary care
Medicines concordance: full guideline DRAFT (July 2008) page 76 of 373
DRAFT FOR CONSULTATION
visit. There were significant
differences for those receiving
preventative health care
interventions.
Ross (2004) (Ross,
19. One RCT found the use of a
S. E., Moore, L. A.,
patient accessible online medical
Earnest, M. A. et al ,
record found no significant results
2004)
for self-efficacy, adherence, health
status or patient satisfaction.
General adherence to medication
advice was significantly improved
(no data given).
Loh (2007) (Loh,
20. One RCT found no significant
Andreas, Simon,
differences in treatment adherence.
Daniela, Wills, Celia
E. et al , 2007)
1
Methods of the evidence review
2
4.5.1
3
This paper includes a narrative summary of the included evidence, structured
4
according to the category of the intervention, following the agreed reviewing
5
protocol:
6
Types of studies: Systematic Reviews of Randomised Controlled Trials
7
(RCTs) or Randomised Controlled Trials of interventions involving shared
8
decision-making in the clinical context.
9
Types of participants: people prescribed medication for a medical condition.
10
Duration of studies: no time limit specified.
Medicines concordance: full guideline DRAFT (July 2008) page 77 of 373
DRAFT FOR CONSULTATION
1
Types of interventions: any interventions involving shared decision making
2
in a consultation between a health care professional and patient.
3
Types of outcome measures: Patient-centred communication in the
4
consultation; Consultation process outcomes: patient involvement, question
5
asking, preparedness; Patient care outcomes: satisfaction, knowledge, self-
6
efficacy. Type of interventions involved and type of information.
7
8
It should be noted that the remit is for conditions with prescribed medication
9
and this excludes conditions which require chemotherapy or screening. All
10
RCTs are within this remit, however many of the systematic reviews included
11
populations outside the remit, this is noted where applicable.
12
Evidence review
13
4.5.2
14
The narratives for this question are not grouped because they cover various
15
aspects of the question. The evidence review has been constructed under the
16
above subheadings of patient-centred communication, patient involvement,
17
type of behaviour, type of intervention and type of information. It should also
18
be noted that RCTs which were included in the Systematic Review were not
19
narrated separately so as not to duplicate results.
20
Rao (2007) (Rao, Jaya K., 2007) conducted a systematic review of
21
interventions to enhance communication behaviours among physicians and
22
patients in an outpatient setting. They included RCTS of interventions which
23
are designed to improve the communication behaviours of physicians and
24
patients. The primary outcome was an assessment of the patient-centred
25
verbal communication behaviours. They rated the intensity of the interventions
26
based on how often it was delivered and the personnel involved in the
27
delivery. Thirty six RCTs met the inclusion criteria, 18 trained physicians or
28
residents and 15 trained patients, 3 intervened on both. The majority (26
29
RCTs) were conducted in primary care clinics while 10 were conducted in
30
medical specialty settings. Sixteen of the studies were outside of the USA.
31
Half the studies participants were assessed before and after the physician-
32
patient interaction. Most of the studies were conducted on audiotapes or
Medicines concordance: full guideline DRAFT (July 2008) page 78 of 373
DRAFT FOR CONSULTATION
1
videotapes and coders were blinded to group assignment. The
2
physician/resident interventions mainly involved a variety of types (e.g.
3
information, feedback, modelling, practice). Nearly all included written
4
instructional information. Some (10 interventions) included videotapes
5
modelling desirable communication behaviours. Most RCTs showed
6
significant improvements in communication behaviours of the
7
physicians/residents. The higher intensity interventions resulted in physicians
8
asking more open-ended questions and less biomedical questions than the
9
comparison groups. They were more likely to elicit the patients’ concerns they
10
had prior to the visit, and show a more patient-centred communication style
11
overall. There were significant improvements in their information provision to
12
patients. Rao noted that few RCTs (6) checked for patients’ understanding of
13
the information received. 18 studies intervened with patients and showed
14
modest effects. Most of the RCTs were informational (17), often written, some
15
instances the written information included models of desirable communication
16
with examples of questions to ask (8), eight included practicing or coaching
17
sessions and four were feedback. Six were moderately-intense, 2 were rated
18
high-intensity. Seventeen of the RCTs used different measures of patient
19
involvement. 7 assessed the degree to which patients spoke during the visit, 5
20
(moderate intensity interventions) of which showed significant changes in
21
communication patterns. All of these were skills practice interventions. The
22
other two RCTs were low-intensity and showed no significant results. To
23
conclude, the interventions enhanced communication behaviours with
24
physicians compared to controls and there were modest effects with the
25
patient interventions. Intervention intensity had a clear relationship to
26
improving the physicians’ behaviours but this was not so much for patients.
27
Harrington (2004) (Harrington-Jane, Noble Lorraine, 2004) conducted a
28
systematic review which looked at interventions to increase patients’
29
participation in medical consultations. Therefore there were a mix of
30
populations including primary care and outpatient oncology. The inclusion
31
criteria were interventions designed to improve patients’ communication with
32
doctors in any setting; reporting data on the impact of the intervention on the
33
patients’ communication. Most of the studies were RCTs, with the others
Medicines concordance: full guideline DRAFT (July 2008) page 79 of 373
DRAFT FOR CONSULTATION
1
being Quasi-experimental. Twenty five papers were retrieved from the search.
2
Most of the studies were from the USA, 2 from Australia, 5 from the UK and
3
one from the Netherlands. Most of the interventions were written followed by
4
face-to-face coaching and videotape. Written interventions were in booklet or
5
checklist form. The specific behaviours most encouraged were question-
6
asking, raising concerns and requesting clarification or checking
7
understanding. The process of communication was measured mostly using
8
interaction analysis from audio recordings. Researchers coding the
9
interactions were blinded in only six of the studies. Overall the effect of
10
interventions was that they encouraged patients to be more active in their
11
consultations. Of the 16 studies examining variables of participation, 10
12
reported a significant increase and five reported a non-significant increase. All
13
but one of the six face-to-face interventions and all of the video interventions
14
reported significant results in increasing participation. Of the 10 written
15
interventions, only two reported a significant increase. Question-asking was
16
equal in significant increases and non-significant trends. There was a
17
significant increase in requesting clarification on matters. There was no
18
significant increase in satisfaction due to the interventions apart from in two
19
studies. There was overall a high level of satisfaction reported. There was a
20
significant increase in perception of control over health and preferences for an
21
active role in health care, recall of information, adherence to
22
recommendations, attendance and clinical outcomes.
23
It should be noted that Lewis, Butow, Ford, Street and Brown were studies
24
with cancer patients exclusively.
25
Kinnersley (2007) (Kinnersley, P., Edwards, A., Hood, K. et al , 2007)
26
conducted a Cochrane Collaboration systematic review of the effects of
27
interventions before consultations designed to help patients address their
28
information needs. The settings and populations varied, and were from six
29
countries, mainly the USA. They stated that it complemented other Cochrane
30
reviews by Wetzels (2007) (Wetzels, R., Harmsen, M., van, Weel C. et al ,
31
2007) and Lewin (2003). The inclusion criteria were RCTs of interventions
32
intended to help the patients, representatives or carers address their
Medicines concordance: full guideline DRAFT (July 2008) page 80 of 373
DRAFT FOR CONSULTATION
1
information needs in a consultation. This was done by encouraging question
2
asking, to express their information needs, to overcome barriers to
3
communication and to clarify their understanding of the information provided.
4
Outcome measures were the consultation process, consultation outcomes
5
and service outcomes. 33 trials described in 35 studies met the inclusion
6
criteria. From the studies assessing single interventions for patients 15
7
included written materials, four were coaching. The multiple component single
8
interventions studies 4 had coaching and written materials. Results: 17
9
studies measure question asking with 6 finding statistically significant
10
increases and 11 studies finding no effects of the interventions compared to
11
controls. Patient participation was measured in 14 studies, it was increased in
12
8, and no effect in five studies. Patient satisfaction was measured in 23
13
studies, in 14 studies there were no changes and in 5 there was increased
14
satisfaction. Patient knowledge was measured in 5 studies with a reduction in
15
two studies and no changes in 3 studies. According to type of intervention,
16
comparisons between written alone and coaching alone showed similar, small
17
to moderate and statistically significant increases for both types for question
18
asking. Patient satisfaction was borderline statistically significant for written
19
materials, coaching the effect was small and statistically significant. Written
20
materials led to a small and statistically significant increase in consultation
21
length with coaching the increase was smaller but not significant.
22
Interventions immediately before the consultation led to a small statistically
23
significant increase in consultation length and patient satisfaction.
24
It should be noted that many of the studies were from other settings: Brown
25
(1999, 2001), Bruera (2003), Butow (1994, 2004), Davison (1997, 2002), Ford
26
(1995), Oliver (2001) were cancer studies. Finney (1990), Kim (2003), Lewis
27
(1991) were from paediatric and family planning settings.
28
Wetzels (2007) (Wetzels, R., Harmsen, M., van, Weel C. et al , 2007)
29
conducted a Cochrane systematic review assessing the effects of
30
interventions in primary care to improve older patients’ involvement. The
31
inclusion criteria were RCTs and quasi-randomised trials; older participants
32
(65 or over) in primary care either before during or after the consultation and
Medicines concordance: full guideline DRAFT (July 2008) page 81 of 373
DRAFT FOR CONSULTATION
1
the interventions had to aim to improve the patients’ involvement. Studies of
2
other health care professionals were excluded. The outcome measures
3
included use of health care, preparation for contact with a care provider,
4
contact with the care provider (communication), feedback of care, health
5
status and behaviour, treatment outcomes and outcomes related to health
6
professionals. Three studies met all inclusion criteria (Cegala 2001; Kimberlin
7
2001; Tennstedt 2001) were published in English and conducted in the USA.
8
The interventions were either face-to-face sessions to coach patients in
9
question asking and participating in consultations (given before the
10
consultation) or written interventions (booklet or checklist). Questioning
11
behaviour of patients was the primary outcome measure in two of the studies.
12
The other study had self-reported active behaviour as the primary outcome
13
measure. The studies used mostly qualitative analysis to assess the
14
outcomes and none were assessed at a later date. In Cegala (2001) the
15
trained patients asked more medically-related questions, gained more
16
information and provided more information than control patients. They did not
17
verify information more than control patients and appointment length was not
18
longer overall. In Kimberlin (2001) there was more question asking about their
19
medication than the control group. Tennstedt (2000) found older patients are
20
in general not involved in their GP visit. Few prepared questions or identified
21
issues to discuss with their GP. However, only 21% found that the GP
22
dominated the visit. The intervention group reported more behaviours such as
23
bringing a list of problems to the visit (non-significant in intention to treat
24
analysis but significant in those who did attend, and higher satisfaction with
25
the interpersonal aspects of the encounter). In conclusion, there is little
26
evidence of interventions specifically for improving older patients’ involvement
27
and thus they cannot recommend the use of the examined interventions in
28
clinical practice. Those interventions that were included resulted in patients
29
asking more and different questions, and they became more active in
30
consultations due to pre-visit preparation.
31
Wetzels (2005) (Wetzels, R., Wensing, M., van, Weel C. et al , 2005) ran a
32
cluster-randomised trial to assess a consultation leaflet implementation
33
programme in which GPs and older patients were encouraged to improve
Medicines concordance: full guideline DRAFT (July 2008) page 82 of 373
DRAFT FOR CONSULTATION
1
older patients involvement when visiting their GP. This study was conducted
2
in the Netherlands.
3
All patients aged 70 years or above in the intervention practices received a
4
consultation leaflet by mail. This leaflet included a short motivating text on
5
patient involvement and a mixture of open and pre-structured questions to
6
help patients prepare for the next consultation and prioritize which problems
7
they wanted to discuss with their GP. The questions were chosen as they
8
would help to explore patient’s ideas, fears and expectations and encouraged
9
them to address important issues. Pre-intervention 315 patients and post-
10
intervention 263 patients were included in the study.
11
Based on results from a previous qualitative study, the authors concluded that
12
it would be important to include GPs in the implementation of the patient
13
involvement, so GPs in the intervention group received a 30 minute practice
14
visit to motivate them to involve patients and instructed them in the use of the
15
consultation leaflet.
16
The main results reported that subjects were satisfied with their involvements
17
and the GPs behaviour during the consultation, however there was no
18
difference in effect as a result of the leaflet on involvement, enablement or
19
satisfaction were found between the intervention and the control group.
20
Intervention group leaflet users reported more psychological symptoms to
21
their GP compared with non-users of the leaflet (p=0.034).
22
Overall the main findings do not support the use of the implementation
23
programme on improving involvement, enablement or satisfaction of older
24
patients in their care.
25
26
Lewin (2001) (Lewin, S. A., Skea, Z. C., Entwistle, V. et al , 2001) conducted
27
a Cochrane Collaboration systematic review of interventions to promote a
28
patient-centred approach in clinical consultations. The inclusion criteria were
29
RCTs, controlled before and after studies. The interventions promoted patient-
30
centred care in clinical health care consultations, not in social support or
31
social care. Other exclusion criteria were studies that considered cultural,
32
disability, sexuality or other sensitivity training only for health care providers;
33
evaluating training in psychotherapy or counselling for health care providers;
Medicines concordance: full guideline DRAFT (July 2008) page 83 of 373
DRAFT FOR CONSULTATION
1
studies that trained health care providers to deliver a specific, secondary
2
intervention. The outcome measures consultation processes; other health
3
care behaviour; health status and wellbeing including physiological measures;
4
patient and/or carers’ satisfaction with care. Interventions were grouped into
5
the intensity of patient centredness and teaching tactics (weak, medium and
6
strong). 5260 titles and abstracts found, 135 with potential to be included, 17
7
were included. The studies were mainly conducted in North America, others
8
were from the UK, Switzerland, Norway and Trinidad and Tobago. The aims,
9
format, content of interventions and the clinical conditions on which they
10
focused showed heterogeneity. Studies were grouped into broadly similar
11
interventions: Patient-centred training compared with no training for providers
12
(11 studies*); Patient-centred training for providers plus patient-centred
13
training or materials for patients (3 studies**); Patient-centred training for
14
providers plus condition or behaviour-specific training or materials for
15
providers and patients (2 studies***); patient-centred training for providers,
16
patient-centred materials for patients plus condition- or behaviour-specific
17
materials for providers and patients compared with condition- or behaviour-
18
specific materials for both providers and patients (1 study****). The
19
participants were mainly health care primary care physicians and patients
20
were the recipients in six studies. In some of the studies the primary goal was
21
to increase patient centredness of care while, these studies tended to focus
22
on communication skills as important on their own right, others patient-centred
23
care was seen as a method to change patient behaviour or improve the health
24
outcome. Overall there is fairly strong evidence that some interventions which
25
promote patient-centred care in the consultation can lead to significant
26
increases in the patient centredness of the consultation process.
27
The following studies dealt with either specific populations or topics outside
28
our main focus: Clark (1998) paediatric doctors, Cope (1986) CCT, Langewitz
29
(1998) paediatric doctors, Lewis (1991) paediatric residents and fellows,
30
Meland (1997) was lifestyle changes for CHD risk.
31
*Cope (1986), Howe (1996), Langewitz (1998), Levinson (1993), Putnam
32
(1988), Robbins (1979), Roter (1995), Roter (1998), Smith (1995), Smith
Medicines concordance: full guideline DRAFT (July 2008) page 84 of 373
DRAFT FOR CONSULTATION
1
(1998), Thom (1999).**Joos (1996), Lewis (1991), Pill (1998).***Clark (1998),
2
Meland (1997).****Kinmonth (1998).
3
Little (2004) (Little, P., Dorward, M., Warner, G. et al , 2004) conducted a
4
randomised control trial to assess the effect of leaflets in empowering patients
5
in primary care consultations. 636 patients were recruited in the study, aging
6
from 16-80 years and were attendees at one of five general practices in the
7
UK. Participants were randomised to four conditions: receipt of a general
8
leaflet, depression leaflet, both leaflets and no leaflets (control group). The
9
general leaflet which asked patients to list issues they wanted to raise and
10
explained that the doctor wanted them to ask questions, talk and discuss any
11
problems of concern to them. The depression leaflet listed symptoms of
12
depression (without labelling as such) and asking if had them and that the
13
doctor would like to discuss them. The outcomes measured were patient
14
satisfaction (the scores reflected aspects of doctor patient communication),
15
consultation time, prescribing, referral and investigation.
16
The only significant interaction was the increase in satisfaction for those who
17
received the general leaflet, the mean difference was 0.17 (95% CI 0.01 to
18
0.32, p=0.04). Consultation time and the general leaflet were significantly
19
associated with improved satisfaction, and leaflet was significantly more
20
effective when consultations were short (leaflet 0.64, 95% CI 0.19 to 1.08;
21
time 0.31, 0.0 to 0.06; interaction between both showed that consultations of
22
5, 8, and 10 mins increased satisfaction by 14%, 10% and 7%. This was also
23
shown for subscales of satisfaction – comfort from communication 1.02 (0.36
24
to 1.68), relief of distress 0.74 (0.0 to 1.49), intention to comply with
25
management decisions 0.65 (0.06 to 1.23) and rapport 0.81 (0.16 to 1.45).
26
The general leaflet increased the number of investigations by the doctor (OR
27
1.43, 1.00 to 2.05), which was unlikely to be due to chance or confounders
28
after controlling.
29
30
Wilkinson (2002) (Wilkinson, C. R. and Williams, M., 2002) conducted a
31
randomised controlled trial to investigate the relationship between providing
32
patients with an education guidebook designed to encourage and support
33
participation in the health care visit and selected patient and system outcome
Medicines concordance: full guideline DRAFT (July 2008) page 85 of 373
DRAFT FOR CONSULTATION
1
measures. The study population included 277 predominantly male
2
participants, with an average age of approximately 60 years. This study was
3
conducted in the USA.
4
Patients in the intervention group were mailed the appointment guidebook
5
with instructions prior to a scheduled routine visit with their primary care
6
provider. The control group was mailed the standard letter informing them of
7
the upcoming appointments. After the visit, the patients in both groups were
8
sent a short questionnaire with instructions and postage-paid return envelope.
9
No significant differences were reported in the proportion of patients in the two
10
groups that agreed with any of the six statements relevant to primary care visit
11
effectiveness. However, significant differences were reported in the proportion
12
of patients who received preventative health care interventions of influenza
13
vaccination, pneumococcal vaccination and gender specific cancer screening.
14
15
Ross (2004) (Ross, S. E., Moore, L. A., Earnest, M. A. et al , 2004) conducted
16
a randomised controlled trial that assessed the effects of a patient accessible
17
online medical record on patient satisfaction, adherence and health status in a
18
randomised controlled trial conducted in the USA. The version used in the
19
study, System Providing Patients Access to Records Online (SPPARO)
20
provides access to clinical notes and test results, and also provides a method
21
of sending electronic messages to the clinical staff.
22
Patients included in the study were aged 18 years or older. One hundred and
23
seven were enrolled in the study.
24
Participants in the SPPARO group were given a user identification and
25
password to SPPARO and a written user guide to the system. The control
26
group continued to receive standard care in practice. Periodic messages were
27
sent by research staff to all participants and were informed about upcoming
28
surveys and encouraged to contact the research assistant if they had a
29
change of address or telephone number. They were also reminded that they
30
could contact the research assistant if they had problems using SPPARO.
31
No statistically significant results were found for self-efficacy, adherence to
32
medication, health status and patient satisfaction. General adherence to
33
medication advice showed significant improvement (however no statistical
34
significance information was provided in the paper) in the intervention group
Medicines concordance: full guideline DRAFT (July 2008) page 86 of 373
DRAFT FOR CONSULTATION
1
compared with the control group. Although the number of patients who visited
2
the emergency department did not differ significantly, there was a significant
3
increase in the number of overall emergency department visit in the
4
intervention group compared to the control group (again, no statistical
5
significance information was provided).
6
7
Loh (2007) (Loh, Andreas, Simon, Daniela, Wills, Celia E. et al , 2007)
8
conducted a randomised controlled trial that investigated the effects of a
9
shared decision-making intervention in primary care of depression compared
10
to usual care on adherence, satisfaction and clinical outcomes. The study was
11
conducted in Sudbagen, Germany with primary care physicians as the unit of
12
randomisation. The sampling frame (n=148) were sent a letter, 30 accepted
13
the invitation to take part, 20 were randomly assigned to the intervention
14
group and 10 to the control group, after drop out 15 and 8 were left
15
respectively. The physicians had to recruit newly diagnosed depressive
16
patients. The intervention physicians enrolled 263 patients and the control
17
group 142. After their diagnosis the patients completed a questionnaire
18
measuring patient involvement, depression severity and socio-demographic
19
characteristics. After 6-8 weeks the patients completed a second
20
questionnaire measuring adherence and treatment outcome. At the same
21
time, the physicians rated their assessment of the patients’ adherence. The
22
shared decision-making intervention was then implemented with the
23
intervention group. The intervention was a multi-faceted program including
24
physician training, a decision board for use during the consultation given to
25
the patients after the consultation, printed patient information with specific
26
encouragement to be active in the decision-making process. The physicians in
27
the intervention group also completed modules on guideline-concordant
28
depression care which included content on enhancing skills for improving the
29
shared decision-making process. The outcomes measures were patient
30
participation, treatment adherence, patient satisfaction, consultation time and
31
clinical outcomes. There was no difference for the control group in patient
32
participation before and after, whereas the intervention group had significantly
33
higher patient participation from pre to post intervention for the doctor
34
facilitation scale (p=0.001) and there was an increase in the patient
Medicines concordance: full guideline DRAFT (July 2008) page 87 of 373
DRAFT FOR CONSULTATION
1
participation scale (p=0.010). There were no significant differences in
2
treatment adherence. Patient satisfaction was significantly higher in the
3
intervention 29.8 (2.7) than the control group 27.0 (3.6), p=0.14. There were
4
no values taken for satisfaction before the intervention. There was no
5
difference between groups for length of consultation. Neither group had a
6
statistically significant reduction in depression severity from baseline to post-
7
intervention.
8
9
10
11
Medicines concordance: full guideline DRAFT (July 2008) page 88 of 373
DRAFT FOR CONSULTATION
1
4.6
Key Clinical Question: How can practitioners elicit
2
patients’ preferences for involvement in decisions
3
about medicines?
Related
Evidence statements (summary of Evidence into recommendations
references
evidence)
1. No tools designed for use in
No tools for use in clinical practice were
clinical practice were found.
found. The GDG used the evidence from
the literature and their professional
opinion to develop recommendations on
eliciting patients information needs.
Ende (1989)
2. The Autonomy Preference Index
(Ende, J.,
(API) and developments of the API
Kazis, L., Ash, have been created to elicit patients’
A. et al ,
preferences for involvement in
1989);
decision-making and may be
Langewitz
reduced to a 6-item subscale.
(2006)
(Langewitz,
W., Nübling,
M., and
Weber, H.,
2006);
Tortolero
(2006)
(TortoleroLunaGuillermo,
Byrd-Theresa
Groff-Janet,
2006); Neame
(2005)
Medicines concordance: full guideline DRAFT (July 2008) page 89 of 373
DRAFT FOR CONSULTATION
(Neame,
Rebecca,
Hammond,
Alison, and
Deighton,
Christopher,
2005);
Braman
(2004)
(Braman,
Amie C.,
2004);
Doherty
(2005)
(Doherty, C.
and Doherty,
W., 2005);
Schneider
(2007)
(Schneider,
A., Wensing,
M., Quinzler,
R. et al ,
2007); Hill
(2006) (Hill, S.
A. and
Laugharne,
R., 2006)
Cox (2007)
3. A brief pre-consultation
(Cox, K.,
questionnaire may be used to elicit
Britten, N.,
patients’ preferences for
Hooper, R. et
involvement in decisions about
Medicines concordance: full guideline DRAFT (July 2008) page 90 of 373
DRAFT FOR CONSULTATION
al , 2007)
medicines.
Caress (1997) 4. A set of 5 sort cards can be used
(Caress, A.,
to elicit the patients’ preferred role
1997)
and perceived role within the
consultation.
Doherty
5. Two of these tools have been
(2005)
used within routine clinical settings.
(Doherty, C.
and Doherty,
W., 2005)
within a
hospital and
Cox (2007)
(Cox, K.,
Britten, N.,
Hooper, R. et
al , 2007)
before and
after a
general
practice
consultation.
1
Medicines concordance: full guideline DRAFT (July 2008) page 91 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
4.6.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies – A previous search included only randomized controlled
6
trials (RCTs) and systematic reviews (SRs). No studies which met the criteria
7
were found. We widened the search to include any type of studies to find
8
relevant information to meet our inclusion criteria.
9
Types of participants - people prescribed medication for a medical condition.
10
Duration of studies – No time limit specified for the studies.
11
Types of interventions - Any intervention (tool) which elicits patient
12
preferences for involvement in decisions about medicines. The tools had to be
13
brief enough to be utilised within a consultation between the patient and
14
practitioner. Therefore long questionnaires were excluded as they would not
15
be manageable.
16
Types of outcome measures – Any outcome relating to the use of the tool
17
was acceptable as we were looking for a tool which could be utilized within a
18
consultation, rather than looking for specific clinical outcomes.
19
4.6.2
20
No tools designed for use in clinical practice were found.
21
The tools in this review are research tools and not clinical tools designed for
22
use in a consultation. We have included these research studies to illustrate
23
brief questionnaires that have been used in research settings indicating the
24
content of the questionnaires to inform the GDG. We have not reported in
25
detail on development or validation of these questionnaires.
26
Ende (1989) (Ende, J., Kazis, L., Ash, A. et al , 1989) constructed an
27
instrument to measure patient preferences for making medical decisions and
28
their desire to be informed. The instrument, named the Autonomy Preference
Evidence review
Medicines concordance: full guideline DRAFT (July 2008) page 92 of 373
DRAFT FOR CONSULTATION
1
Index (API) was in a questionnaire format and the scales were developed by a
2
group of 13 clinicians, medical sociologists, and ethicists who were interested
3
in patient autonomy. Items on the scales were tested for reliability and validity.
4
The final API consisted of an 8-item scale on information-seeking and a 15-
5
item scale on decision-making. Within this 15-item scale was a 6-item sub-
6
scale which related to general items. The other 9-items were related to three
7
clinical vignettes and were too disease-specific to be useful for this question.
8
The 6-item scale (part A) meets our criteria (see above) but is a research tool
9
rather than a clinical tool.
10
11
The Decision making preference scale (Part A): (responses on a 5-point Likert
12
scale ranging from ‘strongly disagree’ to ‘strongly agree’. The higher the score
13
on the scale, the more patients wished to participate in the decision-making.
14
1.* The important medical decisions should be made by your doctor, not by
15
you.
16
2. You should go along with your doctor’s advice even if you disagree with it.
17
3.* When hospitalized, you should not be making decisions about your own
18
care.
19
4. You should feel free to make decisions about everyday medical problems.
20
5.* If you were sick, as your illness became worse you would want your doctor
21
to take greater control.
22
23
24
6. You should decide how frequently you need a check-up.
25
The API was used within a variety of studies: Langewitz (2006) (Langewitz,
26
W., Nübling, M., and Weber, H., 2006); Tortolero (2006) (Tortolero-Luna-
27
Guillermo, Byrd-Theresa Groff-Janet, 2006); Neame (2005) (Neame,
28
Rebecca, Hammond, Alison, and Deighton, Christopher, 2005); Braman
29
(2004) (Braman, Amie C., 2004); Schneider (2007) (Schneider, A.,
30
Wensing, M., Quinzler, R. et al , 2007) and Hill (2006) (Hill, S. A. and
31
Laugharne, R., 2006). All of the studies used the 6-item subscale as
32
illustrated above, except for Langewitz (2006) (Langewitz, W., Nübling, M.,
33
and Weber, H., 2006) who adapted the instrument into two questions and Hill
34
(2006) (Hill, S. A. and Laugharne, R., 2006) slightly adapted the questionnaire
* Scoring for these items was reversed, and goes from 5 to 1, rather than 1 to 5.
Medicines concordance: full guideline DRAFT (July 2008) page 93 of 373
DRAFT FOR CONSULTATION
1
to apply to psychiatric patients. Most of the studies posted their questionnaires
2
to the participants rather than having them completed in a clinical setting.
3
4
Langewitz (2006) (Langewitz, W., Nübling, M., and Weber, H., 2006)
5
As part of the questionnaire, Langewitz (2006) (Langewitz, W., Nübling, M.,
6
and Weber, H., 2006) adapted the API to 4 point Likert scale: fully agree,
7
slightly agree, slightly disagree, fully disagree. How much do you agree with
8
the following statements:
9

convinced of his ideas (Follow physician’s advice)
10
11
12
One should stick to the physician’s advice even if one is not fully

It should completely be left to physicians to decide on a patient’s
treatment (Physician should decide)
13
This was conducted at the University of Basel in NW Switzerland and sent to
14
the patients after discharge from hospital.
15
16
Degner and Sloan (1992) (Degner, L, 2008) adapted the API and used the
17
questions on sort cards for patients with cancer. Their questions were adapted
18
by Caress (1997) (Caress, A., 1997), Doherty (2005) (Doherty, C. and
19
Doherty, W., 2005) and Cox (2007) {6698).
20
Caress (1997) {1155} conducted a cross-sectional study at a Regional Renal
21
unit in the North of England with 462 participants from a convenience sample
22
over a 12 month period. 155 of the patients were pre-dialysis, 103 were
23
dialysis patients and 147 were transplant patients. Using a set of sort cards
24
(as used by Degner and Sloan, 1992 (Degner, L, 2008)) the patients picked a
25
single card which was closest to their preferred role in decision-making and
26
also picked a single card closest to their perceived role in decision-making.
27
Patients were also asked to give their rationale for their preferred role.
28
The 5 sort cards:
29
Active options
30
Card A: I prefer to make the final decision about which treatment I will receive.
Medicines concordance: full guideline DRAFT (July 2008) page 94 of 373
DRAFT FOR CONSULTATION
1
Card B: I prefer to make the final selection of my treatment after seriously
2
considering my doctor's opinion.
3
Collaborative option
4
Card C: I prefer that my doctor and I share responsibility for deciding which
5
treatment is best for me.
6
Passive options
7
Card D: I prefer that my doctor makes the final decision about which treatment
8
will be used but seriously considers my opinion.
9
Card E: I prefer to leave all decisions regarding my treatment to my doctor.
10
The key points found from the study were that: participation preference was
11
highly individualistic, with a lot of patients wishing to remain passive. Those
12
who did prefer an active role were unlikely to attain this preference; trust in the
13
HCP can influence the preference; desire for information is not synonymous
14
with desire for participation.
15
Doherty (2005) (Doherty, C. and Doherty, W., 2005) used the adapted
16
questionnaire to use in an acute hospital trust in England and was one study
17
which tried to elicit the patients’ responses within an actual clinical situation.
18
19
Cox (2007) (Cox, K., Britten, N., Hooper, R. et al , 2007) included 479
20
patients who were approached in the waiting room in general practitioner
21
surgeries to participate and then given an interview where they completed the
22
pre-consultation questionnaire. They were also administered a questionnaire
23
after the consultation. The G.P was given a questionnaire before, which
24
included their preferred role in decision making with patients and a
25
questionnaire afterwards detailing their perceptions of the decision-making
26
during each consultation. The doctors’ assessment of patients’ preference to
27
be involved in shared decision making was correct in 32% of the
28
consultations, overestimated in 45% of the consultations and underestimated
29
in 23% of the consultations. The patients’ preferences for decision making
Medicines concordance: full guideline DRAFT (July 2008) page 95 of 373
DRAFT FOR CONSULTATION
1
involved: 39% wanting the G.P to share the decision, 45% wanting the G.P to
2
be main (28%) or only (17%) decision-maker and 16% wanting to be the main
3
(14%) or only (2%) decision-maker.
4
The questionnaire given to the patients at pre-consultation included the
5
following 5 statements, of which patients were asked to choose one:
6

problem.
7
8



15
I would prefer that my doctor makes the final decision about medicines
I take for this problem, but seriously considers my opinion.
13
14
I would prefer that my doctor and I share responsibility for deciding
about medicines I take for this problem.
11
12
I would prefer that I make the final decision about medicines I take for
this problem after seriously considering my doctor’s opinion.
9
10
I would prefer that I make the decision about medicines I take for this

I would prefer that my doctor makes all decisions about medicines I
take for this problem.
16
Medicines concordance: full guideline DRAFT (July 2008) page 96 of 373
DRAFT FOR CONSULTATION
1
2
4.7
Key Clinical Question: What tools are available to help
elicit patients’ beliefs about medicines?
Related references
Evidence statements
Evidence into recommendations
(summary of evidence)
Horne (1999) (Horne,
1. No tools designed for
No tools designed for use in clinical
R., 2008); Menckeberg
use in clinical practice
practice were found although the GDG
(2008) (Menckeberg, T.
were found.
were aware of current studies to develop
T., Bouvy, M. L.,
Bracke, M. et al , 2008);
Horne (2007) (Horne,
R., Cooper, V.,
Gellaitry, G. et al ,
2007); Brown (2005)
(Brown, Charlotte,
such tools, in particular studies seeking
The Beliefs about
Medicines (BMQ) –
Specific component is a
research tool which is
used to elicit beliefs about
medicines.
to adapt the BMQ for clinical use. The
GDG reviewed the research tools found
but it was not appropriate to use these
outside their research settings. The GDG
used the information from the research
studies and their professional opinion to
Battista, Deena R.,
make recommendations in relation to
Bruehlman, Richard et
elicitation of patients’ beliefs about
al , 2005); Khanderia
medicines.
(2008) (Khanderia, U.,
Townsend, K. A.,
Erickson, S. R. et al ,
2008); Kumar (2008)
(Kumar, K., Gordon, C.,
Toescu, V. et al , 2008);
Kemp (2007) (KempSteven, Feely-Morgan
Hay-Alastair WildHeather CooperCathryn., 2007); Aikens
(2008) (Aikens, James
E., Nease-Donald-E-Jr,
and Klinkman, Michael
S., 2008); Clifford
Medicines concordance: full guideline DRAFT (July 2008) page 97 of 373
DRAFT FOR CONSULTATION
(2008) {1518}; Jenkins
(2003) (7606);
Theunissen (2003)
(Theunissen-Nicolet, C.
M., 2003)
Hamilton (2007)
2. One RCT that assessed
(Hamilton, W., Russell,
a patient self-completion
D., Stabb, C. et al ,
agenda form on
2007)
prescribing and adherence
showed no significant
results for a prescription,
satisfaction scores or
adherence.
Medicines concordance: full guideline DRAFT (July 2008) page 98 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
4.7.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies – We initially included only Randomised Controlled Trials
6
(RCTs) and Systematic Reviews (SRs), however none of these types of
7
studies were found that met the criteria. We increased the current search to
8
include any type of study in order to find relevant information to meet our
9
inclusion criteria.
10
Types of participants - people prescribed medication for a medical condition.
11
Duration of studies – No time limit specified for studies.
12
Types of interventions - Any intervention (tool) which elicits patient beliefs
13
about their medicines. The tools had to be brief enough to be utilised within a
14
consultation between the patient and practitioner. Therefore long
15
questionnaires were excluded as they would not be manageable.
16
Types of outcome measures – Any outcome relating to the use of the tool
17
was acceptable as we were looking for a tool which could be utilised within a
18
consultation, rather than looking for specific clinical outcomes.
19
4.7.2
20
No tools designed to elicit patients beliefs about medicines for use in clinical
21
practice were found. The tools which we found in this review were research
22
tools. We decided to include studies which used these research tools to
23
illustrate some questions that could be asked to the patient in a consultation.
24
Most studies were in questionnaire form and so we included those which were
25
shortest. We have not reported the parts of the questionnaire which were not
26
relevant to the clinical question.
27
Horne (1999) (Horne, R., 2008) created a questionnaire which explicitly states
28
the intention of assessing patients’ beliefs about medicines. The beliefs about
29
medicines questionnaire (BMQ) included two parts – the BMQ-General, which
Evidence review
Medicines concordance: full guideline DRAFT (July 2008) page 99 of 373
DRAFT FOR CONSULTATION
1
assessed beliefs about medicines in general and the BMQ-Specific, which
2
looks at patients’ specific beliefs towards their medicine. The study states that
3
the two sections of the BMQ can be used together or separately. As the BMQ-
4
Specific answers the question, and we are looking for brevity within the
5
consultation, this part of the study is reported.
6
The BMQ-Specific includes two 5-item factors which assess beliefs of the
7
necessity of medicines prescribed (Specific-Necessity) and concerns about
8
medicines prescribed, based on beliefs of the danger of dependence, long-
9
term toxicity and the disruptive effects of medication (Specific-concerns).
10
The BMQ-Specific items, which are rated ‘strongly agree, agree, uncertain,
11
disagree or strongly disagree’:
12

My health, at present, depends on my medicines.
13

Having to take medicines worries me.
14

My life would be impossible without my medicines.
15

Without my medicines I would be very ill.
16

I sometimes worry about long-term effects of my medicines.
17

My medicines are a mystery to me.
18

My health in the future will depend on my medicines.
19

My medicines disrupt my life.
20

I sometimes worry about becoming too dependent on my medicines.
21

My medicines protect me from becoming worse.
22
23
The BMQ-Specific was used in many other studies to assess beliefs about
24
medicines for a range of conditions (Menckeberg, 2008 (Menckeberg, T. T.,
25
Bouvy, M. L., Bracke, M. et al , 2008); Horne, 2007 (Horne, R., Cooper, V.,
26
Gellaitry, G. et al , 2007); Brown 2005 (Brown, Charlotte, Battista, Deena R.,
27
Bruehlman, Richard et al , 2005); Khanderia, 2008 (Khanderia, U., Townsend,
28
K. A., Erickson, S. R. et al , 2008); Kumar, 2008 (Kumar, K., Gordon, C.,
29
Toescu, V. et al , 2008); Kemp, 2007 (Kemp-Steven, Feely-Morgan Hay-
30
Alastair Wild-Heather Cooper-Cathryn., 2007); Aikens 2008 (Aikens, James
31
E., Nease-Donald-E-Jr, and Klinkman, Michael S., 2008); Clifford, 2008
Medicines concordance: full guideline DRAFT (July 2008) page 100 of 373
DRAFT FOR CONSULTATION
1
{1518}; Jenkins (2003) (Jenkins, Linda, Britten, Nicky, Stevenson, Fiona et al ,
2
2003); Theunissen (2003) (Theunissen-Nicolet, C. M., 2003)).
3
4
We did retrieve one study, Hamilton (2006) (Hamilton, W., Russell, D.,
5
Stabb, C. et al , 2007) , which was a randomised controlled trial conducted to
6
test the effect of patient self-completion agenda forms on prescribing and
7
adherence in general practice. This RCT was considered relevant as one of
8
the items of the self-completion form was related to medication expectations
9
and it was considered a clinical tool.
10
One thousand and six hundred and ten patients at ten general practices in
11
Devon and Dorset (UK) were involved in the trial for up to 12 weeks. All
12
patients were given a letter and an envelope when attending their GP. Half the
13
group received an agenda form which they could fill out while waiting for the
14
doctor. The other half received usual care. The agenda form called the SCAF
15
(self completion agenda form) included five questions:
16
1.
17
problem you have e.g. symptoms or current illness.
18
2.
Your ideas about your illness: what do you think is wrong with you?
19
3.
Your concerns: have you any particular worries about your illness?
20
4.
Your expectations: how do you think your problem should be treated?
21
What do you hope the doctor will do?
22
5.
23
problem?
24
The GP was handed this form when the patient went into their appointment, to
25
use in whichever they thought appropriate. There were no significant
26
differences between proportion of patients who received a prescription, or in
27
satisfaction scores or adherence to prescribed medication.
What made you decide to come to see the doctor? Please describe the
Medication: do you think you should receive a prescription for your
28
Medicines concordance: full guideline DRAFT (July 2008) page 101 of 373
DRAFT FOR CONSULTATION
1
2
Medicines concordance: full guideline DRAFT (July 2008) page 102 of 373
DRAFT FOR CONSULTATION
1
4.8
Key Clinical Question: What tools are available to help
elicit patients’ information needs?
2
Related references
Evidence statements (summary
of evidence)
Evidence into
recommendations
1. No tools designed for use in
No tools for use in clinical practice
clinical practice were found.
were found. The GDG used the
evidence from the literature and
Kinnersley (2007)
2. One systematic review found
(Kinnersley, P.,
that 17 RCTs measured question
Edwards, A., Hood,
asking with 6 finding statistically
K. et al , 2007)
significant increases and 11 finding
their professional opinion to
develop recommendations on
eliciting patients information
needs.
no effects.
Kinnersley (2007)
3. One systematic review found
(Kinnersley, P.,
that patient satisfaction in 14 RCTs
Edwards, A., Hood,
there were no changes and 5 there
K. et al , 2007)
was significant increases in
satisfaction.
Strydom (2001)
2. One questionnaire has been
(Strydom, A.,
developed which can assess
Forster, M., Wilkie,
information needs of people with
B. M. et al , 2001)
learning disabilities.
Agård (2004)
3. One study used 4 open-ended
(Agård, Anders,
questions to elicit patient’s
Hermerén, Göran,
information needs.
and Herlitz, Johan,
2004)
Duggan (1992)
4. Three studies included 5 open
Medicines concordance: full guideline DRAFT (July 2008) page 103 of 373
DRAFT FOR CONSULTATION
(Duggan, C. and
questions from the Intrinsic Desire
Bates, I., 2000);
for Information (IDI) scale which
Astrom (2000)
can elicit information needs from
(Astrom, K.,
patients.
Carlsson, J., Bates,
I. et al , 2000);
Zwaenepoel (2005)
(Cramer, Joyce A.,
2004).
Ende (1989) (Ende,
5. Seven studies use the
J., Kazis, L., Ash, A.
Autonomy Preference Index 8-item
et al , 1989);
scale to elicit information needs of
Langewitz (2006)
patients.
(Langewitz, W.,
Nübling, M., and
Weber, H., 2006);
Tortolero (2006)
(Tortolero-LunaGuillermo, ByrdTheresa Groff-Janet,
2006); Neame
(2005) (Neame,
Rebecca,
Hammond, Alison,
and Deighton,
Christopher, 2005);
Braman (2004)
(Braman, Amie C.,
2004); Doherty
(2005) (Doherty, C.
and Doherty, W.,
2005); Schneider
(2007) (Schneider,
Medicines concordance: full guideline DRAFT (July 2008) page 104 of 373
DRAFT FOR CONSULTATION
A., Wensing, M.,
Quinzler, R. et al ,
2007)
1
Method of the evidence review
2
4.8.1
3
This paper includes a narrative summary of the included evidence, structured
4
according to the category of the intervention, following the agreed reviewing
5
protocol:
6
Types of studies – We initially included only randomized controlled trials
7
(RCTs) and systematic reviews (SRs) however none of these types of studies
8
were found to meet the criteria. We increased the search to include any type
9
of studies to find relevant information to meet our inclusion criteria.
10
Types of participants - people prescribed medication for a medical condition.
11
Duration of studies – No time limit specified for the studies.
12
Types of interventions - Any intervention (tool) which elicits patients’
13
information needs. The tools had to be brief enough to be utilised within a
14
consultation between the patient and practitioner. Therefore long
15
questionnaires were excluded as they would not be manageable.
16
Types of outcome measures – Any outcome relating to the use of the tool
17
was acceptable as we were looking for a tool which could be utilized within a
18
consultation, rather than looking for specific clinical outcomes.
19
4.8.2
20
No tools designed for use in clinical practice were found. The tools in this
21
review were research tools opposed to clinical tools which could be used in a
22
consultation. We decided to include studies which used these research tools
23
to illustrate some questions that could be asked to the patient in a
24
consultation. Most studies were in questionnaire form and so we included
25
those which were shortest – and so could possibly be used in a consultation.
Evidence review
Medicines concordance: full guideline DRAFT (July 2008) page 105 of 373
DRAFT FOR CONSULTATION
1
We have not reported the parts of the questionnaire which were not relevant
2
and useable in a consultation.
3
One Cochrane Review included in section 5.4 addressed ways of eliciting
4
patients information needs. Kinnersley (2007) (Kinnersley, P., Edwards, A.,
5
Hood, K. et al , 2007) conducted a Cochrane Collaboration systematic review
6
of the effects of interventions before consultations designed to help patients
7
address their information needs. The settings and populations varied, and
8
were from six countries, mainly the USA. They stated that it complemented
9
other Cochrane reviews by Wetzels (2007) (Wetzels, R., Harmsen, M., van,
10
Weel C. et al , 2007) and Lewin (2003). The inclusion criteria were RCTs of
11
interventions intended to help the patients, representatives or carers address
12
their information needs in a consultation. This was done by encouraging
13
question asking, to express their information needs, to overcome barriers to
14
communication and to clarify their understanding of the information provided.
15
Outcome measures were the consultation process, consultation outcomes
16
and service outcomes. 33 trials described in 35 studies met the inclusion
17
criteria. Of the studies assessing single interventions for patients 15 included
18
written materials, four were coaching. The multiple component single
19
interventions studies four had coaching and written materials. Results: 17
20
studies measured question asking with 6 finding statistically significant
21
increases and 11 studies finding no effects of the interventions compared to
22
controls. Patient participation was measured in 14 studies, it was increased in
23
8, and no effect in five studies. Patient satisfaction was measured in 23
24
studies, in 14 studies there were no changes and 5 there was increased
25
satisfaction. Patient knowledge was measured in 5 studies with reduction in
26
two studies and no changes in 3 studies. According to type of intervention,
27
comparisons between written alone and coaching alone there were similar,
28
small to moderate and statistically significant increases for both types for
29
question asking. Patient satisfaction was borderline statistically significant for
30
written materials, coaching the effect was small and statistically significant.
31
Written materials led to a small and statistically significant increase in
32
consultation length, coaching the increase was smaller but was not significant.
Medicines concordance: full guideline DRAFT (July 2008) page 106 of 373
DRAFT FOR CONSULTATION
1
Interventions immediately before the consultation led to small statistically
2
significant increase in consultation length and patient satisfaction.
3
It should be noted that many of the studies were from other settings: Brown
4
(1999, 2001), Bruera (2003), Butow (1994, 2004), Davison (1997, 2002), Ford
5
(1995), Oliver (2001) were cancer studies. Finney (1990), Kim (2003), Lewis
6
(1991) were from paediatric and family planning settings.
7
8
9
Strydom (2001) (Strydom, A., Forster, M., Wilkie, B. M. et al , 2001)
10
conducted a study of a service-user questionnaire to find out gaps in
11
medication knowledge and information sources. This study specifically
12
involved finding out the views of those with learning disabilities. 21
13
participants were included who were either currently taking prescribed
14
medication (GP or specialist health services) or had taken in the recent past. It
15
should be noted that two thirds of the subjects received help with taking
16
medication. A questionnaire was designed by the authors using previously
17
published guidelines. They used structured and semi-structured sections,
18
including open questions. The questionnaire was delivered by one of the
19
research team with experience of communicating with people with learning
20
disabilities. The questionnaire was designed to find out their experiences and
21
opinions of using medicines.
22
23
A table was given in the paper to show the questions relating to medication
24
knowledge. Please note that it is unknown as to whether this was exhaustive.
25
The open questions were not reported in the paper.
26

Can you read the medication label? (Yes, no)
27

What is written on the label? (Don’t know, medication name, my name,
Chemist’s name, dose, other)
28
29

Approximate name, description).
30
31
32
What is your medication called? (Don’t know, Brand or generic name,

What are you taking medication for? (Don’t know, knew indication,
approximate indication).
Medicines concordance: full guideline DRAFT (July 2008) page 107 of 373
DRAFT FOR CONSULTATION
1

know, Yes, plus example)
2
3
Is there anything you should not do while taking this medication? (Don’t

Are there any side-effects? (Don’t know, one, two or more).
4
The resulting answers led to the framework for a structure of a patient
5
information leaflet for people with learning disabilities who take medicine for
6
psychiatric conditions. It is not clear as to whether the service users who filled
7
in the questionnaire were taking psychiatric medication or another type of
8
medication. The subjects were ‘selected for their range of experiences of
9
taking medications’. Therefore this study does not elicit patients’ information
10
needs in total, but how to elicit the knowledge gaps of those with a learning
11
disability.
12
13
Duggan (2000) (Duggan, C. and Bates, I., 2000)developed and evaluated a
14
survey tool (intrinsic desire for information) to find out Patients’ perceptions
15
and information needs in regards to their medication. It was tested for
16
reliability and by factor analysis and was used with 2 cohorts of patients in
17
East London (sample of 500).
18
19
Astrom’s (2000) (Astrom, K., Carlsson, J., Bates, I. et al , 2000) paper refined
20
and validated the IDI into a 12-item scale. They included 5 open questions
21
which were a joint construction of the project aims and questions from
22
Lindegren (1999), which was a Masters thesis at the Department of
23
Biopharmaceutics at Uppsala University. The open questions were
24
transcribed at the bedside of 299 patients in the wards of three medical
25
hospitals in London. Astrom (2000) concluded that the desire for information
26
may be more complicated and involve an emotional or behavioural
27
component, which was not included. It should be noted that this is a desire for
28
information which may differ from information needs.
29
30
The 12-item scale was deemed too long to meet our inclusion criteria,
31
however some of the open questions may be of relevance.
32
33
The IDI (for reference only):
Medicines concordance: full guideline DRAFT (July 2008) page 108 of 373
DRAFT FOR CONSULTATION
1
Part 1 – Demographic details.
2
Part 2 – Questionnaire items (scored from strongly agree through strongly
3
disagree on a 5-point Likert scale).
4
1.
I always speak to my pharmacist when I want
information about my medicines
5
6
2.
Sometimes I feel a little inhibited when I ask
7
for information…they might think I should
8
know already.
9
3.
If there is anything I need to know, it’s most
convenient to ask at the surgery.
10
11
4.
It’s not really my place to ask for information,
they have enough to do.
12
13
5.
The people at the hospital can easily give me
information when I go for my appointment.
14
15
6.
I need as much information about my
medicines as possible.
16
17
7.
Too much knowledge is a bad thing.
18
8.
You can never know enough about these
things.
19
20
9.
I don’t need any more knowledge about my
medicines/illness.
21
22
10. I read about my medicines/illness as much as
23
possible.
11. What you don’t know (with respect to
24
medicines/illness) doesn’t hurt you.
25
26
12. I find information about my medicines/illness
27
confusing
28
29
Open questions:
30
13. What kind of information about your
31
medicines do you want? Why?
32
33
14. How do you want your information to be
presented (written, oral, both, other)? Why?
Medicines concordance: full guideline DRAFT (July 2008) page 109 of 373
DRAFT FOR CONSULTATION
15. Who would you like to give you information
1
about your medicines? Why?
2
3
16. When would it be best to have the information
4
about your medicine presented (at hospital, at
5
home, at the community pharmacy, at the
6
GP’s)? Why?
17. Would you like to sit down and talk about your
7
medicines with a pharmacist at the hospital?
8
9
10
Zwaenepoel (2005) (Cramer, Joyce A., 2004) used the IDI scale and 5 open
11
questions in a survey of the need for information of 279 psychiatric in-patients
12
in Flanders.
13
14
Ende (1989) (Ende, J., Kazis, L., Ash, A. et al , 1989) created the Autonomy
15
Preference Index (API) which, as well as a decision-making preference scale
16
had an eight-item information-seeking preference scale. The instrument was
17
in questionnaire format and the scales were developed by a group of 13
18
clinicians, medical sociologists, and ethicists who were interested in patient
19
autonomy. Items on the scales were tested for reliability and validity. The final
20
API consisted of an 8-item scale on information-seeking scale (ISS) and a 15-
21
item scale on decision-making.
22
The 8-item ISS consisted of the following items for information-seeking
23
preference, with responses on a five-point Likert scale from ‘strongly disagree’
24
to ‘strongly agree’:
25
26
27
28
1. As you become sicker you should be told more and more about your
illness.
2. You should understand completely what is happening inside your body
as a result of your illness.
29
3. Even if the news is bad, you should be well informed.
30
4. Your doctor should explain the purpose of your laboratory tests.
31
5. *You should be given information only when you ask for it.
32
6. It is important for you to know all the side effects of your medication.
33
7. Information about your illness is as important to you as treatment.
Medicines concordance: full guideline DRAFT (July 2008) page 110 of 373
DRAFT FOR CONSULTATION
1
8. When there is more than one method to treat a problem, you should be
told about each one.
2
3
* Scoring for this item is reversed and goes from 5 to 1, rather than 1 to 5.
4
5
Five further studies used the Autonomy Preference Index to elicit patients
6
information needs (Langewitz, 2006 (Langewitz, W., Nübling, M., and Weber,
7
H., 2006); Tortolero, 2006 (Tortolero-Luna-Guillermo, Byrd-Theresa Groff-
8
Janet, 2006); Neame, 2005 (Neame, Rebecca, Hammond, Alison, and
9
Deighton, Christopher, 2005); Braman, 2004 (Braman, Amie C., 2004);
10
Schneider, 2007 (Doherty, C. and Doherty, W., 2005); Hill, 2006 (Hill, S. A.
11
and Laugharne, R., 2006). All of the studies used the full length (8-item)
12
information-preference scale except for Langewitz (2006) (Langewitz, W.,
13
Nübling, M., and Weber, H., 2006) who incorporated only one question from
14
the API to target information needs: ‘Even when the news is bad the patient
15
must be informed (information)’. Hill (2006) slightly altered the API questions
16
for psychiatric patients to use.
17
18
Agard (2004) conducted a qualitative analysis of semi-structured interviews in
19
Gothenburg, Sweden on 40 patients 60 years and over who were receiving
20
treatment after a heart failure diagnosis. The semi-structured qualitative
21
interview had 4 open-ended questions as an interview guide. The questions
22
were:
23
1. What is your opinion about the medical information that you have been
24
given?
25
2. What kind of information is lacking?
26
3. What information have you been given about heart failure?
27
4. What is your attitude toward receiving prognostic information?
28
They were also encouraged to speak about the questions and to raise other
29
issues related to them to ensure their major personal concerns really
30
emerged. To avoid respondents feeling ignorant or embarrassed about not
31
being able to adequately answer questions relating knowledge they were
Medicines concordance: full guideline DRAFT (July 2008) page 111 of 373
DRAFT FOR CONSULTATION
1
asked first about the information they had been given, rather than asking
2
directly about their knowledge of diagnosis, treatment and prognosis.
3
4
Many patients had a limited understanding of their disease but said they were
5
still satisfied with the information they received. Some were indifferent to,
6
accept or be unaware of their low level of knowledge.
7
They concluded that 'to inform the patient adequately, physicians and nurses
8
should determine the patient's level of knowledge and explore why those
9
patients who have a limited understanding do not assimilate or request
10
information. The information they provide should also be adapted to the
11
patient's capacity, wishes and emotional reactions.’
12
13
14
15
Medicines concordance: full guideline DRAFT (July 2008) page 112 of 373
DRAFT FOR CONSULTATION
1
2
4.9
Key Clinical Question: How can information about
3
medicines be provided for patients in order to enhance
4
SDM in regard to medicines?
Related references
Trevena (2006) (Trevena,
Evidence statements
Evidence into
(summary of evidence)
recommendations
1. A systematic review of
Information provided to patients
Lyndal J., Davey, Heather systematic reviews and RCTs
about treatments increases their
M., Barratt, Alexandra et
found that communicating with
understanding whatever the
al , 2006)
patients about evidence does
format (verbal, written or video)
increase their understanding
particularly if the information is
regardless of what tools used.
structured or interactive and
There was a greater effect if
especially when tailored to the
information was structured
individual. The GDG considered
(either written, verbal or video)
that health care professionals
or interactive, especially if
need to be aware that individuals
tailored to the individual.
will vary in the amount and type
of information they require and in
Trevena (2006) (Trevena,
2. One systematic review of
Lyndal J., Davey, Heather systematic reviews and RCTs
M., Barratt, Alexandra et
found that probabilistic
al , 2006)
information is best
represented as event rates,
rather than words,
probabilities, or summarised
as effect measures such as
relative risk reduction.
Illustrations, such as cartoons
or graphs appear to aid
understanding.
how they can best access that
information. It was the
professional opinion of the GDG
that undue emphasis is currently
placed on use of leaflets and
written information and there is
inadequate access to pictorial
and graphic information.
Examples of useful websites
were presented to the GDG
showing information presented in
a variety of ways and the GDG
believed it important to widen
Medicines concordance: full guideline DRAFT (July 2008) page 113 of 373
DRAFT FOR CONSULTATION
Wills (2003) (Wills, C. E.
3. One systematic review of
knowledge and access to such
and Holmes, Rovner M.,
information formats concluded
resources.
2003)
that decision support/aids can
address patient information
needs for shared decision
making. They enable patients
to better understand treatment
options, including probability
information.
1
2
Evidence review
3
4.9.1
4
This paper includes a narrative summary of the included evidence, structured
5
according to the category of the intervention, following the agreed reviewing
6
protocol:
7
Types of studies: Systematic Reviews of Randomised Controlled Trials
8
(RCTs) or Randomised Controlled Trials of interventions involving shared
9
decision-making in the clinical context.
10
Types of participants: people prescribed medication for a medical condition.
11
Duration of studies: no time limit specified.
12
Types of interventions: any interventions involving shared decision making
13
in a consultation between a health care professional and patient.
14
Types of outcome measures: Patient-centred communication in the
15
consultation; Consultation process outcomes: patient involvement, question
16
asking, preparedness; Patient care outcomes: satisfaction, knowledge, self-
17
efficacy. Type of interventions involved and type of information.
18
19
It should be noted that the remit is for conditions with prescribed medication
20
and this excludes conditions which require chemotherapy or screening. All
Medicines concordance: full guideline DRAFT (July 2008) page 114 of 373
DRAFT FOR CONSULTATION
1
RCTs are within this remit, however many of the systematic reviews included
2
populations outside the remit, this is noted where applicable.
3
4
Trevena (2006) (Trevena, Lyndal J., Davey, Heather M., Barratt, Alexandra et
5
al , 2006) conducted a systematic review of RCTs and review of reviews
6
structured around three aspects of communication with patients about
7
evidence: patients’ preferences and actions; research evidence; and the
8
clinical state and circumstances. These were then translated into three main
9
questions:
10

understanding of evidence?
11
12

15
What are the most effective formats to represent probabilistic
information to improve patient understanding of evidence?
13
14
What are the most effective communication tools to improve patient

What are the most effective strategies to elicit patient
preferences/beliefs/values relating to evidence?
16
The authors excluded studies that did not address and question; were about
17
patient education; were focused on skills and behaviour outcomes without
18
attempting to increase understanding or knowledge; were concerned with
19
counselling as a therapeutic intervention; or were specific to communication
20
regarding clinical trial participation.
21
Overlap between the trials included in the systematic reviews and those
22
identified independently was verified and duplicated studies were excluded.
23
Ten systematic reviews of RCTs and 30 additional RCTs were retrieved.
24
The review concluded that communicating with patients about evidence does
25
increase their understanding regardless of the tools used. The authors also
26
found that there was a greater effect if information was structured (either
27
written, verbal or video) or interactive (computer, touch screen, question
28
prompts) and particularly if the information was tailored to the individual.
29
Probabilistic information was found to be best represented as even rates in
30
relevant groups of people, rather than words, probabilities or summarized as
31
effect measures such as relative risk reduction. Written information was
32
reported to be more effective if illustrations and graphs were used.
Medicines concordance: full guideline DRAFT (July 2008) page 115 of 373
DRAFT FOR CONSULTATION
1
The authors did however remark that one of the difficulties in generalising
2
from the literature was that the trials were conducted in a wide variety of
3
clinical settings using a range of clinical problems and outcomes.
4
5
Wills (2003) (Wills, C. E. and Holmes, Rovner M., 2003) conducted a
6
systematic review of patient health information provision and use for treatment
7
decision-making. It included research from the past 10 years focusing on
8
testing different formats of information presentation for patient decision-
9
making. The three types of formats looked at were probability presentations,
10
graphic formats and words vs numbers. They found two studies where
11
participants preferred presentation of medication in terms of relative risk rather
12
than absolute risk format. They found that people simplify relative risk
13
information into a simplified format of small or large risks and there is a
14
tendency to seriously under or overestimate their personal risks for health
15
outcomes. There is a need to tailor the format of risk communication to the
16
individual’s level of numeracy. In routine clinical encounters information
17
should be presented balanced, in both positive and negative frames. Graphics
18
can improve the understanding of numerical probability information. However
19
some people may dislike some types of displays or misunderstand them.
20
Consistent finding of individual differences in preferences for probability
21
information in words, numbers of both formats implies a need for routine
22
individualized assessments of patient preferences for format. In conclusion,
23
decision support/aids can address patient information needs for shared
24
decision making. They enable patients to better understand treatment options,
25
including probability information.
26
27
Medicines concordance: full guideline DRAFT (July 2008) page 116 of 373
DRAFT FOR CONSULTATION
1
2
4.10
Key Clinical Question: What information about
3
medicines should be provided for patients in order to
4
enhance SDM in regards to medicine?
Recommendations
Evidence into recommendations including
The GDG considered that the provision of information to
patients is to facilitate informed choice about medicines, and
achieve a clear picture of the benefits and risks. The
information that should be provided to a patient is dependent
on what that patient needs to make a decision and prescriptive
list can therefore not be generated. Patient Information Leaflets
(PILS) provided with medicines often do not help in providing
information about medicines and in any case are only received
after the medicine is dispensed. The GDG did consider that
some broad areas of information that patients might require
and used the evidence from expert reviews, in particular those
with patient involvement to inform this. They also considered
that sources such as MHRA leaflet might be useful for patients.
Health care professionals that questions from patients are
often not declared and patients are not always able to ask the
questions that concern them.
Medicines concordance: full guideline DRAFT (July 2008) page 117 of 373
DRAFT FOR CONSULTATION
1
4.10.1 Evidence review
2
The evidence review is a narrative review. The GDG requested review of
3
current national guidance and reports with a particular emphasis on those
4
where patients views and perspectives were given priority.
5
4.10.1.1
Summary of ‘Always read the leaflet (2005) Report of the
6
Committee on Safety of Medicines Working Group on Patient
7
Information’1
8
The Working Group on Patient Information received advice from patients and
9
experts on the quality of Patient Information Leaflets (PILs) and how to
10
improve them. Patient organisations identified the following with regards to
11
PILs:
12

The quality is variable and language is complex with too much jargon
13

The leaflet is often too busy and the print too small
14

Leaflets are too negative and do not mention enough of the benefits
15

The PIL should complement discussion with the prescriber, ideally
available in the consultation
16
17

organisations for further advice could be given
18
19

Helpline numbers and website addresses for further information should
be mentioned
20
21
One PIL can not meet everyone’s needs so information on patient

Comparative information and information about lifestyle issues can help
in decision making
22
23
24
Expert views regarding PILs:
25

Too much use of jargon
26

The use of capital letters was eye-catching but hard to read
27

Inappropriate punctuation can obscure the message
28

Text in boxes may be skipped over
29

Euphemisms are not helpful when referring to serious side effects
30

Messages should be consistent
Medicines concordance: full guideline DRAFT (July 2008) page 118 of 373
DRAFT FOR CONSULTATION
1

Language should be clear and unambiguous
2
3
Research2 showed that patients prioritise four key points of information about
4
a medicine – side effects, dos and don’ts, what it does and how to take it – but
5
different people prefer different orders of priority.
6
The Working Group recommended the following for an improved readability
7
guideline:
8
Usability – PILs should be clear and understandable to the reader. This
9
incorporates writing style, typeface, design and layout, headings, use of
10
colour, use of symbols and pictograms. The use of templates so that
11
information is presented consistently would be useful. (see annex 6). PILS
12
should not be too long and complex.
13
User involvement in PILs – user testing of patient information is
14
recommended. This should be done under controlled conditions and meeting
15
certain stipulations. User testing of content and impact is important. The
16
production of PILs by Companies often occur at the end of the medicine
17
development process, with little thought of involving patients in writing and
18
testing the information. Views should be at all stages of development.
19
Communicating Risk
20
It is fundamental when making an informed decision to understand and weigh
21
up the risk and benefits of a treatment. The working group suggested:
22

and effectiveness of using the medicine.
23
24
Use of headlines to summarise the most important messages for safety

Information on all the side effects is required by law but must be
25
presented logically and include a description of side effects, estimating
26
frequency and advice on necessary actions.
27

Inclusion of the potential benefits to provide balance is important.
Medicines concordance: full guideline DRAFT (July 2008) page 119 of 373
DRAFT FOR CONSULTATION
1

Provide information about the harmful effect itself; the probability of it
2
occurring and how to minimise risk and what actions to take if a
3
problem arises
4

Put most important information first, include information on benefit, use
5
the right words and use numbers to convey risk. Also a supplementary
6
leaflet of risks and benefits in addition to PILs would be useful.
7
8
Trust in the information source is also important. Harms and benefits should
9
be side by side and drug side effects must legally be provided. Care should be
10
taken to give unbiased and clear statistical information.
11
To increase trust in PILs transparency of data and certainty of risk estimates
12
may be effective. To avoid unnecessary concerns the use of clear information
13
on a true scale and nature of such risks are important, such as using
14
analogies and alternative risk scales to show rarity of risk; describe baseline
15
risk and increased risk; provide further information sources on these risks.
16
Headlines – It is suggested that information could be portrayed in headlines
17
which should include: why the patient should take the product; the maximum
18
dose or duration of treatment; potential side effects/withdrawal reactions;
19
contraindications; important drug interactions; circumstances in which the
20
drug should be stopped; what to do if the medicine does not work or where to
21
find further information. Headlines should also include a firm encouragement
22
for the patient to read the rest of the leaflet.
23
Balance - It is important to be balanced and convey information on benefits
24
as well as risks in order for the information to be credible. The PIL should
25
therefore include the potential benefits of taking the medicine. Research
26
shows that short factual statements on benefits help weigh the risks and
27
benefits. It must also be compatible with the Summary of Product
28
Characteristics, and be useful to the patient but not promotional.
29
Information to give a balanced account would include:
30

Why it is important to treat the disease
Medicines concordance: full guideline DRAFT (July 2008) page 120 of 373
DRAFT FOR CONSULTATION
1

Whether the treatment is for short term or chronic use
2

Whether the medicine is being used to treat the underlying disease or
for controlling symptoms
3
4

Whether the effects will last after medication stopped
5

Where it is to treat two or more discrete indications, all should be
succinctly described as above
6
7

Where to obtain more information on the condition
8
9
10
11
Side effects
Better information about side effects would include:

Logical order - the most important information should be first e.g.
12
situations where need to take action such as stopping medication or
13
getting medical help
14

what to do if encounter serious problems
15

Estimates of frequency should be mentioned – as the most serious side
effects are also the rarest
16
17

seriousness/severity
18
19

Many side effects are dose related and so a warning statement is
needed but should not alarm those prescribed high doses
20
21
Use the right wording – not just describe the side effect but convey

It would be useful to have a glossary of lay terms – so there is
standardised side effect lay terminology across medications.
22
23
24
Expression of risk
25
Expressing statistical risk in PILS:
26

quantifying risk using absolute numbers;
27

verbal descriptors of risk only used if accompanied by equivalent
28
statistical information;
Medicines concordance: full guideline DRAFT (July 2008) page 121 of 373
DRAFT FOR CONSULTATION
1

convey uncertainty around risk estimates; frequency ranges; duration
2
of risk; frequency estimates based on spontaneous adverse drug
3
reaction data; constant denominators;
4
5
Concepts deemed inappropriate by the Working Group were:
6

NNT/NNH
7

positive framing and negative framing – too cumbersome and lengthy
8

use of diagrams – constraints in size
9
10
Supplementary information - a leaflet about risks and benefits in addition to
11
PILs would be able to go into more detail.
12
Meeting the needs of special groups of patients
13
Not everyone finds it easy to access and use information in the PIL, e.g.
14
visually impaired people, people whose English is not their first language,
15
people with poor literacy and numeracy, those with learning difficulties or
16
physical difficulties.
17
Suggestions are made and projects described to help these patient groups:
18
For health literacy:
19

A health search engine for healthcare staff and public;
20

Patient information bank for NHS trusts to print consistent information
for individuals on their care and treatment;
21
22

Power questions to ask in consultations;
23
24
Poor basic skills:
25

Clearly written in plain English
26

Signpost other sources of information
27

Helplines
Medicines concordance: full guideline DRAFT (July 2008) page 122 of 373
DRAFT FOR CONSULTATION
1
2
Patients with sight loss:
3

leaflets in Braille or large print
4

audio version
5

leaflets on the web
6

digital television
7

telephone helplines and automated voice systems.
8
9
10
First language not English:

Provision of leaflets in other languages from the company in written or
web-based format
11
12

Telephone helplines
13

The use of translator services
14
15
16
Medicines for children and young people

and so leaflets should be aimed at adults
17
18
Information for Children should be communicated by parents or carers

Information for young people should take into account their lifestyle of
the age group and likely questions
19
20
21
22
Provision of information for carers

may need training on administration techniques
23
24
25
Carers may not be in the consultation when medication prescribed and

Outside power of the group but use of a telephone helpline could
address some concerns
26
27
The Pharmaceutical Companies
Medicines concordance: full guideline DRAFT (July 2008) page 123 of 373
DRAFT FOR CONSULTATION
1
2
Responsibilities:

to the information on the PILs and other measures.
3
4
It is suggested that the pharmaceutical industry could promote access

Portfolio of Information keys for pharmaceutical companies – use these
5
to help identify additional measures that would promote the
6
dissemination of information on safe use of their products to ensure
7
vulnerable group s can access it.
8

Leaflets in other formats; how to signpost these other formats;
9
translation into other languages; use of information mediators such as
10
helplines; expert sources of advice. The PIL can be a pointer to other
11
sources of information for vulnerable groups e.g. booklets, simplified
12
leaflets, magazines and websites.
13
14
The information format of the patient leaflet is very important and should be
15
clear and understandable. The information needs to be balanced, trustworthy,
16
include benefits as well as side effects, with the most important information
17
highlighted. The communication of risk should be conveyed with seriousness
18
but without alarm for the patient. Where to get extra information should be
19
mentioned, if not a separate detailed booklet given. It would be very good
20
practice to have patients test the leaflet to see it’s appropriateness. Special
21
groups of patients should be taken into account while producing PILs.
22
4.10.1.2
23
Raynor (2007) (Raynor, D. K., Blenkinsopp, A., Knapp, P. et al ,
2007) Summary
24
Raynor (2007) (Raynor, D. K., Blenkinsopp, A., Knapp, P. et al , 2007)
25
conducted workshops to elicit stakeholder perceptions of the key issues
26
surrounding information presentation to patients. The workshop discussions
27
found timing of the delivery of the information important, which was often
28
presented after the medicine was prescribed. Sometimes no leaflet was
29
available at all. They found too much information to be overwhelming, harder
30
to understand, often frightening and can have too much irrelevant information.
31
Readability was the most important part of written medicines information – the
32
size of text and content, meaningful information and not jargon.
Medicines concordance: full guideline DRAFT (July 2008) page 124 of 373
DRAFT FOR CONSULTATION
1
2
Other information of importance:
3

dosage and ingredients
4

when and how long to take it
5

the likelihood of it being successful
6

side-effects e.g. how common or rare they are
7

factors relevant to their personal medical condition
8
9
Also the role of medicines:
10

how to take the drug effectively
11

its potential side-effects and interactions
12

how to reduce potential harm from medicines
13

how long before the medicine will have beneficial effects
14

why it is necessary to finish the course
15

why it was recommended not to drink alcohol
16
17
18
What makes medicines information effective?

Timing of delivery of the information – more effective during the
consultation
19
20

Visually appealing and straightforward to read
21

No jargon
22

Basic information about what the medicine contains
23

Designed for patients or professionals
24
25
What participants feel makes medicines information valuable?
26

Looks and feels important and highlights priority information
27

Permits an informed choice
28

Is reassuring and reduces concern, conflict and anxiety about whether
the medicine is the right one for them
29
30

Gives them confidence in taking medicines
Medicines concordance: full guideline DRAFT (July 2008) page 125 of 373
DRAFT FOR CONSULTATION
1
4.10.1.3
The ‘Medicine use review: Understand your medicine’ NHS report4
summary
2
3
The ‘Medicine use review: Understand your medicine’ NHS report4 suggests
4
questions patients should ask about their medication. These could be adapted
5
(as below) into written information for patients:
6

What the medicine does
7

Why is it important to take the medicine
8

Other treatment options
9

When and how it should be taken
10

How long it should be taken for
11

What medicines, drinks, foods or activities the patient should be aware
of when taking the medicine
12
13

What the patient should do if they do not feel well when taking it
14

How the patient can tell if it’s helping
15

How the patient can be sure it’s safe to take it
16

The possible risks and side effects
17

What to do if they get one of the side effects
18

What happens if the patient stops the medicine or takes a lower dose
19

Is there anything they can do to remember to take their medicines
20

Where to go for more information
21
“Taking Medicines” leaflet summary
22
4.10.1.4
23
The Medicines and Healthcare products Regulatory Agency (MHRA) have
24
produced a leaflet for patients called ‘Taking Medicines – some questions and
25
answers about side effects’. The leaflet has 8 questions and short answers to
26
these questions. Patients are advised that they should receive a patient
27
information leaflet with their medicines, and to ask their doctor, pharmacist or
28
NHS direct if they have further queries. The questions in the leaflet are:
29
1.
What do medicines do?
30
2.
Will my medicine cause side-effects?
Medicines concordance: full guideline DRAFT (July 2008) page 126 of 373
DRAFT FOR CONSULTATION
1
3.
What is meant by a common or rare side
effect?
2
3
4.
How much medicine should I take?
4
5.
How can I reduce the risk of side effects?
5
6.
Do side effects always come on straight
away?
6
7
7.
What should I do if I feel unwell after I take
my medicine?
8
9
8.
Will my medication affect my lifestyle?
10
11
12
1 Always Read the Leaflet: Report of the Committee on Safety of Medicines Working Group on Patient Information
13
(2005). The Stationery Office, London.
14
2 Berry, D.C. (1997) Psychol Health. As referenced in Always Read the Leaflet: Report of the Committee on Safety of
15
Medicines Working Group on Patient Information
16
17
4 The ‘Medicine use review: Understand your medicine’ (2005) NHS Report by MHRA and ASK.
Medicines concordance: full guideline DRAFT (July 2008) page 127 of 373
DRAFT FOR CONSULTATION
1
2
4.11
Key clinical Question: Mental capacity narrative
3
Some concern was expressed by the GDG about the potential conflicts
4
between respecting autonomy of the patient and the duty of care felt by
5
practitioners to patients. The GDG discussed the importance of professionals
6
codes of conduct and the legal system in protecting both patients and
7
professionals. The narrative below brings together the principles discussed
8
and is a repeat of section 3.4 in chapter 3.
9
4.11.1 **Roles and responsibilities of patients and health care
10
professionals
11
Concern has been expressed by practitioners that sharing decisions with
12
patients may conflict with their duty of care to patients or their legal or ethical
13
obligations (Stevenson, F. A., 2003). While the UK General Medical Council
14
(GMC) (2001)9 considered one of the key duties of a doctor to ‘respect the
15
rights of patients to be fully involved in decisions about their care’ for many
16
clinicians there is a legitimate area of concern or indeed of conflict between
17
respect for autonomy of the patient and the duty of beneficence when a
18
clinician feels uncomfortable about a patient’s wishes. The GMC (2008)10 has
19
recently updated guidance for doctors about patient consent which makes
20
explicit the right of competent patients to make decisions about their own
21
healthcare. The guidance emphasises the need for doctors to maximize
22
opportunities and patients abilities to make decisions for themselves and that
23
doctors must respect competent patients’ decisions even if they do not agree
24
with them. Doctors do not have to provide a treatment to patients which they
25
believe is not in the patients’ interest but must under such circumstances
26
explain their reasons and other options, including a second opinion to the
27
patient. It remains important however for health care professionals to do their
28
utmost to ensure that patients’ choices are informed choices as outlined
29
above.
9
http://www.gmc-uk.org/publications/annual_reports/review_archive/report2002.pdf
http://www.gmc-uk.org/publications/business_plans/Business_Plan_2008.pdf
10
Medicines concordance: full guideline DRAFT (July 2008) page 128 of 373
DRAFT FOR CONSULTATION
1
The Mental Capacity Act (2005) 11 governs the making of decisions for people
2
who lack capacity in England and Wales. Under the Act doctors are advised
3
that they must work on the presumption that every adult patient has the
4
capacity to make decisions about their care, and to decide whether to agree
5
to, or refuse, an examination, investigation or treatment. A patient is regarded
6
as lacking capacity once it is clear that, having been given all appropriate help
7
and support, they cannot understand, retain, use or weigh up the information
8
needed to make that decision, or communicate their wishes12.
9
Doctors are advised that assumptions must not be made that a patient lacks
10
capacity to make a decision solely because of their age, disability,
11
appearance, behaviour, medical condition (including mental illness), their
12
beliefs, their apparent inability to communicate, or the fact that they make a
13
decision that health professionals disagree with13.**
14
11
12
http://www.opsi.gov.uk/ACTS/acts2005/pdf/ukpga_20050009_en.pdf
http://www.gmc-uk.org/publications/business_plans/Business_Plan_2008.pdf
13
http://www.gmc-uk.org/publications/business_plans/Business_Plan_2008.pdf
Medicines concordance: full guideline DRAFT (July 2008) page 129 of 373
DRAFT FOR CONSULTATION
1
2
4.12
Key Clinical Question: What information about shared
3
decision making and adherence should be recorded in
4
patients’ notes?
Related
Evidence
references
statements
Evidence into recommendations
(summary of
evidence)
.
The GDG considered it important that a record is kept of
discussions between health care professionals and patients
about medicines. Prescribing and taking of medicines is a
long term dynamic process which may involve multiple
interactions between health care professionals and patients.
Good record keeping aids continuity of care by providing
information for healthcare professionals to review the
discussion they or other health care professionals have had
with patients about their medicines. The GDG made
recommendations based on professional opinion.
5
6
Medicines concordance: full guideline DRAFT (July 2008) page 130 of 373
DRAFT FOR CONSULTATION
1
4.13
Key Clinical Question: What tools are available to
2
support the patient in reaching an informed decision?
3
How effective are these tools?
4
Related References
Evidence Statements (summary
Evidence into recommendations
of evidence)
including
O’Connor (2003)
1. One systematic review found
The literature review found a
(O'Connor, A. M.,
that there a variety of decision aids
number of systematic reviews
Stacey, D., Entwistle, used in studies - audio guides, CD-
concerning decision aids and their
V. et al , 2003)
ROMs, web-based, interactive
use. The results of the trials
video-disc, lecture and handouts.
primarily related to decisional
conflict, satisfaction, involvement i
O’Connor (2003)
2. One systematic review found
decision and participation with little
(O'Connor, A. M.,
that decision aids led to greater
effect on health outcomes overall.
Stacey, D., Entwistle, knowledge, realistic expectations,
The GDG considered the evidence
V. et al , 2003)
lower decisional conflict from
supportive of the importance of
feeling informed, more active in
information in a variety of formats
decision making and less
patients but did not feel it
indecision after the intervention.
appropriate to make specific
recommendations regarding
O’Connor (2003)
3. Decision aids reduced decisional
(O'Connor, A. M.,
conflict (1 systematic review, 4
decisions aids.
Stacey, D., Entwistle, RCTs).
V. et al , 2003);
Montgomery (2003)
(Montgomery, A. A.,
Fahey, T., and
Peters, T. J., 2003);
Weymiller (2007)
(Weymiller, A. J.,
Medicines concordance: full guideline DRAFT (July 2008) page 131 of 373
DRAFT FOR CONSULTATION
Montori, V. M.,
Jones, L. A. et al ,
2007); Thomson
(2007) (Thomson, R.
G., Eccles, M. P.,
Steen, I. N. et al ,
2007); Oakley (2006)
(Oakley, S. and
Walley, T., 2006) *;
O’Connor (2003)
4. Decision aids improved patient
(O'Connor, A. M.,
knowledge (1 systematic review, 3
Stacey, D., Entwistle, RCTs).
V. et al , 2003);
Montgomery (2003)
(Montgomery, A. A.,
Fahey, T., and
Peters, T. J., 2003);
Weymiller (2007)
(Weymiller, A. J.,
Montori, V. M.,
Jones, L. A. et al ,
2007); Thomson
(2007) (Thomson, R.
G., Eccles, M. P.,
Steen, I. N. et al ,
2007)
O’Connor (2003)
5. One systematic review found
(O'Connor, A. M.,
that simpler decision aids
Stacey, D., Entwistle, compared to detailed decision aids
V. et al , 2003)
showed a significant improvement
in knowledge, more realistic
expectations and greater
Medicines concordance: full guideline DRAFT (July 2008) page 132 of 373
DRAFT FOR CONSULTATION
agreement between values and
choices.
Montgomery (2003)
6. One RCT found that decision
{2257}
analysis decreased decisional
conflict more than a video/leaflet
intervention.
Fraenkel (2007)
7. One RCT found that an
(Fraenkel, Liana,
interactive computer tool which
Rabidou, Nicole,
generated personalised feedback
Wittink, Dick et al ,
significantly improved decisional
2007)
self-efficacy and preparedness to
participate in decision-making, with
greatest benefit for older adults.
Oakley (2006)
8. One RCT found that a workshop
(Oakley, S. and
plus a decision aid (identifying own
Walley, T., 2006)
risk and pros and cons) and
worksheet did not significantly
improve adherence and although
were initially satisfied with the
information on medicines this was
non-significant by end of trial.
Hamann (2007)
9. One RCT found that a decision
(Hamann, Johannes,
aid and planned talk with doctor
Cohen, Rudolf,
reduced hospitalisation for
Leucht, Stefan et al ,
schizophrenic outpatients. However
2007)
those who showed a higher
preference for autonomy and better
knowledge showed a significantly
higher re-hospitalisation rate.
Medicines concordance: full guideline DRAFT (July 2008) page 133 of 373
DRAFT FOR CONSULTATION
1
*This was compared to baseline, as was not studied if it differed from the
2
control group.
3
4.13.1 Methods of evidence review
4
This paper includes a narrative summary of the included evidence, structured
5
according to the category of the intervention, following the agreed reviewing
6
protocol:
7
Types of studies: Systematic Reviews or Randomized controlled trials
8
(RCTs) of tools to help the patient reach a decision (decision aids).
9
Types of participants: people prescribed medication for a medical condition
10
faced with a decision.
11
Duration of studies: No time limit specified.
12
Types of interventions: any interventions which aid the patient in making an
13
informed decision.
14
Types of outcome measures: types of decision aid; patient outcomes:
15
decisional conflict, patient knowledge, self-efficacy.
16
17
It should be noted that the remit of the guideline is for conditions with
18
prescribed medication and this excludes conditions which require
19
chemotherapy or screening. All RCTs are within this remit, however many of
20
the systematic reviews included populations outside the remit, this is noted
21
where applicable.
22
4.13.2 Evidence review
23
O’Connor (2003) (O'Connor, A. M., Stacey, D., Entwistle, V. et al , 2003)
24
conducted a Cochrane Collaboration systematic review of decision aids for
25
people facing health treatment or screening. They included studies that were
26
RCTs; involving those over age of 14 making decisions about screening or
27
treatment options for themselves, a child, or significant other; including
28
decision aids and looking at whether the decision aids achieved their
29
objectives. They found 131 decision aids developed in the within the previous
Medicines concordance: full guideline DRAFT (July 2008) page 134 of 373
DRAFT FOR CONSULTATION
1
five years, 94 were web-based, 14 paper-based and 12 videos with print
2
resources, eight were audio-guided print resources, two were CD-ROMs and
3
one a web-based with workbook. Most of the decision aids were intended for
4
use before counselling. Decision aids were better in terms of greater
5
knowledge, realistic expectations, lower decisional conflict related to feeling
6
informed, increased proportion of people active in decision making and
7
reduced proportion of people who remained undecided post intervention.
8
Simpler decision aids were proven to have more significant improvement than
9
more detailed decision aids in knowledge, more realistic expectations and
10
greater agreement between values and choices. There was no improvement
11
compared to comparisons for affecting satisfaction with decision making,
12
anxiety, and health outcomes.
13
Only a few of the studies in this systematic review were relevant to the
14
guideline as the majority of studies included were surgery, screening, or other
15
populations not included in the medicines concordance remit. The studies that
16
were relative to the guideline were seven trials of hormone replacement
17
therapy* (audioguides, booklet and interactive videodisc, mix of lecture and
18
handouts); two trials involving Ischaemic heart disease**, interactive videodisc
19
and a videocassette; and one study of atrial fibrillation treatment***
20
(audioguide deciding whether to change from aspirin to warfarin). All studies
21
except five of the HRT trials compared a decision aid with usual care. The
22
other five studies compared a multiple element design with a simple decision
23
aid.
24
*McBride 2002, Murray HRT 2001, O’Connor, 1998-RCT, Dodin 2001,
25
O’Connor Wells 1999, Rothert 1997, Rostrom 2002.**Morgan 1997, Bernstein
26
1998.***Man-Son-Hing 1999.
27
Montgomery (2003) (Montgomery, A. A., Fahey, T., and Peters, T. J., 2003)
28
conducted a factorial randomized control trial to evaluate two interventions to
29
help hypertensive patients decide whether to start drug therapy to reduce
30
blood pressure. This was carried out in 21 general practices in SW England
31
with 217 patients aged 32 to 80 years (mean age 59 years) newly diagnosed
32
with hypertension. Patients were allocated to decision analysis or no decision
Medicines concordance: full guideline DRAFT (July 2008) page 135 of 373
DRAFT FOR CONSULTATION
1
analysis, they were then further randomised to video/leaflet or no video/leaflet
2
groups. The decision analysis in this case was a decision tree for
3
hypertension which used a computerised self-completed interview to assess
4
patients’ utilities with minimal input from researcher and absolute
5
cardiovascular risk calculated. The video was informational about high blood
6
pressure. The information booklet was four fact sheets from the British
7
Hypertension Society. The primary outcome was the total score on the
8
decisional conflict scale, a questionnaire measuring how uncertain about the
9
course of action to take and factors which could be changed that lead to the
10
uncertainty. Secondary outcomes were subscales of the Decisional Conflict
11
Scale and intentions about starting treatment, state anxiety, knowledge of
12
hypertension, actual treatment decision.
13
Both interventions successfully reduced patients’ total decisional conflict at
14
follow-up. Decision analysis decreased the decisional conflict more than the
15
video/leaflet. Total decisional conflict mean for decision analysis was 27.6
16
(SD=12.1), no decision analysis 38.9(18.3) adjusted difference 95%CI -9.4 (-
17
13.0 to -5.8) p<0.001; video/leaflet 30.3 (13.4) and no video/leaflet was
18
36.8(18.8), 95% CI -4.2(-7.8 to -0.6), p=0.021. The Decisional conflict
19
subscales showed a clear reduction in three of the five subscales -
20
uninformed 23.7 (11.8) compared to no decision analysis 40.7 (23.1) adjusted
21
difference 95% CI -15.7 (-20.2 to -11.2), unclear values 28.4 (14.7) vs 43.8
22
(24.3) adjusted difference -13.1 (-18.0 to -8.1) and unsupported 24.4 (13.4 vs
23
34.8 (18.3) adjusted difference -8.7 (-12.8 to -4.7) and some evidence for
24
reduction in uncertainty and no evidence for decision quality. The video/leaflet
25
intervention showed no evidence in these last two subscales and there was
26
only clear evidence on the uninformed subscale. For the intention to start
27
treatment when followed up the adjusted risk ration (95% CI): Yes versus
28
unsure 1.19 (0.59 to 2.40) for decision analysis and 1.80 (0.89 to 3.63) for the
29
video/leaflet. No versus unsure 3.15 (0.91 to 10.98) and 0.52 (0.15 to 1.77)
30
respectively. The overall p values were 0.09 and 0.17 respectively. Actual
31
prescription of medication was not different for either intervention or controls.
32
There was a suggestion (p=0.055) that anxiety may be reduced by decision
33
analysis although the evidence there was weak and no evidence of this for the
Medicines concordance: full guideline DRAFT (July 2008) page 136 of 373
DRAFT FOR CONSULTATION
1
video/leaflet intervention. Both interventions significantly increased knowledge
2
of hypertension. Those who received both interventions had the lowest
3
decisional conflict (27.1 compared with 28.2 and 33.3 and 44.2 for decision
4
analysis only, video/leaflet and control). They had a high knowledge score –
5
the same as video/leaflet. Within the regression models there was a
6
significant (antagonistic) interaction between decision analysis and
7
video/leaflet, so the effect of each was reduced by the presence of the other
8
(interaction coefficient 12.5, 95% CI 5.4 to 19.5, p=0.001 for decisional conflict
9
and -9.1, 95% CI-16.3 to -1.9, p=0.013 for knowledge. This study was
10
followed up in 2005 by Emmett, who found that there was no evidence of any
11
difference in blood pressure, cardiovascular disease risk for either intervention
12
or between them. There were also no effects on medication prescribing, self-
13
reported adherence, consulting behaviour or management changes.
14
Weymiller (2007) (Weymiller, A. J., Montori, V. M., Jones, L. A. et al , 2007)
15
conducted an RCT study of the effect of a decision aid on statin drug
16
treatment decision making. The study was conducted in a diabetes referral
17
clinic in Minnesota, USA. 98 participants were included with a mean age of
18
64 (12) for the decision aid group and 66(8) for the control group. Participants
19
were randomized to either receive usual care plus a standard pamphlet on
20
cholesterol management or a statin choice decision aid. The decision aid
21
included name, cardiovascular risk factors, an estimated level of
22
cardiovascular risk (3 levels) and the absolute risk reduction with statins and
23
the potential disadvantages of taking them. A question prompted patients to
24
express whether they were ready to make a decision and if they wanted to
25
take statins, discuss the issues with their clinician or other. After the
26
consultation the participants were given a questionnaire to complete. The
27
outcomes of interest were improvement in patient knowledge and reduction in
28
decisional conflict. 74% would recommend the decision aid to others
29
compared to 53% of control patients recommending the pamphlet, (OR 2.6,
30
95% CI 0.8-8.0), 68% would want to receive similar support for future
31
decisions compared to 58% of control patients (OR 1.5, 95% CI 0.6-3.8).
32
Those receiving the decision aid had higher knowledge scores than the
33
control group and those allocated to receive the intervention during the visit
Medicines concordance: full guideline DRAFT (July 2008) page 137 of 373
DRAFT FOR CONSULTATION
1
achieved better knowledge than those who received before the consultation.
2
The intervention group had significantly less decisional conflict afterwards
3
than the control group, and at 3 months (although not statistically significant.
4
Those in the DA group felt more informed. 30% of the DA group (6/7 were
5
from the DA during the visit group) decided to start statin therapy immediately
6
after, compared to 21% of the control group. At 3 months 63% of the DA
7
group and 63% of the control group reported taking statins (OR, 1.4, 95%CI,
8
0.8-2.4). Overall, there was no difference in adherence to patient choice at 3
9
months.
10
Thomson (2007) (Thomson, R. G., Eccles, M. P., Steen, I. N. et al , 2007)
11
conducted a randomized controlled trial of a decision aid for anti-thrombotic
12
treatment of patients with atrial fibrillation. 109 patients aged over 60 years
13
from 40 general practices in the Northeast of England. The intervention
14
involved the doctors in the clinic delivering either the decision aid or guidelines
15
to the patient. The decision aid was a computerized aid which presented the
16
individualized benefits and potential harms of warfarin treatment and
17
participants were invited to weigh up the advantages and disadvantages of
18
treatment before coming to a shared decision with the doctor. This involved
19
personalized risk assessment using the Framingham equation for stroke risk
20
and the presentation used graphical and numerical formats followed by a
21
shared decision-making component. The evidence-based guidelines group
22
applied decision analysis derived guidelines according to the participants’ risk
23
factor profile and the recommendation made directly to the participant by the
24
clinic doctor. The primary outcome measure was decision conflict scale
25
immediately after the consultation. Secondary outcomes were anxiety,
26
knowledge and decision-making preferences. The computerized decision aid
27
group had lower decision conflict immediately after the clinic (mean -0.18,
28
95%CI -0.34 to -0.01) and mean -0.15 (-0.37 to 0.06) at three month follow-
29
up. Both groups had less decision conflict after the consultation but the
30
difference between groups was significant at 5% level. Subscales suggested
31
this was due to feeling better informed and clearer of their personal values for
32
the risks and benefits of alternative options. The reduction in anxiety fell
33
significantly but was not different between groups. Knowledge scores
Medicines concordance: full guideline DRAFT (July 2008) page 138 of 373
DRAFT FOR CONSULTATION
1
improved slightly after the consultation but at three months were back at
2
baseline level. Participants in the decision aid group were less likely to start
3
warfarin than those in the guideline arm (39/53, 73.6%) compared to
4
guidelines (50/56, 81.7%), RR 0.82, 95% CI 0.68 to 0.99, this was however
5
almost completely due to participants not already on warfarin, here the
6
differences was 4/6, 25% compared to guidelines 15/16, 93.8%, RR 0.27,
7
95%CI 0.11 to 0.63. There was no difference in health outcomes 3 months
8
after the clinic.
9
Fraenkel (2007) (Fraenkel, Liana, Rabidou, Nicole, Wittink, Dick et al , 2007)
10
conducted a randomized control trial which tested the efficacy of a computer
11
tool to improve informed decision-making for patients with knee pain in an
12
outpatient clinic. The trial was conducted in a primary care outpatient clinic in
13
the USA. 87 Participants over the age of 60 were randomised to receive an
14
Arthritis Foundation information pamphlet (control group) or to perform an
15
Adaptive Conjoint Analysis (ACA)(intervention group). The ACA is an
16
interactive computer tool which could generate immediate feedback to the
17
participant and helps them construct treatment preferences by means of
18
tradeoffs by rating tasks. The treatment characteristics in the ACA task
19
included route of administration, likelihood of expected benefit, and risk of
20
adverse effects. Questionnaires were given to assess outcomes. Decisional
21
self-efficacy and preparedness to participate in decision-making remained
22
significantly higher in the intervention group than the control group after
23
controlling for race and health status. Arthritis self-efficacy was of borderline
24
significance. Outcomes by age and education suggest older adults may be the
25
most likely to benefit. 98% of the participants thought the ACA task was very
26
easy/easy to do. The patients in the intervention group had greater self-
27
confidence in their ability, felt more prepared to participate in shared decision-
28
making and felt they had greater self-efficacy over arthritis than the control
29
group.
30
Oakley (2006) (Oakley, S. and Walley, T., 2006) assessed the effectiveness
31
of a decision aid on patient adherence to oral bisphosphonate medication.
32
They conducted an RCT with postmenopausal women over the age of 65 who
Medicines concordance: full guideline DRAFT (July 2008) page 139 of 373
DRAFT FOR CONSULTATION
1
were diagnosed with osteoporosis. 33 participants were included in the study,
2
with a mean age of 77 years (range 61-90). The intervention group attended
3
an osteoporosis workshop, where they were given a decision aid and group
4
discussion on a worksheet. This included working through the decision-
5
making process and identifying their own lifetime risk of hip fracture, and
6
personal and family health issues which may influence their decision in
7
addition to self-care options already using or willing to try. They were to
8
consider pros and cons and their personal values regarding therapy and noted
9
any questions for the GP on a worksheet. Two weeks later they returned with
10
their worksheets for a follow up with a GP osteoporosis specialist.
11
Questionnaires were then administered to establish compliance. The
12
difference in adherence improvement was not significant. Satisfaction with
13
information about medicines improved initially in the intervention group but
14
reduced by the final questionnaire to non-significance. The decisional conflict
15
scale showed a reduction in decisional conflict from baseline measures (total
16
DCS score pre-intervention, median 2.5 (1.8-3.4) v 2.0 (1.0-2.4) post-
17
intervention, p<0.001, however this cannot be compared to the control group
18
as they were not assessed for this measure. The decision aid improved their
19
ability to make a decision about which treatment was best and to discuss their
20
medication with the GP but had no obvious effect on adherence.
21
Hamann (2007) {3478} conducted an RCT to assess whether shared decision
22
making in antipsychotic drug choice would influence long-term outcome. 86
23
patients with a diagnosis of schizophrenia were included and followed up,
24
58% female with mean age of 38 years (SD 11.4), mean duration of illness 9.2
25
years (SD 8.5). The study was conducted in Germany. The intervention group
26
received a decision aid program and the control group received usual care.
27
The decision aid was a 16 page booklet with the pros and cons of oral versus
28
depot formulation, first versus second generation antipsychotics,
29
psychoeducation and a type of socio-therapeutic intervention. 24 hours after
30
working through the booklet with the trained nurses the patients consulted
31
with the psychiatrists on further treatment. The patients then filled out
32
questionnaires relating to patient involvement, satisfaction and
33
psychopathology. Patients were followed up at 6 and 18 months after
Medicines concordance: full guideline DRAFT (July 2008) page 140 of 373
DRAFT FOR CONSULTATION
1
discharge from hospital. Univariate analysis found no significant differences
2
between groups. When multivariate analysis was conducted to control for the
3
re-hospitalisation rate it showed that there was a positive trend for the
4
decision aid and planned talk in reducing rehospitalisation. Higher
5
participation preferences (OR= 1.06, p=0.03) and better knowledge (OR
6
=1.23, p=0.03) rates significantly predicted rehospitalisation. No other effects
7
were shown.
8
Medicines concordance: full guideline DRAFT (July 2008) page 141 of 373
DRAFT FOR CONSULTATION
1
2
4.14
Key Clinical Question: How can a practitioner elicit
3
whether a patient agrees with the prescription
4
recommended by the practitioner?
Related
Evidence statements
references
(summary of evidence)
Evidence into recommendations
1. No evidence was found
No tools designed for use in clinical practice
on specific clinical tools
were found. The GDG used the information
that can aid the
from the research studies and their
practitioner in eliciting
professional opinion to make
whether patient agrees
recommendations in relation to elicitation of
with the recommended
agreement with decision to prescribe.
prescription
5
6
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
Medicines concordance: full guideline DRAFT (July 2008) page 142 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
4.14.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies – no restrictions on study design.
6
Types of participants - people prescribed medication for a medical condition.
7
Duration of studies - No time limit specified.
8
Types of interventions - Any interventions intended to assess tools that can
9
aid a practitioner in eliciting whether a patient agrees with recommended
10
prescription
11
Types of outcome measures – Agreement with prescription, patient
12
satisfaction with issues prescription.
13
4.14.2 Evidence review
14
We did not find any study that assessed a clinical tool that could aid a
15
practitioner in eliciting whether a patient agrees with the recommended
16
prescription. We did find one prospective observational study (Bikowski 2001)
17
(Bikowski, R. M., Ripsin, C. M., and Lorraine, V. L., 2001) that aimed to
18
characterize the degree of disparity between Physicians’ perceptions of older
19
patients’ medication regime and patients’ perceptions of their regime within a
20
community family medicine residency setting.
21
Eligible patients were age 65 and older, non-institutionalised, visiting the clinic
22
on the index day for a routine visit, had seen the index physician at least three
23
times in the past calendar year, and, by brief review of the medication flow
24
sheet, were taking at least four prescriptions medications. Study sample
25
comprised 50 physician-patient pairs.
26
Physicians were given the patients chart’s, with a request to complete a
27
questionnaire that asked for information on all prescriptions and
28
nonprescription medications, with dosages and frequencies of administration.
Medicines concordance: full guideline DRAFT (July 2008) page 143 of 373
DRAFT FOR CONSULTATION
1
Patients were interviewed at home by first year medical students who
2
received specific training for the study.
3
Percentage congruence- defined as agreement between physician and patient
4
regarding all prescriptions medications, dosages and frequency, was
5
calculated for each pair. Complete congruence was showed for 14% of the 50
6
physician-patient pairs; 74% had at least one medication that either the
7
physician was unaware the patient was taking or the physician thought the
8
patients was taking but that was not part of the patients regime; 12% had
9
dose and/or frequency differences, however they agreed upon the
10
medications in the regime.
11
Antihypertensives were the most commonly prescribed medication,
12
accounting for 36% of the total. The highest congruence was found for
13
diabetic and other endocrine drugs. Pain medications and gastrointestinal
14
medications showed the lowest congruence
15
Medicines concordance: full guideline DRAFT (July 2008) page 144 of 373
DRAFT FOR CONSULTATION
1
4.15
Key Clinical Question: What aspects of consultation
style increase patient involvement in decision-making?
2
3
Related References
Evidence Statements (summary
Evidence into recommendations
of evidence)
including
McKinstry (2006)
1. One high quality systematic
The GDG were aware that there is
(McKinstry, B.,
review found that there is
anecdotal evidence that
Ashcroft, R. E., Car,
insufficient evidence to conclude
practitioners and patients report
J. et al , 2006)
that any intervention may increase
that the quality of the doctor-
or decrease trust in physicians.
patient relationship is important in
decision –making. The quality of
the practitioner-patient relationship
was reported as being important in
Van Dam (2003)
2. One systematic review of RCTs
some studies of patients medicine
(van-Dam, Henk A.,
found that supporting patient
taking behaviour as outlined in
2003)
participation in diabetes care and
chapter 3. The quality of a
self-care behaviour (i.e. by
practitioner –patient relationship is
assistant-guided patient
however likely to be influenced by
preparation for visits to doctors,
a number of factors that relate to
empowering group education,
previous consultations and
group consultations, or automated
problem under discussion. The
telephone management) is more
consultation skill of an individual
effective than changing provider
practitioner is however also likely
consultation style in improving
to be important regardless of
patient self-care and diabetes
length of professional-patient
outcomes.
relationship. It was considered tha
Edwards (2004)
3. One RCT reported significant
(Edwards, A., Elwyn,
effects of the research clinic group
G., Hood, K. et al ,
(which provided more consultation
2004)
time) in confidence in decision and
expectation to adhere to chosen
the level of trust between
practitioner and patient may be a
key factor in this relationship and
the GDG wished to review whethe
this could be specifically increased
in practitioner-patient encounters.
Medicines concordance: full guideline DRAFT (July 2008) page 145 of 373
DRAFT FOR CONSULTATION
treatments.
The evidence from a recent
systematic review suggested that
there is insufficient evidence to
Cohen (2004)
4. Two studies from the same
recommend any specific
(Cohen, D., Longo,
randomised controlled trial found
intervention. The GDG felt that
M. F., Hood, K. et al , that training GPs in SDM or
consultation style should be
2004); Edwards
combined with risk communication
tailored to individual patients to
(2004) (Edwards, A.,
yielded conflicting results in the
allow full communication.
Elwyn, G., Hood, K.
probability of a prescription being
et al , 2004)
issued to patients.
Cohen (2004)
5. Two studies from the same
(Cohen, D., Longo,
randomised controlled trial found
M. F., Hood, K. et al , that training GPs in SDM or
2004); Edwards
combined with risk communication
(2004) (Edwards, A.,
yielded no effect on the probability
Elwyn, G., Hood, K.
of investigations, referrals or follow-
et al , 2004)
up GP visits for any of the
conditions.
Savage (1990)
6. One RCT found that a directing
(Savage, R. and
style of consultation yielded
Armstrong, D., 1990)
significantly higher levels of
satisfaction on almost all the
outcome measures compared to a
sharing style. This was particularly
relevant for patients with physical
problems.
Shields (2005)
7. One RCT reported a statistically
(Shields, C. G.,
significant likelihood of a physician
Epstein, R. M.,
to promote collaboration in
Medicines concordance: full guideline DRAFT (July 2008) page 146 of 373
DRAFT FOR CONSULTATION
Fiscella, K. et al ,
treatment decision-making with
2005)
patients rather than with
companions of the patients.
Shields (2005)
8. One RCT reported a statistically
(Shields, C. G.,
significant responsiveness of a
Epstein, R. M.,
physician to explore the disease
Fiscella, K. et al ,
and illness when the issues were
2005)
raised by the patient compared with
the companion of the patient.
1
2
4.15.1 Methods of the evidence review
3
This paper includes a narrative summary of the included evidence, structured
4
according to the category of the intervention, following the agreed reviewing
5
protocol:
6
Types of studies: Systematic Reviews or Randomised controlled trials
7
(RCTs) that focus on aspects of consultation style that may increase patient
8
involvement in decision-making.
9
Types of participants: people prescribed medication for a medical condition
10
faced with a decision.
11
Duration of studies: No time limit specified.
12
Types of interventions: any interventions which assess which aspects of the
13
consultation style may increase patient involvement in decision-making.
14
Types of outcome measures: Patient-centred communication in the
15
consultation; Consultation process outcomes: patient involvement, question
16
asking, preparedness; Patient care outcomes: satisfaction, knowledge, self-
17
efficacy, type of information.
18
Medicines concordance: full guideline DRAFT (July 2008) page 147 of 373
DRAFT FOR CONSULTATION
1
It should be noted that the remit is for conditions with prescribed medication
2
and this excludes conditions which require chemotherapy or screening. All
3
RCTs are within this remit, however many of the systematic reviews included
4
populations outside the remit, this is noted where applicable.
5
6
4.15.2 Evidence review
7
Our searches retrieved two systematic review and 4 RCTs that were
8
considered relevant to the key clinical question.
9
All the studies looked at the patient-provider interaction, either by exploring
10
the impact of different provider styles or by focusing on patient behaviour
11
changes.
12
13
McKinstry (2006) (McKinstry, B., Ashcroft, R. E., Car, J. et al , 2006)
14
conducted a Cochrane review of interventions to improve the trust of patients
15
in their doctors. They searched 10 databases including Cochrane Central
16
Register of CTs, Medline and Embase. Inclusion criteria for studies was
17
RCTs, CCTS, controlled before and after studies and interrupted time series
18
studies. The interventions were any that influenced patients trust in their
19
doctors, or where trust was an outcome of an intervention. The participants
20
were Doctors; adults or children using healthcare or those related to them.
21
Outcome measures were increase or decrease in patients’ trust and the
22
components of trust; other healthcare behaviours; Health status and well-
23
being; Use of resources; Satisfaction with care; Perception of doctors’
24
communication skills; perception of doctors’ humanistic attributes; perception
25
regarding patients’ trust; perceptions of doctors trustworthiness. Studies were
26
excluded if they did not measure change in trust or not the right type of study.
27
Two authors independently assessed whether the titles and abstracts were
28
relevant and four authors assessed the retrieved articles for inclusion. Two
29
authors assessed the quality of each study according to EPOC criteria. A
30
multidisciplinary advisory group was set up to assess whether there was
31
anything that had been left out of the review. 2099 titles and abstracts were
32
found, five met all the criteria, but two of these referred to the same study and
33
one had insufficient data points before and after the intervention, therefore
Medicines concordance: full guideline DRAFT (July 2008) page 148 of 373
DRAFT FOR CONSULTATION
1
leaving three studies. Two of the studies had a primary aim of the impact of
2
the intervention on patient trust. Thom (1999) coached doctors in behaviours
3
known to be associated with trust and Hall (2002) looked at the impact of
4
disclosing financial incentives physicians receive for compliance with
5
managed health care protocols on the trust patients had in physicians. In the
6
third study (Thompson 2001) trust was a secondary outcome and compared
7
the impact of three different types of induction visit for new patients of an
8
HMO to those who received no intervention, the trust was in any health care
9
professional. The review detailed the quality of studies including allocation
10
concealment, blinding and protection from contamination. On assessment of
11
study quality all three being RCTs provided baseline measures and within and
12
between group differences for measures. Thom (1999) used computer
13
allocation to groups but it was not clear if the researcher was blind to this
14
allocation. The interviewer was blind to the status of the physician but it was
15
unclear if the patients were blinded. Hall (2002) conducted a stratified random
16
sample study and used a computer for allocation with no input from
17
researchers. The interviewers were blind to the patients’ status but the
18
patients were aware of their own status. Thompson (2001) did not report how
19
the randomisation was applied. Patients were aware of their status but it is not
20
clear if interviewers were blind to the status of the interviewees. The study by
21
Thom (1999) on showed no effect on trust (74.4 for the intervention and 76.2
22
for the control group, non-significant). Satisfaction or humaneness scores
23
were non-significant also. Hall found a small increase in trust for both groups
24
from baseline when adjusted this was a 1.4% increase in physician trust
25
(p<0.05). Thompson (2001) found the trust in the health plan health
26
professionals rose significantly following the enrolment visit with health
27
personnel compared to control group p<0.001). The author concluded that
28
there is insufficient evidence to conclude that any intervention may increase or
29
decrease trust in physicians.
30
31
Van Dam (2003) (van-Dam, Henk A., 2003) developed a systematic review of
32
RCTs that looked at the effects of interventions on provider-patient interaction
33
on patient diabetes health behaviour, patient self-care, delivered diabetes
Medicines concordance: full guideline DRAFT (July 2008) page 149 of 373
DRAFT FOR CONSULTATION
1
care and health outcomes, and to disentangle those that are the most
2
effective. Eight studies were included after a rigorous methodological quality
3
assessment, and these showed different interventions on different levels of
4
the provider-patient interaction in diabetes care. Four studies focused on
5
provided consulting behaviour modifications (studies 1-4), and four studies
6
focused directly on patient behaviour change (studies 5-8).
7
All studies were conducted in practical diabetes care, three in hospital
8
outpatient clinics and five in general practices.
9
The main findings suggest that the most effective interventions are those with
10
a direct approach to support patient participation (i.e. by assistant-guided
11
patient preparation for visits to doctors, empowering group education, group
12
consultations, or automated telephone management) in diabetes care and
13
self-care behaviour, while interventions which focus on change of provider
14
behaviour were less effective. Thus, the authors advocate a shift from the
15
traditional medical model to a more patient centred, patient participation and
16
empowerment paradigm of delivery of diabetes care.
17
The authors pointed out that the review did present some limitations,
18
illustrated by the small number of reviewed paper; the differences between the
19
studies; and the focus on RCTs.
20
21
Cohen (2004) (Cohen, D., Longo, M. F., Hood, K. et al , 2004) and Edwards
22
(2004) (Edwards, A., Elwyn, G., Hood, K. et al , 2004) conducted a cluster
23
randomized crossover trial with the aim to explore the costs of training GPs in
24
developing SDM competences and in the use of risk communication (RC) aids
25
and to evaluate the effects of such training on a range of service resource
26
variables. Edwards (2004) published the main trial results that focused on the
27
doctor patient interaction, patient outcomes and satisfaction with the decision.
28
Within each cluster, patients were also allocated randomly to consult with the
29
doctor at one of three points in the study. The study comprised three phases.
30
Phase 1 was pre-training. Phase 2 included training for half of the GPs and
31
the other half in RC. In phase 3, each GP received training in the other
32
element making them fully trained in both. The authors argued that in this
33
way, the design offered the greatest potential to gain understanding about the
34
effects of each form of training alone and in combination and if the sequence
Medicines concordance: full guideline DRAFT (July 2008) page 150 of 373
DRAFT FOR CONSULTATION
1
of skill acquisition was important. A further randomisation allocated patients to
2
attend either in usual surgery time or in a research clinic- audiotaped,
3
including fewer interruptions and more time for each consultation (up to 15
4
minutes each).
5
SDM training involved GPs attending two workshops where standardized and
6
previously piloted skill development process was used. SDM competences
7
were described and demonstrated by means of consultation simulation and
8
pre-prepared scenarios involving the four study conditions. RC also involved
9
attendance at 2 workshops, and the aids consisted of tabulated data and
10
visuals displays of risk estimates for the four study conditions. Patients with
11
one of four conditions (menorrhagia, atrial fibrillation, menopausal symptoms
12
or prostatism) were invited by their GP to attend a “review consultation” to
13
discuss their continuing treatment. Twenty GPs from 20 different practices in
14
South Wales were recruited. Costs of training for both RC and SDM included
15
time of trainers, of those being trained and of the simulated patients used as
16
part of the training exercise. Information on prescriptions, investigations and
17
referrals was obtained from questionnaires completed by each clinician at the
18
review consultation.
19
Main results published by Cohen (2004) indicated that Training in SDM or
20
combined with RC significantly affected the probability of a prescription being
21
issue to women with menopausal symptoms and menorrhagia (despite RC
22
alone not having any effect). However, there was no significant change in
23
prescribing for patients with prostatism or atrial fibrillation. There was also no
24
effect on the probability of investigations, referrals or follow-up GP visits for
25
any of the conditions. Training cost was £1218 per GP, resulting in an
26
increase of cost of consultation by £2.89.
27
Edwards (2004) reported significant effects of the research clinic (which
28
provided more time) in confidence in decision (p<0.01) and expectation to
29
adhere to chosen treatments (p<0.05). Anxiety scores approached statistical
30
significance for the RC intervention, as did expectation to adhere to chosen
31
treatment for both interventions. No statistically significant effects of the risk
32
communication or SDM interventions were seen on the whole range of
33
patient-based outcomes.
Medicines concordance: full guideline DRAFT (July 2008) page 151 of 373
DRAFT FOR CONSULTATION
1
Cohen (2004) concluded that due to the explanatory nature of the study, no
2
assessment could be made on how training could affect the length of a
3
consultation.
4
5
Savage (1990) (Savage, R. and Armstrong, D., 1990) sought to compare the
6
effects of a directing and sharing style of consultation by a GP on patient’s
7
satisfaction with the consultation in a deprived inner city area. Patients were
8
aged 16 to 75 years of age and were randomised to receive a directing or
9
sharing style in the part of the consultation regarding treatment, advice and
10
prognosis. Three hundred and fifty nine patients were randomised, however
11
120 patients failed to complete the assessment that took place a week later.
12
There were no significant differences in the mean length of consultations
13
between the two experimental groups. Patients who had the directing style of
14
consultation reported significantly higher levels of satisfaction on almost all the
15
outcome measures, and was particularly strong for patients with physical
16
problems (excellent explanation p<0.02; excellent understanding p=0.04).
17
There was no significant difference in the responses to the directing and
18
sharing styles in longer consultations, where the main treatment was advice
19
and among patients with psychological or chronic problems. Statistical
20
significance values were not reported. This study was conducted in England.
21
22
Shields (2005) (Shields, C. G., Epstein, R. M., Fiscella, K. et al , 2005)
23
evaluated the influence of accompanied visits on physician-patient
24
communication- particularly on patients centred communication. Thirty
25
patients included in the study were aged above 65 years, were not cognitively
26
impaired and had a companion who could accompany them to their next visit.
27
Companions were not assigned a specific role during the session and
28
physicians were not asked to conduct the sessions in any particular way.
29
There were no statistically significant differences between accompanied and
30
unaccompanied visits on the number of issues that patients raised, however
31
patients did raise more issues in unaccompanied visits. No statistically
32
significant differences were observed for levels of patient-centeredness, or
33
satisfaction, even if patients who were accompanied reported being slightly
34
more satisfied.
Medicines concordance: full guideline DRAFT (July 2008) page 152 of 373
DRAFT FOR CONSULTATION
1
Physicians were more likely to promote collaboration in treatment decision
2
making with patients than with companions (p<0.0001). Physicians were also
3
more responsive to issues regarding exploring the disease and illness when
4
the issues were raised by the patient compared with the companion (p<0.03).
5
Medicines concordance: full guideline DRAFT (July 2008) page 153 of 373
DRAFT FOR CONSULTATION
1
4.16
Key Clinical Question: Do interventions to increase
2
patient involvement increase length of the
3
consultation?
4
Related References
Evidence Statements (summary
Evidence into
of evidence)
recommendations
including
Evidence from the
UK
McCann (1996)
1. Two RCTs from the UK found
The GDG were
(McCann-Simon,
that interventions to increase
concerned that
Weinman John,
participation in the consultation
interventions to increase
1996); Middleton
(leaflet and agenda form)
patient involvement in the
(2006) (Middleton, J.
significantly increased consultation
consultation might result
F., McKinley, R. K.,
length.
in longer consultations
and Gillies, C. L.,
and have impact on
2006)
service delivery more
generally. While the
Little (2004) (Little,
2. Two RCTs from the UK found
P., Dorward, M.,
interventions to increase
Warner, G. et al ,
participation in the consultation (a
2004)
prompt to elicit patient concerns
and a leaflet) did not significantly
McLean (2004)
increase consultation length.
evidence indicated that
simple interventions
might result in increase in
consultation length, this
did not always occur
(McLean, Malcolm
which was of some
and Armstrong,
surprise to the GDG. The
David, 2004)
GDG considered it
important to reassure
clinicians that increasing
Middleton (2006)
3. In one RCT that reported a
(Middleton, J. F.,
significant increase in consultation
patient involvement may
not affect consultation
Medicines concordance: full guideline DRAFT (July 2008) page 154 of 373
DRAFT FOR CONSULTATION
McKinley, R. K., and
length, the increase was more
length but that even if it
Gillies, C. L., 2006)
pronounced when using an agenda
did any extra time
form than when using the agenda
invested in the treatment
form in combination with an
process may have
educational intervention. When
benefits later on.
using the educational intervention
alone no difference was found.
Little (2004) (Little,
4. One RCT reported that the use
P., Dorward, M.,
of a general leaflet in the
Warner, G. et al ,
consultation was significantly more
2004)
effective in increasing satisfaction
when consultations were shorter.
Edwards (2004)
5. Adherence was increased in
(Edwards, A., Elwyn,
clinics where more time was
G., Hood, K. et al ,
available than usual surgery times.
2004)
Evidence including
the rest of the
world
Kinnersley (2007)
6. Evidence from two systematic
(Kinnersley, P.,
reviews and three RCTs suggests
Edwards, A., Hood,
that interventions designed to
K. et al , 2007);
improve patient participation in
Harrington (2004)
consultations does not increase
(Harrington-Jane,
overall length of consultations. One
Noble Lorraine,
review (Kinnersley 2007) found 14
2004); Cegala (2001) RCTs with no significant increase in
Medicines concordance: full guideline DRAFT (July 2008) page 155 of 373
DRAFT FOR CONSULTATION
in Wetzels review
consultation length and 3 RCTs
(2007) (Wetzels, R.,
with a significant increase
Harmsen, M., van,
(Hornberger, 1997; McCann 1996
Weel C. et al , 2007); and Middleton 2006 – 2 UK studies
Loh (2007) (Loh,
reported earlier).
Andreas, Simon,
Daniela, Wills, Celia
E. et al , 2007);
Hamann (2006)
(Hamann, J., Langer,
B., Winkler, V. et al ,
2006)
Hornberger (1997) in
7. In one of the RCTs that reported
Kinnersley (2007)
a significant increase in
(Kinnersley, P.,
consultation length, this increase
Edwards, A., Hood,
was due to time spent discussing
K. et al , 2007)
diagnoses and physical
examination
Kinnersley (2007)
8. Written materials had a small
(Kinnersley, P.,
and statistically significant increase
Edwards, A., Hood,
in consultation length compared to
K. et al , 2007)
coaching where the small increase
was not significant.*
Kinnersley (2007)
9. Interventions carried out
(Kinnersley, P.,
immediately before the intervention
Edwards, A., Hood,
had a small and significant increase
K. et al , 2007)
in consultation length compared to
those beforehand.**
Medicines concordance: full guideline DRAFT (July 2008) page 156 of 373
DRAFT FOR CONSULTATION
Kinnersley (2007)
10. There was no difference in
(Kinnersley, P.,
consultation length between RCTs
Edwards, A., Hood,
with or without clinician training.***
K. et al , 2007)
Hornberger in
11. One RCT (from Kinnersley
Kinnersley (2007)
review) reported that overall quality
(Kinnersley, P.,
of care showed a positive
Edwards, A., Hood,
significance in the intervention
K. et al , 2007)
group which had a longer
consultation time than the control
group.
1
2
3
*This result is from a comparison of written materials and coaching for the consultation length
of all studies which included written materials or coaching (thirteen), three of which were not
relevant to the population of interest in this evidence review.
4
5
6
**This result is from a comparison of studies some time before consultation (2) and
immediately before consultation (11), of which three of the immediately before consultation
were not the relevant population.
7
8
9
***This result included studies of Clinician training (2) compared to 12 studies where
Clinicians were not trained. One of the Clinician training studies and three of the Clinicians not
trained studies were not the relevant population
Medicines concordance: full guideline DRAFT (July 2008) page 157 of 373
DRAFT FOR CONSULTATION
1
4.16.1 Methods of the evidence review
2
The aim of the literature review is to identify the most relevant, published
3
evidence to answer the key clinical questions generated by the GDG. Due to
4
time constraints, exhaustive systematic reviews (see the Methods of the
5
Cochrane Review) were not undertaken. However, the evidence reviews were
6
undertaken using systematic, transparent approaches following the Guidelines
7
Manual 2007 (www.nice.org.uk).
8
The titles and abstracts of records retrieved by the searches, suggested by
9
the GDG or submitted by stakeholders were scanned for relevance to the key
10
questions. Any potentially relevant publications were obtained in full text.
11
These were then reviewed to identify the most appropriate evidence to help
12
answer the key questions and to ensure that the recommendations are based
13
on the best available evidence. This process required four main tasks:
14
selection of relevant studies; assessment of study quality; synthesis of the
15
results; and grading of the evidence.
16
This paper includes a narrative summary of the included evidence, structured
17
according to the category of the intervention, following the agreed reviewing
18
protocol:
19
Types of studies: Systematic Reviews or Randomized controlled trials
20
(RCTs) that assess whether interventions to increase patient involvement
21
increase length in consultation.
22
Types of participants: people prescribed medication for a medical condition
23
faced with a decision.
24
Duration of studies: No time limit specified.
25
Types of interventions: any interventions which assess which aim to
26
increase patient involvement and include details of consultation length.
27
Types of outcome measures: Patient-centred communication in the
28
consultation; Consultation process outcomes: patient involvement, question
Medicines concordance: full guideline DRAFT (July 2008) page 158 of 373
DRAFT FOR CONSULTATION
1
asking, preparedness; Patient care outcomes: satisfaction, knowledge, self-
2
efficacy in relation to consultation length.
3
4
It should be noted that the remit is for conditions with prescribed medication
5
and this excludes conditions which require chemotherapy or screening. All
6
RCTs are within this remit, however many of the systematic reviews included
7
populations outside the remit, this is noted where applicable
8
9
10
NOTE: this evidence review is an updated version containing systematic
reviews retrieved from our searches.
11
4.16.2 Evidence review
12
This review was stratified to firstly present the RCTs research from the UK
13
and secondly present systematic reviews and RCTs which include research
14
from other areas of the world.
15
16
4.16.2.1
RCTS from the UK
17
Little (2004) (Little, P., Dorward, M., Warner, G. et al , 2004) conducted a
18
randomised control trial in the UK to assess the effect of leaflets in
19
empowering patients in primary care consultations. Participants were
20
randomised to four conditions: receipt of a general leaflet, depression leaflet,
21
both leaflets and no leaflets (control group). The general leaflet which asked
22
patients to list issues they wanted to raise and explained that the doctor
23
wanted them to ask questions, talk and discuss any problems of concern to
24
them. The depression leaflet listed symptoms of depression (without labelling
25
as such) and asking if had them and that the doctor would like to discuss
26
them. The only significant interaction was the increase in satisfaction for those
27
who received the general leaflet, the mean difference was 0.17 (95% CI 0.01
28
to 0.32, p=0.04). The general leaflet was significantly more effective when
29
consultations were shorter (leaflet 0.64, 95% CI 0.19 to 1.08; time 0.31, 0.0 to
30
0.06; interaction between both showed that consultations of 5, 8, and 10 mins
31
increased satisfaction by 14%, 10% and 7%). The leaflet overall caused a
32
small non-significant increase in consultation time.
Medicines concordance: full guideline DRAFT (July 2008) page 159 of 373
DRAFT FOR CONSULTATION
1
2
Middleton (2006) (Middleton, J. F., McKinley, R. K., and Gillies, C. L., 2006)
3
conducted a randomised controlled trial in the UK of agenda forms completed
4
by the patient and doctors’ education about the agenda on the outcome of the
5
consultation. The intervention group were asked to think of a list of their
6
concerns, arrive five minutes earlier and bring spectacles and an interpreter if
7
those were required. The intervention doctors were given a one day
8
educational workshop to allow the doctors to have awareness of the patient
9
agenda model of the consultation. The model involved identifying the agenda
10
(ideas, concerns, expectations and reasoning). The doctors reflected on their
11
own agenda and negotiation of action with patients. Half of the patients in this
12
group filled in an agenda form in the preceding time before their consultation,
13
half did not. The control group was the GPs not given the educational
14
programme, and this was split into half the patients using the agenda form
15
and half not using it. The consultation length for the control group was 7.1
16
minutes (95% CI 6.5 to 7.7 minutes). The agenda form significantly increased
17
the duration of consultation by 0.9 minutes (0.3 to 1.5, p=0.004) and the
18
combined intervention by 1.9 minutes (1.0 to 2.8, p<0.001). The educational
19
intervention on its own did not significantly change the length of consultation
20
(0.7 minutes, -0.2 to 1.6 minutes). There was a significant increase in both
21
interventions for number of problems identified. The only change in patient
22
satisfaction was increase in depth of doctor-patient relationship from the
23
agenda form group.
24
25
McCann’s (1996) (McCann-Simon, Weinman John, 1996) randomised
26
controlled trial in the UK was of a brief written intervention leaflet ‘Speak for
27
Yourself’ to increase participation in the consultation read before the
28
consultation compared to those given a control leaflet. The first part asked
29
patients to identify the nature of their problems and to consider their ideas to
30
causes, treatment and effects of the problems. They had space to write down
31
ideas. The second part of the leaflet advises to state their ideas and concerns
32
about the illness to the doctor and ask questions. The intervention group had
33
significantly longer consultations (8.43 minutes, SD=2.97 versus 7.22
Medicines concordance: full guideline DRAFT (July 2008) page 160 of 373
DRAFT FOR CONSULTATION
1
minutes, SD=2.42, 95% CI -0.44 (0.08, 0.81) and they asked more questions
2
than controls.
3
4
McLean (2004) (McLean, Malcolm and Armstrong, David, 2004) conducted
5
an open randomised controlled trial to see whether a prompt to elicit patients’
6
concerns for minor illness would be beneficial and the costs of doing so. 110
7
patients from four training semi-rural general practices in the South-East of
8
the UK took part in the study. The written prompts were ‘May I ask if you have
9
any concerns about this “….” (illness/pain) you have come about today?’
10
followed by: ‘Anything in particular about the “…”?’ and, if still unforthcoming:
11
‘What is it about the “…” that concerns you?’ A consultation satisfaction
12
questionnaire regarding the professional care component was given. The
13
doctor had to record the consultation length (estimated to the nearest minute
14
using a clock to note start and end of consultation) and the diagnosis made.
15
The control group received the consultation as normal. The doctor depending
16
on whether the top sheet of a randomly arranged pile of papers said ‘control’
17
or ‘intervention’. [However it must be noted that the same doctor was
18
conducting both control and intervention and the control condition may
19
inadvertently receive a more patient-oriented consultation]. Patient satisfaction
20
was 80.9 for controls and 88.2 for intervention patients (SD=11.8), mean
21
difference 7.2 (95% CI 2.0 to 12.6). The consultation length of intervention
22
consultations was on average 1 minute longer for intervention group than
23
controls (11.0 vs. 10.0 minutes), but this was not significant.
24
25
Cohen (2004) (Cohen, D., Longo, M. F., Hood, K. et al , 2004) and Edwards
26
(2004) (Edwards, A., Elwyn, G., Hood, K. et al , 2004) conducted a cluster
27
randomized crossover trial in the UK with the aim to explore the costs of
28
training GPs in developing SDM competences and in the use of risk
29
communication (RC) aids and to evaluate the effects of such training on a
30
range of service resource variables. Edwards (2004) published the main trial
31
results that focused on the doctor patient interaction, patient outcomes and
32
satisfaction with the decision.
33
Within each cluster, patients were also allocated randomly to consult with the
34
doctor at one of three points in the study. The study comprised three phases.
Medicines concordance: full guideline DRAFT (July 2008) page 161 of 373
DRAFT FOR CONSULTATION
1
Phase 1 was pre-training. Phase 2 included training for half of the GPs and
2
the other half in RC. In phase 3, each GP received training in the other
3
element making them fully trained in both. The authors argued that in this
4
way, the design offered the greatest potential to gain understanding about the
5
effects of each form of training alone and in combination and if the sequence
6
of skill acquisition was important. A further randomisation allocated patients to
7
attend either in usual surgery time or in a research clinic- audiotaped,
8
including fewer interruptions and more time for each consultation (up to 15
9
minutes each).
10
SDM training involved GPs attending two workshops where standardized and
11
previously piloted skill development process was used. SDM competences
12
were described and demonstrated by means of consultation simulation and
13
pre-prepared scenarios involving the four study conditions. RC also involved
14
attendance at 2 workshops, and the aids consisted of tabulated data and
15
visuals displays of risk estimates for the four study conditions. Patients with
16
one of four conditions (menorrhagia, atrial fibrillation, menopausal symptoms
17
or prostatism) were invited by their GP to attend a “review consultation” to
18
discuss their continuing treatment. Twenty GPs from 20 different practices in
19
South Wales were recruited. Costs of training for both RC and SDM included
20
time of trainers, of those being trained and of the simulated patients used as
21
part of the training exercise. Information on prescriptions, investigations and
22
referrals was obtained from questionnaires completed by each clinician at the
23
review consultation.
24
Edwards (2004) reported significant effects of the research clinic (which
25
provided more time) in confidence in decision (p<0.01) and expectation to
26
adhere to chosen treatments (p<0.05). Anxiety scores approached
27
significance for the RC intervention, as did expectation to adhere to chosen
28
treatment for both interventions. No significant effects of the risk
29
communication or SDM interventions were seen on the whole range of
30
patient-based outcomes.
31
However, Cohen (2004) concluded that due to the explanatory nature of the
32
study, no assessment could be made on how training could affect the length
33
of a consultation.
34
Medicines concordance: full guideline DRAFT (July 2008) page 162 of 373
DRAFT FOR CONSULTATION
1
4.16.2.2
RCTs conducted outside the UK
2
3
Kinnersley (2007) (Kinnersley, P., Edwards, A., Hood, K. et al , 2007)
4
conducted a Cochrane systematic review to find interventions which aimed to
5
increase patient involvement by enabling patients to address their information
6
needs within the consultation. Most of the RCTs were from the USA, 2 from
7
Australia, 5 from the UK and one from the Netherlands. Most of the
8
interventions were written followed by face-to-face coaching and videotape.
9
Written interventions were in booklet or checklist form. The specific
10
behaviours most encouraged were question-asking, raising concerns and
11
requesting clarification or checking understanding.
12
13
Seventeen RCTs in the Kinnersley review (2007) looked at consultation
14
length, 3 studies found a significant increase (Hornberger, 1997; McCann
15
1996 and Middleton 2006). While 14 RCTs found no effect. Bolman (2005)
16
found a decrease in the first consultation and an increase in the last
17
consultation. The meta-analysis showed a small but not significant increase in
18
consultation length (SMD 0.10 95% CI -0.05 to 0.25).
19
Fifteen RCTs reported that the use of written materials during the consultation
20
led to a small and significant increase in consultation length (SMD 0.13, 95%
21
CI 0.05 to 0.21). RCTs with coaching found a non-significant increase (SMD
22
0.07 95% CI -0.07 to 0.20). In studies where there was additional clinician
23
training there was little impact on consultation length for written and coaching
24
materials. RCTs with clinician training SMD 0.17 (95% CI 0.01 to 0.32)
25
compared to studies without clinician training SMD 0.17 (95% CI 0.10 to 0.24).
26
It should be noted that of these seventeen RCTs only eleven of these related
27
to our population of interest, the results for these are detailed below.
28
29
30
4.16.2.3
RCTs included in the Kinnersley (2007) (Kinnersley, P., Edwards,
A., Hood, K. et al , 2007) review
31
32
Hornberger (1997) conducted a two-armed, randomised trial of whether a
33
self-administered previsit questionnaire enhanced awareness of patients
Medicines concordance: full guideline DRAFT (July 2008) page 163 of 373
DRAFT FOR CONSULTATION
1
concerns in the USA. They completed the Patient Concerns Form while
2
waiting for their visit. This covered 25 items of concerns of five categories:
3
desire for medical information, psychosocial assistance, therapeutic listening,
4
general health advice and biomedical treatment. After the interview the
5
patients completed a postvisit questionnaire which assessed their perceptions
6
of the concerns addressed by the physician. The net effect of the intervention
7
compared to the control group was a difference of 6.8 minutes (95%CI 0.4,
8
13.3) for total time in consultation. With most of the extra time spent
9
discussing diagnoses (3.35 minutes, 95%CI 0.00, 6.72) and in performing the
10
physical examination (2.7minutes, 95% CI 0.5, 4.9). The number of diagnoses
11
increased by 30% in the intervention group compared to the control group
12
(increase of 1.7 diagnoses per visit). Those in the intervention group had
13
marginally higher satisfaction but this was not significant except for overall
14
quality of care (0.35, +/- 0.23, p,0.05).
15
16
Greenfield (1985) conducted a randomised controlled trial of an intervention
17
to increase patient involvement in their care in the US. The intervention group
18
received a treatment algorithm as a guide to help them read their medical
19
record and a behaviour-change strategy. The participants were coached in
20
appropriate question-asking and negotiation of decisions. The intervention
21
occurred in a 20 minute session before their regular consultation with their
22
GP. The control group also saw a clinic assistant just before their regular
23
appointment for a similar amount of time as the intervention group and they
24
received a standard protocol of receiving information and review of ulcer
25
disease and given copies of these materials, they did not get to see their
26
medical records. There was no significant difference between groups in length
27
of consultation after the interventions, both groups averaged 16 minutes per
28
encounter. The time of the encounter before was 16.8, SD 8.2 whereas the
29
control group was 15.1 (7.6), a difference of 1.7 (95% CI -2.92, 6.32). The
30
time of the encounter after was 15.7 (6.7) for the intervention and 16.3 (9.7)
31
for the control, -0.6 (95% CI -5.49, 4.29). However, they differed in how they
32
spent their time with the intervention patients spent more time involved in the
33
interaction than controls.
34
Medicines concordance: full guideline DRAFT (July 2008) page 164 of 373
DRAFT FOR CONSULTATION
1
Greenfield (1988) conducted a randomised controlled trial in a diabetic clinic
2
in the USA. This intervention was the same as in Greenfield (1985) but
3
delivered twice, before the initial and follow up consultations. There was no
4
change in question asking, patient satisfaction, knowledge or consultation
5
length (30.30 SD=13.80 intervention group versus 32.50 SD 13.90 for the
6
control group). Participation and the preference for active involvement
7
increased.
8
9
Maly (1999) conducted a randomised controlled trial in a family medicine
10
clinic in the US. where patients received copies of their medical record
11
progress notes and produced two main questions to ask their physician.
12
Control group received health education sheets and made suggestion lists for
13
their clinic care. The consultation length did not differ between groups.
14
15
Roter’s (1977) randomised control trial in a family medicine clinic in the US
16
involved 10 minutes with a health educator to identify questions from a
17
question asking protocol. The participants were encouraged to ask questions
18
and took a list in to the consultation. Question asking and patient satisfaction
19
increased and there was no difference between consultation length (29.90,
20
SD=12.70) vs 40.50 (92.70).
21
22
Thompson (1990a) conducted a randomised controlled trial in an obstetric
23
and gynaecologist clinic in the USA. Participants received a question prompt
24
sheet with instructions to write at least 3 questions to take to the consultation.
25
Question asking increased and there was no change in patient satisfaction
26
and consultation length 7.70 (SD=2.90) vs 8.70 (SD=4.70), 95% CI -0.26 (-
27
0.80, 0.29).
28
29
Martinali’s (2001) randomised controlled trial in the Netherlands used a
30
checklist to prepare coronary patients for visiting their cardiologist. The short
31
checklist which was to be completed at home was aimed towards structuring
32
the exchange of information in the consultation and to concentrate on those
33
issues that caused most concern to the patient. A brochure was also
34
developed with instructions for the checklist. A brochure was also given from
Medicines concordance: full guideline DRAFT (July 2008) page 165 of 373
DRAFT FOR CONSULTATION
1
the Dutch Heart Foundation, which both groups received. The consultation
2
length was 12 minutes (SD 4.2) in the experimental group and 10.3 (SD 3.8)
3
in the control group, f=1.82, p=0.18.
4
5
Bolman’s (2005) randomised controlled trial in cardiology clinics in the
6
Netherlands involved a checklist of 49 questions on 10 issues (as Martinali
7
2001). This was mailed to the patient a week before each of three linked
8
consultations. There was no change in patient satisfaction. Consultation
9
length was reduced at first visit but increased at third visit (13.73, SD=3.73 vs
10
16.22, SD=5.84, 95% CI -0.49, -0.88, -0.10)
11
12
4.16.2.4
Other systematic reviews of RCTs
13
14
Harrington’s (2004) (Harrington-Jane, Noble Lorraine, 2004) systematic
15
review, which investigated how to improve communication in a consultation
16
showed that studies overall found that by involving patients there was not a
17
resultant increase in consultation length. Five out of seven studies that
18
included consultation length (and were our population of interest) found there
19
was not a significant increase in the length of consultation except for
20
Hornberger (1997) and McCann (1996). All of these studies are in the
21
Kinnersley (2007) review except for McGee (1998), (a study conducted in the
22
USA) which did not find any difference in consultation length.
23
24
The Wetzels (2007) (Wetzels, R., Harmsen, M., van, Weel C. et al , 2007)
25
systematic review, which looked at interventions to improve older patients’
26
involvement, reported findings related to consultation length. Only one of the
27
three studies meeting the inclusion criteria of the review included consultation
28
length (Cegala 2001). In Cegala (2001) the trained patients asked more
29
medically-related questions, gained more information and provided more
30
information than control patients. They did not verify information more than
31
control patients and appointment length was not longer overall. (18.81 vs
32
22.59, p=0.46) and time engaged in talk 16.25 vs 14.41, p=0.68). This study
33
was conducted in the USA.
Medicines concordance: full guideline DRAFT (July 2008) page 166 of 373
DRAFT FOR CONSULTATION
1
4.16.2.5
RCTs (not included in any systematic reviews)
2
Loh (2007) (Loh, Andreas, Simon, Daniela, Wills, Celia E. et al , 2007)
3
investigated the effects of a shared decision-making intervention in primary
4
care of depression compared to usual care on adherence, satisfaction and
5
clinical outcomes. The study was conducted in Sudbagen, Germany with
6
primary care physicians as the unit of randomisation. The sampling frame
7
(n=148) were sent a letter, 30 accepted the invitation to take part, 20 were
8
randomly assigned to the intervention group and 10 to the control group, after
9
drop out 15 and 8 were left respectively. The physicians had to recruit newly
10
diagnosed depressive patients. The intervention physicians enrolled 263
11
patients and the control group 142. After their diagnosis the patients
12
completed a questionnaire measuring patient involvement, depression
13
severity and socio-demographic characteristics. After 6-8 weeks the patients
14
completed a second questionnaire measuring adherence and treatment
15
outcome. At the same time, the physicians rated their assessment of the
16
patients’ adherence. The shared decision-making intervention was then
17
implemented with the intervention group. The intervention was a multi-faceted
18
program including physician training, a decision board for use during the
19
consultation given to the patients after the consultation, printed patient
20
information with specific encouragement to be active in the decision-making
21
process. The physicians in the intervention group also completed modules on
22
guideline-concordant depression care which included content on enhancing
23
skills for improving the shared decision-making process. The outcomes
24
measures were patient participation, treatment adherence, patient
25
satisfaction, consultation time and clinical outcomes. There was no difference
26
for the control group in patient participation before and after, whereas the
27
intervention group had significantly higher patient participation from pre to
28
post intervention for the doctor facilitation scale (p=0.001) and there was an
29
increase in the patient participation scale (p=0.010). There were no significant
30
differences in treatment adherence. Patient satisfaction was significantly
31
higher in the intervention 29.8 (2.7) than the control group 27.0 (3.6), p=0.14.
32
There were no values taken for satisfaction before the intervention. There was
33
no difference between groups for length of consultation 29.2 (10.7) versus
Medicines concordance: full guideline DRAFT (July 2008) page 167 of 373
DRAFT FOR CONSULTATION
1
26.7 (12.5). Neither group had a significant reduction in depression severity
2
from baseline to post-intervention.
3
4
Hamann (2006) (Hamann, J., Langer, B., Winkler, V. et al , 2006) conducted
5
a randomised controlled trial which aimed to assess an intervention for shared
6
decision-making in patients with acute schizophrenia. 107 patients from 12
7
acute psychiatric wards of two hospitals in Germany were included in the
8
study. Before the consultation participants were given a talk on their treatment
9
options and to prepare them for their GP consultation. A 16 page booklet
10
decision aid covering the pros and cons of oral vs depot formulation, first vs
11
second generation antipsychotics, psycho-education, and type of socio-
12
therapeutic intervention. Trained nurses assisted the patients to work through
13
the booklet and gave answers to any information requests. They had to write
14
down their experiences with antipsychotic medication and to indicate their
15
preferences. They met with physicians within 24 hours of working through the
16
decision aid. The control group received routine care. There was no difference
17
reported in the time spent in individual consultations as reported by the
18
psychiatrists - mean 64 min/weeks for the intervention group compared to 60
19
min/weeks for the control group, p>0.05.
20
21
22
23
Medicines concordance: full guideline DRAFT (July 2008) page 168 of 373
DRAFT FOR CONSULTATION
1
5
Patients’ experience of medicine-taking
2
5.1
Recommendations *listed at the start of the guideline
according to chapter
3
4
5.2
Introduction
5
If health care professionals are to facilitate patient involvement in decisions
6
about medicines it is helpful and necessary to understand how patients
7
approach the taking of medicines, in particular the ongoing appraisal of
8
medicines that continues after a consultation. Investigation into why patients
9
do not take medicines as prescribed indicates that the decision to take
10
medicines and the continuing taking of medicines should be considered as a
11
complex behaviour (Horne, R, Weinman, J., Barber, N. et al , 2005).
12
13
5.3
Key Clinical Question: what are the barriers and
facilitators for individuals in medicine-taking
14
Related
Evidence statements
Evidence into recommendations
references
(summary of evidence)
Pound
1. Patients wish to minimise
The GDG discussed the appropriate
(2005)(Pound, P.,
medication intake where
research methodology to provide
Britten, N.,
possible. They may wish to do
evidence of patients actual medicine
Morgan, M. et al ,
this to decrease adverse
taking behaviours. The GDG
2005); Munro
effects and potential for
considered it important to provide
(2007) (Munro,
addiction, to make the regimen
health care professionals with
Salla A., Lewin,
more acceptable or for financial
evidence of how patients actually use
Simon A., Smith,
reasons. Patients may decide
medicines to sensitize professionals to
Helen J. et al ,
to use prescribed medicine to
issues that may be relevant for
2007); Mills (2006)
alleviate symptoms or
discussion with individual patients.
(Mills, Edward J.,
strategically, to replace or
The GDG accepted the use of
Nachega, Jean B.,
supplement medicines
qualitative evidence for this and
Bangsberg, David
sometimes or all the time with
considered the Pound synthesis
Medicines concordance: full guideline DRAFT (July 2008) page 169 of 373
DRAFT FOR CONSULTATION
R. et al , 2006);
non-pharmacological
provided the type of evidence they
Carrick (2004)
treatments
were looking for. To augment the
(Carrick, Rachael,
Mitchell, Annie,
Powell, Richard A.
et al , 2004);
Deegan (2005)
(Deegan, P. E.,
2005); Lewis
(2006) (Lewis, M.
P., Colbert, A.,
Erlen, J. et al ,
2006); Cooper
(2002) (Cooper,
V., Buick, D.,
Pound synthesis searches were made
Patients will commonly
evaluate prescribed medicines
by trying out the medicines and
weighing up the costs and
benefits. They will consider
adverse effects and
acceptability of regimen. They
may stop the medicine and see
what happens and obtain
information from non-medical
for evidence not included in the Pound
review and for any other systematic
reviews. The description of patient
behaviours and factors influencing
patients medicine taking behaviours
were used to inform the
recommendations. The GDG were
given access to research in progress
supporting the influence of patient
beliefs on actual adherence.
sources and observe the effect
of medicines on others.
Horne, R. et al ,
Both subjective and objective
2002); Ogedegbe
indicators may be used to
(2004) (Ogedegbe, evaluate medicines.
Gbenga, Harrison,
Melanie, Robbins, Patients do not generally
Laura et al , 2004);
disclose their beliefs and
Lukoscheck (2003) change of regimen to HCPs.
(Lukoschek,
Petra., 2003);
Dowell (2002)
(Dowell, Jon,
Jones, Anni, and
Snadden, David,
2002); Wilson
(2002) (Wilson,
Patients may not be able to
recognize the difference
between effects of medication
and effects of disease and
have difficulty in evaluating
long term preventative
medication where there are no
symptoms.
Holly Skodol,
Hutchinson, Sally
Patient on multiple medications
A., and Holzemer,
may make choices between
Medicines concordance: full guideline DRAFT (July 2008) page 170 of 373
DRAFT FOR CONSULTATION
William L., 2002);
medicines.
Bollini (2002)
7589}; Adam
(2002) (Adam, B.
D., Maticka,
Tyndale E., and
Cohen, J. J.,
2003); Scotto
(2005) (Scotto,
Carrie J., 2005);
Alfonso (2006)
(Alfonso, V.,
Bermbach, N.,
Geller, J. et al ,
2006); Sidat
(2007) (Sidat,
Mohsin, Fairley,
Christopher, and
Grierson, Jeffrey,
2007); Lewis
(2006) (Lewis, M.
P., Colbert, A.,
Erlen, J. et al ,
2006); Pyne
92007) (Pyne, J.
M., McSweeney,
J., Kane, H. S. et
al , 2006); Kikkert
(2006) (Kikkert,
Martijn J., Schene,
Aart H., KoeterMaarten, W. J. et
al , 2006); Vinter-
Medicines concordance: full guideline DRAFT (July 2008) page 171 of 373
DRAFT FOR CONSULTATION
Repalust (2004)
(Vinter, Repalust
N., Petricek, G.,
and Katic, M.,
2004); Campero
(2007) (Campero,
Lourdes, Herrera,
Cristina, Kendall,
Tamil et al , 2007);
Reid (2006) (Reid,
M., Clark, A.,
Murdoch, D. L. et
al , 2006); Elliott
(2007) (Elliott,
Rachel A., Ross,
Degnan Dennis,
Adams, Alyce S. et
al , 2007); Aronson
(2005) (Aronson,
B., 2005); PopaLisseanu (2004)
(Garcia-PopaLisseanu, M. G.,
Greisinger, A.,
Richardson, M. et
al , 2005); Erwin
(1999) (Erwin, J.
and Peters, B.,
1999); Mutchler
(2007) (Mutchler,
Jan E.,
Bacigalupe,
Gonzalo, Coppin,
Medicines concordance: full guideline DRAFT (July 2008) page 172 of 373
DRAFT FOR CONSULTATION
Antonia et al ,
2007); Lawton
(2005) (Lawton, J.,
Ahmad, N.,
Hallowell, N. et al ,
2005); Morgan
(2005) (MorganMyfanwy,
Figueroa-MuñozJose, 2005)
1
Methods of the evidence review
2
5.3.1
3
Searches of the literature revealed a large number of studies that set out to
4
explore patients’ experience of medicine taking. The majority of these studies
5
are qualitative studies. One of the current challenges in qualitative research is
6
how to bring together the findings from individual qualitative studies. One
7
approach is to present a narrative of these studies describing their findings
8
individually. More recently the field of qualitative synthesis has attempted to
9
synthesise the findings from different studies into a common set of findings
10
that includes the findings from individual studies but that may also provide
11
additional levels of understanding that may not be apparent when each study
12
is looked at individually. A synthesis of medicine-taking has already been
13
conducted using a meta-ethnography (Pound, P., Britten, N., Morgan, M. et al,
14
2005). The Pound (2005) (Pound, P., Britten, N., Morgan, M. et al , 2005)
15
review is methodologically sound and systematic synthesis developed by a
16
panel of experts within the field of medicines concordance. Following
17
discussion with the GDG it was agree that rather than conduct an alternative
18
summary or synthesis of the qualitative studies of patients experience of
19
medicine-taking we used the synthesis as the basis for our evidence review.
20
The findings of the synthesis were updated by searching for evidence
21
published since the review was conducted and a narrative summary of these
22
presented. The GDG requested additional search for systematic reviews.
Medicines concordance: full guideline DRAFT (July 2008) page 173 of 373
DRAFT FOR CONSULTATION
Evidence review
1
5.3.2
2
Pound (2005) (Pound, P., Britten, N., Morgan, M. et al , 2005) aimed to
3
conduct a synthesis of qualitative studies of lay experience of medicine taking.
4
The study was part of a Health Technology Assessment project to evaluate
5
meta-ethnography as a method of synthesizing qualitative research studies of
6
health and health care. This narrative has used the paper published in Social
7
Science and Medicine (Pound, P., Britten, N., Morgan, M. et al , 2005). The
8
authors of the HTA report kindly allowed us to read a draft copy of their report
9
for the HTA which primarily concerned the detail of their methodology. Studies
10
using both qualitative methods of data collection and qualitative methods of
11
data analysis and published in English were included. Studies published
12
between January 1992 and December 2002 were eligible for inclusion.
13
Medline, Embase, Cinahl, Web of Science, Psychinfo and Zetoc were
14
searched. The electronic search was supplemented by extensive
15
handsearching. Papers were appraised for quality using a version of CASP
16
1988 criteria. Thirty-eight papers were included in the synthesis. Papers were
17
organised into medicine groups. The initial synthesis brought together papers
18
looking at similar medicine/disease groups and these findings were then
19
synthesised. The medicine groups were antiretroviral therapy,
20
antihypertensives, psychotropic medicine, proton-pump inhibitors, asthma,
21
miscellaneous medicines and medicines in general.
22
The paper reports a summary of the findings of the individual studies and the
23
results of the synthesis of these studies. The authors developed the concept
24
of ‘resistance’ to medicines to describe lay peoples response to medicines.
25
One of the main conclusions of the synthesis is that people do not take
26
medicines as prescribed because of patient concerns about medicines
27
themselves. Their interpretation is that non-adherence is not necessarily a
28
result of failures from the doctors, patients or systems, but because of
29
concerns about the medicines themselves. Drawing on issues such as patient
30
reports of testing for adverse effects; worries about dependence; the potential
31
harm from taking medicines on a long term basis; and issues with disclosure
32
and stigma, the authors’ conclude that many patients ‘resist’ the taking of
33
medicines. Patients can be described as ‘acceptors‘ of medicines, some
Medicines concordance: full guideline DRAFT (July 2008) page 174 of 373
DRAFT FOR CONSULTATION
1
uncritically, others following their own enquires and experimentation. People’s
2
medicine taking may change with different medicines and the illness in
3
question, which illustrates the relative nature of the concept of “resistance”.
4
Those that may “resist” a certain type of medicines may “accept” a different
5
medicine.
6
The findings of the synthesis fell into three parts (1) the way people evaluate
7
their medicines and the difficulties they encounter in doing this (2) the
8
interaction between medicines and patient identity (3) the ways people take
9
their medicines. The themes that make up each part are listed under each
10
heading.
11
5.3.3
12
encountered by people in evaluating medicines
The evaluation of medicines and the difficulties
13
14
Trying out the medicines and weighing up the costs and benefits
15
The most common way of evaluating medicines was to try it out and weigh up
16
the benefits of taking it against the cost of doing so. The majority of the
17
studies focused more on the problems associated with medication and less on
18
the benefits of medication, but it was clear that people had hope that their
19
medicines would help with the symptoms; avoid relapse and hospitalization;
20
for minimization of disease progression or for prevention of future illness.
21
Adverse effects
22
Adverse effects were a key criterion in the evaluation of medicines. This was
23
found in several studies including those about treatments for cancer,
24
rheumatoid arthritis, asthma, diabetes, schizophrenia and digestive disorders.
25
Adverse effects were very prominent in studies of patients taking anti-HIV
26
medication. The frequency and nature of adverse effects experienced by
27
these patients, particularly as adverse effects were severe and unpredictable,
28
resulted in distrust and fear of the medicines. These adverse effects affected
29
social activities, relationships and work.
Medicines concordance: full guideline DRAFT (July 2008) page 175 of 373
DRAFT FOR CONSULTATION
1
Acceptability of regimen
2
People reported the evaluation of the suggested regime in terms of how it
3
fitted in with daily schedules and life in general. The frequency of doses and
4
number of pills was also a challenge, as also taste, smell, size and shape of
5
the pills. Regimes that required disruption of schedules resulted in patients
6
reporting that they were no longer in control of their lives and people varied as
7
to whether they fitted the world around their regime or resisted the demands
8
of the regime and missed doses.
9
Weighing and balancing
10
A process of weighing and balancing was carried out by people where the
11
advantages of treatment in terms of symptoms or effect on disease was
12
balanced by side effects and disruption. Adverse effects and disruption would
13
lead people to question if it was worth taking medication or not..
14
Stopping the medicine and seeing what happens
15
In some cases, patients test the medicines by either changing the doses or
16
stopping the medicine as to observe what happens. Some authors suggest
17
that this process can either be explicit or subconscious and tends to occur
18
more frequently in long term conditions.
19
Observing others, obtaining information
20
Patients used a variety of sources of information with some patients relying
21
more on their observations of how other people dealt with medicines such as
22
HIV medicines rather than the advice/information given by the doctor. People
23
use a variety of sources of information (e.g. internet, books, support groups)
24
as well as that provided from their GPs
25
Objective and subjective indicators
26
People used both objective and subjective indicators to evaluate efficacy of
27
medication. In the studies in the synthesis blood pressure monitoring
28
appeared to be used widely as a means of evaluating efficacy of
29
antihypertensive medication. The perception of symptom minimization could
30
be as important as objective indicators such as an increase in T-cell count in
31
the case of HIV patients.
Medicines concordance: full guideline DRAFT (July 2008) page 176 of 373
DRAFT FOR CONSULTATION
1
Gender differences in evaluating medicines
2
Some studies suggested that women with HIV do not belong to certain social
3
networks like gay men or injecting drug users and thus could be less informed
4
about the medicines. Also, some women were reported to show scepticism
5
towards the drugs used for HIV with arguments that the clinical trials had not
6
been conducted with women.
7
Difficulties with evaluating medicines
8
In some studies it was noted that patients could not distinguish between the
9
effects of their illness from the effects of the medicine. This could even lead to
10
patients rejecting treatments mistakenly. Other authors pointed out that
11
evaluation can be dependent on the individuals understanding of how a
12
medicines works and its function and this can be difficult for certain people
13
due to lack of information and understanding. This is particularly relevant in
14
the case of preventative medicine, as there are no immediate symptoms that
15
can used as indicators of efficacy.
16
Worries about medicines that lay testing and evaluation cannot resolve
17
Fear of dependency and tolerance were pointed out as issues for patients.
18
Fear of addiction was reported when taking psychotropic medicines and
19
general concern over taking medicines on a long term basis was present with
20
hypertensives. Some authors noted that those who worried over the long term
21
effects of medicines were those most likely to change their regimen to the
22
lowest possible dose.
23
5.3.4
24
Non-acceptance
25
Since taking medicine is equated with having an illness people may not take
26
medicine if they do not accept their illness. This was particularly strong in the
27
asthma studies in the synthesis. The relationship with acceptance of disease
28
was however complex with some patients using medicines to keep their
29
problem under control and downplay its significance. In the case of
30
neuroleptic medication and HIV medication an acceptance of the diagnosis
31
was crucial in determining whether the patient would take their medicine as
32
prescribed. Medication was seen as a reminder of illness
Medicines and identity
Medicines concordance: full guideline DRAFT (July 2008) page 177 of 373
DRAFT FOR CONSULTATION
1
Disclosure and stigma
2
People with potentially stigmatizing illnesses such as HIV and mental health
3
problems were particularly concerned that their medications marked them out
4
as individuals with those particular health problems. This could lead to
5
avoiding taking medicine in public and either postponing or foregoing their
6
medicine intake. Some people were reported to not initiate treatment which
7
would reveal them as having HIV. Patients with mental health problems also
8
reported feeling stigmatized by their intake of medicines, and some felt
9
ashamed.
Ways people take their medication
10
5.3.5
11
Motivation to minimize intake
12
The majority of studies illustrate how people wish to minimize their intake of
13
medicines with patients spontaneously reporting dislike of drugs to
14
researchers. This was also true in the case of medicines commonly reported
15
as being overused by patients, such as benzodiazepines.
16
To decrease adverse effects and addiction
17
People often reduce or skip doses, or take tablets separately rather that all at
18
once. Others may pause their medicine intake as a way of cleansing their
19
body and minimize toxicity.
20
To make the regimen more acceptable
21
This section related to how people modified their regimen in order to fit with
22
their daily schedule, alleging that to have some clinical gain, complete
23
adherence was not required. Other argued that the optimum regimen was not
24
known anyway, and that strict adherence was not possible.
25
For financial reasons
26
Some people were reported to have decreased their doses as they could not
27
afford the prescribed amounts.
28
Using medicine symptomatically
29
Some patients were reported to use medicines accordingly to symptoms
30
displayed. Patients associated symptoms with their medical problem and
31
when these symptoms indicated that the problem was not controlled they
Medicines concordance: full guideline DRAFT (July 2008) page 178 of 373
DRAFT FOR CONSULTATION
1
would take their medicines. An example was symptoms of tiredness or
2
weakness in treatment for high blood pressure. People with rheumatoid
3
arthritis modified their doses according to their symptoms as did patients on
4
neuroleptics.
5
Using medicine strategically
6
People adjusted their dose or did not take their medication when planning to
7
drink alcohol as they feared possible interactions. This was also reported for
8
those on neuroleptic medicines and antihypertensive medication. PPIs were
9
also altered according to what people planned to eat.
10
Replacing or supplementing medicines with non-pharmacological treatments
11
In varying ways, people often complemented their treatment with home
12
remedies. On a more specific level, some patients who worried about the
13
harmful effects of medicines would sometime take a break and use natural
14
remedies for a certain period.
15
Doctor-patient communication about regimen modifications
16
Some patients reported being scolded by their doctors if making their own
17
decisions about their care. Rather than confronting them, they would then
18
switch doctors. Also, many patients do not reveal their beliefs to the doctors,
19
but once outside the surgery and in control, people would then change or
20
refute their regimen. Some authors noted that patients who had changed their
21
regimen would not disclose this to the HCPs due to previous experience of
22
coercion or recognition of their powerless position. Authors argue that as
23
patients self-regulate anyway, doctors need to recognize this and even
24
support them in the process, as well as helping people feel in control.
25
Imposed compliance
26
One issue that was found exclusively in studies relating to mental health was
27
that of ‘imposed compliance’. Patients with mental health problems felt
28
surveyed for signs of ill health and under pressure or even coercion from
29
friends, relatives and health care professionals to take medicines. Patients felt
30
that medicine was used to make them acceptable to society and was part of
Medicines concordance: full guideline DRAFT (July 2008) page 179 of 373
DRAFT FOR CONSULTATION
1
an unwritten social contract that required the taking of medicines to allow
2
patients to be acceptable to the community.
3
5.4
Update of qualitative evidence synthesis - Pound 2005
synthesis
4
Methodology of update
5
5.4.1
6
The aim is to update the synthesis of qualitative evidence of medicine taking
7
in a similar methodological approach as when updating a Cochrane Review.
8
The titles and abstracts of records retrieved by the searches were scanned
9
and any potentially relevant publications obtained in full text. The studies were
10
reviewed to identify the most appropriate evidence to help answer the key
11
questions and to ensure that the recommendations are based on the best
12
available evidence. Qualitative synthesis is considered a process where the
13
analysis of a number of qualitative studies may result in new findings not
14
contained in the individual studies. It was felt therefore inappropriate to do
15
additional synthesis. This update is a narrative review which discusses the
16
studies found in the update particularly where the findings add to the existing
17
synthesis.
18
Types of studies - studies with both qualitative methods of data collection
19
and analysis published in English.
20
Types of participants - people aged above 16 years prescribed medication
21
for a medical condition. (3 studies included in the synthesis relate to children
22
and/or adolescents).
23
Duration of studies - no limit on duration of studies.
24
Possible challenges - One of the first challenges of the process of
25
developing a systematic review on qualitative evidence is how to find the
26
evidence. Qualitative evidence is catalogued on a wide range of databases or
27
sometimes not at all and indexes and search filters require substantial
28
improvement if they are to be rigorous and systematic. Secondly, is the lack of
29
agreement of appropriate methods for appraising the quality of qualitative
30
evidence. We will use the same criteria as those used in the original
Medicines concordance: full guideline DRAFT (July 2008) page 180 of 373
DRAFT FOR CONSULTATION
1
synthesis. The third challenge relates to the sheer complexity of synthesizing
2
evidence that can have varying approaches to data collection and analysis
3
and issues surrounding generalisability of the findings which is of utmost
4
importance to the evidence based medicine (EBM) discipline. Generalisability
5
may conflict with the underpinning philosophical framework of some
6
qualitative research.
7
5.4.2
8
Update searches
9
Our update searches produced 381 references. Based on abstract
Evidence review
10
information, eighty five studies were ordered. Further exclusions were made if
11
the study turned out not to be relevant for the topic, or did not have qualitative
12
data collection and analysis. CASP criteria were used to assess the quality of
13
the studies. The details of all the studies are in Appendix C. Forty-five studies
14
were included in the update. The studies covered a wide range of medical
15
conditions and patient groups but did cluster around long term conditions.
16
Twelve were concerned with HIV medication, 8 with medication for psychiatric
17
conditions and 4 for patients with diabetes. Five studies focused on patients
18
from low income and/or ethnic groups. A shift could be seen in that most
19
recent studies discussed the issue of patient-health care professional
20
communication more explicitly than those included in the Pound synthesis.
21
Many papers, as in the Pound synthesis, accepted a medical paradigm that
22
medicine-taking was a good thing and some sought to understand patients’
23
beliefs and experience with a view to improving adherence. Patients’
24
readiness to negotiate with health care professionals and to disclose their
25
medicine taking behaviour to health care professionals was discussed (Bane,
26
Catherine, Hughes, Carmel M., Cupples, Margaret E. et al , 2007) as was the
27
challenges facing health care professionals in supporting the integration of
28
patients needs and preferences (Hayes, Risa P., Bowman, Lee, Monahan,
29
Patrick O. et al , 2006). In general the findings fitted well into the categories
30
elaborated in the Pound synthesis although studies often developed their own
31
terminology and categories. As described in the Pound synthesis many
32
studies did not reference each other. The findings in this update are described
33
under the themes as described in the Pound synthesis i.e. (1) the evaluation
Medicines concordance: full guideline DRAFT (July 2008) page 181 of 373
DRAFT FOR CONSULTATION
1
of medicines and the difficulties people encountered in this, (2) medicines and
2
identity and (3) the ways people take their medication.
Medicines concordance: full guideline DRAFT (July 2008) page 182 of 373
DRAFT FOR CONSULTATION
1
2
5.4.2.1
The evaluation of medicines and the difficulties encountered by
people in evaluating medicines
3
The findings of the update searches were similar to the findings of the Pound
4
synthesis in how patients evaluated medicines. The themes of trying out the
5
medicines and weighing up the costs and benefits were present as well as the
6
importance of adverse effects and the acceptability of the regimen. The
7
studies elaborated different terminology. Carrick (2004) (Carrick, Rachael,
8
Mitchell, Annie, Powell, Richard A. et al , 2004) developed a core concept of
9
‘well being’. The study was an interview study of 25 adults taking antipsychotic
10
medication. The findings were that patients sought to maximize ‘well-being’
11
which was normality of function, feeling and appearance. ‘Well being’ was
12
defined personally by patients and was their goal in taking treatment. The
13
achievement of well-being was a net effect of symptoms and side effects.
14
Some patients preferred the effects of their disease to the side effects of
15
treatment. This was achieved by an interplay of evaluating treatment,
16
managing treatment and patients’ understanding of the situation. Patients
17
considered medication in the context of their beliefs about their illness and its
18
causes. While the maximizing of ‘well-being’ was relevant for all patients
19
interviewed there was a spectrum of behaviour in relation to how active the
20
patient was in engaging with their doctors and talking through their views
21
about medication. Deegan (2005) (Deegan, P. E., 2005) again in the field of
22
psychiatric problems developed a concept of ‘personal medicine’. She
23
interviewed 29 patients with psychiatric problems and considered that
24
psychiatric medications are considered in relation to ‘personal medicine’ i.e.
25
non-pharmaceutical activities that gave meaning and purpose to life and that
26
serve to raise self-esteem, decrease symptoms and avoid unwanted
27
outcomes. Examples of personal medicine were the ability to work, and to
28
parent appropriately and to engage in social activities. Her analysis was that
29
medicines that conflict with patients’ ‘personal medicine’ are unlikely to be
30
used by patients.
31
Balancing the benefits of medication against the side effects of treatment was
32
a theme also in studies of patients on medication for HIV. Lewis (2006)
33
(Lewis, M. P., Colbert, A., Erlen, J. et al , 2006) interviewed patients who were
Medicines concordance: full guideline DRAFT (July 2008) page 183 of 373
DRAFT FOR CONSULTATION
1
100% adherent with HAART treatment and found that they performed a trade
2
off between the benefits and side effects of medication. In this sample of
3
adherent patients the interviewees reported that they did not have many other
4
options for treatment and HAART was important in keeping them well.
5
Cooper (2002) (Cooper, V., Buick, D., Horne, R. et al , 2002) interviewed 26
6
patients who had declined HAART treatment and found that they used their
7
own interpretations (which often differed from professional interpretations) of
8
indicators such as CD4 counts to inform their decision, preferred non-
9
pharmacological treatments and also found the lack of symptoms an issue in
10
considering treatment. The paper uses the concept of patients perceived
11
‘personal necessity’ of treatment as a factor in their decision. Patients also
12
had concerns about medicines from previous personal experiences or from
13
seeing and talking to others taking HIV medicines.
14
The methods by which people evaluate medicines and the difficulties
15
experienced by patients in doing so was present in the studies. Ogedegbe
16
(2004) (Ogedegbe, Gbenga, Harrison, Melanie, Robbins, Laura et al , 2004)
17
interviewed 106 African –American patients in urban primary care clinics.
18
Patients had difficulties in evaluating treatment due to the lack of symptoms of
19
raised blood pressure. Many used their own indicators to consider if and when
20
they should take any treatment. . Lukoschek (2003) (Lukoschek, Petra.,
21
2003) interviewed 92 African-American patients about their beliefs and
22
attitudes to hypertension and anti-hypertensive medication. Patients held
23
differing understandings of high blood pressure and hypertension. Patients’
24
beliefs about problem influenced their approach to treatments including diet,
25
exercise and medication. Patients’ weighed beliefs about advantages of
26
medicines against the side effects and many patients preferred herbal and
27
alternative remedies. Dowell (2002) (Dowell, Jon, Jones, Anni, and Snadden,
28
David, 2002) interviewed patients from 3 GP surgeries in Scotland about
29
experiences of medicines. The patients were selected as a group of non-
30
adherent patients. For this group particular problems were again found for
31
non-symptomatic conditions and patients evaluated illness and medicines
32
using information from their own personal experience and that from their own
33
social circle. Patients with HIV interviewed for the study by Wilson (2002)
Medicines concordance: full guideline DRAFT (July 2008) page 184 of 373
DRAFT FOR CONSULTATION
1
(Wilson, Holly Skodol, Hutchinson, Sally A., and Holzemer, William L., 2002)
2
found it difficult to understand and assess their medication as they did not
3
know whether any symptoms were related to their disease or the medication.
4
Bollini (2004) (Bollini, P., Tibaldi, G., Testa, C. et al , 2004) in a study of
5
patients taking anti-depressant medication indicated that patients would test
6
treatment by stopping once they felt better to see what would happen and if
7
they really needed the medication.
8
9
For HIV patients to derive benefit from HAART medication, high adherence to
10
the prescribed regime is required. The acceptability of the regimen and fitting
11
it into schedules was a significant issues in all studies which examined
12
patients experience of taking HAART medication. Adam (2002) (Adam, B. D.,
13
Maticka, Tyndale E., and Cohen, J. J., 2003) interviewed patients who were
14
taking HAART but found the required schedule difficult and altered the dosing
15
regime and associated eating rules to fit the regime into their schedules. This
16
paper concludes that the nature of HAART medication and its regime should
17
be seen as the problem with this medication and non-adherence not seen as
18
a patient problem.
19
20
5.4.2.2
Medicines and identity
21
The meta-ethnography synthesis indicated that for many patients the taking
22
of medicines interacted with issues of identity. Non –acceptance of diagnosis
23
and issues around disclosure and stigma were significant issues in studies
24
found in Pound synthesis. Medicines challenged patients to consider
25
themselves as someone with a disease or could provide external evidence of
26
a stigmatizing disease. These issues recurred in the studies included in the
27
update. Scotto (2005) (Scotto, Carrie J., 2005) interviewed 14 patients who
28
had required hospitalization for a relapse of heart failure symptoms. In this
29
sample the acceptance of the diagnosis of heart failure resulted in an altered
30
self-image for patients but this acceptance and its integration into patients’
31
lives was an important part of managing medication. Behaviours to support
32
adherence worked when the illness and its management could be integrated
33
into ordinary life. Alfonso (2006) (Alfonso, V., Bermbach, N., Geller, J. et al ,
Medicines concordance: full guideline DRAFT (July 2008) page 185 of 373
DRAFT FOR CONSULTATION
1
2006) interviewed 15 people who were HIV positive who were not taking
2
medication and explored their reasons not to take medicine. Many had prior
3
experience of taking HAART. Not taking medication allowed some to deny
4
that they were HIV positive. For some either taking HIV medication per se or
5
the occurrence of side effects of this treatment risked exposing their HIV
6
status. Many already felt isolated and separate and did not want to exacerbate
7
this. Similar issues of taking HAART were reported in an interview study by
8
Sidat (2007) (Sidat, Mohsin, Fairley, Christopher, and Grierson, Jeffrey, 2007)
9
where patients delayed in starting treatment while dealing with issues of
10
identity and denial. Lewis (2006) (Lewis, M. P., Colbert, A., Erlen, J. et al ,
11
2006) interviewed 13 patients who were known to be 100% adherent to
12
HAART treatment and a prominent theme in this sample was transcending
13
their identity as someone with HIV which for the patients was associated with
14
feelings of self-blame and moving on from that to take control of their health
15
and its’ maintenance. In an interview study by Wilson and colleagues (Wilson,
16
Holly Skodol, Hutchinson, Sally A., and Holzemer, William L., 2002) issues of
17
identity are more dynamic and are part of an ongoing appraisal of medicines
18
and medicine taking. Pyne 2007 (Pyne, J. M., McSweeney, J., Kane, H. S. et
19
al , 2006) explored explanatory models of schizophrenia and treatments for
20
schizophrenia held by professionals and patients to provide both provider and
21
patient perspectives. There were significant differences between providers
22
and patients in models of disease, its causes and required treatments. Stigma
23
was identified as a problem by patients but not providers. Kikkert (2006)
24
(Kikkert, Martijn J., Schene, Aart H., Koeter-Maarten, W. J. et al , 2006)
25
explored views of patients with schizophrenia, their carers and professionals
26
on medication for schizophrenia. A significant barrier for this patient group
27
was the social consequences of extra-pyramidal side effects of medication.
28
Patients taking antidepressant medication also reported difficulty in accepting
29
the diagnosis and therefore the need for treatment Bollini (2004) (Bollini, P.,
30
Tibaldi, G., Testa, C. et al , 2004)) Patients with diabetes complained of a
31
stigma of been seen with and using treatment for diabetes in a study by
32
Vinter-Repalust (2004) (Vinter, Repalust N., Petricek, G., and Katic, M.,
33
2004). For these patients coming to terms with the identify of being a patient
34
with diabetes was a significant challenge. Ogedegbe (2004) (Ogedegbe,
Medicines concordance: full guideline DRAFT (July 2008) page 186 of 373
DRAFT FOR CONSULTATION
1
Gbenga, Harrison, Melanie, Robbins, Laura et al , 2004) found that African –
2
American patients reported difficulty coming to terms with a diagnosis of
3
hypertension as a significant problem in taking anti-hypertensive medication.
4
5
5.4.2.3
Ways people take their medication
6
The ways in which patients take their medicines were similar in the update
7
studies as in the Pound synthesis. Patients changed their medicines and used
8
medicines in strategic ways but did not necessarily disclose this to health care
9
professionals Deegan (2005) (Deegan, P. E., 2005)). In the sample
10
interviewed by Ogedegbe (2004) (Ogedegbe, Gbenga, Harrison, Melanie,
11
Robbins, Laura et al , 2004) the cost of prescriptions and the effort involved in
12
getting prescriptions reordered and refilled meant people did not take their
13
medication continuously. Studies of patients prescribed HIV medication
14
similarly indicated patients making adjustments to their regimes in keeping
15
with their own beliefs and experiences and not reporting these to health care
16
professionals Campero (2007) (Campero, Lourdes, Herrera, Cristina,
17
Kendall, Tamil et al , 2007).
18
19
In general patients would not report their treatment modifications to health
20
care professionals unless they saw themselves as expert patients. The
21
ubiquity of patients’ alteration of their regimes is indicated by the findings of
22
Aronson (2005) (Aronson, B., 2005). This was a small study of 11 patients
23
who were described as being completely adherent to medication. They were
24
all prescribed short term courses of antibiotics and all patients took all the
25
medication. These patients however did alter the timing of doses to fit in with
26
their schedules, doses were forgotten and then taken when remembered.
27
Wilson 2002 (Wilson, Holly Skodol, Hutchinson, Sally A., and Holzemer,
28
William L., 2002) describes HIV patients making decisions about how they
29
take their medication almost on a dose-by dose basis. This was for a number
30
of reasons and in this study patients’ medicine taking behaviours are
31
described as a result of reconciling incompatibilities which included illness
32
beliefs, the difficulty of the regime and its impact of life. Patients generally
33
described themselves as adherent to health care professionals. Reid (2006)
Medicines concordance: full guideline DRAFT (July 2008) page 187 of 373
DRAFT FOR CONSULTATION
1
(Reid, M., Clark, A., Murdoch, D. L. et al , 2006) describes the strategic use of
2
diuretic medication by patients with heart failure- patients changing the timing
3
of medication or omitting the dose according to social and other activities.
4
5.4.2.4
5
The findings of the qualitative studies included in the update of the Pound
6
synthesis (2005) fitted largely within the themes found in the original
7
synthesis. A finding not elaborated in the synthesis and not found in other
8
update studies was medicine-taking experience and behaviours in older adults
9
on complex regimes. Elliott (2007) (Elliott, Rachel A., Ross, Degnan Dennis,
10
Adams, Alyce S. et al , 2007) interviewed 20 patients aged 67-90 about their
11
experience of medicine-taking. Patients were all members of one HMO and
12
took 4-12 medicines. The researchers were particularly interested in how
13
patients on multiple medications make decisions about medicines. The
14
general findings from the study did not differ from themes found in synthesis –
15
patients wished to minimize medication overall, they stopped and started
16
medication to take a break from medication, to check if they were working, to
17
determine the cause of side effects and generally did not disclose this
18
behaviour to their physicians. The patients interviewed did report having made
19
choices between which medicines they would decide to take, this included
20
choices between medicines for different disease and choices within diseases.
21
When choosing between medicines for different diseases patients chose to
22
take the medicine for the disease they feared most or that which gave
23
symptomatic relief. Choices otherwise were influenced by symptom control,
24
side effects, medication cost, negative health experience, illness beliefs and
25
acceptability of medicine (i.e. taste etc). Illness beliefs dominated more
26
general factors such as influence of family, friends, health providers and the
27
media. Complexity of regime did not affect choice. Cost was a factor even
28
when not related to financial hardship, patients appeared to resent the cost of
29
medications. Choices were generally influenced by one dominant factor and
30
less likely to be a result of analysis of multiple factors.
31
Two studies reported on structural issues that interfered with patients’ ability
32
to appraise medication and to receive the information they required to do this.
Additional findings from update search
Medicines concordance: full guideline DRAFT (July 2008) page 188 of 373
DRAFT FOR CONSULTATION
1
Popa-Lissenau (2004) (Garcia-Popa-Lisseanu, M. G., Greisinger, A.,
2
Richardson, M. et al , 2005) reported on the difficulties patients from low
3
incomes with rheumatological disorders had in physically accessing
4
appointments with professions. Mutchler (2007) (Mutchler, Jan E.,
5
Bacigalupe, Gonzalo, Coppin, Antonia et al , 2007) reported how for non-
6
English speaking Latinos in the US the difficulty in engaging with health care
7
professionals results in reduced information available to those patients and
8
poorer relationships with professionals which of itself could reduce trust in
9
treatment.
10
5.4.2.5
Experience of medicine taking of minority groups in UK
11
We specifically searched for papers examining experience of patients from
12
minority groups in the UK. Overall their use and experience of medicines was
13
similar to that already described in the evidence review. One additional finding
14
in Erwin (1999) (Erwin, J. and Peters, B., 1999) was a belief by patients of
15
African origin that they had different physiology from white people and that
16
drugs used for treatment of HIV might not be appropriate for them and that
17
they were being discriminated against. This was an issue also raised by
18
African American patients interviewed. Sidat (2004) (Sidat, Mohsin, Fairley,
19
Christopher, and Grierson, Jeffrey, 2007)This patient group were also often
20
involved in church activities and some churches were against use of
21
medication. Lawton (2005) (Lawton, J., Ahmad, N., Hallowell, N. et al , 2005)
22
explored the perceptions of diabetic patients of Indian and Pakistani origin of
23
taking oral hypoglycaemic drugs. These patients beliefs and use of drugs was
24
influenced by their experience of the health system in their country of origin.
25
They distrusted this system and admired the NHS but consequently
26
considered that the drugs available in the UK were likely to be stronger and
27
more efficacious than those available in their country of origin and so they
28
reduced dose and sought to balance effect of drugs by taking in ‘strong’ foods.
29
People of African origin living in South London were interviewed about their
30
use of malaria prophylaxis Morgan (2005) (Morgan-Myfanwy, Figueroa-
31
Muñoz-Jose, 2005). One of the factors influencing use of anti-malarial was the
32
practice of leaving drugs in Africa for family members.
Medicines concordance: full guideline DRAFT (July 2008) page 189 of 373
DRAFT FOR CONSULTATION
1
2
5.5
Systematic reviews of barriers and facilitators for
individuals in medicine taking
3
Methods of the update
4
5.5.1
5
The GDG requested an additional search for any further systematic reviews of
6
barriers and facilitators for individuals in medicine taking. The search strategy
7
used for the Pound updates searches was applied to this review together with
8
a systematic review filter. The titles and abstracts of records retrieved by the
9
searches were scanned and any potentially relevant publications were
10
obtained in full text. Cross-referencing of all the studies was undertaken to
11
ensure that the search is as comprehensive as possible. The studies were
12
then reviewed to identify the most appropriate evidence to help answer the
13
key questions and to ensure that the recommendations are based on the best
14
available evidence.
15
Types of studies: Systematic reviews
16
Types of participants: people aged above 16 years prescribed medication for
17
a medical condition.
18
Duration of studies: no limit on duration of studies
19
5.5.2
20
The search for systematic review of barriers and facilitators of medicine taking
21
produced two eligible reviews in very diverse patient populations. No
22
systematic reviews examining statistical associations between patient
23
reported factors and actual medicine taking were found. The systematic
24
reviews that were found examined medicine taking in particular population
25
subgroups – patients with TB and patients with HIV on retroviral treatment. It
26
should be noted that although these areas overlapped with the Pound
27
synthesis there was little overlap in papers included highlighting the issues
28
raised by authors of Pound synthesis about locating qualitative literature.
Evidence review
Medicines concordance: full guideline DRAFT (July 2008) page 190 of 373
DRAFT FOR CONSULTATION
1
The first of these Munro (2007) (Munro, Salla A., Lewin, Simon A., Smith,
2
Helen J. et al , 2007) is a review of qualitative studies which aimed to
3
understand the factors considered important by patients, caregivers and
4
health care providers in contributing to TB medication adherence. The authors
5
used meta-ethnography as in Pound synthesis. The majority of the studies in
6
this review were conducted in developing countries. The emphasis on
7
adherence and inclusion of carers and health care providers perspectives as
8
well as the methodology resulted in an analysis which included structural
9
factors that influence patients medicine taking as well as patient factors.
10
The primary themes that emerged from the included studies were: 1)
11
Organization of treatment and care including access to care, treatment
12
requirements and relationship with the provider; 2) Interpretation of illness and
13
wellness; 3) Financial burden including impact on work, cost of treatment,
14
general poverty; 4) Knowledge attitudes and beliefs about treatment ; 5) Law
15
and immigration; 6) Personal characteristics and adherence behaviour
16
including substance abuse, gender ,religion, motivation; 7) Side effects ; 8)
17
Family, community and household influence.
18
A systematic review by Mills (2006) (Mills, Edward J., Nachega, Jean B.,
19
Bangsberg, David R. et al , 2006) examined patient reported barriers and
20
facilitators to adhering to antiretroviral therapy. This analysis included 37
21
qualitative studies and 47 surveys using structured questionnaires or
22
structured interviews. Seventy two studies were conducted in developed
23
countries. Fifty six were from the US and only 3 from the UK. A systematic
24
approach was taken to extracting themes from the qualitative studies and
25
synthesizing the quantitative data and pooling the results. Briefly themes were
26
extracted from qualitative studies and the reviewers then abstracted data from
27
the quantitative surveys to determine if the same issues had been addressed
28
in the surveys. The authors used their own criteria to assess the surveys and
29
these related to the development process and face validity of the
30
questionnaire and the population surveyed. The authors abstracted
31
prevalence of issues as reported in the surveys for their statistical methods.
Medicines concordance: full guideline DRAFT (July 2008) page 191 of 373
DRAFT FOR CONSULTATION
1
This technique is called meta-study and is one of a series of methods being
2
developed to bring together findings of qualitative and questionnaire studies. .
3
Barriers identified in both economic settings (developed and developing world)
4
included: fear of disclosure, concomitant substance abuse, forgetfulness,
5
suspicions of treatment, regimens that are too complicated ,number of pills
6
required, decreased quality of life, work and family responsibilities, falling
7
asleep and access to medication. Important facilitators reported by patients in
8
developed nation settings included having a sense of self work, seeing
9
positive effects of anti-retrovirals, accepting their seropositivity, understanding
10
the need for strict adherence, making use of reminder tools, and having a
11
simple regimen. In a further study of adherence rates in sub-Saharan Africa
12
and North America (Mills, Edward J., 2006) Mills comments that the most
13
prevalent barriers to adherence in sub-Saharan Africa are cost, not disclosing
14
HIV status to a loved one or fear of being stigmatized, alcohol abuse and
15
difficulty in following complex drug regimens.
16
Medicines concordance: full guideline DRAFT (July 2008) page 192 of 373
DRAFT FOR CONSULTATION
1
6
Information regarding medicines for patients
2
and practitioners when patients are discharged
3
from inpatient care
4
6.1
Recommendations *listed at the start of the guideline
according to chapter
5
6
6.2
Introduction
7
Patients are frequently started on medications when in hospital as an
8
inpatient or when attending outpatient clinics. Transitions between care
9
settings have been recognised as a time when potential errors in medication
10
can occur. NICE and the National Patient Safety Agency (NPSA) have
11
recently produced joint guidance on medicines reconciliation when adult
12
patients are admitted to hospital (www.NICE.org.uk/PSG001). Literature
13
reviews suggested unintentional variances of 30-70% between medicines
14
patients were taking before admission and those prescribed on admission.
15
The GDG considered that patients have the same rights to information and
16
choice when in hospital and community care settings but acknowledged that
17
actual care does not always allow this.
18
Medicines concordance: full guideline DRAFT (July 2008) page 193 of 373
DRAFT FOR CONSULTATION
1
2
6.3
Key clinical question: What information regarding
3
medicines should be provided for patients and
4
practitioners when patients are discharged from
5
secondary care
Related
Evidence statements
references
(summary of
Evidence into recommendations
evidence)
The GDG recognised that medications may be
initiated in hospital settings in emergency
situations and when patients are unwell. They
recognised that in these situations discussions of
details of medication therapy may not be
possible. However as patients condition
improves and patients are prepared for
discharge they should be offered a full
explanation of their medication. This explanation
should allow patients to make informed choices
about their continued use of medication
prescribed. The experience of the GDG was of
considerable confusion and lack of information
provided to patients and to subsequent providers
of care when patients are discharged. Difficulties
arise not only in knowledge of what medications
have been prescribed for patients but what
information patients have been given about their
illness and medicines. The GDG based these
recommendations on professional opinions and
information from expert sources.
Medicines concordance: full guideline DRAFT (July 2008) page 194 of 373
DRAFT FOR CONSULTATION
1
Methods of the evidence review
2
6.3.1
3
The evidence review is a narrative review. The GDG requested review of
4
available guidance and reports with a particular emphasis on those where
5
patients rights to information and involvement were given priority.
6
6.3.2
7
The Academy of Royal Medical Colleges is currently preparing consensus
8
guidance on what should be included in hospital discharge summaries. This
9
guidance which will include advice on information about medication will be
Evidence review
10
available late in 2008. The WHO produced a report in 2007 on Assuring
11
Accuracy of Medicines at Transitions of Care (www.jcipatientsafety/org/). The
12
emphasis on reports and guidance about medicines reconciliation is on the
13
reduction of medication errors. Patients’ rights to information and involvement
14
in decisions about medicines are not the primary concern of these reports.
15
The WHO report does however state that effective involvement of patients
16
and families in medicines reconciliation is key to reducing errors. They
17
suggest that
18
19
20
 patient is in the best position to be aware of all the medications
prescribed by multiple caregivers.
 Consideration should be given to asking patients to put all their
21
medications in a bag and bring it with them whenever going to
22
the hospital or a doctor visit.
23
 Patients, family, and caregivers should be encouraged to keep
24
and maintain an accurate list of all medications, including
25
prescription and non prescription medications, herbal and
26
nutritional supplements, immunization history, and any allergic or
27
adverse medication reactions. These medication lists should be
28
updated and reviewed with the patient/family/caregiver at each
29
care encounter.
30
 Patients should be taught about the risks of medications, both
31
individually and in combination, with particular attention to
Medicines concordance: full guideline DRAFT (July 2008) page 195 of 373
DRAFT FOR CONSULTATION
1
patients on multiple medications prescribed by multiple
2
caregivers.
3
Medicines concordance: full guideline DRAFT (July 2008) page 196 of 373
DRAFT FOR CONSULTATION
1
2
7
Assessment of adherence
3
7.1
Recommendations *listed at the start of the guideline
according to chapter
4
5
7.2
Introduction
6
In a systematic review of communication between health care professionals
7
and patients Cox (2004) (Cox, F, Stevenson, N, Britten, N et al , 2004) states
8
that for concordance to occur patients and professionals need to have two-
9
way discussions in which they exchange information and views about
10
medicines. Clearly any initial decision to prescribe medication needs to
11
include a two-way discussion about medicines. However many patients take
12
medications over long periods of time and discussions about these
13
medications need to also consider the patients experience of taking the
14
medication. This includes an assessment or discussion about whether or not
15
the patient is taking the medication and if they are doing this exactly as
16
prescribed or in some other way.
17
18
A number of ways of assessing adherence have been developed. These can
19
generally be described as direct methods or indirect methods. Direct methods
20
are examinations of blood, urine or other bodily fluids for the presence of the
21
drug or a metabolite. Indirect methods do not measure the presence of the
22
drug but use methods such as self report from patients, pill counts,
23
prescription reordering, pharmacy refill records, electronic drug monitoring
24
and therapeutic effect to form an assessment of adherence. In the context of
25
routine clinical practice and of involving patients in decisions about medicines
26
the GDG decided that measures such as self-report were most likely to be off
27
use to health care professionals. The GDG therefore wished to consider the
28
advantages and disadvantages self report in routine clinical practice.
Medicines concordance: full guideline DRAFT (July 2008) page 197 of 373
DRAFT FOR CONSULTATION
1
2
7.3
Key Clinical Question: What are the advantages and
3
disadvantages of self-report in assessing patient’s
4
adherence?
Related references
Evidence statements
Evidence into recommendations
(summary of evidence)
The GDG did not consider that
Advantages
direct measures of adherence
Hawkshead (2007)
1. Self-reporting is the most
(Hawkshead, J. and
simple and inexpensive method
Krousel-Wood, M.
of measuring adherence.
A., 2007); Gagne
(2005) (GagnéCamille, Godin
Gaston, 2005);
Paterson (2002)
(Paterson, David L.,
Potoski, Brian, and
Capitano, Blair,
2002); Miller (2000)
(Miller, L. G. and
Hays, R. D., 2000);
Turner (2002)
(Turner, Barbara J.,
were relevant in routine clinical
practice. Indirect methods such as
therapeutic effects and prescription
ordering and refills are methods
which should alert prescribers and
dispensers to problems of
adherence. In these situations and
as part of medication reviews
health care professionals need to
be able to discuss medicine taking
with patients. The GDG made
recommendations on how
professionals should assess
adherence using the review of
advantages and disadvantages of
self-report.
2002); Farmer
(1999) (Farmer, K.
C., 1999); Bender
(1997) (1997)
(Bender, B.,
Milgrom, H., and
Rand, C., 1997);
Medicines concordance: full guideline DRAFT (July 2008) page 198 of 373
DRAFT FOR CONSULTATION
LaFleur (2004) (La,
Fleur J., 2004);
Rand (1994) (Rand,
C. S. and Wise, R.
A., 1994);
Miller (2000) (Miller,
2. Self-reporting is quick and
L. G. and Hays, R.
easy to administer, avoiding the
D., 2000);
use of sophisticated methodology
Farmer(1999)
or equipment.
(Farmer, K. C.,
1999); Paterson
(2002) (Paterson,
David L., Potoski,
Brian, and Capitano,
Blair, 2002); Bender
(Bender, B.,
Milgrom, H., and
Rand, C., 1997);
Rand (1994) (Rand,
C. S. and Wise, R.
A., 1994);
Hawkshead (2007)
3. Self-reporting methods which
(Hawkshead, J. and
are validated can feasibly be
Krousel-Wood, M.
used in clinical settings.
A., 2007);
Bender(1997)
(Bender, B.,
Milgrom, H., and
Rand, C., 1997);
Hawkshead (2007)
4. Self-reporting can identify
(Hawkshead, J. and
those who are not adherent. It is
Medicines concordance: full guideline DRAFT (July 2008) page 199 of 373
DRAFT FOR CONSULTATION
Krousel-Wood, M.
most likely those reporting non-
A., 2007); Paterson
adherence are correct.
(2002) (Paterson,
David L., Potoski,
Brian, and Capitano,
Blair, 2002); Farmer
(1999) (Farmer, K.
C., 1999); Hecht
(1998) (Hecht, F. M.,
1998); Bennett
Johnson (1992)
(Johnson, S. B.,
1992); George
(2007) (George,
Johnson, KongDavid, C. M., and
Stewart, Kay, 2007)
Hawkshead (2007)
5. Self-reporting can gather
(Hawkshead, J. and
social, situational and
Krousel-Wood, M.
behavioural factors including
A., 2007); Miller
revealing patterns of medication
(2000) (Miller, L. G.
use and what leads to
and Hays, R. D.,
noncompliance.
2000); Rand (1994)
(Rand, C. S. and
Wise, R. A., 1994);
Bennett Johnson
(1992) (Johnson, S.
B., 1992);
Disadvantages
George (2007)
6. Self-reporting has the problem
Medicines concordance: full guideline DRAFT (July 2008) page 200 of 373
DRAFT FOR CONSULTATION
(George, Johnson,
of over-estimating adherence.
Kong-David, C. M.,
and Stewart, Kay,
2007); Hawkshead
(2007) (Hawkshead,
J. and KrouselWood, M. A., 2007);
LaFleur (2004) (La,
Fleur J., 2004);
Turner (2002)
(Turner, Barbara J.,
2002); Miller (2000)
(Miller, L. G. and
Hays, R. D., 2000);
Hecht (1998) (Hecht,
F. M., 1998); Bender
(1997) (Bender, B.,
Milgrom, H., and
Rand, C., 1997);
Hawkshead (2007)
7. Inaccurate self-reporting can
(Hawkshead, J. and
be caused by recall bias, social
Krousel-Wood, M.
desirability bias and errors in
A., 2007);
self-observation.
Gagne(2005)
(Gagné-Camille,
Godin Gaston,
2005); LaFleur
(2004) (La, Fleur J.,
2004); Turner (2002)
(Turner, Barbara J.,
2002); Farmer(1999)
(Farmer, K. C.,
1999); Bennett
Medicines concordance: full guideline DRAFT (July 2008) page 201 of 373
DRAFT FOR CONSULTATION
Johnson (1992)
(Johnson, S. B.,
1992)
Paterson (2002)
8. The timeframe of the
(Paterson, David L.,
adherence recollection can affect
Potoski, Brian, and
the accuracy of the recall.
Capitano, Blair,
Specifying the time period can
2002); Hecht (1998)
help.
(Hecht, F. M., 1998);
Bennett Johnson
(1992) (Johnson, S.
B., 1992);
Hawkshead (2007)
9. Wording of questions, the way
(Hawkshead, J. and
a question is asked and the skills
Krousel-Wood, M.
of the interviewer can either
A., 2007); Farmer
facilitate or be detrimental to
(1999) (Farmer, K.
gaining accurate responses.
C., 1999); Hecht
(1998) (Hecht, F. M.,
1998);
Turner (2002)
10. Being non-judgmental, giving
(Turner, Barbara J.,
a preamble before adherence
2002); Bennett
questions, and asking about
Johnson (1992)
specific behaviours can help
(Johnson, S. B.,
validity.
1992);
Medicines concordance: full guideline DRAFT (July 2008) page 202 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
7.3.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies – We included literature reviews and systematic reviews
6
only.
7
Types of participants - People prescribed medication for a medical
8
condition.
9
Duration of studies – No time limit was specified.
10
Types of interventions - Any interventions intended to change adherence to
11
prescribed medication which reviews studies which focus on self-report
12
advantages and disadvantages.
13
Types of outcome measures - No outcome measures specified.
14
7.3.2
15
The searches mainly returned literature reviews, rather than systematic
16
reviews, therefore details (of the various studies mentioned in these reviews)
17
were not always given and some only mentioned the studies briefly.
18
Garber (2004) (Garber, Mathew C., Nau, David P., Erickson, Steven R. et al ,
19
2004) produced a systematic review on the concordance of self-report with
20
other measures of medication adherence. They searched a number of
21
databases and identified 2757 articles. The inclusion criteria were to include
22
studies where at least 2 adherence measures were used, one of which being
23
a self-report measure, the other a non self-report measure. The self-report
24
measures included questionnaires, diaries or interviews and were categorised
25
under these. They found 86 unique comparisons, mostly interviews (57%),
26
questionnaires (27%) and diaries (17%). The nonself-report measures were
27
electronic measures (35%), pill count or canister weight (26%), a plasma drug
28
concentration (20%) a claims-based measure (13%) and a clinical opinion
29
(6%). 43% of the pairings of self-report and nonself-report measures were
Evidence review
Medicines concordance: full guideline DRAFT (July 2008)
page 203 of 373
DRAFT FOR CONSULTATION
1
highly concordant. In the majority (45/59) of those not highly concordant the
2
self-report measure showed higher adherence compared to the nonself-report
3
measure but this varied widely depending on type of self-report measure. 31%
4
of the interviews were highly concordant with nonself-report measures. Diaries
5
(71%) and questionnaires (55%) were much more likely to be highly
6
concordant with non-self report measures. It also depended on the nonself-
7
report measures. Self-reporting had higher concordance with other types of
8
nonself-report measures (58%) than electronic measures (17%). Interviews
9
showed the least concordance with electronic measures. It should be noted
10
that this is a comparison between measures which can not fully evaluate the
11
accuracy of any of the measures.
12
In summary, questionnaires and diaries were more concordant with other
13
measures. However this is a comparison with other measures which cannot
14
fully evaluate the accuracy of any of the measures.
15
George (2007) (George, Johnson, Kong-David, C. M., and Stewart, Kay,
16
2007) conducted a literature review to assess adherence of COPD patients
17
with disease management programs. They searched OVID and International
18
Pharmaceutical Abstracts. They did not report the inclusion/exclusion criteria
19
or how many studies were retrieved.
20
They found that self-reporting of missed doses (by questionnaire)
21
underestimated non-adherence compared to more objective measures e.g.
22
capsule count (Dompeling, 1992), inhaler weights (Rand, 1995) and electronic
23
monitoring (Rand, 1992; Braunstein, 1996; Simmons, 2000). Self-report was
24
shown to have moderate reliability compared to objective measures such as
25
canister weight (Rand, 1995) and electronic monitoring (Gong, 1988; Nides,
26
1993; Bosley,1995).
27
Self-reporting of non-adherence of medication for COPD has shown
28
satisfactory reliability, when compared to objective measures (Dolce, 1991;
29
Nides, 1993; Rand, 1995). Self-report is commonly criticised for
30
overestimating adherence and poor reliability yet those who report non-
Medicines concordance: full guideline DRAFT (July 2008)
page 204 of 373
DRAFT FOR CONSULTATION
1
adherence are likely to be telling the truth (Haynes, 1980; Inui, 1981; Choo,
2
1999; Erickson, 2001).
3
In summary, self-reporting questionnaires underestimate non-adherence but
4
have shown reliability and are usually correct for those who say they are non-
5
adherent.
6
Hawkshead (2007) (Hawkshead, J. and Krousel-Wood, M. A., 2007)
7
presented a narrative review of the advantages and limitations of methods for
8
measuring adherence in hypertensive patients. No mention is given to how
9
they searched for these studies or decided to include/exclude. They state that
10
self reporting is the simplest method for assessing medication adherence and
11
can include patient diaries, interviews during office visits and adherence-
12
specific questionnaires. ‘Several multi-item questionnaires have been
13
developed and tested in outpatient settings with the explicit aim of
14
ascertaining valid and reliable estimates of adherence to antihypertensive
15
medications’, of which many have reported high measures of validity and
16
reliability (Morisky, 1986; Kim, 2000; Shea, 1992; Hyre, 2007). There are
17
three previously validated self-reported medication-adherence instruments –
18
the Medication Adherence Survey (MAS), the Brief Medication Questionnaire
19
(BMQ) and the Medical Outcomes Study (MOS).
20
Validated self-report measures can feasibly be used in clinical settings and
21
help to identify those who are non-adherent, and intervene to increase this
22
(Harmon, 2006). The advantages they state are that self-report is simple and
23
economical; it can also gather social, situational, and behavioural factors
24
which can impact on adherence. The disadvantages are the possibility that
25
there could be recall bias, over-estimation of compliance and responses which
26
are socially acceptable. Validity can also depend on the skills of the
27
interviewer as well as the question construction and timeframe (Farmer, 1999
28
and Wang, 2004). It is suggested that self-report could be combined with
29
objective information, e.g. prescription-fill data, to improve adherence
30
measurement.
Medicines concordance: full guideline DRAFT (July 2008)
page 205 of 373
DRAFT FOR CONSULTATION
1
In summary, some self-reporting questionnaires have been validated and can
2
be simple and feasible to use in clinical settings and identify non-adherers.
3
However they can have biases and overestimate adherence.
4
Gagne (2005) (Gagné-Camille, Godin Gaston, 2005) reported on how to
5
improve self-report measures for non-adherence to HIV medications, with
6
particular attention to techniques that can be applied with questionnaires
7
administered in clinical practice. Questionnaires are inexpensive and
8
convenient and can be conducted in clinical and research settings. But can
9
vary in terms of accuracy. According to many authors, forgetfulness (Brooks,
10
1994; Hayes & DiMatteo, 1987; Holzemer, 1999; Rand, 2000; Svarstad, 1999)
11
and social desirability (Felkey, 1995; Gordis, 1969; Gray, 1998; Rand, 2000;
12
Svarstad, 1999) are main factors leading to inaccurate self-reporting of non-
13
adherence. Social desirable answers can depends on how much the patient
14
perceives the desirability of the behaviour to be. Those behaviours perceived
15
as undesirable are under-reported and behaviours perceived as desirable can
16
be over-reported (Cannell 1979; Fowler, 1995). There are techniques
17
suggested for minimising forgetfulness and social desirability (Cannell, 1979;
18
Fowler, 1995; Sudman & Bradburn, 1974; Sudman & Bradburn, 1982)
19
although methods to reduce these are not well-documented, are often derived
20
from clinical practice than controlled experimental studies and their reported
21
effectiveness is inconsistent.
22
Suggestions were made to reduce socially desirable answers:
23

HCPs (Eldred, 1998; Gordillo, 1999).
24
25

28
Explaining that there are no right or wrong answers (Des Jarlais, 1999;
Chesney, 1990).
26
27
Assuring confidentiality and that information will not be available to

How the question is asked (Ickovics, in Eldred, 1998; Chesney, 1999;
Svarstad, 1999).
Medicines concordance: full guideline DRAFT (July 2008)
page 206 of 373
DRAFT FOR CONSULTATION
1

Wording the question to increase the likelihood of gaining certain
2
desired answers, such as non-adherence (loading the question)
3
(Sudman, 1982; Bradburn, 1982; Allaire, 1988).
4
5

Open-ended questions can avoid the pitfalls of response categories
(Schwarz, 1985; Sudman, 1982).
6
Open-ended questions have been used in studies of HIV (e.g. Chesney,
7
1990) and for measuring adherence/non-adherence (e.g. Svarstad, 1999).
8
Open-ended answers have shown to be less affected by social desirability
9
than close-ended answers (Sudman, 1974). Sudman (1974) also found
10
that open-ended questions were less affected by forgetfulness and
11
remembering it happening more than it actually did.
12
Recall can be aided by:
13

meaning and one idea (Sudman, 1982);
14
15


Aided-recall techniques such as memory cues may be useful
(Sudman, 1982).
18
19
Words should not have blame implications (Averitt, in Eldred,
1998).
16
17
Item wording, using familiar words and words that have only one

Specifying a reference time period, especially a recent and short
20
time frame can aid forgetfulness (Brooks, 1994; Chesney, 1999;
21
Holzemer, 1999; Sudman, 1982).
22
However there is the problem of the time period being too short and not
23
accurately representing the adherence level, as adherence varies over time
24
(Chesney, 1997b; Gray, 1998; Kastrissios, 1998). This could be solved by
25
using a short period of time and administering the questionnaire a number of
26
times over the period. However, this could lead to less motivation and could
27
be costly. Shorter periods of reference could be used when administering the
28
questionnaire only once. According to episodic and semantic memory it may
Medicines concordance: full guideline DRAFT (July 2008)
page 207 of 373
DRAFT FOR CONSULTATION
1
be best to ask more precise information about the past few days and less
2
specific information from a longer time period.
3
In summary, self-reporting by questionnaire can have biases such as social
4
desirable responses and recall bias. These biases can be minimised using
5
certain techniques.
6
LaFleur (2004) (La, Fleur J., 2004) conducted a brief narrative review of
7
methods to measure compliance with medication regimens. No search or
8
inclusion/exclusion criteria were given. They state that self-report is the most
9
popular method for assessing compliance as it is inexpensive but is often
10
unreliable (Myers, 1998). Self-report can include patient interviews or self-
11
report surveys. When compared to objective measures e.g. electronic
12
monitoring devices or drug level monitoring of compliance self-reporting has
13
shown to over-report compliance over 50% of the time (Spector, 1986; Gordis,
14
1969; Waterhouse, 1993; Straka, 1997). It is also often inaccurate for those
15
reporting non-compliance with medication-taking. In Kwon (2003) a
16
comparison of self-reporting of antidepressant use with prescription claims
17
showed a 20% difference in those reporting non-adherence to
18
antidepressants. The reasons for any discrepancies with other measures
19
could be that patients do not understand regimens, know indications for their
20
medicine, or not report behaviours perceived as not socially-acceptable, or
21
forgetting of non-compliance. No references were given for these assertions.
22
In summary, self-report by interviews or surveys can be inexpensive but can
23
be unreliable and over-report compliance. Those who report non-compliance
24
can also be inaccurate. There could be biases such as social desirability,
25
recall and not understanding medication regimes.
26
Turner (2002) (Turner, Barbara J., 2002) reviewed literature to compare
27
various measures of adherence to Antiretroviral Therapy. This was a narrative
28
review with no details of search/inclusion criteria. They state that self-reports
29
are less complex but that there can be problems with recall over long time
30
periods. Many studies use self-report over the past 4 days but additional
Medicines concordance: full guideline DRAFT (July 2008)
page 208 of 373
DRAFT FOR CONSULTATION
1
questions may be needed, e.g. about weekends, as this tends to be a difficult
2
time for adherence.
3
All types of self-reporting overestimate adherence compared to other
4
measures (Arnsten, 2001; Golin, 1999; Melbourne, 1999). Even those who
5
report missing doses tend to overestimate adherence compared to other
6
measures (Wagner, 2000). Social desirability biases can contribute. Those
7
who report problems with adherence usually have poorer adherence with
8
other measures (Haynes, 1980). Those who report non-adherence appear
9
responsive to interventions, and are important to identify (Haynes, 1980).
10
The validity can be increased with a preamble before questions about
11
adherence in order to reassure patients that information will not be held
12
against them and that non-adherence is common. Audio computer-assisted
13
self-interviewing is suggested for more sensitive topics (Metzger, 2000;
14
Gribble, 2000).
15
In summary, all types of self-report overestimate adherence, even with those
16
who report non-adherence and biases such as social desirability can occur.
17
Certain techniques could be used to minimise these biases.
18
Paterson (2002) (Paterson, David L., Potoski, Brian, and Capitano, Blair,
19
2002) conducted a brief narrative review to ascertain how adherence to
20
antiretroviral medicine should be measured. The methods reported were
21
electronic monitoring, pill counts, pill recognition, review of pharmacy records,
22
patient self-report, biological parameters, therapeutic drug monitoring and
23
provider prediction of adherence. They noted that how a question is asked
24
can influence self-report of adherence (i.e. in face-to face inquiry or patient-
25
completed questionnaires). A non-judgemental stance can help and this can
26
be achieved by a preamble before the questions to show that they are not
27
being judged and are looking for honest answers (Turner, 2001).
28
Another disadvantage of self-report (face-to-face interview) is that periods
29
shorter than 7 days are not long enough to determine the percentage of
30
adherence likely, however some patients may not correctly report adherence
Medicines concordance: full guideline DRAFT (July 2008)
page 209 of 373
DRAFT FOR CONSULTATION
1
for 7 day periods. They state that additional questions may be necessary to
2
counteract this e.g. about adherence at the weekend.
3
One method to counteract problems with gaining honest answers is computer-
4
assisted self-interviewing (Bangsberg, 2001) or diary. Diaries hold advantage
5
as they can be inexpensive and accurate. Their disadvantage is that some
6
may complete them retrospectively or not at all.
7
Paterson (2002) asserts that self-report is ‘likely to be the simplest means of
8
assessing adherence’ and so the reliability is important to assess. Adherence
9
was found to be ‘considerably higher’ than that measured by electronic
10
monitoring or pill count (Liu, 2001). Self-report overestimates adherence. It is
11
most useful in those who admit to being poor adherers (Murri 2000). They
12
conclude that electronic monitoring devices are the closest to a gold standard
13
in adherence measurement.
14
In summary, various self-reporting measures were reported and interviews
15
may be too late for recall or may be too early to gain useful adherence
16
information. Diaries are inexpensive and can be more accurate as there is no
17
recall bias however they may not be completed or completed retrospectively.
18
Self-report can overestimate adherence but can identify those who report non-
19
adherence.
20
Miller (2000) (Miller, L. G. and Hays, R. D., 2000) reviewed current literature
21
on measuring adherence to Antiretroviral Medications in clinical trials. They
22
report that the simplest method of measuring adherence is self-report. But
23
there is no standardised instrument. Self-reported surveys are quick and avoid
24
sophisticated methodology or equipment and are inexpensive compared to
25
other methods of measurement. They have limitations, such as significantly
26
exceeding adherence measured by other objective methods (Bond, 1991;
27
Stratka, 1997; Cramer, 1991). HIV studies also confirm this (Golin, 1999;
28
Arnsten, 2000; Paterson, 1999; Bangsberg, 1999). Interviews and surveys
29
often promote socially acceptable responses (DiMatteo, 1982). Less adherent
30
patients report higher adherence than they actually had (Bond, 1991).
31
Memory can also affect the accuracy of reporting adherence. Most surveys
Medicines concordance: full guideline DRAFT (July 2008)
page 210 of 373
DRAFT FOR CONSULTATION
1
use broad response categories to report the proportion of pills taken, thus
2
small degrees of nonadherence is hard to distinguish with self-report. The
3
information is useful, but accuracy is limited and biased towards higher
4
adherence.
5
However, self-reported non-adherence has been associated with worse
6
virologic outcomes (Demasi, 1999; Bangsberg, 1999; Duong, 1999; Murri,
7
1999; Le Moing, 1999) and as an independent predictor of clinical response to
8
HAART when controlling objective virologic and immunologic markers
9
(Montaner, 1999). Therefore it can provide information that explains variation
10
in clinical response to antiretroviral therapy which is not explained by other
11
clinical factors.
12
In summary, self-report surveys are simple and inexpensive but can
13
overestimate adherence. Interviews and surveys can have social desirability
14
and recall biases. Also as categories are large, small degrees of non-
15
adherence are hard to detect. There is no standardised instrument. However it
16
can explain variation in clinical responses to ART.
17
Farmer (1999) (Farmer, K. C., 1999) conducted a review of methods for
18
measuring and monitoring medication regimen adherence in clinical trials and
19
clinical practice. They searched Medline for the years 1990 to 1999 and
20
retrieved 2630 articles regarding patient compliance. They found that forms of
21
self-report included questioning/interrogation and the use of diaries and
22
survey instruments. They tabulated the various methods for assessing
23
adherence and their advantages and disadvantages. Patient interviews are
24
easy to use and inexpensive but the patient can be influenced by question
25
construction and interviewer’s skill. Adherence questionnaires are easy to
26
administer (on site, mail, telephone), can be validated and may explain patient
27
behaviour. However there is a lack of continuous data and the accuracy is
28
instrument dependent.
29
Patient interviews are considered the most unreliable for assessing adherence
30
(Grymonpre, 1998; Matsui, 1994; Craig, 1985; Straka, 1997; Park, 1964; Inui,
31
1981; Gordis, 1969). Those who report non-adherence are usually correct,
Medicines concordance: full guideline DRAFT (July 2008)
page 211 of 373
DRAFT FOR CONSULTATION
1
whereas those who say they are adherent may not be (Cramer, 1991).
2
However it can depend on the method used and how it is used. Assessing
3
self-reporting is difficult mainly because there are so many methods. The
4
interviewer’s skill and the construction of the questions can affect the
5
accuracy and validity of self-report. The relationship and communication
6
between the HCP and patient have shown to significantly affect compliance
7
(Davis, 1969). Highest compliance was found with those who joked, laughed
8
and sought suggestions from their GP. The wording of questions can affect
9
the response, and implications of blame can encourage biased responses
10
(Ross, 1991). Some answers are socially desirable and concealed their real
11
behaviour (Sherbourne, 1992). It is hard to assess studies of interviews as the
12
way they are asked could bias the result. Stewart (1987) looked at 2
13
compliance questions in an interview to assess medication-taking behaviour.
14
Comparing the results to pill counts, the questions had a specificity of 69.8%
15
and sensitivity of 80%, therefore an overall 74.5% accuracy. The time frame
16
used for recall can differ, some researchers do not specify, others are 7-10
17
days and some are a month (Grymonpre, 1998; Dirks, 1982; Straka, 1997).
18
To correct these problems some researchers have tried to construct a
19
standardised questionnaire for measuring adherence. For example Morisky
20
(1986) developed a 4-item questionnaire specific to medication regimen
21
adherence. It was assessed on unidimensionality and reliability and
22
concurrent validity with blood pressure control. The instrument’s sensitivity
23
was 81% and specificity 44%. It was not found to be efficient at predicting
24
poor adherence (Morisky, 1986).
25
In summary, a few methods of self-report were looked at. Interviews are
26
simple and inexpensive, but can depend on the interviewer. Questionnaires
27
can be administered in a variety of methods, but are considered the most
28
unreliable. Those who say they are non-adherent are usually correct but many
29
who say they were adherent may not be.
30
Bender (1997) (Bender, B., Milgrom, H., and Rand, C., 1997) conducted a
31
literature review to assess non-adherence in asthmatic patients. A search of
32
Medline was made from 1990 to 1997 of all pertinent articles, preferably
Medicines concordance: full guideline DRAFT (July 2008)
page 212 of 373
DRAFT FOR CONSULTATION
1
controlled studies. Self-report measures can be collected by interview, diaries
2
and questionnaires but o validated adherence-specific questionnaire is
3
commonly used as they are often too specific. Self-report measures are
4
simple, inexpensive and usually brief and so they are commonly used to
5
measure adherence. Especially in the clinical setting they are the best
6
measure for collecting information of beliefs, attitudes and experiences with
7
medication regimes. Accuracy with other measures is highly variable. Spector
8
(1986), Coutts (1992) and Gibson (1995) compared asthmatics self-reporting
9
of inhaler usage with electronic medication monitoring devices and they
10
showed that asthma diaries usually overestimate adherence. Demands of the
11
setting can influence the usefulness and reliability of the information gained
12
from self-reporting. These can be a desire to please on the part of the patient
13
and HCP skill and sensitivity in eliciting self-reports. When collected well it can
14
give good insight into patients’ problems with adherence. And as they are
15
unlikely to identify themselves as nonadherers, this helps identify the
16
nonadherers (Coutts, 1992; Spector, 1986; Dolce, 1991; Morisky, 1990).
17
In summary, self-report measures are simple, inexpensive, brief and the best
18
way of collecting information in the clinical setting. However diaries
19
overestimate adherence and the demands of the setting can influence the
20
usefulness and reliability of the measure.
21
Rand (1994) (Rand, C. S. and Wise, R. A., 1994) reported in a narrative
22
review on measuring adherence to asthma medication regimens. They did not
23
state search or inclusion criteria.
24
They state that self-report is the most inexpensive and quick way of
25
measuring adherence (Soutter, 1974). The possible advantage of diary cards
26
is that they can measure adherence across time and can reveal patterns
27
between the disease exacerbation and compliance with the medication. As
28
there are many drugs used within asthma prescribing, it can help to see the
29
adherence of certain drugs rather than just overall. It can also specifically
30
assess overuse, inappropriate use or erratic use of medications as well as
31
triggering events for the need for medication e.g. in Kesten (1991). Asthma
32
diaries may share commonalities but there is no standardised diary as such in
Medicines concordance: full guideline DRAFT (July 2008)
page 213 of 373
DRAFT FOR CONSULTATION
1
research. A disadvantage of asthma diaries can be they may be complex and
2
time-consuming. Also criteria of acceptable adherence may differ from patient
3
to patient. One way to evaluate the level of adherence is to use trained,
4
masked, medical personnel to score the compliance. It is preferable to
5
develop standardised compliance criteria for all raters and train them by a
6
standardised protocol and make sure there is interrater reliability.
7
Many studies have used questionnaires to collect clinic or follow-up data of
8
patient adherence (Bailey, 1987; Kinsman, 1980; Dolce, 1991), mainly
9
designed for a particular research project. Many have adherence questions
10
within a larger questionnaire, such as the 76-item Revised Asthma Problem
11
Behaviour Checklist for adults. Rand (1994) points out that both asthma
12
diaries and self-report are the most common for assessing asthma medication
13
adherence but that these instruments, because they are not standardised or
14
not published so they rarely have validity and reliability assessed. Except for
15
the Medication Adherence Scale and Inhaler Adherence Scale (Kinsman,
16
1980; Dolce, 1991; Bailey, 1990), which are six-item scales based on
17
Morisky’s work (1990). This instrument was found to have a Chronbach’s
18
alpha of 0.76 and 0.69 and was concordant with outcome measures in the
19
UAB adult asthma study.
20
The limitations of self-report have been mentioned by many authors (Masur,
21
1981; Mawhinney, 1991; Cramer, 1989; Rand, 1992). When compared to
22
objective measures it varies highly on the degree of accuracy (Gordis, 1966;
23
Mattar, 1974). Diary self-reports were compared to electronic medication
24
monitoring device to measure adherence to asthmatic medication by Spector
25
(1986). The findings were that all patients self-reported using the inhaler on
26
certain days, whereas the measured medication suggested just over half did
27
so. Adding a diary can add more complexity to the patient regime than there
28
all ready is. It has been shown that the greater the complexity of a regime the
29
lower the compliance (Masur, 1981). Some participants alter their records of
30
medication use to appear compliant (Mawhinney, 1991; Rand, 1992). This can
31
be improved if they also have reporting by the family/partner of the patient
32
(Paulson, 1977).
Medicines concordance: full guideline DRAFT (July 2008)
page 214 of 373
DRAFT FOR CONSULTATION
1
Self-reporting can also depend on the individual patient or practitioner. For
2
example elderly patients may have memory impairment, especially when
3
taking many medications and not report accurately. Long-term usage may be
4
forgotten but able to recall recent usage. The skill and sensitivity of the HCP
5
can also play a role in how much information is given and the reliability of it.
6
When collected carefully it could be very god insight into the problems of a
7
patient’s adherence. Also it is unlikely that patients will represent themselves
8
as non-adherers (Gordis, 1976) so it will identify non-adherers correctly.
9
In summary, self-report is inexpensive and quick and diaries can measure
10
adherence across time and reveal any patterns and assess overuse of
11
medication. However there is no standardised diary and it can be complex
12
and time consuming. If there is no standardised questionnaire of diary then no
13
validity or reliability are assessed. Therefore there is variation on accuracy,
14
depends on the individual or practitioner.
15
Bennett Johnson (1992) (Johnson, S. B., 1992) conducted a narrative
16
literature review of adherence measurement in diabetes management. No
17
search or inclusion criteria was given.
18
They point out that self-report of regimen adherence are often mistrusted.
19
Patients may say one thing but do something completely different, often
20
because of what they think the doctor wants to hear. However noncompliance
21
self-reporting appears more valid than self-reporting of compliance (Diehl,
22
1987). Asking about specific behaviours can lead to better adherence data
23
(Cerkoney, 1980; Cox, 1984; Shlenk, 1984; Brownlee-Duffeck, 1987; Hanson,
24
1987; Hanson, 1987; Hanson, 1987; Hanson, 1988; Hanson, 1990). There
25
have only been a few that have looked at the reliability of these reports
26
(Hanson, 1987 and Hanson, 1988). If asked to report their specific behaviours
27
over a certain time period, the data can be good quality (Glasgow, 1987;
28
Johnson 1986). Multiple interviews are recommended to ensure
29
representation of adherence behaviours.
Medicines concordance: full guideline DRAFT (July 2008)
page 215 of 373
DRAFT FOR CONSULTATION
1
One disadvantage with self-reporting is problems of memory recall. Where
2
possible a significant other should additionally be interviewed regarding the
3
patient’s behaviour.
4
The advantages of self-report are numerous, as reliable information can be
5
obtained; interviews can be done over the telephone making them accessible;
6
the patient does not have to do very much apart from give their time for an
7
interview. They however do need trained interviewers, or with multiple
8
interviews and multiple patients the process can take a lot of time and effort.
9
No references were made for these assertions.
10
In summary, self-reporting of non-compliance is likely to be more valid,
11
whereas compliance reporting is not valid. They can ask about specific
12
behaviours and find out about what leads to non-compliance. It is easy for the
13
patients to do and interviews can be done by phonecall. However there are
14
biases with recall and people may say one thing but do another and there can
15
be errors in reporting e.g. self-observation skills.
16
Dunbar (1989) (Dunbar, J., Dunning, E. J., and Dwyer, K., 1989) reviewed the
17
methods to assess adherence to arthritis medication with a review that
18
included ‘16 representative studies of compliance’. No inclusion/exclusion
19
criteria or search details were given.
20
They noted that a major problem is the accuracy of reporting, with poor
21
compliance usually underreported. One issue is the memory decay when
22
assessing adherence (Farr, 1987). Effects, such as not realising the
23
diminishment of higher adherence levels has occurred and moving past
24
events forward in perception can all lead to inaccuracy. Motivational factors
25
are also important, errors in reporting can be due to self-observation skills,
26
especially when the compliance behaviour is itself variable. Misconceptions of
27
the regimen may lead to errors through inaccurately labelling events compliant
28
or noncompliant.
29
Self report has advantages in that it can identify some noncompliance in a
30
cost-efficient manner and permits an in-depth study of the types of errors that
Medicines concordance: full guideline DRAFT (July 2008)
page 216 of 373
DRAFT FOR CONSULTATION
1
patients can make which leads to non-compliance. It also has reasonable
2
sensitivity and specificity in discriminating compliers from non-compliers.
3
In summary, self-reporting can mean poor compliance is underreported, there
4
can be recall bias and self-observation skills may be erroneous. It is cost-
5
efficient and can identify non-compliers.
6
Hecht (1998) (Hecht, F. M., 1998) reported briefly with a narrative review on
7
measures for HIV adherence in clinical practice. Sackett (1975) compared
8
self-report to pill counts. Of those that reported having less than 80%
9
adherence, 95% were found non-adherent by pill count. Those reporting that
10
they were adherent over 80% of the time, 34% were shown to be non-
11
adherent by pill count. Gilbert and Sackett’s studies, suggest that self-report is
12
more accurate than physician assessment. Thus if HCPs want to know if
13
patients are taking ART, they need to ask them rather than relying on their
14
judgement. When they say they are missing medication, believe them, as this
15
is mostly the truth. Patient self-report tends to overestimate adherence. Those
16
who report missing doses infrequently may have a significant problem of non-
17
adherence.
18
Hecht (1998) says that what matters is how HCPs ask the questions. Stating it
19
should be in a specific, non-judgmental way and one that allows them to
20
disclose non-adherence. Therefore, questions should not imply that they are
21
wrong if they do not take their medication the way they are ‘supposed to’. A
22
time period must also be specified. No references given for these
23
conjectures.
24
Self-report is more accurate than physician’s judgement alone. It tends to
25
overestimate adherence. It depends on how the questions are asked and a
26
time period must be specified.
27
28
29
Medicines concordance: full guideline DRAFT (July 2008)
page 217 of 373
DRAFT FOR CONSULTATION
1
8
prescribed medication
2
3
Interventions to increase adherence to
8.1
Recommendations *listed at the start of the guideline
according to chapter
4
5
8.2
Introduction
6
Adherence is defined as ‘the extent to which the patient’s behaviour matches
7
agreed recommendations from the prescriber’ (Horne, R, Weinman, J.,
8
Barber, N. et al , 2005). Adherence describes patient behaviour in the actual
9
taking of medicines. This definition of adherence presumes that the patient
10
has reached some agreement with the health care professional about the
11
prescribed medication. The Guideline Development Group were interested in
12
interventions that would support patients in taking of medicines following
13
agreement with the health professional.
14
Non-adherence can be intentional or unintentional. Non-adherence is
15
unintentional if the patient does not take the medication, for example, due to
16
forgetfulness or not being able to access the medication because of problems
17
with packaging and dexterity. Non adherence is intentional when the patient
18
makes a decision to either not take the medicine as previously agreed or to
19
take it in a way other than recommended by the prescriber because of their
20
own beliefs and appraisals of the medicine and medicine taking. Both
21
intentional and unintentional non-adherence can occur regarding the amount
22
or duration of missed medication e.g. a single dose may be missed, a patient
23
may miss several days of medication or a patient may permanently stop
24
taking medication. In some patients non-adherence takes the form of the
25
patient reducing or increasing the dose of prescribed medication rather than
26
omitting it.
27
Medicines concordance: full guideline DRAFT (July 2008)
page 218 of 373
DRAFT FOR CONSULTATION
1
8.3
Methods
2
The aim of the evidence review is to identify the most relevant, published
3
evidence to answer the key clinical questions generated by the GDG. Due to
4
time constraints, exhaustive systematic reviews (see the Methods of the
5
Cochrane Review) were not undertaken. However, the evidence reviews were
6
undertaken using systematic, transparent approaches following the Guidelines
7
Manual 2007.The following evidence reviews were developed as an update to
8
the 2005 Cochrane Review14 titled: “interventions for enhancing medication
9
adherence”. The Haynes review aimed to summarise the results of
10
randomised controlled trials (RCTs) of interventions to help patients follow
11
prescriptions for medications for medical problems, including mental disorders
12
but not addictions. For short-term treatments, not enough studies were
13
retrieved on any individual disease condition to allow grouping by disease.
14
The main results were: For short-term treatments, four out of nine
15
interventions (reported in eight RCTs) showed an effect on both adherence
16
and at least one clinical outcome, while one intervention reported in one RCT
17
significantly improved patient compliance, but did not enhance the clinical
18
outcome. For short-term treatments, three of the interventions were related to
19
counselling (mainly about the importance of adherence and information on the
20
medicines). For long-term treatments, 26 out of 58 interventions reported in 49
21
RCTs were associated with improvements in adherence, but only 18
22
interventions led to improvement in at least one treatment outcome. Almost all
23
of the interventions that were effective for long term care were complex,
24
including combinations of more convenient care, information, reminders, self-
25
monitoring, reinforcement, counselling, family therapy, psychological therapy,
26
crisis intervention, manual telephone follow-up, and supportive care. Even the
27
most effective interventions did not lead to large improvements in adherence
28
and treatment outcomes. Six studies showed that telling patients about
29
adverse effects of treatment did not affect their adherence.
30
These results led the authors to conclude that for short-term treatments
31
several quite simple interventions increased adherence and improved patient
14
first published in 2002
Medicines concordance: full guideline DRAFT (July 2008)
page 219 of 373
DRAFT FOR CONSULTATION
1
outcomes. Current methods of improving adherence for chronic health
2
problems are mostly complex and not very effective, so that the full benefits of
3
treatment cannot be realised and that further research concerning innovations
4
to assist patients to follow medication prescriptions for long-term medical
5
disorders is required.
6
A more recent version of this Cochrane Review has now been published in
7
2008. The main results were: for short-term treatments, four out of ten
8
interventions reported in nine RCTs showed an effect on both adherence and
9
at least one clinical outcome, while one intervention reported in one RCT
10
significantly improved patient adherence, but did not enhance the clinical
11
outcome. For long-term treatments, 36 out of 81 interventions reported in 69
12
RCTs were associated with improvements in adherence, but only 25
13
interventions led to improvement in at least one treatment outcome. Almost all
14
of the interventions that were effective for long-term care were complex,
15
including combinations of more convenient care, information, reminders, self-
16
monitoring, reinforcement, counselling, family therapy, psychological therapy,
17
crisis intervention, manual telephone follow-up, and supportive care. Even the
18
most effective interventions did not lead to large improvements in adherence
19
and treatment outcomes. In this update the authors also for short-term
20
treatments where several quite simple interventions increased adherence and
21
improved patient outcomes the challenge is that there are inconsistent results
22
from study to study with less than half of studies showing benefits.
23
The evidence in the Cochrane Review was organized by disease, however as
24
this guideline is intended to be a generic document we re-arranged the
25
evidence according to the type of intervention. We also conducted additional
26
specific searches for several questions. This is described below.
27
The titles and abstracts of studies retrieved by electronic searches were
28
scanned for relevance to the topic on interventions to increase adherence.
29
Any potentially relevant publications were obtained in full text. These were
30
then reviewed to identify the most appropriate evidence and were then
31
allocated to the relevant key clinical questions. Following this, the assessment
Medicines concordance: full guideline DRAFT (July 2008)
page 220 of 373
DRAFT FOR CONSULTATION
1
of study quality; synthesis of the results; and grading of the evidence was
2
undertaken.
3
While the GDG were interested in any intervention that might be useful in
4
supporting adherence, the key clinical questions agreed by the GDG included
5
questions on a number of specific interventions. We initially conducted one
6
broad search in order to find evidence on interventions to increase adherence.
7
One broad search allowed us to pick up any intervention designed to increase
8
adherence and then allocate the evidence to the respective clinical question.
9
This reduced the duplication of sifting and reviewing of the evidence. This
10
search was supplemented by specific searches in areas considered important
11
by the GDG where we also included some observational studies. These were
12
areas where the broad search left considerable uncertainty but the
13
interventions were considered either potentially important in clinical practice or
14
areas where popular preconceptions may exist.
15
These were:
 Does change in dosing regime affect adherence to prescribed
16
medication?
17
 Does drug formulation/packaging affect adherence to prescribed
18
medication?
19
 Do medication reviews affect adherence to prescribed
20
medication?
21
 Do prescription charges/costs affect adherence to prescribed
22
medication?
23
 Do dosette boxes affect adherence to prescribed medication?
24
25
26
8.4
Evidence to recommendations : difficulties in
27
interpreting studies on interventions to improve
28
adherence
29
Given the advances of medical therapies and the increase in prescribing and
30
drug use in the last decades, it could be expected that this has been
Medicines concordance: full guideline DRAFT (July 2008)
page 221 of 373
DRAFT FOR CONSULTATION
1
accompanied by a greater understanding of the processes of adherence and
2
non adherence and the effectiveness of interventions to promote adherence.
3
In general however the current body of literature on adherence is of poor
4
methodological quality and was considered inadequate to answer many of the
5
GDG questions on interventions to increase adherence. Medicine taking does
6
not appear to be recognized by many researchers as a complex human
7
behaviour which should be studied using complex interventions in line with the
8
MRC framework (2000) (Medical Research Council, 2000). Specific issues
9
affecting quality of the studies appraised for interventions to promote
10
adherence are outlined below
11
(a) The content and method of delivery of the components of the intervention
12
are not well described and differ in different studies.
13
(b)The lack of distinction between content of intervention and how it was
14
delivered makes it impossible to understand the overall poor and contradictory
15
outcomes
16
(c)The majority of the studies does not assess a single intervention but
17
include multi-component interventions. In a trial the two interventions being
18
compared may each have a different set of components. This means it is not
19
possible to compare one trial with another and even when an intervention
20
works it is not possible to know which component or combination of
21
components is effective.
22
(d) Studies do not report whether the planned interventions took place as
23
intended in the study protocol.
24
(e) No standard method of assessing adherence is used.
25
(f) There is infrequent justification of the relevance of certain interventions in
26
improving adherence e.g. there is no theoretical framework informing the
27
studies and this precludes development of understanding of phenomena of
28
adherence/non-adherence.
Medicines concordance: full guideline DRAFT (July 2008)
page 222 of 373
DRAFT FOR CONSULTATION
1
(h) Studies had frequently small sample sizes and inadequate description of
2
settings and existing rates of adherence; generalising of these interventions to
3
routine clinical practice is therefore not possible.
4
(i) Many of the interventions are extremely complex and labour-intensive and
5
are carried out by paid research staff and thus it is unknown whether they
6
could be carried out in a non-research setting.
7
(i ) Even the most apparently effective interventions did not lead to substantial
8
improvements in adherence and treatment outcomes.
9
(k) Interventions are not targeted to causes of non-adherence – it is likely
10
therefore that some interventions may be more effective than evidence
11
suggests if directed to actual cause of non–adherence in individual patients.
12
(l) Comment is not made on patient involvement in decision to prescribe
13
medicines
14
(m) Few studies mention whether raters of outcome or adherence were blind
15
to whether subjects were in the intervention or control group
16
(n) Despite the comprehensive and detailed searching, some trials that met
17
our criteria may have been missed. The literature on patient adherence is not
18
well indexed as it sprawls across the traditional disease areas and as a result
19
of all these, there is little information available to fully understand why one
20
intervention works and other very similar ones did not. Our concern about the
21
evidence in this area is mirrored in other key reviews and trials (Haynes, R.
22
B., Ackloo, E., Sahota, N. et al , 2008).
23
The results of the reviews need to be interpreted with caution, as some of the
24
elements that have worked within some of the trials are present in other
25
studies that have not yielded significant improvements. We were interested in
26
simple interventions that might be targeted to individuals but the majority of
27
the studies are complex interventions with an extremely wide range of
28
components. Aggregation of studies was extremely difficult and has required
29
subjective judgment.
30
Medicines concordance: full guideline DRAFT (July 2008)
page 223 of 373
DRAFT FOR CONSULTATION
1
2
3
8.5
Key Clinical Question: Does change in dosing regime
affect adherence?
Related references
Evidence statements
Evidence into recommendations
(summary of evidence)
Shi (2007) (Shi, L.,
1.Systematic reviews and
For general discussion of limitations
Hudges, M., Yurgin, N. et
RCTs show that
evidence see section 7.3
al , 2007); Schroeder
simplification of dosing
(2004) (Urien, A. M.,
frequency can increase
Guillen, V. F., Beltran, D.
adherence to prescribed
O. et al , 2004); Iskedjian
medication.
(2002) (Iskedjian, M.,
Einarson, T. R.,
MacKeigan, L. D. et al ,
2002); Claxton (2001)
(Claxton, A. J., Cramer, J.,
and Pierce, C., 2001)
The GDG were interested in whether
there was evidence to indicate that
changes in dosing regime would imp
adherence. The findings of the evide
review were that reducing the comple
of a regime can increase adherence
the quality of evidence was low. The
evidence from the qualitative intervie
indicated that the difficulty for patient
integrating the regime into their lives
Baird (1984) (Baird, M. G.,
rather than dose complexity per se a
Bentley-Taylor, M. M.,
the GDG recommendation is that
Carruthers, S. G. et al ,
changes to dosing regime need to be
1984); Brown (1997)
tailored to needs of individual patient
(Brown, B. G., Bardsley, J.,
Poulin, D. et al , 1997);
Girvin (1999) (Girvin, B.,
McDermott, B. J., and
Johnston, G. D., 1999);
Portsmouth 2005 [1216};
Molina (2007) (Molina,
Jean Michel, Podsadecki,
Thomas J., Johnson,
Medicines concordance: full guideline DRAFT (July 2008)
page 224 of 373
One area of interest for the GDG was
use of drugs by injection particularly
antipsychotic drugs and contraceptive
This option could be classified as
changes to dose regime or drug
formulation. The GDG were clear tha
using drugs by injection in this way m
be an appropriate choice where patie
DRAFT FOR CONSULTATION
Margaret A. et al , 2007)
have non-intentional adherence i.e. t
forget to take their medicines. As su
Schroeder (2004) (Shi, L.,
2. Two Systematic reviews
Hudges, M., Yurgin, N. et
and four RCTs show that
al , 2007); Iskedjian (2002)
reducing twice-daily to
(Iskedjian, M., Einarson, T.
once-daily dosing may
R., MacKeigan, L. D. et al ,
increase adherence to
2002); Portsmouth 2005;
prescribed medication.
Molina (2007) (Molina,
this choice should be offered to patie
The GDG were aware of evidence fro
qualitative synthesis that patients with
mental health problems can feel coer
to take medication and wished it mad
clear that the aim of our
recommendations is to support inform
Jean Michel, Podsadecki,
adherence.
Thomas J., Johnson,
Margaret A. et al , 2007);
Baird (1984) (Baird, M. G.,
Bentley-Taylor, M. M.,
Carruthers, S. G. et al ,
1984); Girvin (1999)
Claxton (2001) (Claxton,
3. Evidence from one low
AL, Cramer, J, and Pierce,
quality systematic review
C, 2001); Rudd (2004)
and two RCTs showed
(Rudd, P., Miller, N. H.,
that once daily dosing
Kaufman, J. et al , 2004);
compared to twice daily
Parienti (2007) (Parienti,
did not increase
Jean Jacques, Massari,
adherence to prescribed
Véronique, Reliquet,
medication.
Véronique et al , 2007).
1
Methods of the evidence review
2
8.5.1
3
This paper includes a narrative summary of the included evidence, structured
4
according to the category of the intervention, following the agreed reviewing
5
protocol:
Medicines concordance: full guideline DRAFT (July 2008)
page 225 of 373
DRAFT FOR CONSULTATION
1
Types of studies – We initially included only randomized controlled trials
2
(RCTs) of interventions to increase adherence. The excluded studies list from
3
the Cochrane review was checked as we have included those studies with
4
less than 80% follow-up of participants. As with the Cochrane review we found
5
only a small number of studies that fulfilled our criteria. For this evidence
6
review we excluded any randomised controlled trials that evaluated changes
7
in dosing regimes but did not assess the same medication in all comparative
8
groups. The GDG requested further search to pick up any systematic review
9
published after the Cochrane Review search cut-off. However, the included
10
systematic reviews did not follow this criterion.
11
Types of participants - people prescribed medication for a medical condition.
12
Duration of studies - six months follow up from the time of patient entry for
13
long-term regimens for the RCTs. No time limit specified for short-term
14
conditions.
15
Types of interventions - any interventions intended to change adherence to
16
prescribed medication. As the Cochrane review is presented by condition, we
17
have used the evidence extracted in that review and reconfigured it by
18
intervention.
19
Types of outcome measures – inclusion criteria (as defined in the Cochrane
20
review) were expanded by including studies that used adherence as the only
21
outcome variable as opposed to adherence and treatment outcome variables.
22
The excluded studies list of the Cochrane review was cross-referenced to
23
ensure that no potentially relevant study was missed out.
24
8.5.2
25
8.5.2.1
26
Systematic Reviews
27
One recent review of systematic reviews and empirical studies (Shi 2007)
28
(Shi, L., Hudges, M., Yurgin, N. et al , 2007) looked at the impact of dose
29
frequency on compliance and health outcomes, particularly for injectables.
Evidence review
Does change in dosing regime affect adherence?
Medicines concordance: full guideline DRAFT (July 2008)
page 226 of 373
DRAFT FOR CONSULTATION
1
Inclusion criteria were that the studies should compare different dose
2
frequencies, including injectable medications and be published in a peer
3
reviewed journal. Exclusion criteria were that the article should not focus
4
solely on dosage forms, dose administration or dose timing.
5
Full text reviews were conducted on a total of 64 empirical studies and 25
6
literature/systematic reviews. No details were given on overall methodological
7
quality of the studies.
8
Results were presented through five main areas: cardiovascular diseases,
9
diabetes, nephrology/urology, neurology/psychiatry and rheumatoid/muscle.
10
Of the 21 studies that measured compliance, 17 reported a positive impact
11
(no details of significance given) of reducing dose frequency on compliance,
12
whilst inconclusive results were seen in four. Details of the dose frequency
13
reductions contained in the studies were not provided by the review.
14
Articles not measuring compliance as the main outcome looked at efficacy
15
and other outcomes of extended-release medications in comparison to the
16
immediate-release forms. The studies also supported the general benefits of
17
reducing dosing frequency on improved quality of life or patients’ satisfaction
18
(6 studies), greater control over side effects (5 studies) and improved
19
economic outcomes using extended-release formulation (2 studies).
20
Schroeder (2004) (Urien, A. M., Guillen, V. F., Beltran, D. O. et al , 2004)
21
used the Cochrane methodology to review dosing regimes and adherence in
22
hypertensive patients. The methodological quality of the primary included
23
studies was assessed to be generally low. Many RCTs showed marked
24
heterogeneity in terms of participants, interventions and outcomes. A pooled
25
analysis was considered inappropriate as results on adherence were reported
26
in many different ways. Simplifying dosing regimens improved adherence in 7
27
of 9 studies with relative improvement in adherence increasing by 8% to
28
19.6%. All of the studies that used objective outcome measurement
29
(Medication Event Monitoring System) showed an improvement in adherence
30
through the use of once daily instead of twice daily dosing regimens, although
31
4 of these compared 2 different drugs. Only 1 study showed an increase in
Medicines concordance: full guideline DRAFT (July 2008)
page 227 of 373
DRAFT FOR CONSULTATION
1
adherence (905 vs. 82%; p<.01) together with a reduction in systolic blood
2
pressure of 6 mm Hg (p<.01).
3
A weak meta-analysis conducted by Iskedjian (2002) (Iskedjian, M.,
4
Einarson, T. R., MacKeigan, L. D. et al , 2002), combined comparative studies
5
of different research designs including prospective trials (RCTs and cohort
6
studies), retrospective chart reviews and database analyses. Adherence was
7
defined differently in various studies and different instruments were used to
8
measure patient adherence. In this meta-analysis, all variables that could
9
affect adherence other than daily dose frequency, were assumed to be equal
10
between comparators.
11
Eight studies involving a total of 11,465 observations were included (1830 for
12
daily [QD] dosing, 4405 for twice a day dosing [BID] and 4147 for dosing >2
13
times daily [>BID] and 9655 for multiple daily dose [MDD]). The primary
14
objective was to assess adherence. The average adherence rate for QD
15
dosing (91.4%, SD=2.2%) was significantly higher than for MDD (83.2%,
16
SD=3.5%; p<0.001). The difference between adherence rates for QD dosing
17
(92.7%) and BID dosing (87.1%) was also statistically significant (p=0.026),
18
although the difference in this analysis was smaller than in the QD-versus-
19
MDD analysis (5.7% vs. 8.2%). The difference in adherence rates between
20
BID dosing (90.8%, SD 4.7%) and >BID dosing (86.3%, SD=6.7%) was not
21
significant (p=0.069). However, a subgroup analysis using a stricter definition
22
of adherence (≥90% intake) did reveal a statistically significant difference
23
between BID and >BID dosing (respective adherence rates of 76.1% and
24
67.0%, p<0.001).
25
Another systematic review published by Claxton (2001) (Claxton, A. J.,
26
Cramer, J., and Pierce, C., 2001) also found that simpler, less frequent dosing
27
regimes resulted in better compliance. This systematic review appeared to
28
include several study designs. This review showed strong methodological
29
limitations particularly in terms of data analysis. The study did not give full
30
details of inclusion/exclusion criteria and thus possibly including studies that
31
compared different dosing regimes in different medicines. The results should
32
therefore be viewed with caution.
Medicines concordance: full guideline DRAFT (July 2008)
page 228 of 373
DRAFT FOR CONSULTATION
1
Seventy-six studies were included in the review. By combining all data it was
2
found that increasing the number of daily doses was significantly related to a
3
decline in compliance (p<0.001 among dose schedules). Comparisons
4
between dose regimens showed that compliance was significantly higher with
5
once daily regimens vs. 3 times daily (p=0.008) or 4 times daily regimens
6
(p<0.001). Compliance with twice daily dosing was significantly higher than 4
7
times daily dosing (p=0.001). There were no significant differences in
8
compliance between once daily and twice daily regimens or between twice
9
daily and 3 times daily.
10
Randomised controlled trials
11
Three RCTs Baird (1984) (Baird, M. G., Bentley-Taylor, M. M., Carruthers,
12
S. G. et al , 1984); Brown (1997) (Brown, B. G., Bardsley, J., Poulin, D. et al ,
13
1997) and Girvin (1999) (Girvin, B., McDermott, B. J., and Johnston, G. D.,
14
1999) from the Cochrane Review (2005) , assessed the effect of the
15
simplification of a dosing frequency. Baird (1984) (Baird, M. G., Bentley-
16
Taylor, M. M., Carruthers, S. G. et al , 1984) compared twice a day 100mg
17
Betaloc tablets to once daily 200mg Betaloc Durules, in a sample comprising
18
389 participants. Mean age of the participants was 52.7 years for the twice
19
daily group, compared to 54.5 years. Over the total study period, compliance
20
exceeded 80% in 96.4% of patients in the Durules group and 90.0% of patient
21
in the Betaloc tablets group (p=0.0591). When to 90% levels of compliance
22
were compared, overall compliance exceeded this level in 92.8% of patients
23
on Durules and in only 81.5% of patients on tablets (p=0.009). A significant
24
effect in increasing adherence was reported. Brown (1997) (Brown, B. G.,
25
Bardsley, J., Poulin, D. et al , 1997) tested controlled-release niacin twice
26
daily to regular niacin, four times daily, in the treatment of hyperlipidemia and
27
coronary artery disease, in 29 male participants aged ≤65 years. Compliance
28
was 95% with the controlled- release niacin versus 85% with regular niacin
29
(p< 0.001). Girvin (1999) (Girvin, B., McDermott, B. J., and Johnston, G. D.,

Study information indicates that duration is less than 6 months, however this is not stated in Cochrane
Review.

Study with less than 6 months duration that was included in the Cochrane Review as results for blood
pressure outcomes were negative.
Medicines concordance: full guideline DRAFT (July 2008)
page 229 of 373
DRAFT FOR CONSULTATION
1
1999) tested enalapril 20 mg once daily versus enalapril 10 mg twice daily in
2
the treatment of high blood pressure. Sample size comprised of 27 patients.
3
Mean age of participants was 62 years. Overall medication adherence was
4
improved with once-a-day dosing. The difference in percentage of doses
5
taken by pill count between the two periods was significantly in favour of the
6
once daily regimen (p<0.01), as was the percentage of doses taken as
7
measured by a pill container that recorded lid openings (MEMS) (p< 0.001).
8
One RCT Portsmouth (2005) (Portsmouth, S. D., Osorio, J., McCormick, K.
9
et al , 2005) included in the 2008 revision of the Cochrane Review assessed
10
whether virologically controlled HIV-1-infected individuals switched from a
11
twice-daily antiretroviral regimen to a once daily regimen demonstrate
12
improved adherence and quality of life while maintaining virological control.
13
Forty-three patients were included in this study, with 22 in the once daily
14
(intervention) group, and 21 in the twice daily (control) group.
15
The once daily group (intervention): the prolonged release capsule group
16
(PRC) were assigned to take d4T PRC/3TC/EFV all once-daily (24 h apart);
17
Twice daily (control group): participants in the control group were assigned to
18
continue either d4T IR/3TC/EFV or Combivirs/EFV as per their screening
19
regimen. Note: participant weighing less than 60 kg were prescribed either 30
20
mg of d4T IR or 75 mg of d4T PRC.
21
After randomization, patients allocated to the PRC (intervention) maintained
22
this high adherence, while those allocated to IR (control) showed a
23
significantly reduced adherence in ‘taking compliance’ (P=0.0237)
24
(percentage of prescribed number of doses taken), ‘correct dosing
25
compliance’ (P=0.0104) (percentage of days with correct number of doses
26
taken) and ‘timing compliance’ (P=0.028) (percentage of doses taken within 3
27
hours of the prescribed dosing intervals) at both weeks 12 and 24.
28
One RCT, Rudd (2004) (Rudd, P., Miller, N. H., Kaufman, J. et al , 2004),
29
from the Cochrane Updated Review that was included in the evidence review
30
on the effects of self-monitoring on adherence reported some results in regard
31
to once-daily regimens compared to twice-daily regimens. This RCT
Medicines concordance: full guideline DRAFT (July 2008)
page 230 of 373
DRAFT FOR CONSULTATION
1
assessed a system for patients to monitor their own blood pressure. Seventy
2
six patients received routine care while the intervention group (n=74) received
3
an automated blood pressure device for use at home with management by a
4
nurse care manager. The patients recorded their own blood pressure then the
5
device printed these which were mailed to the nurse care manager in order to
6
guide drug therapy. The adherence measures by a drug event monitor was
7
found to be significant (80% for the intervention group and 69% for the control
8
group, p=0.03). One of the outcomes found that once-daily regimens had
9
higher adherence 82% (SD=28%) than twice-daily 69% (SD=34%) or more
10
frequently 49% (SD=41%). None of these differences were statistically
11
significant.
12
Update searches
13
From the conducted update searches, we retrieved two RCTs that were
14
considered important as they would contribute to modifying the
15
recommendations drafted for the topic of the impact of changes of dosing
16
regimes on adherence to prescribed medication. These were Molina (2007)
17
and Parienti (2007).
18
The safety, efficacy and adherence to lopinavir/ritonavir (LPV/r) dosed QD or
19
BID in antiretroviral-naive, HIV-1-infected subjects was evaluated in an RCT
20
Molina (2007) (Molina, Jean Michel, Podsadecki, Thomas J., Johnson,
21
Margaret A. et al , 2007). A randomized, open-label, multicenter comparative
22
study was conducted through 96 weeks of treatment. A total of 190
23
antiretroviral-naive subjects with plasma HIV-1 RNA above 1000 copies/ml
24
and any CD4(+) T cell count were enrolled. Subjects were randomized (3:2) to
25
LPV/r 800/200 mg QD (n = 115) or 400/100 mg BID (n = 75). Subjects
26
received TDF 300 mg and FTC 200 mg QD. Adherence to LPV/r through 96
27
weeks was measured using MEMS((R)) monitors. Median baseline VL and
28
CD4(+) T cell count were 4.8 log(10) copies/ml and 216 cells/mm(3),
29
respectively. Prior to week 96, 37% (QD) and 39% (BID) of subjects
30
discontinued, primarily due either to adverse events (17% QD, 9% BID) or to
31
loss to follow-up or nonadherence (12% QD, 17% BID). The proportion of
32
subjects with VL <50 copies/ml (57% QD, 53% BID; p = 0.582 (ITT NC = F)),
Medicines concordance: full guideline DRAFT (July 2008)
page 231 of 373
DRAFT FOR CONSULTATION
1
change in CD4 count (244 cells/mm(3) QD, 264 cells/mm(3) BID; p = 0.513),
2
and evolution of resistance did not differ between groups through 96 weeks.
3
Diarrhoea (17% QD, 5% BID, p = 0.014) was the most common moderate or
4
severe, study drug-related adverse event. Adherence to LPV/r was higher for
5
the QD group than the BID group and declined over time in both groups.
6
Parienti (2007) (Parienti, Jean Jacques, Massari, Véronique, Reliquet,
7
Véronique et al , 2007) aimed to determine the effect of once-daily dosing on
8
adherence to nevirapine. This RCT was comprised of three-phase (3-month
9
observational, 4-month randomized, 5- month interventional) open-label,
10
clinical trial at four French academic medical centres during 2005-2006
11
among 62 chronically HIV-1- infected subjects with long-lasting viral
12
suppression under a twice-a- day nevirapine-based antiretroviral combination.
13
Participants were randomly assigned to switch to nevirapine 400 mg once-
14
daily (n = 31) or continue nevirapine 200 mg twice-a-day (n = 31). After the
15
randomized phase, participants had an opportunity to choose their
16
antiretroviral dosage.
17
Fifty-two patients qualified for electronic data analysis. During the randomized
18
phase, the mean adherence rate was non-significantly superior by 0.5% in
19
once-daily versus twice-a-day dosing (P = 0.68), adjusting for previous twice-
20
a-day adherence rate (P < 0.0001). Once-daily group increased days without
21
dose (odds ratio (OR) 1.7; 95% confidence interval (CI) 1.0, 2.8; P = 0.04),
22
adjusting for previous drug interruptions (P < 0.0001). In the longitudinal
23
analysis, once-daily dosing was significantly associated with at least two
24
consecutive days without dose (OR 4.4; 95% CI 1.9, 10.3; P < 0.001). The
25
authors concluded that changing from twice to once-daily nevirapine did not
26
improve adherence
27
28
Medicines concordance: full guideline DRAFT (July 2008)
page 232 of 373
DRAFT FOR CONSULTATION
1
2
8.6
Key Clinical Question : Effect of prescription
charges/costs on adherence to prescribed medication
3
Related
Evidence statements
references
(summary of evidence)
Hirth (2008)
1. Some UK patients may have
Most of the evidence for cost as a barrier
(Hirth, Richard
difficulty affording medicines.
to adherence comes from US. Only a few
A., Greer, Scott
L., Albert,
Justin M. et al ,
2008); Atella
(2005) (Atella,
Vincenzo,
Schafheutle,
Ellen, Noyce,
Peter et al ,
2005)
Evidence into recommendations
studies have been conducted in the UK.
2. The most common strategies
for patients with problems
affording medicines is to delay
the dispensing of medicines, to
These indicate that for some patients this
is a concern. Cost concerns may also
indicate doubts by the patient about the
value of the prescription.
not visit the GP and to lower the
dose below that prescribed to
When cost is a concern for patients a
extend the duration of the
variety of options are available each of
prescription.
which have advantages and
disadvantages. Prescription length may be
increased giving the patient longer
prescription for same cost but this may
reduce opportunity for review. Quite short
dispensing time frames may be important
for example when patients are suicidal or
need careful monitoring of medication and
its effects. Instalment dispensing is
possible for certain drug items but in
general it seems unreasonable to ask a
patient to pay for each dispensing point.
However the GDG was also mindful of the
fact, that it might also be unreasonable for
the pharmacist to make serial dispensings
for a single dispensing fee. Costs and
value of prescriptions should be
Medicines concordance: full guideline DRAFT (July 2008)
page 233 of 373
DRAFT FOR CONSULTATION
considered not just at the point of
prescribing but at all stages of the process
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
3
Medicines concordance: full guideline DRAFT (July 2008)
page 234 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
8.6.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies – no restrictions on study design. However, due to the
6
nature of the question, the requirement was that the studies needed to be
7
conducted in the UK.
8
Types of participants- people prescribed medication for a medical condition.
9
Duration of studies - no time limit specified.
10
Types of interventions - any interventions intended to assess the correlation
11
between prescription charges/costs and the impact on adherence to
12
prescribed medication.
13
8.6.2
14
Types of outcome measures – adherence to prescribed medication, cost
15
reducing strategies.
16
We retrieved one observational study (Hirth 2008) (Hirth, Richard A., Greer,
17
Scott L., Albert, Justin M. et al , 2008) that examined out of pocket medication
18
spending and cost-related medication nonadherence among dialysis patients
19
in twelve countries including the UK.
20
Data were gathered from 2002 to 2004 as part of the dialysis outcomes and
21
practice patterns study (DOPPS), an observational study of haemodialysis
22
practices and outcomes in twelve countries- Australia, New Zealand, Belgium,
23
Canada, France, Germany, Italy, Spain, Sweden, United Kingdom, Japan,
24
and the United States. A random sample of patients was selected, totalling
25
N=7.766. Of the selected 83 per cent who agreed to enrol and have their
26
medical records abstracted, 85 per cent of these enrolled patients also
27
completed the patient questionnaire. A total of 70 per cent of patients provided
28
both medical and questionnaire data. Local currencies were converted to US
29
Dollars.
Evidence review
Medicines concordance: full guideline DRAFT (July 2008)
page 235 of 373
DRAFT FOR CONSULTATION
1
Questionnaires and medical record abstraction techniques were standardised
2
across countries and languages. Patient questionnaires were administered
3
soon after recruitment. They were asked about the total out-of-pocket
4
spending for prescription and over the counter (OTC) medications in the
5
previous month. They were also asked “Do you sometimes decide not to
6
purchase medications because of cost?” and to report their out-of-pocket
7
spending for haemodialysis treatments.
8
Whilst the United States reported 86 per cent of out-of-pocket spending for
9
medications, only patients in Australia/New Zealand, Belgium, and Sweden
10
were significantly more likely to face out-of-pocket spending, while those in
11
France, Japan, Spain and the UK were significantly less likely to do so.
12
Mean monthly spending for prescription and OTC medications ranged from $8
13
in the UK to $114 to the United States. Among patients with medication
14
spending, only 10 per cent faced monthly costs greater than $30 in the United
15
Kingdom, whereas 10 per cent incurred costs greater than $310 in the United
16
States.
17
Observed cost-related nonadherence, indicated by the proportion of patients
18
who reported that they sometimes did not purchase medications because of
19
cost, was significantly less than expected in France, Japan, Spain, Sweden
20
and the UK.
21
Nonadherence was associated with the percentage of patients reporting any
22
out-of-pocket spending and the average out-of-pocket cost. Although the US
23
had high out-of-pocket spending burdens, their nonadherence was still clearly
24
higher than would be expected on the basis of the percentage facing any
25
costs or the mean cost burden. On the other hand, Sweden and Belgium had
26
lower levels of nonadherence than would be expected given either measure of
27
out-of-pocket spending burden. The lowest nonadherence rates existing in
28
France, Japan, Spain and the UK were correlated with low out-of-pocket
29
spending.
30
Atella 2005 (Atella, Vincenzo, Schafheutle, Ellen, Noyce, Peter et al , 2005)
31
aimed to explore how and to what extent costs incurred by patients influence
Medicines concordance: full guideline DRAFT (July 2008)
page 236 of 373
DRAFT FOR CONSULTATION
1
their decision-making behaviour in accessing medicines, both in the UK and in
2
Italy.
3
Based on findings from focus groups, a questionnaire was designed to assess
4
medication cost issues. As such, several hypotheses were tested regarding
5
patients’ decision-making behaviour and how it was influenced by health and
6
sociodemographic status and the novel concept of a self-rated affordability
7
measure. Patients were eligible if they had either dyspepsia or mild
8
hypertension. They were sampled as successive patients who visited 51
9
physicians in Italy and 21 community pharmacists in the UK. Samples were
10
drawn from the areas of Manchester and Rome. Of the 550 dyspepsia and
11
600 hypertension questionnaires distributed, 122 and 153 were returned- a
12
response rate of 22.2% and 25.5%, respectively. In the UK, 296 dyspepsia
13
and 277 hypertension questionnaires were distributed, targeting dyspepsia
14
patients who bought OTC medicines, and dyspepsia and hypertension
15
patients who had to pay prescription charges; 110 dyspepsia and 134
16
hypertension questionnaires were returned, giving a response rates of 37.5%
17
and 48.4%. In both countries the majority of the respondents were not
18
exempt.
19
The self-rated affordability measure showed that 70.3 per cent of the UK
20
sample and 66.5 percent of the Italian sample had to think about the cost of
21
medicines at least sometimes. Also, 24.3 per cent and 16.3 per cent,
22
respectively said they always have to think about how much money they have
23
available to spend when they obtain medicines. According to the results, the
24
patient initiated strategy most commonly used by UK respondents with
25
affordability problems is (1) to delay the dispensing of drugs until they get
26
paid, (2) not visiting the GP to avoid incurring the cost of prescribed
27
medication and (3) reducing the dose below that prescribed to extend the
28
course of medication.
29
Affordability issues were also strong when examining the use of self-
30
medication strategies. The UK respondents were particularly cost conscious
31
when considering the price of an OTC product before buying it, or they would
32
ask for something cheaper if they could not afford a particular OTC product.
Medicines concordance: full guideline DRAFT (July 2008)
page 237 of 373
DRAFT FOR CONSULTATION
1
The authors point out that affordability seemed to play a more important role
2
in the UK sample than in the Italian, however they do point out that Italian
3
patients with dyspepsia were sampled only through GPs and may be those
4
more severely affected and/or less likely to be disposed towards self
5
medication. Also, OTC products are much more expensive in relation to the
6
prescription charge that they are in the UK where the prescription charge is
7
high.
8
9
10
11
12
13
14
Medicines concordance: full guideline DRAFT (July 2008)
page 238 of 373
DRAFT FOR CONSULTATION
1
2
8.7
Key Clinical Question: Does drug formulation and/or
packaging affect adherence?
3
Related references
Evidence statements (summary
Evidence into
of evidence)
recommendations
For general discussion of
Drug packaging
limitations of evidence see
Orton (2005) (Orton, L. and
1. Two systematic reviews do not
Barnish, G., 2005) and
convincingly support the use of
Connor (2004) (Connor, J.,
unit-dose packaging to improve
The GDG considered the
Rafter, N., and Rodgers,
adherence to prescribed
evidence review did not
A., 2004)
medication
provide convincing
section 7.3
evidence that drug
packaging per se increases
Lee (2006)(Lee, J. K.,
2. There is conflicting evidence on
adherence. Health care
Grace, K. A., and Taylor, A. the value of special packaging in a
professionals should
J., 2006); Becker
multicomponent intervention in
explore with patients
(1986)(Becker, L. A.,
increasing adherence to
whether the way in which a
Glanz, K., and Sobel, E.,
prescribed medication.
drug is packaged causes
1986); Henry (1999)(Henry,
difficulty and respond to
A. and Batey, R. G., 1999)
individual problems.
Schneider (2008)
3. One RCT found that the use of
(Schneider, P. J., Murphy,
blister packaging (Pill Calendar)
J. E., and Pedersen, C. A.,
compared to medication in a bottle
2008)
significantly increased adherence.
Lee (2006)(Lee, J. K.,
4. One RCT showed that special
Grace, K. A., and Taylor, A. packaging in a multicomponent
J., 2006)
intervention given to an elderly
population (≥65 years) may
Medicines concordance: full guideline DRAFT (July 2008)
page 239 of 373
DRAFT FOR CONSULTATION
increase adherence to prescribed
medication.
Drug formulation
Bangalore (2007)
5. One highly biased systematic
For general discussion of
(Bangalore, Sripal,
review suggests that fixed dose
limitations of evidence see
Kamalakkannan, Gayathri,
combination compared to free drug section 7.3
Parkar, Sanobar et al ,
component regimen may increase
2007)
adherence to prescribed
medication.
The GDG considered the
evidence review did not
provide convincing
Brown 1997(Brown, B. G.,
6. One RCT showed that
evidence that changes to
Bardsley, J., Poulin, D. et al controlled release medication
drug formulation will
, 1997)
along with simplified dosing
improve adherence.
compared to regular medication
Changes to drug
may increase adherence to
formulation should be
prescribed medication.
considered with changes to
dosing as a response to
individual patient problems
only.
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements, feasibility,
patient perspective
Medicines concordance: full guideline DRAFT (July 2008)
page 240 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
8.7.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies - We initially included only randomized controlled trials
6
(RCTs) of interventions to increase adherence. The excluded studies list from
7
the Cochrane review was cross-referenced as we included studies with less
8
than 80% follow-up of participants. After the GDG voiced concerns over the
9
possibility of missing out important studies by only having included a small
10
amount of studies, the search was redone to pick up any systematic review
11
published after the Cochrane Review search cut-off.
12
Types of participants - people prescribed medication for a medical condition.
13
Duration of studies - six months follow up from the time of patient entry for
14
long-term regimens for the RCTs. No time limit specified for short-term
15
conditions.
16
Types of interventions - any interventions intended to change adherence to
17
prescribed medication. As the Cochrane review is presented by condition, we
18
have used the evidence extracted in that review and reconfigured it by
19
intervention.
20
Types of outcome measures - inclusion criteria (as defined in the Cochrane
21
review) were expanded by including studies that used adherence as the only
22
outcome variable as opposed to adherence and treatment outcome variables.
23
The excluded studies list of the Cochrane review was cross-referenced to
24
ensure that no potentially relevant study was missed out.
25
8.7.2
26
8.7.2.1
27
Systematic Reviews
Evidence review
Effect of drug packaging on adherence
Medicines concordance: full guideline DRAFT (July 2008)
page 241 of 373
DRAFT FOR CONSULTATION
1
One Cochrane Review Orton (2005) (Orton, L. and Barnish, G., 2005) that
2
aimed to assess the effects of unit-dose packaged treatment and treatment
3
adherence in people with uncomplicated malaria was retrieved. Any type of
4
programme that included unit-dose packaging of antimalarial drugs packed in
5
units of a single dose was incorporated in the review. Treatment adherence
6
was a secondary outcome, however all four included studies measured it.
7
Interventions and control arm groups had to received the same antimalarial
8
drug and any other intervention. The interventions that were assessed in this
9
systematic review ranged from labelled and boxed blister packs of chloroquine
10
and primaquine tablets and capsules, simple labelled and sectioned polythene
11
bags of chloroquine tablets, tablets or capsules in paper envelopes or loose
12
and chloroquine syrup in bottles.
13
Three quasi RCTs and one cluster RCT met the inclusion criteria, and overall
14
trials were of poor methodological quality.
15
A meta-analysis of two trials (with 596 participants) showed that participant
16
reported treatment adherence was higher with blister-packed tablets
17
compared with tablets in paper envelopes (RR 1.18, 1.12 to 1.25). Two trials
18
using tablets in sectioned polythene bags as the intervention also reported an
19
increase in participant reported treatment adherence: in one study (cluster
20
RCT) it was compared with the tablets in paper envelopes whilst the other trial
21
compared it with syrup in bottles (RR 2.15, 1.76 to 2.61; 299 participants).
22
It appears that unit-dose packaging drugs (in combination with prescriber
23
training and patient information) was associated with higher participant
24
reported treatment adherence, however this conclusion is drawn from trials
25
with methodological limitations.
26
The Bulletin of the World Health Organization published a systematic review
27
in Connor (2004) (Connor, J., Rafter, N., and Rodgers, A., 2004) which
28
evaluated the evidence on fixed dose combination pills and unit-of use
29
packaging with relation to adherence. Fifteen randomized or quasi
30
randomized trials met inclusion criteria: fixed dose on cure combination pills
31
were investigated in three of these while unit-of-use packaging was studied in
Medicines concordance: full guideline DRAFT (July 2008)
page 242 of 373
DRAFT FOR CONSULTATION
1
12 trials. Eligible studies varied substantially in their settings, participant
2
selection, medical conditions, interventions, adherence measures and clinical
3
outcome measures, as well as in study quality. For these reasons, a meta-
4
analysis was not undertaken. The results of the trials suggested that there
5
were trends towards improved adherence which reached statistical
6
significance in seven out of thirteen trials reporting medication adherence.
7
Measures of adherence were however heterogeneous and interpretation was
8
further limited by methodological issues, particularly small sample size, short
9
duration and loss to follow up. The authors therefore concluded that
10
uncertainty remains about the size of the benefits of drug formulation and
11
packaging.
12
Randomised Controlled Trials
13
Lee (2006) (Lee, J. K., Grace, K. A., and Taylor, A. J., 2006) compared a
14
comprehensive pharmacy care program which was delivered to one group for
15
3-8 months and a second group for 3-14 months in 200 patients aged 65
16
years or over, taking 4 or more chronic medications daily with positive results.
17
The first group returned to usual care after 8 months. The care program
18
consisted of 3 elements, including individualised medication education,
19
medications dispensed using an adherence aid (blister packs) and regular
20
follow-up with clinical pharmacists every 2 months. This study was conducted
21
in the USA.
22
Mean baseline adherence overall was 61.2% (SD 13.5%) with an overall level
23
of adherence of 96.9% at 8 months of intervention. At 14 months, medication
24
adherence was 95.5% (SD 7.7%) in the continued intervention group and
25
69.1% (SD 16.4%) in the control group (p < 0.001). Proportions of people who
26
had at least 80% adherence rates were 97.4% in the intervention group and
27
21.7% in the control group (p < 0.001).
28
Henry (1999) (Henry, A. and Batey, R. G., 1999) from the Cochrane Review
29
delivered an intervention where verbal advice on medication use and possible
30
side-effects were employed along with information sheets on the treatments
31
and medication with dose-dispensing unit. Control group were only given
Medicines concordance: full guideline DRAFT (July 2008)
page 243 of 373
DRAFT FOR CONSULTATION
1
treatment, along with verbal advice and information sheets. A total of 119
2
patients were studies. Mean age of patients was 58 years for the control
3
group and 57 for the intervention group. Compliance in intervention group
4
patients was also encouraged by a phone call 2 days after the start of therapy.
5
This study was conducted In Australia. No significant effects on adherence
6
were reported.
7
Becker (1986) (Becker, L. A., Glanz, K., and Sobel, E., 1986) from the
8
Cochrane review delivered an intervention whereby patients aged 20 to 80
9
years were assigned to the experimental group received all their medications
10
in the special packaging format (all pills taken together were packaged in a
11
single plastic blister sealed with a foil backing on which was printed the day of
12
the week and the time of day at which each medication was to be taken). One
13
hundred and eighty patients were included in the study. Patients in the control
14
group received all of their antihypertensive medications in the conventional pill
15
vials (separate vials for each pill that were labelled with the drug name, the
16
dosage, the medication instructions, and the physician’s name). All
17
medications for both groups were provided free of charge to ensure that all
18
patients would receive their medications. This study was conducted in the
19
USA. No significant effects on adherence were reported.
20
Schneider (2008) (Schneider, P. J., Murphy, J. E., and Pedersen, C. A.,
21
2008) conducted a randomised controlled trial to assess the impact of one
22
medication packaging type on adherence and treatment outcomes of older
23
patients. The study was conducted at 3 sites in Tucson and Columbus in the
24
USA. 85 participants aged 65 years or older, prescribed lisinopril
25
(antihypertensive medication) were randomised to receive daily-dose blister
26
packaged medication (pill calendar) as the intervention compared to traditional
27
bottles of loose tablets as the control group. Patients returned for refills every
28
28 days during a 12 month period where the pharmacist would record the time
29
between prescription refills for the medication and any study-related problems.
30
At 6 and 12 months after enrolling the patients visited the physician to find out
31
blood pressure management; the occurrence of morbidity in the past 6 months
32
e.g. angina, myocardial infarction and stroke; and any medical services they
Medicines concordance: full guideline DRAFT (July 2008)
page 244 of 373
DRAFT FOR CONSULTATION
1
had required in the past 6 months e.g. hospitalisations or emergency
2
department visits. Medical charts were reviewed by two pharmacists to gather
3
this information. The percentage of times prescriptions were refilled on time
4
(within 5 days before or after due date) were significantly higher 80.4%
5
(SD=21.2) for the intervention group than the control group, 66.1% (SD=28),
6
p=0.012. The Medication possession rate (the sum of the day’s supply for all
7
prescriptions received during the study divided by the number of days
8
between the first and last prescription dispensed) was also significantly higher
9
for the intervention group, 0.93 (SD=11.4) and 0.87 (SD 14.2) for the control
10
group, p=0.039. No differences were found between the groups for systolic
11
blood pressure and diastolic blood pressure measures.
12
13
8.7.2.2
Does drug formulation affect adherence
14
Systematic Reviews
15
The possibility of bias was assessed to be high in a systematic review by
16
Bangalore (2007) (Bangalore, Sripal, Kamalakkannan, Gayathri, Parkar,
17
Sanobar et al , 2007) which included RCTs and retrospective reviews of data
18
bases. Nine studies were combined in a meta-analysis which included three
19
RCTs and four retrospective data bases of pharmacy claims. There was
20
marked heterogeneity in the compliance measures among the studies
21
evaluated and the patient group had different conditions which were being
22
treated. In the meta-analysis a total of 11,925 patients on fixed dose
23
combination were compared against 8317 patients on free drug component
24
regimen. Fixed dose combination resulted in a 26% decrease in the risk of
25
non compliance compared with free drug component regimen (pooled RR
26
0.74 [CI 0.69-0.80]; p<0.0001). There was no evidence of heterogeneity in this
27
analysis (p=.07). A subgroup analysis of the four studies on hypertension
28
showed that fixed dose combination (pooled RR 0.76 [CI 0.71-0.81];
29
p<0.0001) decreased the risk of medication non-compliance by 24%
30
compared with free drug combination regimens. Due to methodological
31
concerns about the conduct of the meta-analysis, the results of this study
32
should be viewed with caution.
Medicines concordance: full guideline DRAFT (July 2008)
page 245 of 373
DRAFT FOR CONSULTATION
1
Randomized Controlled Trials
2
Brown (1997) (Brown, B. G., Bardsley, J., Poulin, D. et al , 1997) from the
3
Cochrane Review looked at the effect of different formulations as they
4
compared regular niacin versus polygel controlled release niacin. All patients
5
received lovastatin 20 mg, colestipol 10 g, and niacin 500 mg for 12 months,
6
with dosage adjustment to target cholesterol of 150 to 175 mg/dl, and to
7
minimize side effects. Twenty-nine male participants were enrolled aged ≤65
8
years At 12 months, patients were randomly assigned to 1) continue with
9
regular niacin at a dose identical to that established during the 12 month
10
dose-finding period, or 2) change to polygel controlled-release niacin at that
11
daily dosage, but given twice rather than 4 times/day. At 20 months, groups 1)
12
and 2) were reversed (crossover). This study was conducted in the USA.
13
Adherence was significantly greater for the controlled-release preparation.
Medicines concordance: full guideline DRAFT (July 2008)
page 246 of 373
DRAFT FOR CONSULTATION
1
8.8
Key Clinical Question: Is there any evidence on
2
interventions that aim to minimize side-effects in order
3
to increase adherence?
Related references
Evidence statements
Evidence into
(summary of evidence)
recommendations
Rickles (2005) (Rickles,
1. There is conflicting
For general discussion of
N. M., Svarstad, B. L.,
evidence with regard to
limitations of evidence see section
Statz-Paynter, J. L. et al ,
whether discussing side
7.3
2005); Collier (2005)
effects, as part of a multi
(Collier, A. C., Ribaudo,
component intervention,
H., Mukherjee, A. L. et al
increases adherence.
Interventions relating to side
effects primarily involve providing
information for patients on side
, 2002); Kemp (1998)
effects, No single way of providing
(Kemp, R, Kirov, G,
information on side effects with a
Everitt, B et al , 1998).
view to increase adherence to
prescribed medication can be
recommended. This is mainly due
Rathbun (2005)
2. There is conflicting
(Rathbun, R. C., Farmer,
evidence with regard to
K. C., Stephens, J. R. et
whether educating patients
al , 2005); Chaplin (1998)
about side effects, as part of
(Chaplin, R. and Kent, A.,
a multi component
1998); Canto De Cetina
intervention, increases
The GDG considered that there
(2001) (Canto-De-Cetina,
adherence.
are a number of ways managing
to the evidence not assessing the
impact of minimizing side effects
independently, thus not being able
to ascertain their true effect.
T. E., Canto, P., and
side effects to support patient
Ordoñez, Luna M., 2001);
adherence. These include
Tuldra (2000) (Tuldrà, A.,
adequately informing patients
Fumaz, C. R., Ferrer, M.
about side effects, exploring how
J. et al , 2000); Peveler
patient wants to manage side
(1999) (Peveler, R.,
effects, reducing the dose of drug
George, C., Kinmonth, A.
and changing the drug to an
L. et al , 1999).
alternative.
Medicines concordance: full guideline DRAFT (July 2008)
page 247 of 373
DRAFT FOR CONSULTATION
Howland (1990)
3. One RCT showed that
(Howland, J. S., Baker,
informing patients about side
M. G., and Poe, T.,
effects did not have a
1990).
significant effect on
adherence.
Vivian (2002) (Vivian, E.
4. There is conflicting
M., 2002); Finley (2003)
evidence with regard to
(Finley, P. R., Rens, H.
whether giving an intervention
R., Pont, J. T. et al ,
deliverer (e.g. a pharmacist)
2003); Katon (2002)
the power to adjust a patient’s
(Katon, W., Russo, J.,
medication and/or dosage, as
Von, Korff M. et al ,
part of a multi component
2002).
intervention, increases
adherence.
Chisholm (2001)
5. There is conflicting
(Chisholm, M. A., Mulloy,
evidence with regard to
L. L., Jagadeesan, M. et
whether giving an intervention
al , 2001); Adler (2004)
deliverer (e.g. a pharmacist)
(Adler, D. A., Bungay, K.
the power to make
M., Wilson, I. B. et al ,
recommendations about the
2004).
treatment to the patient’s
practitioner, as part of a multi
component intervention,
increases adherence.
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
3
Medicines concordance: full guideline DRAFT (July 2008)
page 248 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
8.8.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies - randomised controlled trials (RCTs) of interventions to
6
increase adherence. The excluded studies list from the Cochrane review was
7
checked as we have included those studies with less than 80% follow-up of
8
participants.
9
Types of participants - people prescribed medication for a medical condition.
10
Duration of studies - six months follow up from the time of patient entry for
11
long-term regimens. No time limit specified for short-term conditions.
12
Types of interventions - any intervention intended to change adherence to
13
prescribed medication. As the Cochrane review is presented by condition, we
14
have used the evidence extracted in that review and reconfigured it by
15
intervention.
16
Types of outcome measures – inclusion criteria (as defined in the Cochrane
17
review) were expanded by including studies that used adherence as the only
18
outcome variable as opposed to adherence and treatment outcome variables.
19
The excluded studies list of the Cochrane review was cross-referenced to
20
ensure that no potentially relevant study was missed out.
21
8.8.2
22
Although a number of RCTs were found which addressed side effects we did
23
not find many where the interventions’ sole purpose was to address side
24
effects. Within the RCTs which did address side effects there was a lot of
25
variability in how they did this and to what extent side effects were a focus of
26
the intervention. What follows is a summary of the RCTs which addressed
27
side effects with information on interventions limited to those parts of the study
28
which addressed side effects. For further details please see evidence tables
29
for the RCTs retrieved from update searches, or the Cochrane review.
Evidence review
Medicines concordance: full guideline DRAFT (July 2008)
page 249 of 373
DRAFT FOR CONSULTATION
1
8.8.2.1
Discussing side effects with patients
2
RCTs were excluded from this section if side effects were only discussed, and
3
if no further details were given on how to educate the patients in managing
4
their side-effects with a view to increase adherence. Other RCTs were
5
excluded if addressing side effects was potentially part of a multi-component
6
intervention but when the actual decision to address side effects was the
7
choice of the intervention provider (therefore meaning it was impossible to tell
8
how many patients in the intervention group had side effects addressed and
9
how many did not have this issue addressed).
10
3 RCTs addressed side effects by clearly discussing them with patients.
11
Rickles (2005) (Rickles, N. M., Svarstad, B. L., Statz-Paynter, J. L. et al ,
12
2005) had pharmacists address adverse events during telephone calls in 98
13
patients with a mean age of 38 years. The pharmacist could probe or explain
14
issues not understood by the patient and make recommendations. The
15
intervention had positive effects. There was significant difference at six
16
months in adherence with the rate of missed doses significantly lower in the
17
intervention group (30.3%, SD 36.4 vs. 48.6%, SD 39.2, p = < 0.05). This
18
study was conducted in the USA. Collier (2005) (Collier, A. C., Ribaudo, H.,
19
Mukherjee, A. L. et al , 2002) had nurses address participants’ medication-
20
related behaviour and barriers to adherence during telephone calls in 282
21
patients. Advice around side effects was offered. Over 24 months, rates of
22
adherence were high in both groups (>72% reported at least 95% adherence).
23
No difference was seen between groups (OR 0.86, 95% CI 0.57 to 1.29). This
24
study appears to have been conducted in the USA. Kemp (1998) (Kemp, R,
25
Kirov, G, Everitt, B et al , 1998), an RCT from the Cochrane review, as part of
26
“compliance therapy”, had 2 sessions where intervention group participants
27
focused on symptoms and the side effects. This study had 74 participants.
28
The mean age in the intervention group was 34 years (SD: 10.6) and 37 years
29
(SD: 11.9) in the control group. This study was conducted in the UK.
30
Patients receiving compliance therapy demonstrated higher adherence ratings
31
(p < 0.001) than control group patients.
Medicines concordance: full guideline DRAFT (July 2008)
page 250 of 373
DRAFT FOR CONSULTATION
1
8.8.2.2
Educating patients and follow up
2
5 RCTs addressed side effects through educating the patient about them and
3
then following up on this education.
4
Rathbun (2005) (Rathbun, R. C., Farmer, K. C., Stephens, J. R. et al , 2005)
5
provided patients with a mean age of 38.0 years with education about
6
adverse-event management strategies, among other things. A total of 43
7
patients were included in this study. Telephone follow-up followed to identify
8
early problems. At week 28, adherence rates were 74% (SD 31%) in the
9
intervention group and 51% (SD 41%) in the control group (p = 0.080,
10
difference between groups 23%, 95% CI: 1% to 44%). Mean decline in
11
adherence between weeks 4 and 28 were 12% (p = 0.15) in the intervention
12
group and 22% (p = 0.002) in the control group. Patients in the intervention
13
group were more likely to take their medication at the prescribed dosing
14
schedule: at 4 weeks, 69% in the intervention group vs 42% in the control
15
group (p = 0.025) and at 28 weeks, 53% in the intervention group vs 31% in
16
the control group (p = 0.046). No significant difference was seen between the
17
groups based on patient self-report (94% vs 89% intervention vs control,
18
p=0.51). This study was conducted in the USA.
19
Chaplin (1998) (Chaplin, R. and Kent, A., 1998), an RCT from the Cochrane
20
review, had intervention group participants, “…participate in a discussion
21
about the risks and benefits of neuroleptic medications based on individual
22
semi-structured educational sessions with reference to a standardised
23
information sheet. The patients were asked whether they had heard of tardive
24
dyskinesia. The common movements of TD were modelled and the patients
25
were asked whether they thought they had the condition or had seen others
26
with it. They were informed that they were receiving an antipsychotic drug and
27
were given information about extrapyramidal symptoms and TD, its risk
28
factors, prevalence, treatment, potential irreversibility and the 1% risk of TD in
29
non-antipsychotic-treated patients. They were told that gradual discontinuation
30
of antipsychotic medication was the best way to prevent the condition but if
31
done abruptly carries a high risk of relapse and of precipitating TD. It was
32
stated that the optimum maintenance treatment, taking into account its risks
Medicines concordance: full guideline DRAFT (July 2008)
page 251 of 373
DRAFT FOR CONSULTATION
1
and benefits, was to use the lowest dose of antipsychotic drug that would
2
keep them well. Most importantly, they were asked not to make any changes
3
to their treatment without discussion with their psychiatrist. Finally, they were
4
given the opportunity to ask questions in an informal interactive session
5
lasting 30 minutes, and were given an information sheet for reference”
6
(Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study included 56
7
participants (age range not given). This study was conducted in the UK. The
8
intervention did not increase adherence relative to the control condition.
9
Canto De Cetina (2001) (Canto-De-Cetina, T. E., Canto, P., and Ordoñez,
10
Luna M., 2001), an RCT from the Cochrane review, had women in their
11
intervention group (counselling group) receive, “…a structured pretreatment
12
counselling with indications about the mode of action of DMPA, the common
13
side effects of the drug, including the possibility of irregular menstrual periods,
14
heavy bleeding, spotting, and amenorrhea. To mentally prepare users for
15
potential side effects, it was stressed that these side effects would be not
16
detrimental to their health. These indications were repeated at each follow-up
17
visit” (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study included 350
18
participants. The mean age in the counselling group was 33.9 years with a 20-
19
35 range and 34 years in the control group with also a 20-35 range. This study
20
was conducted in Mexico. There was a positive effect of the intervention in
21
terms of cumulative termination rates.
22
Tuldra (2000) (Tuldrà, A., Fumaz, C. R., Ferrer, M. J. et al , 2000), an RCT
23
from the Cochrane review, investigated a psycho-educative intervention part
24
of which involved participants being taught how to manage medication and
25
tackle problems such as forgetting, delays, side effects and changes in the
26
daily routine. During follow-up participants were provided with skills to deal
27
with minor adverse effects. This study had 116 participants. The mean age in
28
the intervention group was 39 years (SD: 10) and 38 years (SD: 7) in the
29
control group. It appears this study was conducted in Spain. “In an intention to
30
treat (ITT) analysis, no improvements were found in adherence (the p-values
31
were slightly above the 0.05 significance level). However, when a per protocol
32
analysis was conducted, the intervention resulted in improvements in
Medicines concordance: full guideline DRAFT (July 2008)
page 252 of 373
DRAFT FOR CONSULTATION
1
compliance to HAART at 48 weeks. The lack of statistical significance
2
observed using the ITT analysis might be a reflection of a low power to detect
3
differences due to the relatively small sample size for each arm (n = 55 for
4
intervention, n = 61 for control). The per protocol analysis is suspect in any
5
adherence study as it ignores patients who dropped out, the most severe form
6
of non-adherence.” (Haynes, R. B., Yao, X., Degani, A. et al , 2005).
7
Peveler (1999) (Peveler, R., George, C., Kinmonth, A. L. et al , 1999), an
8
RCT from the Cochrane review, compared four treatment groups,
9
“…treatment as usual, leaflet, drug counselling, or both interventions. The
10
information leaflet contained information about the drug, unwanted side
11
effects, and what to do in the event of a missing dose. Patients were given
12
drug counselling by a nurse at weeks 2 and 8, according to a written protocol.
13
Sessions included assessment of daily routine and lifestyle, attitudes to
14
treatment, and understanding of the reasons for treatment...The importance of
15
drug treatment was emphasized, and side effects and their management
16
discussed.” (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study had
17
213 participants. The mean age of participants was 45.3 years with a range of
18
21-83. This study was conducted in the UK. “The treatment leaflets had no
19
effect on adherence…This study was only 12 weeks in duration, which is
20
shorter than our usual 6 months follow-up criterion. However, because the
21
results were negative for adherence and clinical outcomes with the leaflet
22
intervention, the paper was included for this review. (Counselling about drug
23
treatment, however, did result in significant improvements in adherence and
24
clinical outcomes. Nonetheless, because the follow-up was less than six
25
months in duration, the results for counselling are not considered in the
26
conclusions of this review.)” (Haynes, R. B., Yao, X., Degani, A. et al , 2005).
27
Howland (1990) (Howland, J. S., Baker, M. G., and Poe, T., 1990), an RCT
28
from the Cochrane review, informed intervention group patients of six possible
29
side-effects of treatment with erythromycin, while control (uninformed) patients
30
were not made aware of potential side effects of treatment. This study had 98
31
participants. The mean age in the intervention group was 50 years and 48
32
years in the control group. It appears this study was conducted in the USA.
Medicines concordance: full guideline DRAFT (July 2008)
page 253 of 373
DRAFT FOR CONSULTATION
1
The intervention did not increase adherence relative to the control condition
2
nor did it decrease adherence.
3
4
8.8.2.3
Adjusting medication and/or dosage
5
3 RCTs addressed side effects by giving the intervention deliverer power to
6
adjust medication and/or medication dosage.
7
In a study by Vivian (2002) (Vivian, E. M., 2002) intervention patients saw
8
clinical pharmacists who could make changes in the prescribed drugs and
9
dosages and provided medication counselling centred around the discussion
10
of side effects, lifestyle and adherence (note we have made an assumption
11
here that the discussion of side effects and changes in medication are related,
12
this is not explicitly stated in the study). Fifty seven patients aged above 18
13
years were included in this study. The majority of the study population was
14
African American (77%). There were no significant differences in compliance
15
(from self report measure) between (p > 0.25, mean and SD not given for
16
adherence) or within (p = 0.07) the two groups at baseline or at the end of the
17
study. This study was conducted in the USA.
18
Finley (2003) (Finley, P. R., Rens, H. R., Pont, J. T. et al , 2003) had
19
pharmacists provide a detailed explanation of the role of antidepressants
20
(including potential therapeutic effects and adverse effects). Care managers
21
were permitted to titrate antidepressant drugs in a fashion consistent with the
22
HMO’s clinical guidelines and current recommended practices (note we have
23
made an assumption here that the discussion of side effects and changes in
24
medication are related, this is note explicitly stated in the study). . During
25
follow-up phone calls and clinic appointments pharmacists followed a
26
standardized set of questions that assessed adverse effects, among other
27
things. One hundred and twenty five patients were included in the study. The
28
majority of the study population was female ( 42 (84) % in the control group
29
and 64 (85) % in the intervention group; and the mean ages (s.d.) were 54.1 ±
30
17.3 years in the control group and 54.4± 14.1 years in the intervention group.
31
After 6 months, the intervention group demonstrated a significantly higher
Medicines concordance: full guideline DRAFT (July 2008)
page 254 of 373
DRAFT FOR CONSULTATION
1
drug adherence rate than that of the control group (67% vs. 48%, p = 0.038).
2
This study was conducted in the USA.
3
As part of a multifaceted intervention, Katon (2002) (Katon, W., Russo, J.,
4
Von, Korff M. et al , 2002) scheduled two sessions for intervention patients
5
with a psychiatrist in a primary care clinic. The study included 228 patients
6
aged between 18 and 80 years. Mean ages were 47.2 ± 14.0 years in the
7
intervention group and 46.7 ± 13.4 years in the usual care group. When
8
severe side effects or inadequate response to treatment occurred, the
9
psychiatrist helped the patient and primary care physician alter the dosage or
10
choose an alternative medication. There were no differences between the four
11
study groups in either adherence to the care suggestions, combined or
12
individually. There were no inter-group differences in medication adherence.
13
This study was conducted in the USA.
14
8.8.2.4
15
2 RCTs addressed side effects by giving the intervention deliverer power to
16
make recommendations to other health care professionals involved in the
17
patients care.
18
Chisholm (2001) (Chisholm, M. A., Mulloy, L. L., Jagadeesan, M. et al , 2001)
19
examined an intervention which included a pharmacist taking medication
20
histories and reviewing medications with the patient, with an emphasis on
21
optimizing medication therapy to achieve compliance outcomes while
22
minimizing adverse events related to medication. Twenty four patients aged
23
between 18 and 60 years were included in the study. Majority of the study
24
population was male (75%). The clinical pharmacist also provided
25
recommendations to the nephrologists with the goal of achieving desired
26
outcomes. Counselling involved discussion of patients concerns around their
27
medication therapy and instructing them how to properly take their
28
medications. Counselling was both verbal and/or in writing. At 12 months the
29
mean compliance rate in the intervention group was 96.1% (SD 4.7%)
30
compared to 81.6% (SD 11.5%) in the control group (p < 0.0001). For 6 of the
31
12 months, higher rates of compliance were seen in the intervention group (p
32
< 0.05). Also, 75% (n = 9) of the intervention patients were compliant each
Medicines concordance: full guideline DRAFT (July 2008)
page 255 of 373
Recommendations to health care professionals
DRAFT FOR CONSULTATION
1
month compared to 33.3% (n = 4) of the control group. This study was
2
conducted in the USA.
3
Adler (2004) (Adler, D. A., Bungay, K. M., Wilson, I. B. et al , 2004) employed
4
a pharmacist intervention which emphasised, among other things; assessing a
5
patient's medication regimen for drug-related problems (such as side effects
6
or drug interactions); monitoring drug efficacy and toxicity and educating
7
patients about depression and antidepressants. This study included 533
8
patients aged above 18 years. The mean age was 42.3 years, and the
9
majority was female. After an initial appointment with the patient, pharmacists
10
provided the patients PCP with a thorough medication history (including
11
adherence to prescribed medications and drug-related problems) and
12
whatever recommendations the pharmacist may have suggested to improve
13
the regimen. For patients using antidepressants at study entry (n = 227) there
14
were no significant differences in antidepressant usage between the
15
intervention and control groups either at 3 (90.7% vs. 87.2, p = 0.50) or 6
16
months (83.4% vs. 78.4%, p = 0.33). This study was conducted in the USA.
17
18
19
20
Medicines concordance: full guideline DRAFT (July 2008)
page 256 of 373
DRAFT FOR CONSULTATION
1
8.9
Key Clinical Question: How does the way the
2
information is presented (e.g. pictorial vs. written)
3
affects adherence?
Related
Evidence statements
references
(summary of evidence)
Raynor (2007)
1. One high quality systematic
For general discussion of limitations of
(Raynor, D. K.,
review of quantitative and
evidence see section 7.3
Blenkinsopp, A.,
qualitative research on the role
Knapp, P. et al ,
and effectiveness of written
2007)
information available to patients
about individual medicines stated
that no robust evidence was
found that the information
(delivery) had an effect on patient
satisfaction or compliance.
Schaffer (2004)
2. One RCT showed that written
(Schaffer, S. D.
information alone and written and
and Tian, L.,
verbal information resulted in
2004)
greater improvements in
Evidence into recommendations
While there is no conclusive evidence
about the effectiveness of the mode of
delivery of information affects
adherence, in certain cases/diseases it
made a difference. Thus, it should be
tailored to the individual’s choices and
preferences.
adherence to prescribed
medication compared to verbal
information alone.
Segador (2005)
3. One RCT showed that verbal
(Segador, J., Gil-
and written information compared
Guillen, V. F.,
to verbal information alone
Orozco, D. et al ,
resulted in significantly greater
2005)
improvements in adherence to
prescribed medication.
Medicines concordance: full guideline DRAFT (July 2008)
page 257 of 373
DRAFT FOR CONSULTATION
Atherton-Naji
4. One RCT showed that simple
(2001) (Atherton-
tailored information (mailed
Naji , A.,
leaflets with written and pictorial
Hamilton, R.,
information) did not significantly
Riddle, W. et al ,
improve adherence to prescribed
2001)
medication compared to usual
care.
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
3
Medicines concordance: full guideline DRAFT (July 2008)
page 258 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
8.9.1
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies - randomised controlled trials (RCTs) of interventions to
6
increase adherence. The excluded studies list from the Cochrane review was
7
checked as we have included those studies with less than 80% follow-up of
8
participants.
9
Types of participants - people prescribed medication for a medical condition.
10
Duration of studies - six months follow up from the time of patient entry for
11
long-term regimens. No time limit specified for short-term conditions.
12
Types of interventions - any interventions intended to change adherence to
13
prescribed medication. As the Cochrane review is presented by condition, we
14
have used the evidence extracted in that review and reconfigured it by
15
intervention.
16
Types of outcome measures - inclusion criteria (as defined in the Cochrane
17
review) were expanded by including studies that used adherence as the only
18
outcome variable as opposed to adherence and treatment outcome variables.
19
The excluded studies list of the Cochrane review was cross-referenced to
20
ensure that no potentially relevant study was missed out.
21
8.9.2
22
A health technology assessment report of a “Systematic review of quantitative
23
and qualitative research on the role and effectiveness of written information
24
available to patients about individual medicines” (Raynor, D. K., Blenkinsopp,
25
A., Knapp, P. et al , 2007) was retrieved. This report aimed to address the role
26
and value of written information given to patients; and how effective this
27
information is in improving patient’s knowledge of their treatment and health
28
outcomes. The inclusion criteria of this review was broader than those applied
29
in our reviews, as members of the public not currently taking medication;
Evidence review
Medicines concordance: full guideline DRAFT (July 2008)
page 259 of 373
DRAFT FOR CONSULTATION
1
general public using over the counter medicines and some addiction therapies
2
were included. Also, studies did not necessarily need to have an aim to
3
increase adherence. Despite these differences, it was felt that due to the high
4
quality of this document, it would be relevant to refer to it in this section to
5
support the evidence included in the reviews and inform the decision making
6
process.
7
Key findings of the report show that
 the majority of people do not value the written information they
8
receive, and
9
 no robust evidence was found that the information had any effect
10
on patient satisfaction or compliance.
11
12
Most patients did not value the current package insert patient information
13
leaflets (PILS) and did not consider information written by medicine
14
manufacturers to be sufficiently independent.
15
The (PILS ) supplied had deficiencies in the content (e.g. complexity of
16
language) and layout (e.g. print size). However, it did show that patients
17
valued written information that contained condition-based details along with
18
the medicines information, in addition to alternative treatments for the
19
condition.
20
In addition, the qualitative evidence included in the report did not show that
21
patients perceive improvement of compliance as a function of PILs. This can
22
be explained by how an informed decision not to take medication is a
23
legitimate and acceptable outcome. In contrast, some health care
24
professionals viewed that the increase of compliance was one of the main PIL
25
uses.
26
The key points for improvement of written medicines information outlined by
27
the review were:
28
29
 The need to involve patients in all stages of the process, as to
reflect better their needs.
Medicines concordance: full guideline DRAFT (July 2008)
page 260 of 373
DRAFT FOR CONSULTATION
 To incorporate the findings from the review to improve future
1
information design and content
2
 To present risk information numerically instead of verbal
3
descriptions.
4
5
8.9.2.1
Verbal vs. written vs. verbal and written vs. usual care
6
Schaffer (2004) (Schaffer, S. D. and Tian, L., 2004), a study from the
7
Cochrane review, compared two interventions and a combination of the two
8
control groups in patients with asthma aged 18-65 (n = 46) with results
9
dependant on the measure of adherence used. There were 4 groups,
10
“…standard provider education (control group); (b) audiotape alone; (c)
11
National Heart Lung and Blood Institute (NHLBI) booklet alone; and (d)
12
audiotape plus NHLBI booklet”. This study was conducted in the United
13
States.
14
“The results showed a significant increase in adherence by pharmacy-refill
15
measure (but not by self-report) for NHLBI booklet versus control, and for
16
NHLBI booklet plus audiotape versus control, but not for audiotape versus
17
control at six months”.
18
8.9.2.2
19
Segador (2005) (Segador, J., Gil-Guillen, V. F., Orozco, D. et al , 2005)
20
compared the effect of written information in addition to verbal information in
21
patients receiving antibiotic treatment for acute sore throat (n = 158) with
22
significant results. Patients in the written information group were given written
23
information at the time of their first visit to their GP. The written information
24
emphasized the importance of completing the antibiotic treatment, of
25
respecting intervals between doses and the drawbacks of an early drop-out,
26
and was given only at the time of initial consultation. The control group was
27
given verbal information only. This study was conducted in Spain.
28
The pill count average was 87.4+/-25.2% and it was higher in the intervention
29
group (93.7+/-24.5%) than in the control group (81.1+/-24.5%) (P < 0.05).
Verbal and written information vs. verbal information alone
Medicines concordance: full guideline DRAFT (July 2008)
page 261 of 373
DRAFT FOR CONSULTATION
1
8.9.2.3
Written and/or pictorial information given vs. usual care
2
Atherton-Naji (2001) (Atherton-Naji , A., Hamilton, R., Riddle, W. et al , 2001)
3
compared an educational intervention to routine care in patients (n = 45) with
4
depression with non-significant results. Patients in the intervention group
5
received simple tailored information (mailed leaflets with written and pictorial
6
information) at 1, 6 and 16 weeks after the initial prescription. The leaflets
7
contained basic information about the condition, treatment and general
8
problems people may have with adherence to the treatment. Leaflets were
9
personalised for each patient and their specific drug. This study was
10
conducted in the UK. Over 6 months, 35.6% (n = 16) collected prescriptions at
11
each month (no significant difference between groups). The proportion
12
decreased over time from month 1 (intervention group: 95.8% (n = 23) vs.
13
control group: 100% (n = 21)) to month 6 (intervention group: 58.3% (n = 14)
14
vs. control group: 52.4% (n = 11)).
15
Medicines concordance: full guideline DRAFT (July 2008)
page 262 of 373
DRAFT FOR CONSULTATION
1
2
8.10
Key Clinical Question: Do specific forms of therapy
(e.g. CBT) affect adherence?
Related references
Evidence
Evidence into recommendations
statements
(summary of
evidence)
All studies included in
1. There is some
For general discussion of limitations of evidence
evidence review
evidence that
see section 7.3
elements of CBT or
psychobehavioural
can help but the
quality of the
evidence does not
allow us to
generalise these
results.
The evidence concerning specific forms of
therapy and their effect on adherence is
inadequate. Conclusions from a variety of single
studies with ill defined content and delivery are
inconclusive. The GDG noted that definitions of
CBT and other therapies were often unclear. It
was also argued that some of these intervention
may not be salient to medicine-taking behaviou
Gray 2006 (Gray, R.,
2. The majority of
The available evidence is that patients make
Leese, M., Bindman,
evidence suggests
their own appraisal of medicines based on
J. et al , 2006); Wyatt
that CBT approaches
factors important to them and in this context the
2004 (Wyatt, G. E.,
do not improve
behaviour is rational and coherent and as such
Longshore, D., Chin,
adherence relative to
not appropriate for CBT
D. et al , 2004);
other forms of
Bechdolf 2004 and
treatment.
2005 (Bechdolf, A.,
Knost, B.,
Kuntermann, C. et al ,
Therapies which worked with patients and
families addressing social and cultural issues d
provide evidence of specific principles of
engaging with patients that may be of value.
2004) (Bechdolf, A.,
Köhn, D., Knost, B. et
al , 2005); Weber
2004 (Weber, R.,
Christen, L., Christen,
Medicines concordance: full guideline DRAFT (July 2008)
page 263 of 373
DRAFT FOR CONSULTATION
S. et al , 2004);
Antoni, 2006 (Antoni,
M. H., Carrico, A. W.,
Duran, R. E. et al ,
2006); Wagner 2006
(Wagner, G. J.,
Kanouse, D. E.,
Golinelli, D. et al ,
1909); Lam, 2003
(Lam, D. H., Watkins,
E. R., Hayward, P. et
al , 2003);
Strang (1981) (Strang, 3. One RCT showed
J. S., Falloon, I.-R. H.,
that family therapy
and Moss, H. B.,
increased adherence
1981)
to prescribed
medication when
compared to
individual support
sessions.
Xiong (1994) (Xiong,
4. Two RCTs showed
W., Phillips, M. R.,
that family therapy
Hu, X. et al , 1994);
did not increase
Zhang (1994) (Zhang,
adherence to
M., Wang, M., Li, J. et
prescribed
al , 1994)
medication when
compared to
standard care.
Miklowitz, 2003
5. One RCT showed
(Miklowitz, D. J.,
that family therapy
Medicines concordance: full guideline DRAFT (July 2008)
page 264 of 373
DRAFT FOR CONSULTATION
George, E. L.,
and
Richards, J. A. et al ,
pharmacotherapy
2003);
increased adherence
to prescribed
medication when
compared to crisis
management and
pharmacotherapy.
Razali, 2000 (Razali,
6. One RCT showed
S. M., Hasanah, C., I,
that culturally
Khan, U. A. et al ,
modified family
2000)
therapy increases
adherence when
compared to
behavioural family
therapy.
Remien, 2005
7. One RCT showed
(Remien, R. H.,
that couple based
Stirratt, M. J., Dolezal,
therapy increases
C. et al , 2005)
adherence compared
to usual care.
Ruskin, 2004 (Ruskin,
8. One RCT showed
P. E., Silver-Aylaian,
that telepsychiatry
M., Kling, M. A. et al ,
did not increase
2004)
adherence compared
to face to face
psychiatry.
Kemp 1996, 1998
9. There is conflicting
(Kemp, R., Hayward,
evidence with
Medicines concordance: full guideline DRAFT (July 2008)
page 265 of 373
DRAFT FOR CONSULTATION
P., Applewhaite, G. et
regards to whether
al , 1996) (Kemp, R,
compliance therapy
Kirov, G, Everitt, B et
increases adherence
al , 1998); O’Donnell
compared to
2003 (O'Donnell,
counselling
Colin, Donohoe, Gary,
Sharkey, Louise et al ,
2003).
Pradier, 2003
10. There is
(Pradier, Christian,
conflicting evidence
Bentz, Laurence,
to suggest that
Spire, Bruno et al ,
multicomponent
2003); Van Servellen
interventions mainly
(2005) (van,
based on
Servellen, G,
motivational
Nyamathi, A., Carpio,
principles increase
F. et al , 2005)
adherence
Weber 2004 (Weber,
11. One RCT
R., Christen, L.,
showed that CBT
Christen, S. et al ,
and usual care did
2004)
not increase
adherence compared
to usual care alone.
Gray, 2006 (Gray, R.,
One RCT showed
Leese, M., Bindman,
that CBT and health
J. et al , 2006).
education did not
increase adherence
compared to health
education alone.
Medicines concordance: full guideline DRAFT (July 2008)
page 266 of 373
DRAFT FOR CONSULTATION
Bechdolf, 2004 and
12. One RCT
2005 (Bechdolf, A.,
showed that CBT
Knost, B.,
and health education
Kuntermann, C. et al ,
did not increase
2004) (Bechdolf, A.,
adherence compared
Köhn, D., Knost, B. et
to psycho-education.
al , 2005).
Antoni 2006 (Antoni,
13. One RCT
M. H., Carrico, A. W.,
showed that
Duran, R. E. et al ,
cognitive behavioural
2006)
stress management
in addition to
antiretroviral
medication
adherence training
did not increase
adherence compared
to medication
adherence training
alone.
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
3
4
8.10.1 Methods of the evidence review
5
This paper includes a narrative summary of the included evidence, structured
6
according to the category of the intervention, following the agreed reviewing
7
protocol:
8
Types of studies - randomised controlled trials (RCTs) of interventions to
9
increase adherence. The excluded studies list from the Cochrane review was
Medicines concordance: full guideline DRAFT (July 2008)
page 267 of 373
DRAFT FOR CONSULTATION
1
checked as we have included those studies with less than 80% follow-up of
2
participants.
3
Types of participants - people prescribed medication for a medical condition.
4
Duration of studies - six months follow up from the time of patient entry for
5
long-term regimens. No time limit specified for short-term conditions.
6
Types of interventions - any interventions intended to change adherence to
7
prescribed medication. As the Cochrane review is presented by condition, we
8
have used the evidence extracted in that review and reconfigured it by
9
intervention.
10
Types of outcome measures – inclusion criteria (as defined in the Cochrane
11
review) were expanded by including studies that used adherence as the only
12
outcome variable as opposed to adherence and treatment outcome variables.
13
The excluded studies list of the Cochrane review was cross-referenced to
14
ensure that no potentially relevant study was missed out.
15
8.10.2 Evidence review
16
8.10.2.1
17
Miklowitz (2003) (Miklowitz, D. J., George, E. L., Richards, J. A. et al , 2003)
18
compared family focused therapy and pharmacotherapy with crisis
19
management and pharmacotherapy (serving as control group) in patients with
20
bi-polar disorder with positive results. The study included 101 participants with
21
ages that ranged from 18 to 62 years (mean age 35.6 ± 10.2 years). Family
22
focused therapy involved three modules: 1/ psycho-education, which involved
23
passing on information about the disorder, its aetiology, signs, symptoms and
24
also information on how to prevent relapse; 2/ communication training where,
25
through role play, skills of listening, offering feedback, and requesting
26
changes in behaviour were passed on; and 3/ problem solving skills, where
27
participants identified potential problems, came up with and evaluated various
28
solutions. Family focused therapy involved approximately 21 sessions over a
29
nine month period and was conducted with the whole of the patient’s family at
30
the patients/family’s home. This study was conducted in the USA.
Medicines concordance: full guideline DRAFT (July 2008)
page 268 of 373
Family Therapy
DRAFT FOR CONSULTATION
1
Patients in the intervention group had higher mean drug adherence scores
2
during follow up (2.77 +/- 0.43) than patients in the control group (2.56 +/-
3
0.48, p = 0.04). This study was conducted in the USA.
4
Razali (2000) (Razali, S. M., Hasanah, C., I, Khan, U. A. et al , 2000), a RCT
5
from the Cochrane review, compared the effects of “culturally modified family
6
therapy“(CMFT) to the effects of ”behavioural family therapy“ (BFT serving as
7
the control condition) in patients with schizophrenia. This study included 166
8
participants, with ages ranging from 17 to 55 years. The majority of the
9
patients came from a low socio-economic background. The CMFT was
10
delivered by a psychiatrist and sessions were given monthly for the first 3
11
months and then every 6 weeks in the following months. The CMFT consisted
12
of a “….Socio-cultural approach of family education, drug intervention
13
programme and problem-solving skills. The socio-cultural approaches to
14
family education include explanations of the concept of schizophrenia from a
15
cultural perspective and an attempt to correct negative attitudes toward
16
modern treatment. The family education and drug intervention was delivered
17
as a package. The drug intervention programme included drug counselling,
18
clear instruction about dose, frequency and possible side effects, the role of
19
carers in supervision of medication at home, and close monitoring of
20
compliance by a drug intake check-list presented in every follow-up visit
21
(Haynes, R. B., Yao, X., Degani, A. et al , 2005).” This study was conducted in
22
Malaysia.
23
At six months and one year, patients in the intervention group (CMFT) had
24
significantly higher compliance than those in the control (BFT) group.
25
Strang (1981) (Strang, J. S., Falloon, I.-R. H., and Moss, H. B., 1981), a RCT
26
from the Cochrane review, compared family therapy or individual support
27
sessions in patients with schizophrenia with positive results. Thirty two
28
patients were enrolled in this study. No information on the ages of the patients
29
was given. All patients had scheduled therapy and monthly medication
30
appointments. Patients were allocated to family therapy or individual support
31
sessions. This study was conducted in the UK.
Medicines concordance: full guideline DRAFT (July 2008)
page 269 of 373
DRAFT FOR CONSULTATION
1
The patients in the family therapy group were significantly more adherent than
2
those in the individual support group.
3
Xiong (1994) (Xiong, W., Phillips, M. R., Hu, X. et al , 1994), a study from
4
the Cochrane review, compared a family based intervention with standard
5
care in patients with schizophrenia with negative results. Sixty three families
6
were enrolled in this study and mean age was 31 years (ranging from 17 to 54
7
years). The family based intervention “…included monthly 45 minute
8
counselling sessions focused on the management of social and occupational
9
problems, medication management, family education, family group meetings,
10
and crisis intervention”. (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This
11
study was conducted in China. There was no difference in terms of adherence
12
between the two groups.
13
Zhang (1994) (Zhang, M., Wang, M., Li, J. et al , 1994), a study from the
14
Cochrane review, compared a family intervention with no addition care above
15
standard care in patients discharged after their first admission to the hospital
16
for schizophrenia with negative results. This study included 83 patients. Mean
17
ages were 23.5 ± (7.6) (s.d.) for the intervention group and 24.1 ± (8.1) (s.d.)
18
for the control group Families and patients in the family intervention group
19
were, “…assigned to one of two counsellors for their ongoing care, were
20
invited to come to a discharge session that focused on education about the
21
management of the patient’s treatment, asked to come to a family group
22
counselling session with other families three months after discharge, and then
23
attend three-monthly group sessions with other families with similar patient
24
problems. Non-attendance triggered a visit from study staff. Each family was
25
contacted at least once during the 18-month follow-up.” (Haynes, R. B., Yao,
26
X., Degani, A. et al , 2005). This study was conducted in China. There was no
27
difference in terms of adherence between the two groups.
28
29
8.10.2.2
Couples Therapy
30
Remien (2005) (Remien, R. H., Stirratt, M. J., Dolezal, C. et al , 2005)
31
compared a couple-based ART adherence intervention with usual care in HIVMedicines concordance: full guideline DRAFT (July 2008)
page 270 of 373
DRAFT FOR CONSULTATION
1
serodiscordant couples with positive results. The intervention included
2
structured discussions and instruction, as well as specific problem-solving and
3
couple-communication exercises. A total of 215 couples aged >18 years were
4
enrolled in the study. The mean age of the participants was 42 years. The
5
study sample mainly consisted of lower-income racial/ethnic minorities. Key
6
components included education about the importance of adherence to avoid
7
viral resistance and maintain health, identifying patterns of non-adherence,
8
developing communication and problem-solving strategies to overcome
9
adherence barriers, optimising partner support and building confidence in the
10
couple for achieving and maintaining improved adherence. In addition the
11
intervention sought to help couples to address issues of sex and intimacy. The
12
intervention was administered to each couple by a nurse practitioner through
13
45-60 min sessions held over 5 weeks. The study was conducted in the USA.
14
At 6 months there were significant differences in adherence change between
15
the 2 groups.
16
Significant group differences in adherence change from baseline to week 8 in
17
terms of proportion of prescribed doses taken (p = 0.021) and proportion of
18
doses taken within specified windows (p< 0.001). At 3 months, only the
19
proportion of doses taken within specified time windows was significant (p =
20
0.028).
21
8.10.2.3
22
Ruskin (2004) (Ruskin, P. E., Silver-Aylaian, M., Kling, M. A. et al , 2004)
23
compared patients being seen by a psychiatrist, either in person or by means
24
of telepsychiatry, who had one of the following five diagnoses: major
25
depressive disorder, dysthymic disorder, adjustment disorder with depressed
26
mood, mood disorder due to a general medical condition, or depressive
27
disorder not otherwise specified with negative results. One hundred and thirty
28
one patients were enrolled in the study. Mean age of participants was 49.7
29
(12.8) years. Treatment sessions lasted approximately 20 minutes and
30
consisted of antidepressant medication management, psycho-education, and
31
brief supportive counselling. Treatment consisted of eight sessions with a
32
psychiatrist over a 6-month period. This study was conducted in the USA.
Medicines concordance: full guideline DRAFT (July 2008)
page 271 of 373
Telepsychiatry
DRAFT FOR CONSULTATION
1
There was no difference in the percentage of adherent patients between the
2
two treatment groups.
3
8.10.2.4
4
Kemp’s (1996, 1998) (Kemp, R., Hayward, P., Applewhaite, G. et al , 1996)
5
(Kemp, R, Kirov, G, Everitt, B et al , 1998) RCTs from the Cochrane review,
6
compared “compliance therapy” with supportive counselling sessions (serving
7
as the control group) in patients with psychotic disorders with significant
8
results. Forty-seven patients aged 18 to 65 years were included in the 1996
9
study and 74 patients also aged 18 to 65 years in the 1998 study. Compliance
Compliance Therapy
10
therapy consisted of 4 to 6 sessions and was defined as, “…a strategy that
11
borrows from motivational interviewing. During session 1 and session 2,
12
patients reviewed their illness and conceptualized the problem. In the next 2
13
sessions, patients focused on symptoms and the side effects of treatment. In
14
the last 2 sessions, the stigma of drug treatment was addressed (Haynes, R.
15
B., Yao, X., Degani, A. et al , 2005)”. This study was conducted in the UK.
16
At 12 months patients receiving compliance therapy received higher
17
adherence ratings (p < 0.001) than those patients receiving non-specific
18
counselling.
19
O’Donnell (2003) (O'Donnell, Colin, Donohoe, Gary, Sharkey, Louise et al ,
20
2003), a RCT from the Cochrane review compared “compliance therapy” with
21
non-specific counselling (as the control group) in patients with schizophrenia,
22
with negative results. The study included 94 patients aged between 18 and 65
23
years. The mean age for both of the groups was 32 years (SD=9). The
24
intervention lasted 5 sessions, each session lasting 30-60 minutes. It is
25
reported that, “…the sessions covered a review of the patient’s illness history,
26
understanding of the illness and his or her ambivalence to treatment,
27
maintenance medication and stigma. Compliance therapy is a cognitive
28
behavioural intervention with techniques adapted from motivational
29
interviewing, other cognitive therapies and psycho-education.” (Haynes, R. B.,
30
Yao, X., Degani, A. et al , 2005). This study was conducted in Ireland.
31
There was no difference in terms of adherence between the two groups.
Medicines concordance: full guideline DRAFT (July 2008)
page 272 of 373
DRAFT FOR CONSULTATION
1
8.10.2.5
Multicomponent intervention
2
Pradier (2003) (Pradier, Christian, Bentz, Laurence, Spire, Bruno et al ,
3
2003), a RCT from the Cochrane review, compared a combined educational
4
and counselling intervention with a control condition in patients with HIV with
5
positive results. The study included 244 patients aged >18 years. Median age
6
of the participants was 40 years in the intervention group and 38 years in the
7
control group. The intervention consisted of 3 individual sessions delivered by
8
nurses lasting 45-60 minutes. The intervention was, “…founded on the
9
principles of motivational psychology, client centred therapy and the use of an
10
”empathic therapeutic to enhance participants’ self efficacy“. The intervention
11
focused on cognitive, emotional, social and behavioural determinants affecting
12
adherence.” (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study was
13
conducted in France.
14
Self-reported adherence between baseline and six months was significantly
15
improved in the intervention group, versus control.
16
Van Servellen (2005) (van, ServellenG, Nyamathi, A., Carpio, F. et al , 2005)
17
compared an enhanced adherence intervention with standard clinical care in
18
patients (n = 85) taking antiretroviral medications for at least 3 months with
19
negative results. Of note is that to be eligible to take part in this study
20
participants had to be able to speak Spanish. The enhanced adherence
21
intervention consisted of two parts the first being modular instruction which
22
was aimed at increasing patients HIV knowledge and ability to communicate
23
with medical staff and was delivered over 5 sessions by health educators and
24
nurses. These were followed up case management sessions, delivered either
25
face to face or via a telephone by a nurse, which concentrated on addressing
26
patient’s potential or actual risks for non adherence using motivational
27
interviewing techniques. Content involved going over things misunderstood in
28
the modular instruction stage, identifying barriers to adherence and finding
29
strategies to challenge these and helping to find community, treatment and
30
social support/referrals to help address adherence barriers. This study was
31
conducted in the USA.
Medicines concordance: full guideline DRAFT (July 2008)
page 273 of 373
DRAFT FOR CONSULTATION
1
There where no significant differences between the group at 6 months in: self
2
efficacy of adherence management (control group, -0.06, SD 0.59,
3
intervention group, 0.12, SD 0.95); 2 or more doses missed in last 4 days
4
(control group, 6.79%, intervention group, -5.69%); 2 or more doses missed
5
pasted 24 hours (control group, 18.21%, intervention group, -32%); average
6
doses missed in last 4 days (control group, 0.04, SD 0.13, intervention group,
7
0.02, SD, 0.14); proportion over 95% adherent in last four days (control group,
8
-4.85% ,intervention group, 1.71%); proportion > 90% adherent in last four
9
days (control group, -11.47% intervention group, -0.49%); follow medication
10
special instructions for 4 days (control group, 0.06, SD 0.34, intervention
11
group, -0.07, SD 0.36) and following medication schedule (control group, -
12
0.09, SD 1.60, intervention group, 0.33, SD 1.58).
13
8.10.2.6
14
Antoni (2006) (Antoni, M. H., Carrico, A. W., Duran, R. E. et al , 2006)
15
compared cognitive behavioural stress management (CSBM) in addition to
16
antiretroviral medication adherence training (MAT) with MAT alone in patients
17
with HIV with negative results. CSBM sessions included a didactic
18
component, as well as group discussion, with opportunities provided to apply
19
newly learned techniques. One hundred and thirty patients aged between 18
20
and 65 years were included in the study. The mean age was 41.6 (SD=8.3)
21
years. Homework was assigned to provide opportunities for participants to
22
practice techniques and increase their self-efficacy. The treatment was
23
focused extensively on eliciting participant experiences with adherence and
24
medication side effects. Throughout the 10-week, 135-minute group sessions
25
(90 minute stress management and 45 minute relaxation), facilitators
26
encouraged participants to examine potentially distorted cognitions and how
27
these may influence adherence to HAART (as well as other relevant self-care
28
behaviours). During cognitive restructuring exercises, participants were asked
29
to examine medication-relevant thoughts both in session and through
30
homework exercises. Adherence was also a key target during the skills
31
training sessions. This study appears to have been conducted in the USA (not
32
explicitly stated).
Cognitive Behavioural Treatment
Medicines concordance: full guideline DRAFT (July 2008)
page 274 of 373
DRAFT FOR CONSULTATION
1
The experimental conditions did not differ significantly in participant-reported
2
medication adherence throughout the 15-month investigation period.
3
Bechdolf (2004 and 2005) (Bechdolf, A., Knost, B., Kuntermann, C. et al ,
4
2004) (Bechdolf, A., Köhn, D., Knost, B. et al , 2005) compared group CBT
5
with group psycho-education (PE) in patients who had suffered an episode of
6
a schizophrenia or a related disorder with negative results. Eighty-eight
7
patients aged between 18 and 64 years were included in the study. Only a
8
minority of patients were employed. Group CBT focused on assessment and
9
engagement (sharing information about voices and delusions, models of
10
psychosis), improving self-esteem, formulation of key-problems and
11
developing interventions directed at reducing the severity and the occurrence
12
of key problems, relapse prevention/keeping well and enhancing medication
13
compliance. There was a specific focus on the component "improving self-
14
esteem" to foster feelings of hope and engagement with therapy. Group CBT
15
involved 16 sessions in 8 weeks by a psychiatrist or clinical psychologist. This
16
study was conducted in Germany.
17
Compliance was high initially (group CBT mean: 3.9 (SD 0.3) vs. PE group:
18
3.8 (SD 0.5)). At 8 weeks post-treatment, there was no significant difference
19
between groups (3.9 (SD 0.3) vs. 3.7 (SD 0.7)) nor at 6 months (3.5 (SD 0.9)
20
vs. 3.2 (SD 1.0). This remained non-significant when corrected for pre-
21
treatment scores. At 24 month follow-up again no significant differences were
22
seen (3.4 (SD 0.7) vs. 2.9 (SD 1.1).
23
Weber (2004) (Weber, R., Christen, L., Christen, S. et al , 2004), a RCT from
24
the Cochrane review, compared cognitive behavioural therapy in addition to
25
usual care to usual care alone in patients with HIV with negative results. The
26
study included 60 patients, and the median age was 41 years. The
27
intervention was delivered by a psychotherapist. The Cochrane Review
28
informs that, “…protocol defined a minimum of three and a maximum of 25
29
sessions within the 1- year study period. Participant and psychotherapist
30
determined the frequency of appointments and set their own goals for future
31
interventions. The intervention had to be based on concepts of cognitive
Medicines concordance: full guideline DRAFT (July 2008)
page 275 of 373
DRAFT FOR CONSULTATION
1
behavioral therapy” (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This
2
study was conducted in Switzerland.
3
There was no significant difference in mean adherence between the two
4
groups, but both groups had very high mean adherence rates (92.8% versus
5
88.9%), and a higher proportion of intervention group patients were at or
6
above 95% adherence (70% versus 50%, p = 0.014).
7
There was no difference in terms of adherence between the two groups.
8
Lam (2003) (Lam, D. H., Watkins, E. R., Hayward, P. et al , 2003) compared
9
cognitive therapy and minimal psychiatric care v minimal psychiatric care
10
alone in patients with bi-polar disorder with positive results. One hundred and
11
three patients with ages ranging from 18 to 70 years were enrolled in the
12
study. Mean age was 46.4 ± (12.1) (s.d.) years for the intervention group and
13
41.5± (10.8) (s.d.) for the control group. Traditional cognitive therapy for
14
depression was provided by clinical psychologists with new elements
15
highlighting the need for combined psychological and drug treatment, CBT
16
skills for monitoring mood and preventing relapse and highlighting the
17
importance of sleep and routine. The therapy also addressed illness beliefs.
18
Cognitive therapy involved 12 to 18 individual sessions within the first 6
19
months and 2 booster sessions in the second 6 months. This study was
20
conducted in the UK.
21
At 14 months, medication adherence was 95.5% (SD: 7.7%) in the cognitive
22
therapy group and 69.1% (SD 16.4%) in the control group (p < 0.001).
23
Proportions of people who had at least 80% adherence rates were 97.4% in
24
the cognitive therapy group and 21.7% in the control group (p < 0.001).
25
Gray (2006) (Gray, R., Leese, M., Bindman, J. et al , 2006) compared
26
adherence therapy (AT) with health education (HE) (serving as the control
27
group) in patients with schizophrenia with negative results. Adherence therapy
28
is a brief, individual CBT approach. 6 elements formed the core of the therapy:
29
assessment, medication problem solving, medication timeline, exploring
30
ambivalence, discussing beliefs and concerns about medication and using
31
medication in the future. Three hundred patients were included in the study,
Medicines concordance: full guideline DRAFT (July 2008)
page 276 of 373
DRAFT FOR CONSULTATION
1
and the mean age (s.d.) was 41.5 (11.5) years. Key therapy skills that the
2
therapists use included exchanging information, developing discrepancies
3
between participant’s thoughts and behaviours about medications and
4
working with resistance to discussing psychiatric medication and treatment.
5
The overall aim of process was to achieve a joint decision about the
6
medication. Participants were offered a maximum of 8 sessions lasting 30-50
7
minutes over a 5 month period and the intervention was delivered by 9
8
therapists (four psychologists, three psychiatrists and 2 mental health nurses).
9
The study was conducted in 4 countries: The Netherlands, Germany, England
10
and Italy.
11
At 12 months, there were no significant differences between the groups using
12
either patient assessment (AT group: 3.20 (SD 1.07), HE group: 3.33 (SD
13
1.02), 95% CI -0.35 to 0.08) or clinical assessment (AT group: 5.22 (SD 1.57)
14
HE group: 5.03 (SD 1.55), 95% CI -0.12 to 0.52) of adherence.
15
Wagner (2006) (Wagner, G. J., Kanouse, D. E., Golinelli, D. et al , 1909)
16
compared a cognitive behavioural treatment with an enhanced condition of the
17
treatment (a 2 week pre-treatment practice trial) and a control group, for the
18
effect on adherence to a new regime of antiretroviral therapy. The study
19
included 230 patients with a mean age of 39 (ranging from 21 to 70 years),
20
80% were male, 49% were Latino(a) and 65% were unemployed. The study
21
was set in the USA. The intervention involved five sessions of cognitive
22
behavioural therapy. Questionnaires were administered and blood was drawn
23
at screening (four weeks before treatment baseline), and periodically up to 48
24
weeks from the start of treatment. There was no difference in adherence
25
between the intervention and the enhanced intervention group. There was
26
initially a significant increase in attaining ‘good’ adherence (90% of prescribed
27
dose) for the intervention groups compared to the control group (82% versus
28
65%, p=0.01). The difference reduced in the following weeks and was not
29
significant. At week 48 the difference was reversed to 57% (intervention
30
group) versus 65% (control group), but this was also non-significant (p=0.52).
31
Wyatt (2004) (Wyatt, G. E., Longshore, D., Chin, D. et al , 2004) compared a
32
cognitive behavioural approach (the Enhanced Sexual Health Intervention) to
Medicines concordance: full guideline DRAFT (July 2008)
page 277 of 373
DRAFT FOR CONSULTATION
1
usual care for risk reduction and treatment adherence for 147 women who
2
had HIV and a history of childhood sexual abuse. The mean age was 41 (8.2),
3
25-65 years, 51% were African American and 49% were Latina and primarily
4
unemployed. The study was set in the USA. The intervention involved 11
5
weekly sessions for 2.5 hours per week of psycho-educational content relating
6
to child sexual abuse and HIV status. They were followed up at the end of the
7
11 weeks and then again at 3 and 6 months. Although an effect was found for
8
risk reduction there was no increase in adherence to medication in the
9
intervention group (75.6% versus 73.3%, OR=1.13, p=0.41). However a
10
significant effect was found for adherence for those who attended at least
11
eight sessions (91.3%) compared to seven or fewer (49.7%), OR=4.90,
12
p=0.044. The difference in adherence of the high attendees was 91.3%
13
compared to the control group 74.7%, this was statistically significant.
14
15
8.11
Key Clinical Question: Would a contractual agreement
between HCP and patient affect adherence?
16
Related
Evidence statements
Evidence into recommendations
references
(summary of evidence)
Bosch-
1.Evidence from one high
For general discussion of limitations of evidence
Capblanch
quality systematic review of
see section 7.3
(2007)
RCTS suggests that the
(Bosch-
use of contracts within a
Capblanch,
healthcare setting does not
X., Abba,
appear to increase
K., Prictor,
adherence to prescribed
M. et al ,
medication.
There is no evidence to show that contractual
arrangements have any impact in improving
adherence and the GDG did not wish to make a
recommendation in this area.
2007)
17
18
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective.
Medicines concordance: full guideline DRAFT (July 2008)
page 278 of 373
DRAFT FOR CONSULTATION
1
8.11.1 Methods of the evidence review
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies - Randomized controlled trials (RCTs) of interventions to
6
increase adherence. The excluded studies list from the Cochrane review was
7
checked as we have included those studies with less than 80% follow-up of
8
participants.
9
Types of participants - people prescribed medication for a medical condition.
10
Duration of studies - six months follow up from the time of patient entry for
11
long-term regimens. No time limit specified for short-term conditions.
12
Types of interventions - any interventions intended to change adherence to
13
prescribed medication. As the Cochrane review is presented by condition, we
14
have used the evidence extracted in that review and reconfigured it by
15
intervention.
16
Types of outcome measures - inclusion criteria (as defined in the Cochrane
17
review) were expanded by including studies that used adherence as the only
18
outcome variable as opposed to adherence and treatment outcome variables.
19
The excluded studies list of the Cochrane review was cross-referenced to
20
ensure that no potentially relevant study was missed out.
21
8.11.2 Definitions of contracts
22
Contracts can be generically viewed as reciprocal agreements between two or
23
more parties in which one or more will need to do something.
24
From a behavioural strategy perspective, a contract to increase patient’s
25
adherence can be defined as “a process of specifying a set of rules regarding
26
some behaviour of interest and formalizing a commitment to adhere to them”
27
(Haynes, T. B., Taylor, D. W., and Sackett, D. L., 1979).
Medicines concordance: full guideline DRAFT (July 2008)
page 279 of 373
DRAFT FOR CONSULTATION
1
Contracts can be written or verbal. Most contracts are between healthcare
2
practitioners and patients, but they may also occur between practitioners and
3
carers, carers and patients or by a patient with him/herself.
4
8.11.3 Evidence review
5
We retrieved one Cochrane Review that aimed to assess whether contracts
6
between healthcare practitioners and patients had an effect on patients'
7
adherence to treatment, prevention and health promotion activities (Bosch-
8
Capblanch, X., Abba, K., Prictor, M. et al , 2007). Although this review
9
included settings other than clinical settings, we decided to include this high
10
quality systematic review of RCTs on the grounds that the results were
11
reported in groups, thus allowing us to make conclusions from those settings
12
relevant to the guideline. It also included other treatment groups that are
13
outside the remit of the guideline, such as substance addiction treatments and
14
interventions for hypertension and overweight without prescribed medication.
15
Other areas that were included were acne, acute bacterial infections, arthritis,
16
asthma, breast self examination, contact lens care, depression, diabetes,
17
phobias, promotion of healthy diet and exercise and tuberculosis.
18
The Cochrane review also assessed the effects of contracts on other
19
outcomes, including patient participation and satisfaction, health practitioner
20
behaviour and views, health status, harms, costs, and ethical issues.
21
Seven trials assessed contracts between HCPs and patients, nine trials
22
assessed contracts between patients and carers, peers or others, and
23
between HCPs and carers in one trial. Four trials assessed contracts between
24
HCPs, patients and carers, two trials assessed self-contract and the other
25
seven trials did not report which type of contract was being used. Twenty one
26
trials included some type of financial incentive
27
Several of the trials were of poor quality and included small numbers of
28
people. Most were conducted in the USA and were conducted in specialised
29
services.
Medicines concordance: full guideline DRAFT (July 2008)
page 280 of 373
DRAFT FOR CONSULTATION
1
Two trials that examined the effects of contracts in the context of hypertension
2
management reported adherence outcomes. However, only one showed
3
statistically significant results in favour of the group with contracts.
4
In the miscellaneous section, six trials in the contexts of acne, acute bacterial
5
infection, asthma, depression, diabetes and tuberculosis reported adherence
6
outcomes. However, in some cases it was not possible to determine whether
7
adherence was also related to prescribed medication or with an overall
8
treatment regime (depression, and diabetes). From these, five trials did not
9
report any statistical significance in favour of the contracts groups, whilst the
10
acute bacterial infection trial reported significantly better results (based on pill
11
count) in the contract group. However, there was no difference between
12
groups in self-reported adherence, nor in the number of additional
13
prescriptions to finalise the treatment.
14
Based on the results for the trials that included an assessment of patients’
15
adherence to medication, the use of contracts does not appear to improve
16
adherence.
17
Overall, the conclusions from the Cochrane authors state that there is limited
18
evidence that contracts can have a positive effect in improving adherence. In
19
addition they argue that there is insufficient evidence from large, good quality
20
studies to routinely recommend contracts for improving adherence to
21
treatment or preventive health regimens (Bosch, Capblanch, X and Garner,
22
P., 2006).
23
24
25
26
Medicines concordance: full guideline DRAFT (July 2008)
page 281 of 373
DRAFT FOR CONSULTATION
1
2
8.12
Key Clinical Question: Do reminders (and what types of
3
reminders, text messaging etc) increase adherence to
4
prescribed medication?
Related
Evidence statements
references
(summary of
Evidence into recommendations
evidence)
Heneghan (2006)
1. A high quality
For general discussion of limitations of evidence
(Heneghan, C. J.,
systematic review
see section 7.3
Glasziou, P., and
suggests that
Perera, R., 2006)
medication packaging
that incorporates a
reminder system may
improve adherence for
patients taking longterm medications.
The available evidence suggests that reminders
can be of value in improving adherence for some
patients in some situations. A wide range of
reminders were found from reminders related to
packaging to telephone reminders. The GDG did
not consider the evidence sufficient to make
general recommendations about reminders such
Stewart (2005)
2. Three RCTs show
as telephone and text reminders but did consider
(Stewart, A.,
that a reminder given
that packaging may have a role in reminding
Noakes, T.,
via a telephone call in
patients. This overlaps recommendation
Eales, C. et al ,
a multicomponent
regarding drug packaging.
2005);
intervention can
Urien(2004)
increase adherence to
(Urien, A. M.,
prescribed medication.
Guillen, V. F.,
Beltran, D. O. et
al , 2004); Piette
(2000) (Piette, J.
D., Weinberger,
M., McPhee, S. J.
et al , 2000)
Medicines concordance: full guideline DRAFT (July 2008)
page 282 of 373
DRAFT FOR CONSULTATION
Beaucage (2006)
3. One RCT showed
(Beaucage, K.,
no significant results
Lachance-
on the effect of a
Demers, H., Ngo,
reminder given via a
T. T. et al , 2006)
telephone call in a
multicomponent
intervention in
increasing adherence
to prescribed
medication
Hamet
4. There is conflicting
(2003)(Hamet, P.,
evidence on the effect
Campbell, N.,
of reminders via mail in
Curnew, G et al ,
a multicomponent
2003); Peterson
intervention in
(1984) (Peterson,
increasing adherence
G. M.,
to prescribed
Fitzmaurice, K.
medication.
D., Naunton, M. et
al , 2004)
Vrijens (2006)
5. There is conflicting
(Vrijens, B.,
evidence on the effect
Belmans, A.,
of electronic reminders
Matthys, K. et al ,
in a multicomponent
2006);
intervention in
Mannheimer
increasing adherence
(2006)
to prescribed
(Mannheimer, S.
medication. (
B., Morse, E.,
Matts, J. P. et al ,
Medicines concordance: full guideline DRAFT (July 2008)
page 283 of 373
DRAFT FOR CONSULTATION
2006); Sackett
(1975) (Sackett,
D. L., Gibson, E.
S., and Taylor, D.
W., 1975)
Guthrie (2001)
6. One RCT showed
(Guthrie, R. M.,
that using a
2001)
combination of postal
and telephone
reminders in a
multicomponent
intervention showed no
effect on adherence to
prescribed medication.
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
Medicines concordance: full guideline DRAFT (July 2008)
page 284 of 373
DRAFT FOR CONSULTATION
1
8.12.1 Methods of the evidence review
2
This paper includes a narrative summary of the included evidence, structured
3
according to the category of the intervention, following the agreed reviewing
4
protocol:
5
Types of studies - randomized controlled trials (RCTs) of interventions to
6
increase adherence. The excluded studies list from the Cochrane review was
7
checked as we have included those studies with less than 80% follow-up of
8
participants.
9
Types of participants - people prescribed medication for a medical condition.
10
Duration of studies - six months follow up from the time of patient entry for
11
long-term regimens. No time limit specified for short-term conditions.
12
Types of interventions - any interventions intended to change adherence to
13
prescribed medication. As the Cochrane review is presented by condition, we
14
have used the evidence extracted in that review and reconfigured it by
15
intervention.
16
Types of outcome measures – inclusion criteria (as defined in the Cochrane
17
review) were expanded by including studies that used adherence as the only
18
outcome variable as opposed to adherence and treatment outcome variables.
19
The excluded studies list of the Cochrane review was cross-referenced to
20
ensure that no potentially relevant study was missed out.
21
8.12.2 Evidence review
22
One high quality systematic review by Heneghan (2006) (Heneghan, C. J.,
23
Glasziou, P., and Perera, R., 2006) aimed to determine the effects of reminder
24
packaging to increase patient adherence with self-administered long-term
25
medications.
26
The systematic review included eight studies containing 1137 participants. All
27
types of setting were included and no age limits were set. Studies where
28
direct observation of therapy occurred through a health professional were
29
excluded. Interventions that were included required a reminder system for the
Medicines concordance: full guideline DRAFT (July 2008)
page 285 of 373
DRAFT FOR CONSULTATION
1
day of the week or the time that the medication was to be taken, and it had to
2
form part of the packaging. Reminders that were separate to the intervention
3
(e.g. mailed reminders) were excluded. Electronic systems were also
4
excluded. Packaging aids were included irrespective of whether the
5
medication required a prescription or not.
6
The primary outcome of importance was adherence to medication which was
7
measured by pill counts and/or self-reporting.
8
Reminder packaging showed a significant increase in the percentage of pills
9
taken, weighted mean difference 11% (95% confidence interval (CI) 6% to
10
17%). Thus, the authors conclude that reminder packaging may represent a
11
simple method for improving adherence for patients in certain conditions.
12
8.12.2.1
13
Stewart (2005) (Stewart, A., Noakes, T., Eales, C. et al , 2005) compared four
14
once monthly educational sessions, the prescription of a home based walking
15
program and once monthly phone calls with four once monthly educational
16
sessions, the prescription of a home based walking program without once
17
monthly phone calls in patients attending a hypertension clinic with positive
18
results at week 24 but not 36. Sample size was comprised 83 participants.
19
During the phone calls patients (or a family member) were asked about the
20
exercise program and reminded about diet and medication. In total 5 (pairs) of
21
telephone calls (to patient and family member) were made once monthly over
22
24 weeks by a physiotherapist. This study was conducted in South Africa.
23
At week 24 significantly more patients in the group receiving telephone calls
24
(65%) were taking their medications as prescribed compared to the group not
25
receiving telephone calls (44.7%, p = 0.05), however, there was no difference
26
between the groups at week 36 (82.4% vs 86.7%) (Stewart, A., Noakes, T.,
27
Eales, C. et al , 2005).
28
Urien (2004) (Urien, A. M., Guillen, V. F., Beltran, D. O. et al , 2004)
29
compared a telephone-delivered intervention plus educational advice with
30
educational advice alone in patients receiving antibiotic treatment. The sample
Telephone reminders
Medicines concordance: full guideline DRAFT (July 2008)
page 286 of 373
DRAFT FOR CONSULTATION
1
was comprised of 128 participants aged ≥18 years. The telephone call was
2
undertaken on the 4th day after the start of treatment, when the first box of
3
antibiotics should have been finished. The patient was advised to continue the
4
treatment according to the dosage and number of days that had been
5
prescribed. The patient was also reminded that although he or she may feel
6
better or even cured, the treatment was to be continued for 10 days. This
7
study was conducted in Spain.
8
Adherence was significantly higher in the intervention group (78.3%) than in
9
the control group (54.1%) (p = 0.005) (Urien, A. M., Guillen, V. F., Beltran, D.
10
O. et al , 2004).
11
Beaucage (2006) (Beaucage, K., Lachance-Demers, H., Ngo, T. T. et al ,
12
2006) compared a pharmacist telephone intervention with usual care in
13
patients on antibiotic treatment with negative results. A telephone call was
14
made to patients in the intervention group by a pharmacist 3 days into
15
treatment. The pharmacist asked about the patient’s general condition, on the
16
presence of adverse effects and the participants understanding of dosing. The
17
pharmacist emphasized the importance of adherence and offered motivation
18
to the patient. The patients were offered an opportunity to ask questions and
19
were given the pharmacists contact details in case they wanted to make
20
contacted their pharmacist at a later time. This study was conducted in
21
Canada.
22
23
Mean adherence to treatment was 94% +/- 9% and 94% +/- 12% in the
24
intervention and control groups respectively (p = 0.803). The proportion of
25
patients with less than 80% adherence was similar in the two groups
26
(Intervention group: 8%, control group: 9%).
27
Piette (2000) (Piette, J. D., Weinberger, M., McPhee, S. J. et al , 2000)
28
compared a telephone intervention with usual care in patients with diabetes (n
29
= 280) with positive results. Patients were excluded if they were above 75
30
years of age. The intervention consisted of, “…a series of automated
31
telephone assessments designed to identify patients with health and self-care
Medicines concordance: full guideline DRAFT (July 2008)
page 287 of 373
DRAFT FOR CONSULTATION
1
problems (TeleminderModel IV automated telephone messaging computer).
2
Calls were made on a biweekly basis, up to 6 attempted calls, and involved a
3
5 to 8-minute assessment. During each assessment, patients used the touch-
4
tone keypad to report information about self-monitored blood glucose
5
readings, self-care, perceived glycaemic control, and symptoms of poor
6
glycaemic control, foot problems, chest pain, and breathing problems, with
7
automated prompts for out-of-range errors. The automated telephone calls
8
were also used to deliver, at the patient’s option, 1 of 30 targeted and tailored
9
self-care education messages at the end of each telephone session. Patients
10
only received a 1-page instruction sheet on the use of the phone. Each week,
11
the automated assessment system generated reports organized according to
12
the urgency of the reported problems, and a diabetes nurse educator used
13
these reports to prioritize contacts for a telephone follow-up. During follow-up
14
calls, the nurse addressed problems reported during the assessments and
15
provided more general self-care information. After several months,
16
intervention group patients were offered additional automated self-care calls
17
that focused on glucose self-monitoring, foot care and medication adherence.
18
In the medication adherence part of these sessions, patients were asked
19
about their adherence to insulin, oral hypoglycaemic medications,
20
antihypertensive medications, and antilipidemic medications. For each type of
21
medication, patients without adherence problems received positive feedback
22
and reinforcement. Patients reporting less than optimal adherence were asked
23
about specific barriers and were given advice from the nurse about
24
overcoming each barrier. The nurse was located outside the clinic and had no
25
access to medical records other than the baseline info collected at enrolment
26
and her own notes. She did not have any face-to-face contact with patients.
27
The nurse addressed problems raised by patients in the automated calls and
28
also gave general self-care education. The nurse also checked on patients
29
who rarely responded to automated calls. A small number of patients initiated
30
calls to the nurse by toll free no. She referred these to the primary care
31
physician as appropriate.” (Haynes, R. B., Yao, X., Degani, A. et al , 2005).
32
This study was conducted in the United States.
Medicines concordance: full guideline DRAFT (July 2008)
page 288 of 373
DRAFT FOR CONSULTATION
1
Compared with usual care, patients in the intervention group reported fewer
2
problems with medication adherence (P < 0.003).
3
8.12.2.2
4
Hamet (2003) (Hamet, P., Campbell, N., Curnew, G et al , 2003) compared
5
the avapromise intervention (designed to modify behaviour by medication
6
adherence through reinforcement and lifestyle modification) with usual care in
7
patients with high blood pressure. This was a study that comprised a total of
8
4864 participants. The ages of participants ranged from 16 to 89 years. The
9
intervention was made up of two elements that were delivered together. The
Mail reminders
10
first element attempted to reinforce medication behaviours by using
11
medication reminder letters, blood pressure diaries and telephone nurse
12
counselling sessions. The second element addressed lifestyle management
13
through educational brochures. Patients assigned to the intervention group
14
were mailed the material at one, two, three, four, six and 12 months. This
15
study was conducted in Canada.
16
A total of 25% of patients discontinued their treatment from the intervention
17
group and 25.5% from the control group (p = 0.94). There was no statistically
18
significant difference in the duration of irbesartan compliance between the
19
treatment groups (Hamet, P., Campbell, N., Curnew, G et al , 2003).
20
In Peterson’s (1984) (Peterson, G. M., Fitzmaurice, K. D., Naunton, M. et
21
al , 2004) (from the Cochrane review) RCT, the intervention group received
22
several adherence-improving strategies: patients were counselled on the
23
goals of anticonvulsant therapy and the importance of good adherence in
24
achieving these goals; a schedule of medication-taking was devised that
25
corresponded with the patient's everyday habits; patients were given a copy of
26
an educational leaflet; each patient was provided with a 'dosette' medication
27
container and counselled on its utility; patients were instructed to use a
28
medication/seizure diary; and patients were reminded by mail of upcoming
29
appointments and of missed prescription refills. The control group received
30
none of these interventions. Patient compliance improved significantly with the
31
intervention group patients. Fifty three adults and teenagers were enrolled in
Medicines concordance: full guideline DRAFT (July 2008)
page 289 of 373
DRAFT FOR CONSULTATION
1
the study, and the median age was 35 years for the control group and 28
2
years for the intervention group. The study was conducted in Australia.
3
8.12.2.3
4
Guthrie (2001) (Guthrie, R. M., 2001) delivered an intervention involving
5
postal and telephone reminders regarding coronary risk reduction and
6
medication compliance, which were sent during the first 2 months of
7
pravastatin treatment, or usual care. This was a large study that comprised a
8
total of 13,100 participants. Mean age was 58.0 years. Both groups received
9
reminder postcards at 4 and 5 months, in addition to counselling by physicians
Mail and telephone reminders
10
about coronary risk reduction. At study discontinuation, patients completed
11
and mailed (to the program-coordinating centre) questionnaires concerning
12
compliance with care, as well as self-reported adoption of other lifestyle
13
modifications. This study was conducted in the USA. Neither early reminders
14
nor baseline patient characteristics were significantly associated with reported
15
pravastatin compliance rates.
16
8.12.2.4
17
Vrijens (2006) (Vrijens, B., Belmans, A., Matthys, K. et al , 2006) compared a
18
supportive intervention program with usual care in patients who had been
19
taking atorvastatin with positive results. Four hundred and twenty nine
20
participants aged >18 years entered the study. As part of the supportive
21
intervention program participants were supplied with a ‘beep-card’ that
22
reminded the patient of the dosing time, and also gave educational reminders.
23
The supportive intervention program also provided patients with a medication
24
review by the patients’ pharmacist of their electronically compiled dosing
25
history (through MEMs). At each follow-up visit the pharmacist delivered an
26
educational message, updated the patients‘ compliance passport and
27
analysed with the patient their electronically compiled dosing history over the
28
previous month/3 months (depending on the gap between follow-up
29
appointments). Baseline adherence was significantly higher in the intervention
30
group compared to the control group (96.43% vs. 94.33%, p = 0.003). At 12
31
months, the intervention group had an increased adherence of 6.5%
32
compared to the control group (95.89% vs. 89.37%, p < 0.001). The analyses
Electronic reminders
Medicines concordance: full guideline DRAFT (July 2008)
page 290 of 373
DRAFT FOR CONSULTATION
1
were stratified by baseline compliance and language region. Over time, the
2
difference between groups increased, with approximately 17% difference in
3
adherence between groups at 300 days. 13% (n = 25) of the intervention
4
group discontinued medication before 300 days, compared to 25% (n = 51) in
5
the control group. Persistence was significantly higher in the intervention
6
group compared to the control group (87% vs. 74%, p = 0.002). This remained
7
significant when adjusted for multiple baseline variables. This study was
8
conducted in Belgium (Vrijens, B., Belmans, A., Matthys, K. et al , 2006).
9
In a study conducted by Sackett (1975) (Sackett, D. L., Gibson, E. S., and
10
Taylor, D. W., 1975) (from the Cochrane Review), subjects in one of the
11
interventions (augmented convenience) groups visited company physicians,
12
rather than their family physicians, for hypertensive and follow-up care during
13
paid working hours. Two hundred and thirty Canadian steelworkers were
14
enrolled. The ages of the participants were not reported in the study. The
15
second intervention, mastery learning, aimed to give the facts about
16
hypertension, its effects upon target organs, health, and life expectancy, the
17
benefits of antihypertensive therapy, the need for adherence with medications
18
and some simple reminders for taking pills (which was provided in a slide-tape
19
format, and reinforced by a secondary-school graduate). No statistically
20
significant results were reported. This study was conducted in Canada.
21
Mannheimer (2006) (Mannheimer, S. B., Morse, E., Matts, J. P. et al , 2006)
22
conducted cluster randomised controlled trial (2x2 factorial design) to assess
23
two interventions to increase adherence in 928 patients who were taking Anti-
24
retroviral Therapy in the USA. One intervention is the Medication Manager
25
(MM) which involved a trained research staff member working with the
26
participants individually to provide tailored adherence support according to a
27
standardised protocol, identifying and addressing information, motivation and
28
skills for antiretroviral adherence (using detailed questionnaires). The second
29
intervention was the electronic medication reminder system, a small portable
30
alarm (ALR) which was programmed to flash and sound when antiretroviral
31
doses were due. Participants were followed up with assessments at 1 and 4
32
months after randomisation and 4 months thereafter. The MM group had
Medicines concordance: full guideline DRAFT (July 2008)
page 291 of 373
DRAFT FOR CONSULTATION
1
significantly higher reporting of 100% adherence over time compared to non-
2
MM interventions (OR=1.42, p<0.001). There were no significant differences
3
between the ALR group and the non-ALR groups for adherence.
4
8.12.2.5
5
De Geest (2006) (De, Geest, S, Schafer-Keller, P., Denhaerynck, K. et al ,
6
2003) compared a nurse-led intervention and usual care with usual care alone
7
in patients who had undergone a kidney transplant with negative results. This
8
was a small study that comprised a total of 18 participants aged > 18 years.
9
The intervention group received one home visit and three telephone
Type of reminders used not stated
10
interviews, one at the end of the month for three consecutive months. During
11
the home visit printouts were discussed with patient for problem detection,
12
and adherence goals were made. All patients received self-efficacy
13
interventions. Nurses also implemented additional educational, behavioural
14
(e.g. the use of reminders) and/or social support interventions if they felt this
15
would help the patient. Telephone calls served to discuss adherence in
16
previous month, to check on health status, and discuss and change, if
17
appropriate, adherence interventions. This study was conducted Switzerland.
18
Adherence increased in both groups over the first 3 months (p = 0.04). The
19
overall difference between groups was non-significant at 3 months (p = 0.31)
20
and at 9 months (p = 0.58) with a gradual decline over the total 9 months to a
21
level still higher than initial levels (De, GeestS, Schafer-Keller, P.,
22
Denhaerynck, K. et al , 2003).
23
24
Medicines concordance: full guideline DRAFT (July 2008)
page 292 of 373
DRAFT FOR CONSULTATION
1
2
8.13
Key Clinical Question : Does being involved in self-
3
monitoring (e.g. of own blood pressure) increase
4
adherence to prescribed medication?
Related
Evidence statements
references
(summary of
Evidence into recommendations
evidence)
Haynes 1976
1. The majority of
For general discussion of limitations of evidence se
(Haynes, R. B.,
evidence suggests that
section 7.3
Sackett, D. L.,
involving patients in the
and Gibson, E.
self-monitoring of their
S., 1976); Sadik
medication adherence
2005 (Sadik, A.,
(e.g. through recording
Yousif, M., and
adherence in diary
McElnay, J. C.,
logs) appears to
2005); Tsuyuki
increase adherence as
2004 (Tsuyuki,
part of a multi
R. T., Fradette,
component
M., Johnson, J.
intervention.
The evidence of the value of self-monitoring in
increasing adherence was conflicting. The GDG
considered that this was perhaps not surprising
given that qualitative evidence had indicated that
patients may use such measures to inform their ow
decisions and evaluations of treatments rather tha
to ensure they follow a previous decision. Self
monitoring is also used in conditions where the
patient can alter treatment according to results and
this group will have different characteristics and
A. et al , 2004);
needs than a general patient group.
Sadik 2005
2. There is conflicting
(Sadik, A.,
evidence in regard to
Yousif, M., and
whether having
McElnay, J. C.,
patients record
2005); Bailey
information (e.g. in a
1990 (Bailey, W.
diary log) relevant to
C., Richards-Jr,
their condition (e.g.
J. M., Brooks, C.
symptoms) can
M. et al , 1990);
increase adherence as
Medicines concordance: full guideline DRAFT (July 2008)
page 293 of 373
DRAFT FOR CONSULTATION
Cote 1997 (Coté, part of a multi
J., Cartier, A.,
component
Robichaud, P. et
intervention.
al , 1997); Cote
2001(Côté, J.,
Bowie, D. M.,
Robichaud, P. et
al , 2001);
Friedman 1996
(Friedman, R. H.,
Kazis, L. E.,
Jette, A. et al ,
1996); Haynes
1976 (Haynes,
R. B., Sackett, D.
L., and Gibson,
E. S., 1976);
Johnson 1978
(Johnson, A. L.,
Taylor, D. W.,
and Sackett, D.
L., 1978); Morice
2001(Morice, A.
H. and Wrench,
C., 2001);
Peterson (1984)
(Peterson, G. M.,
Fitzmaurice, K.
D., Naunton, M.
et al , 2004)
Haynes 1976
3. There is conflicting
(Haynes, R. B.,
evidence with regard to
Medicines concordance: full guideline DRAFT (July 2008)
page 294 of 373
DRAFT FOR CONSULTATION
Sackett, D. L.,
whether providing
and Gibson, E.
participants with
S., 1976); Cote
information on how to
1997 (Coté, J.,
adjust their treatment
Cartier, A.,
based on their own
Robichaud, P. et
self-monitoring effects
al , 1997);
adherence.
Morice 2001
(Morice, A. H.
and Wrench, C.,
2001); Bailey
1990 (Bailey, W.
C., Richards-Jr,
J. M., Brooks, C.
M. et al , 1990);
Bailey (1999)
(Bailey, W. C.,
Richards-Jr, J.
M., Brooks, C.
M. et al , 1990); ;
Cote, 2001
(Cote, J., Bowie,
D. M.,
Robichaud, P. et
al , 2001).
Morice 2001
4. There is conflicting
(Morice, A. H.
evidence that involving
and Wrench, C.,
patients more in their
2001); Cote
care through the self-
1997 (Coté, J.,
monitoring of
Cartier, A.,
respiratory function
Robichaud, P. et
through measurement
Medicines concordance: full guideline DRAFT (July 2008)
page 295 of 373
DRAFT FOR CONSULTATION
al , 1997); Bailey
of PEF increases
1990 (Bailey, W.
adherence as part of a
C., Richards-Jr,
multi component
J. M., Brooks, C.
intervention.
M. et al , 1990);
Cote 2001(Cote,
J., Bowie, D. M.,
Robichaud, P. et
al , 2001); Levy
(2000) (Levy, M.
L., Robb, M,
Allen, J et al ,
2000).
Friedman 1996
5. There is conflicting
(Friedman, R. H., evidence with regard to
Kazis, L. E.,
whether involving
Jette, A. et al ,
patients in the self-
1996); Haynes
monitoring of their
1979 (Haynes,
blood pressure
R. B., Sackett, D. improves adherence as
L., and Gibson,
part of a multi
E. S., 1976);
component
Johnson 1978
intervention.
(Johnson, A. L.,
Taylor, D. W.,
and Sackett, D.
L., 1978);
MarquezContreras (2006)
(MarquezContreras, E.,
Medicines concordance: full guideline DRAFT (July 2008)
page 296 of 373
DRAFT FOR CONSULTATION
Martell-Claros,
N., Gil-Guillen,
V. et al , 2006);
Rudd (2004
(Rudd, P., Miller,
N. H., Kaufman,
J. et al , 2004)
1
2
*e.g. relative value place on different outcomes, trade-offs between benefits and harms, uncertainties in Evidence, likely impact if implements,
feasibility, patient perspective
3
4
8.13.1 Methods of the evidence review
5
This paper includes a narrative summary of the included evidence, structured
6
according to the category of the intervention, following the agreed reviewing
7
protocol:
8
Types of studies - randomized controlled trials (RCTs) of interventions to
9
increase adherence. The excluded studies list from the Cochrane review was
10
checked as we have included those studies with less than 80% follow-up of
11
participants.
12
Types of participants - people prescribed medication for a medical condition.
13
Duration of studies - six months follow up from the time of patient entry for
14
long-term regimens. No time limit specified for short-term conditions.
15
Types of interventions - any interventions intended to change adherence to
16
prescribed medication. As the Cochrane review is presented by condition, we
17
have used the evidence extracted in that review and reconfigured it by
18
intervention.
19
Types of outcome measures – inclusion criteria (as defined in the Cochrane
20
review) were expanded by including studies that used adherence as the only
21
outcome variable as opposed to adherence and treatment outcome variables.
Medicines concordance: full guideline DRAFT (July 2008)
page 297 of 373
DRAFT FOR CONSULTATION
1
The excluded studies list of the Cochrane review was cross-referenced to
2
ensure that no potentially relevant study was missed out.
3
8.13.2 Evidence review
4
Although a handful of studies were found which addressed self monitoring we
5
did not find any study where the study interventions sole purpose was to
6
address self monitoring (although arguably a few appear in the Cochrane
7
Review studies). Within the studies which did address self monitoring there
8
was variability in how they did this and to what extent self monitoring was a
9
focus of the intervention. Narratives of each study, which give full details of
10
the entire intervention used in each individual study, are summarised below.
11
Tsuyuki (2004) (Tsuyuki, R. T., Fradette, M., Johnson, J. A. et al , 2004)
12
compared a patient support program with usual care in patients with heart
13
failure with negative results. Seven hundred and sixty six patients aged above
14
18 years were enrolled in the study. The mean age of the patients was 74
15
years. Patients in the support group received educational material consisting
16
of information about heart failure, non-drug treatment, medication information
17
(with special emphasis on proven benefits of therapies) and self-monitoring,
18
all written at a grade 8 reading level. Patients also received adherence aids
19
including a medication organizer, medication administration schedule, and
20
daily weight log. Community follow-up in the patient support program
21
consisted of telephone contact by the local research coordinator at 2 and 4
22
weeks and then monthly there after for 6 months after discharge The
23
telephone contact was to reinforce education and adherence relating to heart
24
failure and other self-care activities, focusing on the 5 essential components:
25
salt and fluid restriction, daily weighting, exercise alternating with rest periods,
26
proper medication use and knowing when to call their physician. This study
27
was conducted in Canada. ACE inhibitor adherence over 6 months after
28
discharge was 86.2 ± 29% in the usual care group vs. 83.5 ± 29% in the
29
patient support group (p = NS).
30
Sadik (2005) (Sadik, A., Yousif, M., and McElnay, J. C., 2005) compared a
31
pharmacist delivered intervention with usual care in patients with heart failure
Medicines concordance: full guideline DRAFT (July 2008)
page 298 of 373
DRAFT FOR CONSULTATION
1
with positive results. A total of 221 patients were enrolled in the study. Mean
2
age of participants was 58 years. For the intervention group patients the
3
research pharmacist discussed with their physicians if rationalization of drug
4
therapy or simplification of dosage regimens were considered appropriate.
5
Intervention patients were also educated (in a structured fashion) on HF, their
6
prescribed medication and the management of HF symptoms by the research
7
pharmacist. A printed booklet developed for this type of education programme
8
was used and each patient was given a copy to take home. The booklet
9
contained information on HF, its symptoms, the aims of treatment, the types of
10
medication used and their possible side-effects, diet and lifestyle changes,
11
advice to stick to one brand of digoxin (it having a narrow therapeutic index)
12
and information on the action to take if doses of medication were missed.
13
Intervention group patients were also instructed on a self-monitoring
14
programme (signs and symptoms of HF; compliance with prescribed
15
medication) in which they were asked to become involved; a monitoring diary
16
card (covering 1 month) was used. Patients were asked to complete their
17
monitoring diary cards at home and to show them to their physicians when
18
attending an appointment. The patients were asked to return their completed
19
diary cards to the research pharmacist for review when they visited the
20
hospital to receive medication refills. Reinforcement of the educational
21
message was carried out by the pharmacist as deemed necessary. This study
22
was conducted in the United Arab Emirates.
23
The number of intervention group patients vs. control patients who exhibited
24
self-reported compliance with the prescribed medicines (85 vs. 35) and
25
lifestyle adjustment (75 vs. 29) was higher than in control group patients at 12
26
months (p < 0.05).
27
Bailey (1990) (Bailey, W. C., Richards-Jr, J. M., Brooks, C. M. et al , 1990), a
28
study from the Cochrane review, compared a multi-faceted intervention with
29
standard care in patients with recurrent episodes of wheezing with positive
30
results. A total of 267 patients aged above 18 years were enrolled in the
31
study. Patients in the intervention group were, “…provided with the
32
standardized asthma pamphlets and were provided with a skill-oriented selfMedicines concordance: full guideline DRAFT (July 2008)
page 299 of 373
DRAFT FOR CONSULTATION
1
help workbook, a one-to-one counselling session, and were subject to several
2
adherence-enhancing strategies, such as attending an asthma support group
3
and receiving telephone calls from a health educator. Physicians emphasized
4
these skills at regular clinic visits.” (Haynes, R. B., Yao, X., Degani, A. et al ,
5
2005). The Cochrane review also states that the intervention included
6
involving patients more in their care through self-monitoring of their respiratory
7
function. This study was conducted in the USA. The intervention group were
8
significantly more adherent than the control group.
9
Bailey (1999) (Bailey, W. C., Richards-Jr, J. M., Brooks, C. M. et al , 1990), a
10
study from the Cochrane review, conducted a randomised controlled trial of
11
asthma self-management. 236 patients stratified by moderate or severe
12
asthma in Alabama, USA, were randomised to the University of Alabama
13
Birmingham (UAB) asthma self-management group, the UAB core-elements
14
group and usual care group. The core components involved a skill-oriented
15
self-help asthma workbook, which the patients were counselled in for one
16
hour. The participants also received 2 telephone calls and a letter at 1, 2 and
17
4 weeks to discuss problems and peak flow readings (see Bailey 1990). The
18
core elements program involved a shorter version of the workbook of which
19
counselling was given briefly (15-20 minutes). They were trained to use
20
inhalers and peak flow meters. Participants were followed up by telephone a
21
week later and a letter 2 weeks later. The usual care group received usual
22
education from their GP and informational leaflets. There were no significant
23
differences between groups in adherence.
24
Cote (1997) (Coté, J., Cartier, A., Robichaud, P. et al , 1997), a study from the
25
Cochrane review, compared two intervention groups and one control group in
26
patients with moderate to severe asthma with negative results. A total of 188
27
patients aged above 16 years were enrolled in the study. The intervention was
28
an “…asthma education program with an action plan based on peak-flow
29
monitoring (Group P) or an action plan based on asthma symptoms (Group
30
S). The Control group (Group C) received instructions from their
31
pulmonologists regarding medication use and influence of allergenic and non-
32
allergenic triggers. They were taught how to use their inhaler properly by the
Medicines concordance: full guideline DRAFT (July 2008)
page 300 of 373
DRAFT FOR CONSULTATION
1
educator. A verbal action plan could be given by the physician. Groups P and
2
S received the same education as the Controls plus individual counselling with
3
the specialised educator during a 1-hour session. All participants received a
4
book titled ”Understand and Control Your Asthma“ at no extra charge. Group
5
P received a self-management plan based on peak expiratory flow (PEF).
6
They were asked to continue measuring PEF twice a day and to keep a diary
7
of the results. Each time, subjects only recorded the best of three
8
measurements. Every attempt was made to ensure that patients knew how to
9
interpret the measurement and how to respond to a change in PEF. At each
10
follow-up visit, the patient’s diary card was reviewed, and if the action plan
11
had not been implemented when required, further explanations were given
12
regarding when treatment should be modified. Group S received a self-
13
management plan based on asthma symptom monitoring. These patients
14
were asked to keep a daily diary of asthma symptom scores, using a scale of
15
0 (no symptoms) to 3 (night time asthma symptoms, severe daily symptoms
16
preventing usual activities), and adjust their medications according to the
17
severity of respiratory symptoms using the guidelines of the action plan.”
18
(Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study was conducted in
19
Canada. Neither intervention had a significant effect on participants’
20
adherence.
21
Cote (2001) (Cote, J., Bowie, D. M., Robichaud, P. et al , 2001), a study from
22
the Cochrane review, compared two different educational interventions with
23
usual care for adult patients consulting with an acute asthma exacerbation
24
with negative results. A total of 126 patients aged above 18 years entered the
25
study. Patients in the Group Limited Education (LE) group were given a self-
26
action plan that was explained by the on-call physician. The action plan used
27
”traffic lights“ (green, yellow, red) to describe specific states of asthma control
28
based on Peak Expiratory Flow and symptoms and actions that the patient
29
should take for each state. Patients in a ”Structured Educational group (SE)“,
30
in addition to what patients in Group LE received, participated in a structured
31
asthma educational program based on the PRECEDE model of health
32
education. This model took into consideration three different issues that were
33
important when dealing with health-related behaviours: predisposing factors
Medicines concordance: full guideline DRAFT (July 2008)
page 301 of 373
DRAFT FOR CONSULTATION
1
(belief, attitude, knowledge), enabling factors (community resource, family
2
support), and reinforcement…Reinforcement was provided at the 6-month
3
follow-up visit…the teaching was provided individually or in small groups
4
according to patient preference…Both intervention groups also received usual
5
care.” (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study was
6
conducted in Canada.
7
Neither intervention had a significant effect on participants’ adherence.
8
Levy (2000) (Levy, M. L., Robb, M, Allen, J et al , 2000) compared a nurse
9
delivered intervention with usual care in patients with asthma (n = 211) with
10
positive results. The intervention group were, “…invited to attend a 1h
11
consultation with one of the nurses beginning 2 weeks after entry to the study,
12
followed by two or more lasting half an hour, at 6-weekly intervals. The
13
second and third could be substituted by a telephone call. Patients were
14
phoned, by the nurse before each appointment in order to improve attendance
15
rates. Patient’s asthma control and management were assessed followed by
16
education on recognition and self-treatment of episodes of asthma. The
17
patients were taught to step-up medication when they recognized uncontrolled
18
asthma using PEF or symptoms. The advice was in accordance with national
19
guideline. Prescriptions were obtained from one of the doctors in the clinic or
20
by providing the patient with a letter to their general practitioner. Patients
21
presenting with severe asthma (severe symptoms of PEF below 60% of their
22
best/normal)were referred immediately to the consultant.” (Haynes, R. B.,
23
Yao, X., Degani, A. et al , 2005). This study was conducted in the UK
24
Self-reported compliance was significantly higher in the intervention group for
25
use of inhaled topical steroids and rescue medication for severe asthmatic
26
attacks, but there was no significant difference between the groups for use of
27
these medications for mild attacks.
28
Friedman (1996) (Friedman, R. H., Kazis, L. E., Jette, A. et al , 1996), a study
29
from the Cochrane review, compared a telephone-linked computer system
30
(TLC) intervention for monitoring and counselling patients with usual care in
31
patients with hypertension with positive results. Two hundred and sixty seven
Medicines concordance: full guideline DRAFT (July 2008)
page 302 of 373
DRAFT FOR CONSULTATION
1
patients aged ≥ 60 years were recruited for the study. The mean age was 76
2
years for the TLC group and 77 years for the usual care group. TLC is, “…an
3
interactive computer-based telecommunications system that converses with
4
patients in their homes, using computer-controlled speech, between office
5
visits to their physicians. The intervention patients would call the TLC on a
6
weekly basis. Before calling, subjects would record their own blood pressure
7
using an automated sphygmomanometer with a digital readout. During the
8
conversation, subjects would answer a standard series of questions and the
9
TLC would provide education and motivational counselling to improve
10
medication adherence. The TLC then transmitted the reported information to
11
the subject’s physician”. (Haynes, R. B., Yao, X., Degani, A. et al , 2005). This
12
study appears to have been conducted in the USA.
13
The unadjusted results did not demonstrate significant improvement in
14
compliance or clinical outcome in patients using TLC as compared to those
15
patients receiving usual care. However, when the data were adjusted for age,
16
sex, and baseline adherence, the patients using TLC demonstrated a greater
17
improvement in medication adherence than those receiving usual care (p <
18
0.05). Sub-group analysis showed, in people who were non-adherent at
19
baseline, patients using TLC had greater improvement in medication
20
compliance (p < 0.05) than those receiving usual care. In people who were
21
adherent at baseline, TLC showed no significant difference in adherence
22
between the two groups over the course of the trial.
23
Haynes (1976) (Haynes, R. B., Sackett, D. L., and Gibson, E. S., 1976), a
24
study from the Cochrane review, compared a multi-component intervention
25
with usual care in patients with high blood pressure with positive results. Thirty
26
nine patients were enrolled in the study. It is not clear from the study the ages
27
of the patients. Patients in the intervention group were, “…all taught the
28
correct method to measure their own blood pressures, were asked to chart
29
their home blood pressures and pill taking, and taught how to tailor pill taking
30
to their daily habits and rituals. They also visited fortnightly at the worksite a
31
high-school graduate with no formal health professional training who
32
reinforced the experimental manoeuvres and rewarded improvements in
Medicines concordance: full guideline DRAFT (July 2008)
page 303 of 373
DRAFT FOR CONSULTATION
1
adherence and blood pressure. Rewards included allowing participants to
2
earn credit, for improvements in adherence and blood pressure, which could
3
be applied towards the eventual purchase of the blood pressure apparatus
4
they had been loaned for the trial”. (Haynes, R. B., Yao, X., Degani, A. et al ,
5
2005). This study was conducted in Canada. There was a significant increase
6
in adherence associated with the intervention.
7
Johnson (1978) (Johnson, A. L., Taylor, D. W., and Sackett, D. L., 1978), a
8
study from the Cochrane review, compared four groups (1) self-recording and
9
monthly home visits, (2) self recording only, (3) monthly home visits, and a
10
control group consisted of (4) neither self-recording nor home visits with
11
negative results. One hundred and forty patients aged between 35 and 65
12
years were included in the study. Patients receiving antihypertensive
13
medications were studied. Participants, “…in groups (1) and (2) received a
14
blood pressure kit and instruction in self-recording. Patients in the self-
15
recording groups were to keep charts of their daily blood pressure readings
16
and were instructed to bring these charts to their physician at each
17
appointment. Subjects in groups (1) and (3) had their blood pressure
18
measured in their homes every four weeks, and the results were reported to
19
both the patient and the physician”. (Haynes, R. B., Yao, X., Degani, A. et al ,
20
2005). This study was conducted in Canada. There was no effect on
21
adherence from either intervention.
22
Marquez-Contreras (2006) (Marquez-Contreras, E., Martell-Claros, N., Gil-
23
Guillen, V. et al , 2006), included in the Cochrane Updated Review,
24
conducted a randomised controlled trial of a programme of home blood
25
pressure management (HBPM) in patients with mild-to-moderate arterial
26
hypertension. This study was conducted in 40 primary care centres in Spain.
27
250 patients were included with data for 226. Mean age of participants was 59
28
years, and around 50% females/males. The no. of diseases was significantly
29
higher in the intervention group 2.6 (SD=1.6) vs 2.2 (SD=1.2), p=0.023).
30
Patients in the control group received usual GP care and the intervention
31
group received the intervention from GPs plus an OMRON automatic monitor
32
for HBPM. The programme was measuring their blood pressure 3 days a
Medicines concordance: full guideline DRAFT (July 2008)
page 304 of 373
DRAFT FOR CONSULTATION
1
week (Tuesdays, Thursdays and Saturdays), twice before breakfast and twice
2
before supper and record the results on a card. 74% of the control group and
3
92% of the intervention group were compliant (measured by MEMS) (95% CI
4
63.9-84.1 and 86.7-97.3, p=0.0001); the mean percentage compliances were
5
87.6% for the control group and 93.5% for the intervention group (95% CI
6
81.2-94.0 and 88.7-98.3, p=0.0001); the percentage of days the drug was
7
taken correctly were 83.6% and 89.4; the percentages of participants taking
8
medication at the correct time was 79.89 vs. 88.06.
9
Rudd (2004) (Rudd, P., Miller, N. H., Kaufman, J. et al , 2004) (included in
10
the Cochrane Updated Review) conducted a RCT of a system for patients to
11
monitor their own blood pressure. Patients of two medical clinics in California
12
were randomised to receive routine care (n=76) or an automated blood
13
pressure device at home with management by a nurse care manager (n=74).
14
The mean age for the intervention was 59 and 60 for the control group. 50%
15
of the intervention and 56% of the control group were female. Patients
16
recorded their blood pressure twice a day at the same time using the semi-
17
automated portable device. At the end of the week the device printed a report
18
of up to 14 measurements and every two weeks patients were to mail the
19
values on the printout to the nurse care manager who used the data to guide
20
drug therapy. Any new blood pressure drug initiation was requested from the
21
doctor. The primary outcome was change of BP at 6 months. Secondary
22
analyses were made for frequency of drug changes and adherence.
23
Adherence was measured by a drug event monitor (a microchip in the pill
24
bottle lid of the most frequently used medication) and the patients were
25
required to return to the clinic at 3 and 6 months so this data could be
26
downloaded, although this was not used as feedback to patients, physicians
27
or nurse care managers. The mean daily adherence rate was 80% (SD= 23%)
28
for the intervention group and 69% (SD=31%) for the control group, p=0.03).
29
Morice (2001) (Morice, A. H. and Wrench, C., 2001), a study from the
30
Cochrane review, compared an asthma-nurse led intervention with routine
31
care in patients with asthma with negative results. A group of 80 patients (53
32
women) aged between 16 and 72 years were included in the study. Mean age
Medicines concordance: full guideline DRAFT (July 2008)
page 305 of 373
DRAFT FOR CONSULTATION
1
was 36.1 years. Compared with the control group, patients in the educational
2
intervention group had a minimum of two separate sessions, lasting on
3
average 30 minutes each. These were carried out on an individual basis. The
4
first session involved discussion on the basic mechanisms of asthma,
5
including common triggers and an explanation of the changes which occur to
6
the airways resulting in the symptoms experienced by the patient. This was
7
supported by illustrations in the ‘Regular Therapy with Asthma’ booklet which
8
was given to each intervention group patient. Lifestyle influences, such as
9
occupation and leisure activities were discussed where appropriate to the
10
individual. The need for ‘preventer’ and ‘reliever’ medication was also
11
emphasized during this session. Patients were encouraged to actively
12
participate in the session and relatives were included at the patients’ request.
13
The second session took place on the following day. Previously given
14
information was briefly summarized with input from the patient as a means of
15
checking understanding. An agreed individualized self-management plan was
16
determined, with written instructions using the ‘Sheffield Asthma Card’. This
17
also contained a telephone contact number. Each patient was given a peak
18
low meter to take home and instructions on monitoring, with documentation of
19
predicted peak low measurement and parameters for altering treatment, as
20
well as clear written guidelines on when to seek emergency care. Home
21
intervention was based upon a combination of symptoms and peak low
22
recordings and all guidance offered throughout the educational programme
23
was based on the BTS guidelines for the management of asthma in adults”
24
(Haynes, R. B., Yao, X., Degani, A. et al , 2005). This study was conducted in
25
the UK. There were no significant improvements in compliance at six months.
26
Medicines concordance: full guideline DRAFT (July 2008)
page 306 of 373
DRAFT FOR CONSULTATION
1
2
8.14
Key Clinical Question : Does the use of dosette boxes
increase adherence to prescribed medication?
Related
Evidence statements
references
(summary of
Evidence into recommendations
evidence)
1. No evidence was
The GDG were aware of regular use in the
found on the effects of
community of compliance aids such as dosette
dosette boxes in
boxes in the community. When general searches
increasing adherence
did not locate evidence regarding their use the GD
to prescribed
asking for specific searches. The GDG recognized
medication
the potential use of these devices for patients who
may have wish to organize their treatment regimen
but did not consider it appropriate to make a gener
recommendation about their use.
3
4
8.14.1 Methods of the evidence review
5
This paper includes a narrative summary of the included evidence, structured
6
according to the category of the intervention, following the agreed reviewing
7
protocol:
8
Types of studies – no restrictions on study design.
9
Types of participants- people prescribed medication for a medical condition.
10
Duration of studies - no time limit specified.
11
Types of interventions - any interventions intended to assess the correlation
12
between the use of dosette boxes and the impact on adherence to prescribed
13
medication.
14
8.14.2 Evidence review
15
No studies that were relevant to this clinical question were retrieved.
16
Medicines concordance: full guideline DRAFT (July 2008)
page 307 of 373
DRAFT FOR CONSULTATION
1
9
Reviewing medicines
2
9.1
Recommendations *listed at the start of the guideline
according to chapter
3
4
9.2
5
Review of medicines and medicine taking is seen as an important aspect of
6
health care. Professionals involved in prescribing and dispensing of medicines
7
are currently reimbursed for reviewing medicines. General practitioners in the
8
UK are remunerated for medication review via the Quality and Outcomes
9
Framework (QOF). Community pharmacists are reimbursed for carrying out
10
reviews which are called Medicines Use Reviews (MURs). The Dispensing
11
Review of Use of Medicines (DRUM) is part of the Dispensing Services
12
Quality Scheme for GP surgeries.
13
The terminology in this area is not standardised and is subject to change. The
14
Medicines Partnership Programme 15 defined medication review as ‘a
15
structured, critical examination of a patient’s medicines with the objective of
16
reaching an agreement with the patient about treatment, optimising the impact
17
of medicines, minimising the number of medication-related problems and
18
reducing waste’. It is implicit in this definition that the patients is involved. In
19
‘Room for Review’ in 2002 they suggested four levels of medication review –
20
level 0 which is an ad-hoc opportunistic review; level 1 a prescription review
21
which is a technical review of a patients list of medicines; level 2 is a
22
treatment review which is a review of medicines with the patients full notes
23
and level 3 which is a clinical medication review which is a face-to face review
24
with patients of medicine and condition. A review with the patient’s notes but
25
not necessarily with the patient (as in level 2 as described above) fulfils the
26
criteria for QoF. An MUR is described as a one-one conversation between
27
people and pharmacists that are designed to identify any problems a person is
28
experiencing with their medicines (Pharmacy in England White paper 2008).
29
Community pharmacists carrying out these reviews will not generally have
15
Introduction
http://www.npc.co.uk/med_partnership/assets/room_for_review.pdf
Medicines concordance: full guideline DRAFT (July 2008)
page 308 of 373
DRAFT FOR CONSULTATION
1
access to clinical information about patients. The recent Pharmacy in England
2
White Paper (2008)16 reports that many people report satisfaction with this
3
service but longer term impacts can not be assessed. The White Paper
4
reports that government plans for MUR services to be prioritised to meet
5
health needs and ensuring funding rewards health outcomes.
6
The National Prescribing Centre has recently revisited the topic in A Guide to
7
Medication Review (2008). The guide aims to advise those providing and
8
commissioning medication reviews. This characterises 3 types of medication
9
review with an emphasis on the purpose of the review: Type 1 Prescription
10
review; Type 2 Concordance and compliance review and Type 3 Clinical
11
medication review. The three types of medication review replace the earlier
12
levels of medication review. This reclassification appears to make clearer the
13
role of the review and the place of the patient and clinical information in
14
different types of review.,
15
16
The GDG were interested in whether there was any evidence that medication
17
review improved either shared decision-making or adherence. In this context
18
medication review has to involve a face to face meeting with professionals
19
and patient. The professional involved was not pre-defined. The evidence
20
search used ‘medication review’ as a generic term.
21
16
http://www.official-documents.gov.uk/document/cm73/7341/7341.pdf
Medicines concordance: full guideline DRAFT (July 2008)
page 309 of 373
DRAFT FOR CONSULTATION
1
2
3
9.3
Key Clinical Question: does medication review increase
shared decision–making or adherence?
Related references
Evidence
Evidence into recommendations
statements
(summary of
evidence)
All retrieved
1. There is
The GDG considered that review of prescribed
evidence.
conflicting evidence
medicines is most commonly undertaken in
with regards to
clinical settings as part of management of
whether medication
patients and their medical problems. In this
review increases
setting it is seen as integral to continuing care
adherence.
and not separate from it. The GDG considered
that all levels of medication review as described
Lowe (2000) (Lowe,
2. Five RCTs
C. J., Raynor, D. K.,
conducted in the UK
Purvis, J. et al ,
shows that
2000); Sturgess
medication review
(2003) (Sturgess, I.
increased
K., McElnay, J. C.,
adherence to
Hughes, C. M. et al ,
prescribed
2003); Nazareth
medication.
in ‘Room for Review’ take place in this setting
and have a role. Revisiting a decision to
prescribe medicines and exploring patients
medicine-taking behaviour was considered by
the GDG to be part of the dynamic process that
long term medicine prescribing required.
The research evidence primarily addresses
(2001) (Nazareth, I.,
medication reviews that take place separate
Burton, A., Shulman,
from the delivery of clinical care, often by
S. et al , 2001);
practitioners who do not have access to clinical
Bernsten (2008)
history and notes. These have been a recent
(Bernsten, C,
development and have primarily involved
Bjorkman, I,
pharmacists. Most of the evidence on reviews
Caramona, M et al ,
by pharmacists comes from studies that
2001); Begley (1997)
targeted older adults on multiple medications.
(Begley, S.,
Many studies include quite complex
Medicines concordance: full guideline DRAFT (July 2008)
page 310 of 373
DRAFT FOR CONSULTATION
Livingstone, C.,
pharmaceutical care programmes where the
Hodges, N. et al ,
interventions consists of a number of
1997).
components including education and follow up
which the GDG considered more intensive than
Lipton (1994) (Lipton,
3. There is
H. L. and Bird, J. A.,
conflicting evidence
1994); Hanlon (1996)
from six RCTs
(Hanlon, J. T.,
conducted outside
Medication review can have benefits for the
Weinberger, M.,
the UK that
patient but evidence was conflicting whether
Samsa, G. P. et al ,
medication review
this led to improvements in adherence to
1996); Chisholm
increased
prescribed medication.
(2001) (Chisholm, M.
adherence to
A., Mulloy, L. L.,
prescribed
Jagadeesan, M. et al
medication.
is currently provided in any type of medication
review provided in the UK.
The GDG were particularly concerned that
reviews of medication carried out remote from
the clinical settings needed to feed back to
, 2001); Taylor (2003)
clinicians who were involved in prescribing and
(Taylor, C. T., Byrd,
other aspects of care. Increasing the number of
D. C., and Krueger,
medication reviews and the personnel involved
K., 2003); Grymonpre
in carrying them out might not be effective if
(2001) (Grymonpre,
communication and follow up is not achieved.
R. E., Williamson, D.
A., and Montgomery,
The GDG were clear that the lack of research
P. R., 2001);
on reviews conducted as part of clinical care
Sookanekun (2004)
should not indicate that these were not of value.
(Sookaneknun, P.,
Review of medications will continue to be part
Richards, R. M.,
of delivery of health care. As responses to
Sanguansermsri, J. et
medication can change over time, both in terms
al , 2004)
of patient behaviour (adherence) and clinical
Grymonpre
4. Medication review
(2001)(Grymonpre, R. was carried out by
E., Williamson, D. A.,
pharmacists in all of
and Montgomery, P.
the RCTs, except for
R., 2001)
one RCT where a
trained volunteer
outcomes, a process of medication review is
likely to be necessary and be part of on-going
processes of decision-making and medicinetaking. An informal review of medication should
continue as part of good clinical practice but it is
not possible to be recommend precise timings
Medicines concordance: full guideline DRAFT (July 2008)
page 311 of 373
DRAFT FOR CONSULTATION
undertook the
for formal medication reviews outside the
review which was
clinical setting.
then reviewed by a
pharmacist
consultant.
Lowe (2000)(Lowe,
5. Most of the RCTs
C. J., Raynor, D. K.,
included only
Purvis, J. et al ,
participants over 65
2000); Hanlon
years old.
(1996)(Hanlon, J. T.,
Weinberger, M.,
Samsa, G. P. et al ,
1996); Grymonpre
(2001)(Grymonpre, R.
E., Williamson, D. A.,
and Montgomery, P.
R., 2001); Nazareth
(2001)(Nazareth, I.,
Burton, A., Shulman,
S. et al , 2001);
Taylor (2003)(Taylor,
C. T., Byrd, D. C.,
and Krueger, K.,
2003); Bernsten
(2008) (Bernsten, C,
Bjorkman, I,
Caramona, M et al ,
2001); Begley (1997)
(Begley, S.,
Livingstone, C.,
Hodges, N. et al ,
1997); Sturgess
Medicines concordance: full guideline DRAFT (July 2008)
page 312 of 373
DRAFT FOR CONSULTATION
(2003) (Sturgess, I.
K., McElnay, J. C.,
Hughes, C. M. et al ,
2003)
Medicines concordance: full guideline DRAFT (July 2008)
page 313 of 373
DRAFT FOR CONSULTATION
Methods of the evidence review
1
9.3.1
2
The titles and abstracts of studies retrieved by an electronic search for
3
medication review were scanned for relevance to the question of whether
4
medication review increases adherence to medications. Any potentially
5
relevant publications were obtained in full text. These were then reviewed to
6
identify the most appropriate evidence to help answer the question and to
7
ensure that the recommendations are based on the best available evidence.
8
This process required four main tasks: selection of relevant studies;
9
assessment of study quality; synthesis of the results; and grading of the
10
evidence.
11
This paper includes a narrative summary of the included evidence, following
12
the agreed reviewing protocol:
13
Types of studies - randomized controlled trials (RCTs) of medication review
14
interventions to increase adherence.
15
Types of participants - people prescribed medication for a medical condition.
16
Medication review performed by any healthcare professional or trained
17
personnel.
18
Setting- Carried out in the community.
19
Duration of studies - No time limit specified for this evidence review.
20
Types of interventions - any medication review ( as implying face to face
21
meeting between the patient and the health care professional doing the
22
review) interventions intended to change adherence to prescribed medication.
23
The content and delivery of interventions are not standardized in the literature.
24
The term ‘pharmaceutical care programme’ is used and this applies to
25
pharmacist led programmes which assess medication use, develop an
26
intervention and provide long term follow up to patients including liaison with
27
the prescriber. Some of these interventions provide intensive support.. Some
28
subjective assessment of studies was required as content is often not well
29
defined. Many of studies on interventions to increase adherence used
Medicines concordance: full guideline DRAFT (July 2008)
page 314 of 373
DRAFT FOR CONSULTATION
1
pharmacists to carry out the intervention but we have included here only
2
studies that were carried out in the community and were providing general
3
review rather than disease specific support.
4
Types of outcome measures - any prescribed medication adherence
5
outcomes which changed as a result of the medication review. Outcomes
6
relevant to patient involvement were reported as part of the evidence review.
7
9.3.2
8
Of the RCTs found relating to medication review, many had to be excluded. as
9
they did not have adherence outcomes. Instead they focused on hospital re-
10
admissions, care home admissions, death and cost effectiveness. One high
11
quality RCT conducted by Zermansky (2002) (Zermansky, A. G., Petty, D. R.,
12
Raynor, D. K. et al , 2002) for HTA could not be included as there were no
13
specific adherence outcomes Zermansky (2002) studied whether a trained
14
pharmacist could conduct effective clinical medication reviews of elderly
15
patients who were on repeat prescriptions from their GP. The participants
16
were 65 years or over on repeat medication, who were not resident in a
17
nursing or residential home and were not terminally ill. The study lasted 12
18
months and the intervention involved the pharmacist assessing the patient,
19
their illnesses and their medication regimen and making recommendations.
20
The primary outcome measure was the number of repeat medication changes
21
per patient, which was 2.2 in the intervention group and 1.9 in the control
22
group (Difference of 0.31, 95% CI 0.06 to 0.57), p=0.02). The secondary
23
outcome was the effect on cost of medication. There was a rise in repeat
24
medication items for both groups, but this was significantly less for the
25
intervention group (intervention mean 0.2, SD 1.55; control mean 0.4, SD
26
1.53, difference -0.2, 95% CI, -0.4 to-0.1). The cost saving for the intervention
27
group compared to the control group was £4.75 per 28-day month, a total of
28
£61.75 per patient per year.
29
A systematic review (Holland 2007) focusing specifically on pharmacist-led
30
medication review was found. However the primary outcome of interest was
31
reduction of hospital admissions and deaths in older people and adherence
Evidence review
Medicines concordance: full guideline DRAFT (July 2008)
page 315 of 373
DRAFT FOR CONSULTATION
1
was a secondary outcome. The studies included all forms of medication
2
review for checking and optimising the patients’ drug regimens apart from
3
those with only knowledge and/or adherence outcomes. They reported that 14
4
of the trials included adherence, with 7 reporting a significant effect and 7
5
reporting a non-significant positive effect.
6
The term ‘pharmaceutical care programme’ is also used in the literature and
7
this generally applies to pharmacist led programmes which assess medication
8
use, develop an intervention and provide long term follow up to patients.
9
Although more intensive that any programmes currently delivered in the UK
10
we included these studies as it was important to assess whether such
11
structured and intensive support was either clinically or cost effective.
12
Some subjective assessment of the studies was necessary as the content of
13
the reviews and pharmaceutical programmes is not always clearly defined.
14
9.3.2.1
15
Sturgess (2003) (Sturgess, I. K., McElnay, J. C., Hughes, C. M. et al , 2003)
16
measured a structured pharmaceutical care programme provided to elderly
17
patients by community pharmacists 191 elderly patients with a mean age of
18
73.1 ± 5.0 for the intervention group and 74.2 ± 6.3 for the control group . This
19
RCT was conducted in Northern Ireland. In the intervention pharmacists
20
assessed patients to identify drug-related problems. A number of information
21
sources were used by intervention pharmacists during this assessment
22
procedure including: the patient (via informal questioning), the patient’s GP,
23
study questionnaires and computerised medication records. During the
24
assessment, pharmacists were asked to document any identified drug-related
25
problems and to form with the patient an intervention and monitoring plan e.g.
26
education, implementation of adherence improving strategies. Pharmacists
27
visited patients at home to assess storage of medicines where problems were
28
identified.
29
Self reported compliance: between-group analysis at each assessment point
30
indicated that a significantly higher proportion of intervention patients were
31
compliant with their medicine at 12 (intervention group: 40.4%, control group:
RCTs conducted in the UK
Medicines concordance: full guideline DRAFT (July 2008)
page 316 of 373
DRAFT FOR CONSULTATION
1
24.4%) and 18 (intervention group: 47.3%, control group: 14.7%) months
2
compared to control patients (p < 0.05) (6 months: intervention group: 34.5%,
3
control group: 29.4%). Analysis of change in compliance during the study
4
(change in compliance status compared to that reported at baseline) showed
5
that a significantly higher proportion of intervention patients changed from
6
non-compliant to compliant compared to control patients (intervention 13.4%
7
vs. control 9.1%) and a significantly higher proportion of control patients
8
changed from compliant to non-compliant compared to intervention patients at
9
18 months (control 36.4% vs. intervention 4.5%).
10
Lowe (2000) (Lowe, C. J., Raynor, D. K., Purvis, J. et al , 2000) determined
11
whether a medicine review and education programme influenced elderly
12
patients’ compliance and knowledge compared to a control group in a RCT.
13
161 participants, mean age 77.5 (SD 65-96) for the intervention group and 75
14
(SD 65-88) for the control group, mainly female (67%), living with spouse or
15
relative 55% (intervention group) and 57% (control group) and prescribed an
16
average of 4 medicines (ranging from 1 to 8). The RCT was conducted in a
17
GP practice in Leeds, UK. An investigator visited patients and filled in a
18
structured questionnaire regarding their medicines, which medicine had been
19
used and patients’ understanding and ability to take medications. The
20
investigator then reported the findings to doctors where there was a need to
21
reduce dosage and discontinue medication, then liaised with the pharmacist
22
for modifications to medicine containers. At the second visit after a month they
23
delivered 1 months supply of medication and removed any other prescribed
24
medications. They discussed the regimen, the purpose of the medications
25
and the correct way to take them, with the use of a reminder chart if needed.
26
At 3 weeks follow-up participants were given a further months supply and
27
assessed on their knowledge and compliance, by counting the medications
28
left from the last visit. The mean compliance score was 91.3% for the
29
intervention group (95% CI, 89%-94%) and 79.5% for the control group (75%-
30
84%), which was significantly different (p<0.0001).
Medicines concordance: full guideline DRAFT (July 2008)
page 317 of 373
DRAFT FOR CONSULTATION
1
N.B It should be noted that that this study was under 6 months duration but
2
the patients were followed up twice at 3 week to monthly intervals and the
3
study was of particular relevance.
4
Nazareth (2001) (Nazareth, I., Burton, A., Shulman, S. et al , 2001)
5
compared patients who had been discharged from hospital with a discharge
6
plan with those who had a standard discharge letter. This RCT included 362
7
patients from four hospitals in central London. The participants had a mean
8
age was 84 years in both groups (SD 5.2 and 5.4 respectively), mainly female
9
(62% and 66%), white (97%) with a mean of three chronic medical conditions
10
and prescribed a mean of 6 drugs (SD=2). The discharge plan included
11
assessing the prescribed medication, rationalising the drug treatment and
12
assessing patients’ medication management, knowledge and support. The
13
participants were then followed up 7 to14 days later at home by community
14
pharmacists who compared medicine taking with prescribed medications and
15
their understanding and adherence to the medication regimen. They
16
intervened when necessary and provided medication counselling, disposal of
17
excess medications and liaison with GPs. There was no significant difference
18
in adherence to medicines for either group at 3 months or at 6 months (49%
19
intervention group versus 50% for the control group, with a mean difference of
20
0.02 (95% CI -0.106 to 0.126).
21
Begley (1997) (Begley, S., Livingstone, C., Hodges, N. et al , 1997)
22
assessed the influence of domiciliary pharmacy visits on medication
23
management in sample of elderly people recently discharged from hospital to
24
their own homes. Patients were aged 75 years or older. The study included
25
one intervention group receiving home visits and counselling, in which
26
structured patient interviews were conducted during the domiciliary visits and
27
consisted of six sections: patient information; drug knowledge; Patient
28
dexterity; abbreviated mental test; medication management; and compliance
29
with medication regimen. Patients were seen during 12 months. There were
30
two control groups: one which was the control and received visits only (called
31
V group), and other which was the control group that received traditional
Medicines concordance: full guideline DRAFT (July 2008)
page 318 of 373
DRAFT FOR CONSULTATION
1
pharmaceutical services with no visits except for the beginning and the end of
2
the study (NV group).
3
At each visit there were significant differences between the groups in terms of
4
distribution of patients at the various levels of compliance
5
(p<0.001).Compliance was higher at 3 months and 12 months for the
6
intervention group compared to the other control groups (p<0.001), despite
7
the low compliance value for the intervention group at the 12 month visit.
8
Patients in the intervention group who increased their compliance rates
9
between visits also increased their drug knowledge scores (p<0.005). Mean
10
scores for drug knowledge did not differ significantly between the groups at
11
any of the visits, although the mean score for the intervention group increased
12
significantly between the initial and the two weeks visits (p=0.001). There
13
were no changes for patient dexterity scores between groups at any point of
14
the study. Contacts with GP and health workers was lower for the intervention
15
group than for the control (V) in each of the four time periods (p<0.01).
16
Bernsten (2008) (Bernsten, C, Bjorkman, I, Caramona, M et al , 2001)
17
conducted a multicentre RCT in seven European Countries including the UK
18
that evaluated a pharmaceutical care programme provided to elderly patients
19
(aged 65 or older) taking 4 or more medications by community pharmacists. A
20
total of 1290 intervention patients and 1164 control patients were recruited.
21
The programme interventions included: 1) educating the patient about their
22
drug regimen and their condition; 2) implementing compliance-improving
23
interventions such as drug reminder charts; 3) rationalising and simplifying
24
drug regimens in collaboration with the patients GP. This was a continuous
25
process throughout the 18 months of the study.
26
Generally, the programme had some positive effects on humanistic health
27
outcomes such as satisfaction with treatment, and sign and symptom control,
28
and on economic outcomes, but had less impact than anticipated on drug
29
therapy, drug knowledge and compliance with medication. An analysis of
30
changes in compliance during the study indicated that at 18 months a
31
significantly higher proportion of the intervention patients changed from being
32
noncompliant to compliant compared with the control groups (p=0.028).
Medicines concordance: full guideline DRAFT (July 2008)
page 319 of 373
DRAFT FOR CONSULTATION
1
Intervention patients rated the services provided higher that the control at 6
2
and 18 months (p<0.05). There was a small statistically significant increase in
3
satisfaction in the intervention group over time (baseline vs. 12 months
4
p=0.039).
5
9.3.2.2
6
Lipton (1994) (Lipton, H. L. and Bird, J. A., 1994) assessed the impact of
7
clinical pharmacists' consultations on drug regimens, compliance, and health
8
service use of 706 geriatric hospitalized patients discharged on 3 or more
9
medications. The RCT was conducted in the USA. Mean age was 74.6 in the
10
experimental group and 74.4 in the control group. Pharmacists consulted with
11
experimental patients at discharge and 3 months thereafter, and with
12
physicians as needed. Controls received usual care. At 6-8 weeks after
13
enrolment, experimental patients were more knowledgeable about regimens
14
than controls. At 12-14 weeks, they were on fewer medications and less
15
complex regimens, and had better compliance scores p<0.001). There was no
16
effect on service use or charges, perhaps due to inadequate sample size and
17
lack of targeted drug groups analysis.
18
Hanlon (1996) (Hanlon, J. T., Weinberger, M., Samsa, G. P. et al , 1996)
19
was an RCT which compared the effects on elderly outpatients who had an
20
additional pharmacist intervention with those who received usual care from
21
their physician. Most of the patients were male (98% in the intervention and
22
100% in the control group), white (79% and 75% respectively), married
23
(65.7% intervention, 85.4% control), with baseline compliance rates of 73%
24
and 74% respectively. The mean age of participants was 70 years old. The
25
RCT was conducted in Durham, North Carolina. Before attending the
26
physician the pharmacist reviewed their medical records and medication lists
27
to ascertain their current medication use, drug-related problems and to
28
evaluate their needs by applying the Medication Appropriateness Index. Their
29
findings were then reported to the physician. The Pharmacist educated the
30
patient on drug-related problems and encouraged compliance through
31
strategies such as medication reminders and written patient materials. They
32
reviewed safe medicine use and the importance of discussing medications
RCTs conducted outside the UK
Medicines concordance: full guideline DRAFT (July 2008)
page 320 of 373
DRAFT FOR CONSULTATION
1
with physicians. There were no significant differences between the groups at
2
the end of the follow-up period with regard to medication compliance (77.4%
3
of intervention group and 76.1% of control group complied, p=0.88).
4
Chisholm (2001) (Chisholm, M. A., Mulloy, L. L., Jagadeesan, M. et al ,
5
2001) studied the compliance rates of patients who received a clinical
6
pharmacist intervention in addition to usual care compared to control patients
7
after a renal transplant. The RCT included 24 participants, 75% male with a
8
mean of 49 years (+/-10 years) and 58.3% Caucasian, 37.5% African-
9
American and 1 Hispanic. The RCT was conducted in Augusta, Georgia,
10
USA. The Pharmacist obtained medication histories and reviewed
11
medications monthly. They made recommendations to the nephrologists and
12
counselled the patients on their medication, including instructing how to use
13
the medication. Patients were encouraged to call them with any questions.
14
The patients were assessed on their understanding of their medication and
15
advised on how to enhance compliance.
16
At 12 months the compliance rate was statistically significant for the
17
intervention group 96.1% (+/- 4.7%) compared to the control group 81.6% (+/-
18
11.5%), p<0.001. For 6 of the 12 months there were differences in compliance
19
rates (64 to 100% for control group and 89 to 100% for the intervention group)
20
with the intervention group always at a higher rate (p<0.05). The duration that
21
patients complied for also differed with the intervention group remaining 75%
22
compliant each month compared to only 33.3% of the control group (p<0.05).
23
Taylor (2003) (Taylor, C. T., Byrd, D. C., and Krueger, K., 2003) conducted
24
an RCT of patients attending a community-based physician and compared
25
those who additionally received a Pharmacist intervention to a control group.
26
The majority of participants were women (63.6% in the intervention group vs
27
72.2% in the control group, p=0.445), most were white (60.6% vs 61.1%,
28
p=.966) with a mean age of 64.4 and 66.7 years respectively (p=0.467), the
29
majority were married 75.8% vs 72.2 (p=0.935) with 12 years mean
30
education. They were taking on average six medications each. The RCT was
31
set in medicine clinics in Alabama, USA.
Medicines concordance: full guideline DRAFT (July 2008)
page 321 of 373
DRAFT FOR CONSULTATION
1
The pharmacists evaluated the patients’ medication, reviewed medical
2
records and examined medication history to determine compliance and
3
complications with medication. Therapeutic recommendations were made to
4
the physicians and they made follow-up visits to patients and gave
5
individualised education and were available to answer questions. Patients'
6
responses to drugs were monitored and their medication regimens
7
consolidated, dosage frequency was reduced and medication reminders and
8
techniques for using certain devices were taught.
9
The number within the intervention group to have compliance scores of 80-
10
100% increased by 15% but there was no change for the control group (time
11
period not stated). By 12 months this difference was not significant, 100% of
12
patients in the intervention group versus 88.9% (SD 6.3) of the control group
13
had compliance scores of 80-100%, p=0.115). At baseline this was 84.9%
14
(SD=6.7) and 88.9% respectively (SD 5.8, p=0.728).
15
The most frequently sited reasons for not complying with medication were
16
forgetting to take the medications (n=10), having too many to take (n=9),
17
finding it hard to read or understand the directions (n=4) and too much trouble
18
to take (n=4).
19
Grymonpre (2001) (Grymonpre, R. E., Williamson, D. A., and
20
Montgomery, P. R., 2001) compared the impact on geriatric patients who
21
received pharmaceutical care compared to those patients who did not in a
22
RCT. Most of the patients were female (75% intervention vs 83% control,
23
p=0.254), aged 77 (SD 8.0 to 9.0), Caucasian (100%) and lived alone 61% vs
24
77% respectively, p=0.018. The RCT was conducted in a community-based
25
health clinic in Manitoba, Winnipeg, Canada. Volunteers and staff were
26
trained to conduct a comprehensive medication review which was utilised by
27
the pharmacist to identify and document potential and actual drug-related
28
issues and to address these with the patient and they physician. Their use of
29
prescribed and non-prescribed medicines, social drugs, home remedies,
30
regimen, adherence and communication with GPs, problems or side effects
31
with drugs were all assessed. The recommendations were given in a letter to
Medicines concordance: full guideline DRAFT (July 2008)
page 322 of 373
DRAFT FOR CONSULTATION
1
physicians and the patients were followed up by the pharmacist when required
2
to monitor therapeutic endpoints and sort out any problems that had arisen.
3
The mean number of mediations adhered to at follow-up was 87 (+/-46) for
4
the intervention and 85 (+/-41) for the control group, p=0.895, showing no
5
significant difference in adherence.
6
Sookaneknun (2004) (Sookaneknun, P., Richards, R. M.,
7
Sanguansermsri, J. et al , 2004) compared hypertensive patients assigned
8
to a pharmacist-involved group with those who had no pharmacist
9
involvement, with the objective of stabilising blood pressure. The participants
10
of the RCT included 235 patients, mean age 63 years old and mainly female
11
(64% in intervention and 71% in control group). The RCT was conducted in
12
Thailand. The intervention group’s blood pressure was measured every month
13
by the pharmacist and they assessed the patients understanding of
14
medications, adherence and reviewed adverse effects from the medications.
15
Medication counselling was given and drug-related problems were identified,
16
resolved and prevented. The recommendations for change of medication
17
regimen were given to the physicians. The adherence at pre-test was not
18
significantly different but at post-test the treatment group had significantly
19
increased adherence compared to the control group, Pre-test adherence of
20
80% or more was found in 51% of the treatment group and 56% of the control
21
group. At post-test this had increased to 63% for the treatment group and had
22
remained constant (55%) for the control group (p=0.014).
23
Quality of studies
24
It should be noted that the quality of many of the RCTs was low. This was
25
mainly due to the possibility of bias occurring within the methodology.
26
27
28
Medicines concordance: full guideline DRAFT (July 2008)
page 323 of 373
DRAFT FOR CONSULTATION
1
2
10
Health Economics and Medicines
Concordance
3
4
10.1
Introduction
5
Health economics is about improving the health of the population through the
6
efficient use of resources, so it necessarily applies at all levels, including
7
individual clinical decisions. Clinicians already take resources and value for
8
money into account when making clinical decisions, and the incorporation of
9
good-quality health-economic evidence into clinical guidelines can help make
10
this more consistent. The guideline development group (GDG) is required to
11
make decisions based on the best available evidence of both clinical and cost
12
effectiveness.
13
This chapter will outline the methodology, approach and findings of health
14
economic evidence that were used to inform the GDG in their decision
15
making.
16
10.1.1 How terminology of compliance, concordance and
17
adherence is used in HE literature
18
The terminology relevant to medicine taking i.e. the term compliance,
19
adherence and concordance have a different meaning when using a health
20
economic perspective. In Chapter 3, the relevant concepts have been
21
discussed. In a recent economic review paper (Elliott, Rachel A., Barber, Nick,
22
and Horne, Rob, 2005), the authors state that “compliance and adherence
23
imply patient behaviour being congruent with healthcare providers’
24
recommendation”. We found that most economic papers use adherence and
25
compliance interchangeably, despite the evident conceptual difference in
26
meaning. In accordance with this guideline, this chapter will use “shared
27
decision making” (SDM) to describe “concordance as a patient centred
28
process where health care professional (HCP) make a therapeutic alliance
29
with a patient, one result of which may be increased compliance” and
Medicines concordance: full guideline DRAFT (July 2008)
page 324 of 373
DRAFT FOR CONSULTATION
1
adherence for describing “medicine taking behaviour congruent with
2
prescriber’s recommendation”.
3
4
As discussed in chapter 3.6, the structure chosen for the guideline was to look
5
separately at decision –making about medicines and medicines taking. This
6
understanding suggests two separate processes - firstly the consultation with
7
the health care provider (HCP) where shared decision making is possible and
8
secondly the medicine taking after the consultation. This includes additional
9
contacts with dispensers and HCPs other than the prescriber as represented
10
in the care pathway. This division will be reflected in the structure of this
11
chapter.
12
13
14
10.2
Health Economics methods
15
Economic evaluation provides a formal comparison of benefits and harms as
16
well as the costs of alternative health programmes. It helps to identify,
17
measure, value and compare costs and consequences of alternative
18
treatment options. These outcomes are usually synthesised in cost
19
effectiveness (CEA) or cost utility analysis (CUA), which reflect the principle of
20
opportunity costs. For example, if a particular treatment strategy were found to
21
yield little health gain relative to the resources used, then it could be
22
advantageous to re-deploy resources to other activities that yield greater
23
health gain.
24
25
To assess the cost-effectiveness of interventions to increase adherence, we
26
conducted a comprehensive systematic review of the economic literature
27
relating to medicines and nonadherence.
28
29
In accordance with the NICE social value judgement the primary criteria
30
applied for an intervention to be considered cost effective were either:
31
Medicines concordance: full guideline DRAFT (July 2008)
page 325 of 373
DRAFT FOR CONSULTATION
1
a) The intervention dominated other relevant strategies (that is it is both less
2
costly in terms of resource use and more clinically effective compared with the
3
other relevant alternative strategies); or
4
5
b) The intervention cost less than ￡20,000 per quality-adjusted life-year
6
(QALY) gained compared with the next best strategy (or usual care)
7
8
10.2.1 Health Economic evidence review methodology
9
10
The following information sources were searched:
11
•
Medline (Ovid) (1966-June 2006)
12
•
Embase (1980-June 2006)
13
•
NHS Economic Evaluations Database (NHS EED)
14
•
PsycINFO
15
•
Cumulative Index to Nursing and Allied Health Literature (CINAHL)
16
17
The electronic search strategies were developed in Medline and adapted for
18
use with the other information databases. The clinical search strategy was
19
supplemented with economic search terms. Titles and abstracts retrieved
20
were subjected to an inclusion/exclusion criterion and relevant papers were
21
ordered. No criteria for study design were imposed a priori. In this way the
22
searches were not constrained to randomised controlled trials (RCTs)
23
containing formal economic evaluations. Papers included were:
24
•
Full/partial economic evaluations
25
•
Considered patients over 16 years of age
26
•
Written in English, and reported health economic information that could
27
be generalised to UK.
28
29
The full papers were critically appraised by a health economist using a
30
standard validated checklist. A general descriptive overview of the studies,
31
their quality, and conclusions was presented and summarised in the form of a
32
narrative review.
Medicines concordance: full guideline DRAFT (July 2008)
page 326 of 373
DRAFT FOR CONSULTATION
1
Each study was categorized as one of the following: cost-effectiveness
2
analysis or cost utility analysis (i.e. cost-effectiveness analysis with
3
effectiveness measured in terms of QALYs or life year gained). Some studies
4
were categorised as ‘cost consequences analysis’ or ‘cost minimisation
5
analysis’. These studies did not provide an overall measure of health gain or
6
attempt to synthesise costs and benefits. Such studies were considered as
7
partial economic evaluations.
8
9
10.2.2 Cost-effectiveness modelling methods
10
De novo modelling was considered for aspects of medicine taking. However,
11
due to heterogeneity in the population covered by this guideline this was not
12
possible. This is discussed in more detail in below..
13
14
15
16
10.3
Shared decision making and medicine taking from a
health economic perspective
17
10.3.1 Patient involvement and shared decision making
18
From an economic view, one aim of facilitating shared decision making (SDM)
19
on measurable outcomes is adherence to the joint decision taken. As
20
discussed previously in Chapters 3 and 4 of this guidance, the decision would
21
be taken jointly between HCP and patient and/or carer and based on the
22
appropriate information shared between them. The decision may result in the
23
agreement to prescribe and take a medication, or equally, to not prescribe and
24
take a different medication or no medication, depending on the individual
25
patient case. A programme that facilitates shared decision making between a
26
HCP and a patient can be seen as an intervention to increase adherence to
27
joint decisions including prescribed medication, and thereby health of the
28
population.
29
30
An alternative view focuses on the notion of sharing the decision itself. The
31
process of shared decision making can increase patient wellbeing by
Medicines concordance: full guideline DRAFT (July 2008)
page 327 of 373
DRAFT FOR CONSULTATION
1
improving patient satisfaction with the consultation, as well as the wellbeing of
2
possibly doctors and carers. This change in satisfaction and related wellbeing
3
can be summarised by the concept used in health economics of process utility
4
(Brouwer, Werner B. F., Exel, N. Job, Berg, Bernard van den et al , 2005).
5
Generally, being informed and involved in important decisions concerning our
6
lives, for example in an occupational setting, often increases well being and
7
satisfaction. As a result, terms such as enrichment and involvement were
8
introduced into management studies (Bambra, Clare, Egan, Matt, Thomas,
9
Sian et al , 2007). In health care, similar effects exist for patients. Thus, the
10
process and meaning of shared decision making can have an effect on well
11
being. However, it should be noted that, although many people may perceive
12
involvement in their care decisions as beneficial, not everybody will value
13
shared decision making in this way. As a result, a change in process utility
14
may both be positive or negative. Health economic research has so far mostly
15
concentrated on health-related consequences. More work is needed to
16
understand the magnitude of effects and value of process utility better.
17
18
There is a body of literature which describes and analyses peoples’
19
preferences for various combinations of attributes. Discrete choice
20
experiments, also referred to as conjoint analyses, evaluate what people
21
actually value in a product or service. In recent years health economic
22
research has started to adapt this methodology to health care. Some papers
23
exist that investigate the relative importance of consultation features. This
24
evidence has not been systematically reviewed and appraised as part of he
25
health economic evidence reviews, however, the following papers were
26
identified as examples of conjoint analyses (Albus-Christian, Schmeisser-
27
Norbert Salzberger-Bernd Faetkenheuer, 2005); (Aristides, M., Weston, A. R.,
28
FitzGerald, P. et al , 2004); (Beusterien, K. M., Dziekan, K., Flood, E. et al ,
29
2005); (Frileux-Stéphanie, Sastre-María-Teresa-Muñoz Mullet-Etienne Sorum,
30
2004); (Goldring, A. B., Taylor, S. E., Kemeny, M. E. et al , 2002); (Honish, A.,
31
Westerfield, W., Ashby, A. et al , 2006); (Kellett, N., West, F., and Finlay, A.
32
Y., 2006); (King, Madeleine T., Hall, Jane, Lancsar, Emily et al , 2007);
33
(Lamiraud, Karine and Moatti, Jean Paul, 2006); (Lancsar, E. J., Hall, J. P.,
Medicines concordance: full guideline DRAFT (July 2008)
page 328 of 373
DRAFT FOR CONSULTATION
1
King, M. et al , 2007); (Mahadevia, Parthiv J., Shah, Shailen, Leibman,
2
Christopher et al , 2004); (Mark, T. L. and Swait, J., 2004); (Reginster, J. Y.,
3
Rabenda, V., Neuprez, A. et al , 2006); (Richter, J., Eisemann, M., and
4
Zgonnikova, E., 2001); (Ryan, Mandy and Farrar, Shelley, 2000); (Stone,
5
Valerie E., Jordan, Jamie, Tolson, Jerry et al , 2004); (Vick, S. and Scott, A.,
6
1998).
7
8
Despite limitations of the evidence, it seems likely that the shared decision
9
making process would improve cost-effectiveness by enabling patients to
10
make a prediction of their individual valuation of harms and benefits and
11
subsequently opt in or out of treatment. Theoretically it is likely to be of
12
economic benefit to enable patients to decline a suggested prescription as it
13
prevents people from accepting and filling prescriptions they would otherwise
14
had not intended to take.
15
16
This analysis does not include any anticipated impact on the resources used
17
to inform patients as well as on the overall length of the consultation. A
18
systematic review requested by the development group sought to answer the
19
question whether interventions to increase patient involvement increase the
20
length of the consultation between HCP and patient and other health care
21
resource use. The included studies resulted in a variety of results. Please
22
refer to Chapter 4 for more information on the clinical narrative. Overall, it was
23
found that simple interventions do not always increase overall length of
24
consultation. The GDG felt it important to look beyond the initial consultation
25
process when evaluating the benefits of health programmes to improve
26
patient involvement. Extra time invested in the initial consultation where
27
medicines are suggested and/or prescribed was deemed likely to have
28
benefits later on. Specifically, time of follow up consultation could be reduced
29
since concerns have been addressed and the perceived need to seek
30
information from other professionals could be reduced. The group
31
emphasized that an improvement of skill and knowledge on both sides can
32
enable the HCP to be more effective and efficient in the consultation process,
33
which may have a positive effect on the cost effective use of resources.
Medicines concordance: full guideline DRAFT (July 2008)
page 329 of 373
DRAFT FOR CONSULTATION
1
2
10.3.2 Medicine-taking
3
Medicine taking is a complex behaviour, specific to individuals. People are
4
understood to differ in their valuation of benefits and risks, thus making
5
choices not only based on information but also on their own preferences.
6
Apart from these more conscious, or intentional, choices, non adherence to
7
medication may be unintentional, when, for example, patients forget to take
8
their tablets or have physical difficulties to do so.
9
10
Interventions to raise adherence need to address underlying factors or
11
barriers that influence patients in their decisions about medicine taking in
12
order to improve their clinical outcome. It should be noted that the relationship
13
between adherence and clinical benefit is not necessarily positively linear, i.e.
14
it can not be presumed that the more adherent the patient the better the
15
outcome, as, for example, medications can also induce side effects and
16
adverse events. From a cost effectiveness perspective such interventions to
17
increase adherence would have to be effective in raising adherence and
18
improve survival and/or quality of life for patients enough to warrant any
19
additional costs. The effect of adherence to medicines on health is also
20
dependent on factors such as the natural history of the individual disease as
21
well as specific drug pharmacology and pharmacokinetics in particular. For
22
example, missing a small proportion of drugs that lower cholesterol levels may
23
have only marginal impact on cardiovascular risk. However, missing a similar
24
proportion of immunosuppressant drugs in the short term after renal
25
transplantation may cause renal rejection or even death. This ability of a drug
26
to sustain its pharmacological action after a dose has been missed has been
27
termed ‘drug forgiveness’ in the literature and has been defined as “the
28
postdose duration of action of the drug minus the prescribed interval between
29
doses” (Urquhart, J., 1996) (Girvin, B. G. and Johnston, G. D., 2004). The
30
variability of these factors imply significant heterogeneity in the population and
31
interventions covered in this guideline.
32
Medicines concordance: full guideline DRAFT (July 2008)
page 330 of 373
DRAFT FOR CONSULTATION
1
Introducing an adherence enhancing intervention aimed at raising adherence
2
rates with a medication regimen will entail effects on costs and benefits. The
3
following sections will look in more detail at these theoretical effects.
4
5
Costs consequences
6
Costs include the costs of providing the intervention to increase adherence
7
comprising service costs incurred by the health service; knock on costs of
8
prescriptions that had formerly not been collected but as a result of the
9
intervention will be; associated opportunity costs to filled prescriptions that are
10
not taken and whose benefit is thus foregone. Also, if patients were to make
11
an informed decision jointly with the HCP, and were open to decide against
12
medication during consultation this could prevent a suggested prescription
13
from being filled. As a result, this could result in a possible monetary saving.
14
15
Health consequences
16
The effectiveness of receiving an intervention to increase adherence in form
17
of health benefits can be grouped in two main categories: treatment effects
18
from a particular medication and possible process utility gains or losses from
19
receiving the intervention.
20
21
Differences in medication treatment effect will depend on the effectiveness of
22
the adherence enhancing intervention, as well as the dose-response related
23
efficacy of the medication (cf. drug ‘forgiveness’, as mentioned in section
24
10.3.2). Assuming the intervention is effective and raises adherence rates, it
25
still remains uncertain what the incremental treatment effect of the medication
26
will be. The intervention may result in a potential QALY loss due to side
27
effects of the medication (where before non compliance would have prevented
28
them).
29
30
As discussed in above, there may be benefits attributable to an effect of
31
health outcome, but there may also be changes in process utility. Such
32
changes are dependent on individual preferences respective to attributes of
Medicines concordance: full guideline DRAFT (July 2008)
page 331 of 373
DRAFT FOR CONSULTATION
1
the provided intervention. The mentioned literature on conjoint analysis or
2
discrete choice experiments may shed light on such preferences. Thus,
3
receiving the intervention may result in a process utility gain or loss. Changes
4
in utility may also incur during the process of taking medicines, through
5
practical (swallowing a tablet) or psychological issues (cognitive dissonance).
6
These changes may be positive or negative depending on the individual
7
valuation of processes and agreement with diagnosis and choice of therapy.
8
9
10
11
10.4
Cost effectiveness literature review of Interventions
to increase adherence
12
10.4.1 Methods of the review
13
The aim of this literature search was to provide an overview of the health
14
economic evidence base of the cost effectiveness of interventions to increase
15
adherence. For details on the methods please refer to section 2.6. The
16
chosen study methods, reported outcomes and results from the analyses
17
were presented. Titles and abstracts of records were retrieved by the
18
searches, suggested by the GDG or submitted by stakeholders. They were
19
then scanned for relevance to the key questions. Any potentially relevant
20
publications were obtained in full text. Papers were then reviewed to identify
21
the most appropriate evidence to help answer the key questions and to
22
ensure that the recommendations are based on the best available evidence.
23
This process required three main tasks: selection of relevant studies;
24
assessment of study quality; synthesis of the results. From this, three
25
systematic reviews were identified and selected by the GDG for subsequent
26
reviewing. Evidence reviews were undertaken using systematic, transparent
27
approaches following the Guidelines Manual 2007 (www.nice.org.uk).
28
29
Types of studies: Systematic reviews of cost effectiveness studies or
30
comparative economic analyses based on modelling or randomised controlled
31
trials (RCTs) of interventions to increase adherence.
Medicines concordance: full guideline DRAFT (July 2008)
page 332 of 373
DRAFT FOR CONSULTATION
1
Types of participants: people prescribed medication for a medical condition
2
from healthcare professionals in any health service setting.
3
Duration of studies: No time limit was applied.
4
Types of interventions: any interventions intended to change adherence to
5
prescribed medication.
6
Types of outcome measures: adherence/compliance levels as well as clinical
7
outcome.
8
9
10.4.2 Interventions to increase adherence: cost effectiveness
10
systematic reviews narrative
11
Three systematic reviews were found and included. First, the evidence from
12
these three systematic reviews will be presented in section 10.4.3 and an
13
update review of the latest systematic review (Elliott, Rachel A., Barber, Nick,
14
and Horne, Rob, 2005) will be presented as case studies in Section 10.5.
15
16
10.4.3 Summary of findings and methodological issues
17
highlighted by systematic reviews
18
A UK systematic review by Elliott (2005) (Elliott, Rachel A., Barber, Nick, and
19
Horne, Rob, 2005) found 42 papers, 30 from the US, two from the UK and the
20
remainder from a range of countries. This review does not report the cost
21
effectiveness findings of the individual studies, but concentrated on critiquing
22
the variability and prevailing inadequacy of the methods used. The authors
23
used a variety of minimum quality criteria comprising economic evaluation
24
quality criteria, standard hierarchies of evidence, and adherence-specific
25
design issues.
26
27
The second included systematic review by Hughes (2001) (Hughes, D. A.,
28
Bagust, A., Haycox, A. et al , 2001) was from the UK. They only included
29
pharmacoeconomic evaluations that applied sensitivity analysis to non-
30
adherence rates in order to evaluate the impact of non-adherence on the cost-
31
effectiveness of different drug therapies. 22 evaluations were included for
32
reviewing of which 13 were from the US, 5 were Canadian and 2 UK papers.
Medicines concordance: full guideline DRAFT (July 2008)
page 333 of 373
DRAFT FOR CONSULTATION
1
Hughes et al. reviewed only two studies that were clearly described as
2
specifically investigating interventions that aim at increasing adherence, which
3
had mixed results.
4
5
A third, Dutch systematic review by Cleemput (2002) (Cleemput, I, Kesteloot,
6
K, and DeGeest, S, 2002) included 18 studies on the economics of
7
therapeutic non-adherence, which were assessed according to their definition
8
and measurement of medication non-adherence, study design, and
9
identification and valuation of costs and outcomes. The majority of articles
10
dealt exclusively with medication non-adherence. Eight studies examined the
11
economics of interventions to increase adherence. Of these, three were
12
excluded from this review on methodological grounds.
13
14
15
Expectedly, there was high heterogeneity. All three systematic reviews found
16
that the reviewed analyses were conducted in different clinical areas,
17
including chronic, acute and infectious, and studied a wide variety of
18
interventions. For example, Elliott (2005) reviewed programmes to improve
19
convenience of care, information, counselling, reminders, self-monitoring,
20
reinforcement, family therapy, and other forms of additional supervision or
21
attention. Elliott (2005) and Cleemput (2002) described considerable variation
22
between studies in terms of the form of delivery, such as telephone or postal
23
communication. Most interventions contained the provision by a specified
24
health professional, an educational component, and often used more than one
25
component. Cleemput (2002), Hughes (2001) and Elliott (2005) also
26
described heterogeneity in terms of study design. Of the 42 papers reviewed
27
by Elliott (2005), 21 were based on RCTs, of which 4 used a modelling
28
methodology. According to Hughes et al., a decision analytic model was
29
employed in most instances and some of the evaluations modelling chronic
30
illnesses adopted a Markov-type decision analysis approach. It was not
31
specified if models were based on one trial exclusively or on multiple sources.
32
The papers reviewed by Cleemput (2002) included different types of economic
33
evaluations, ranging from cost effectiveness over cost benefit to cost outcome
Medicines concordance: full guideline DRAFT (July 2008)
page 334 of 373
DRAFT FOR CONSULTATION
1
descriptive analyses. Time horizons varied considerably between studies and
2
ranged from 2 weeks to lifetime, with 3 studies spanning time up to 1 year and
3
10 papers of over 10 years. Cleemput (2002) found time horizons to be less
4
than 18 months in all studies, and said that long-term benefits of compliance-
5
enhancing interventions could thus rarely be shown.
6
7
All three systematic reviews found inadequate definitions given for non-
8
adherence. Some studies used previously unvalidated adherence measures.
9
Often no attempt at defining the measure of adherence was made.
10
Alternatively, adherence was defined arbitrarily at some cut off point, or simply
11
assumed to be increased based on opinion.
12
13
All three reviews found effectiveness measures to also be substantially
14
heterogeneous. They were mostly disease specific, and ranged from QALYs
15
to ‘fractures avoided’. Some studies did not report an outcome at all.
16
The link between adherence and health benefit is important, however, was
17
described by both reviews as particularly problematic. Some analyses
18
provided no indication of the differences in health benefits which would be
19
observed when patients were non-adherent. Hughes found that, although
20
some studies were found to link non-adherence to changes in risk
21
probabilities or outcomes, only few referenced an evidence based source. Of
22
those studies which disclosed sources for values and assumptions, only very
23
few used sources other than opinion. Not surprisingly, none of the five studies
24
reviewed by Cleemput (2002) seemed to have linked adherence to clinically
25
relevant outcomes, such as health outcomes represented in QALYs.
26
27
The majority of studies conducted insufficient sensitivity analysis, particularly
28
of adherence rates. For example, Elliott (2005) found that of the 42 studies
29
reviewed only nine conducted sensitivity analysis to quantify the measure of
30
uncertainty which is an important methodological omission. Moreover, no
31
study assessed how the use of a particular adherence measure may have
32
affected ICERs or how sensitive the results were to changes in adherence
33
rates. This limits the generalisability of any results reported.
Medicines concordance: full guideline DRAFT (July 2008)
page 335 of 373
DRAFT FOR CONSULTATION
1
2
The reviews also noted other methodological problems of the studies
3
reported. Cleemput (2002) and Elliott (2005) found costing methods and
4
appropriate statistical analysis for cost data to be often invalid. Further
5
problems included stating the perspective of evaluation, appropriate
6
discounting and quantification of uncertainty, as well as incremental analysis.
7
8
The Elliott review concluded that a meaningful comparison across studies was
9
not possible due to heterogeneity. Hughes (2001) concluded that it was not
10
possible for the review to compare the magnitude of the impact of non-
11
adherence among different drug-disease combinations. Of the two studies
12
that specifically investigated interventions to increase adherence, reviewed by
13
Hughes (2001), one found that the intervention with a higher associated
14
adherence rate was cost effective compared to that with lower adherence,
15
whereas the other found the intervention with lower adherence to be cost
16
effective. Cleemput (2002) could not conclude on a certain approach to have
17
a clear advantage compared with another. Irrespective of the methodological
18
shortcomings, some of the reviewed studies on Interventions to increase
19
adherence show an improvement in terms of cost savings or reduced
20
noncompliance. They also found that one study on an intervention to increase
21
adherence for antimalarials and one for antihypertensives showed some
22
increase in the efficacy of treatment. However, the cost-effectiveness of an
23
intervention to increase adherence will depend upon the costs and health
24
effects associated with usual care and the intervention’s own costs and health
25
effects.
26
27
The systematic reviews did, however, raise some other important issues.
28
Hughes (2001) found that the nature of non-adherence, the severity and
29
pathophysiology of the disease, and the extent to which a drug ‘forgives’ to
30
non-adherence all contribute in determining the extent of the clinical and
31
economic consequences of non-adherence. Drug forgiveness describes the
32
ability of a drug to sustain its pharmacological action after a dose has been
33
missed.
Medicines concordance: full guideline DRAFT (July 2008)
page 336 of 373
DRAFT FOR CONSULTATION
1
In terms of the review findings, studies showed that non-adherence generally
2
results in a reduction in efficacy. Such a relationship was not as clear in cost
3
terms. While eight of the reviewed studies show that costs increase as
4
adherence decreases, six found the opposite trend. This discrepancy did not
5
appear to be related to the nature of the disease, the measure of non-
6
adherence or the assumptions relating to the health benefits experienced by
7
non-compliers.
8
The authors emphasise the need to systematically include non-adherence in
9
pharmacoeconomic evaluations. The vast majority of such studies were based
10
upon trials designed to demonstrate efficacy, and not effectiveness (i.e.
11
protocol effectiveness versus effectiveness in clinical practice). While the
12
randomised clinical trial remains the ‘gold standard’ for comparing alternative
13
treatments, the high internal validity required to demonstrate efficacy comes at
14
the expense of external validity, that is, generalisability of results to the ‘real
15
world’ in medical practice.
16
17
The review by Cleemput (2002) described the important nonlinearity in the
18
relationship between quality of life and non-adherence. Non-adherence may
19
improve patients’ quality of life, for instance when they deliberately adapt their
20
medication schedule to their own lifestyle, or it may decrease their quality of
21
life due to increased morbidity, adverse events and/or side-effects. From a
22
pharmacological perspective, under-dosing or extended time intervals
23
between two medication intakes may increase morbidity (and subsequently
24
costs), whereas over-dosing or shorter intervals between two medication
25
intakes may increase unpleasant side-effects or toxicity of the medication.
26
From an economic view, buying medication without actually taking it implies a
27
pure economic production cost.
28
The authors highlight the importance for clinicians, policymakers in healthcare
29
as well as patients to take non-adherence on the cost effectiveness of health
30
interventions into account. Methodologies to do so adequately need to be
31
improved and used in a standardised way.
32
Medicines concordance: full guideline DRAFT (July 2008)
page 337 of 373
DRAFT FOR CONSULTATION
1
10.4.4 Conclusion
2
It is not possible to conclude which intervention to increase adherence is most
3
cost effective on the basis of the reviewed evidence. The disease specific
4
evaluation of interventions to increase adherence is important to establish
5
what works best for whom and in what circumstance. This is paramount for
6
the cost effectiveness profile of different interventions to increase adherence.
7
The reviews emphasised the importance of standardising the methods to take
8
non-adherence into account when assessing the effectiveness and cost
9
effectiveness of medicines. Moreover, the necessity to measure adherence
10
and to establish a link to clinical outcome appropriately were highlighted. The
11
need to distinguish persistence from compliance/adherence outcomes was
12
described by Hughes and they define medication persistence as “the length of
13
time from initiation to discontinuation of therapy and is measured in units of
14
time”(Hughes, Dyfrig, Cowell, W, Koncz, T et al , 2007).
15
The evidence from the three reviews confirmed that for the analysis of the
16
cost–effectiveness of Interventions to increase adherence it is paramount to
17
look at both costs of the intervention and outcomes, not only in terms of
18
adherence, but also in terms of the subjective value of the clinical outcome for
19
the patient. If the intervention offers a better health outcome than usual care
20
by effectively increasing adherence, or if its cost is more than offset by the
21
cost-savings from reduced non-adherence, the intervention is worthwhile. It is
22
unlikely to be possible to identify a single intervention to increase adherence
23
that fits all patients with their differing health situations, beliefs and
24
preferences in a generic format.
25
26
10.5
Update of the systematic review by Elliott
27
10.5.1 Methods of the review
28
The aim of this literature search is to update the systematic review by Elliott
29
(2005) and colleagues (Elliott, Barber, & Horne 2005) with relevant papers
30
published from 2004 onwards. The titles and abstracts of records retrieved by
31
the searches, suggested by the GDG or submitted by stakeholders were
32
scanned for relevance to the key questions. Any potentially relevant
Medicines concordance: full guideline DRAFT (July 2008)
page 338 of 373
DRAFT FOR CONSULTATION
1
publications were obtained in full text. This yielded eight papers, which were
2
then reviewed to identify the most appropriate evidence to help answer the
3
key questions and to ensure that the recommendations are based on the best
4
available evidence. This process required three main tasks: selection of
5
relevant studies; assessment of study quality; synthesis of the results. Three
6
of the eight references were subsequently excluded as they either failed to
7
describe and impute adherence specifically as outcome measure or did not
8
meet the methodological requirements. Five papers were reviewed. Evidence
9
reviews were undertaken using systematic, transparent approaches following
10
the Guidelines Manual 2007 (www.nice.org.uk).
11
Types of studies: Systematic reviews of cost effectiveness studies or
12
comparative economic analyses based on modelling or randomized controlled
13
trials (RCTs) of interventions to increase adherence.
14
Types of participants: people aged 16 and over prescribed medication for a
15
medical condition from healthcare professionals in any health service setting.
16
Duration of studies: No time limit was applied.
17
Types of interventions: any interventions intended to change adherence to
18
prescribed medication.
19
Types of outcome measures: adherence/compliance levels as well as clinical
20
outcome.
21
22
10.5.2 Update review narrative
23
Five papers were found and included. One Dutch paper by Bosmans
24
(Bosmans 2007) assessed the cost effectiveness of a pharmacy based
25
coaching programme to improve adherence to antidepressants. Another
26
Dutch paper by Cleemput (2004) (Cleemput et al. 2004) compared renal
27
transplantation with haemodialysis for patients with renal failure. A paper by
28
Brunenberg (2007) (Brunenberg-Danielle et al. 2007) examined the cost
29
effectiveness of an adherence improving programme comprising monitoring
30
system and adherence training for patients with hypertension receiving
31
statins. One US paper by Edwards (2005) {8837} assessed the cost
32
effectiveness of long acting risperidone compared to other oral agents in
33
patients with schizophrenia in a decision model. Finally Munakata (2006)
Medicines concordance: full guideline DRAFT (July 2008)
page 339 of 373
DRAFT FOR CONSULTATION
1
(Munakata et al. 2006) analysed in a theoretical scenario how much an
2
adherence enhancing intervention could cost for it to be cost effective when
3
provided for HIV positive patients who receive HAART.
4
5
The first Dutch based cost effectiveness analysis by Bosmans (2007)
6
(Bosmans, J. E., Brook, O. H., Van-Hout, H.-P. J. et al , 2007) was based on a
7
RCT of 151 patients with a prescription for non-tricyclic antidepressants from
8
their GP for depressive complaints. They were randomised to either receiving
9
an intervention consisting of three personal coaching contacts with a
10
pharmacist and an educational video to take home, or usual care including
11
standard oral and written information.
12
Adherence was measured using an electronic pill container (eDEM) and was
13
primary outcome, with the Hopkins depression 13 item subscale (SCL) used
14
as secondary outcome for depressive symptoms. Mean adherence did not
15
differ significantly between the intervention group (88%) and the control group
16
(86%) at six months (mean difference 2.1%, 95% CI -5.6, 9.8). In respect to
17
SCL subscale, there was no statistically significant difference between the
18
groups either despite a slight improvement in the pharmacist intervention
19
group (-0.15, 95% CI -0.54, 0.23).
20
The ICER for coaching and education by pharmacists compared with usual
21
care was €149 per 1% improvement in adherence and €2550 per point
22
improvement in the SCL depression mean item score. Uncertainty was
23
considerable, reflected by insignificance of mean differences. Pairs of costs
24
and effects were distributed in all four quadrants of the cost effectiveness
25
plane. The Cost effectiveness acceptability curve (CEAC) for adherence
26
showed extreme uncertainty, guiding decision makers to have little belief that
27
coaching and education by pharmacists is cost effective as a means of
28
increasing adherence to antidepressants compared with usual care.
29
30
The cost utility analysis by Cleemput (2004) (Cleemput, Irina, Kesteloot,
31
Katrien, Vanrenterghem, Yves et al , 2004) compared haemodialysis with
32
renal transplantation and was based on a decision analytic model. The model
33
drew on data from a prospective study of 126 adults with chronic renal failure
Medicines concordance: full guideline DRAFT (July 2008)
page 340 of 373
DRAFT FOR CONSULTATION
1
and varying adherence levels. Of these, 23 received renal transplant.
2
Adherence was measured using an electronic event monitoring (EEM) device.
3
Five study subjects were defined nonadherent with medication which gives a
4
rate of nonadherence of 21% in this population.
5
6
Lifetime costs after transplantation in the adherent patient group are higher
7
than lifetime costs in the non adherent group, mainly because adherent
8
patients live longer after transplantation. Compared with dialysis, renal
9
transplantation offers better outcome in both adherent and nonadherent
10
patients. Transplant dominated haemodialysis on all adherence levels and
11
was therefore found to be more cost effective. When full adherence is
12
assumed, transplant generates a cost saving relative of dialysis and 5.19
13
additional QALYs. In a heterogeneous group of adherent and nonadherent
14
patients, the saving was greater but fewer QALYs were generated (5.06). This
15
was mainly due to a reduced life expectancy. Among transplant patients,
16
adherence with immunosuppressants after transplantation is associated with a
17
QALY gain, albeit at a higher cost which was mainly due to a longer overall
18
life span. Mean costs per QALY in adherent patients relative to nonadherent
19
patients after transplantation amounted to €35 021 (95%CI 26 959 - 46 620).
20
In 2004, the authors concluded that this leaves scope for an effective
21
adherence enhancing intervention assuming a willingness to pay threshold
22
similar to that currently used by NICE. This study illustrates the effect
23
nonadherence can have on the findings of an economic evaluation. Assuming
24
full or good adherence, which seems common in RCTs, has the tendency to
25
overestimate cost effectiveness by producing more effect and fewer costs in a
26
scenario like this study.
27
28
The cost utility study by Brunenberg (2007) (Brunenberg-Danielle, E. M.,
29
Wetzels-Gwenn, E. C., Nelemans, Patricia J. et al , 2007) compared a
30
medication events monitoring system MEMS) plus adherence training with
31
usual care alone over a 5 month follow up. The MEMS is a drug container and
32
cap equipped with a microchip that registers the date and time of each
33
opening. This study was based on a randomised controlled trial. There were
Medicines concordance: full guideline DRAFT (July 2008)
page 341 of 373
DRAFT FOR CONSULTATION
1
164 hypertensive patients in the MEMS arm and 89 in usual care group and
2
they had a systolic blood pressure (BP) >160mm Hg and/or diastolic BP
3
>95mm Hg despite being antihypertensive drug eligible.
4
Adherence was defined as intake minimum 85% of days as prescribed. From
5
the healthcare perspective, electronic monitoring led to a cost saving of €100
6
and an additional 3.1% of patients achieved normal blood pressure than in the
7
usual care arm and was therefore dominating. Sensitivity analysis revealed
8
considerable uncertainty. Using the QALY as an outcome measure, this
9
probability was between 45% and 51%, albeit from a societal perspective.
10
Overall, effect sizes were small and not statistically significant. From both,
11
healthcare and societal perspectives, the CEA bootstrap replicates on the CE
12
plane covered the origin. This means that the analysed pairs of cost and
13
effectiveness estimates result in both cost effective and not cost effective
14
decision rules, thereby disabling the decision maker to reach a certain
15
conclusion.
16
For statins as a long term treatment of hypertension, the length of follow up of
17
five months appears insufficiently short to anticipate the long term effect of the
18
intervention on adherence, and the effect of adherence on outcome.
19
Theoretically, if in the UK the costs for electronic monitoring do not exceed
20
those of a potential drug cost saving, even a moderate increase in adherence
21
would be cost effective. This would have to be proven in a different trial set
22
up.
23
24
Edwards (1999) (Edwards, G. and Anderson, I., 1999) conducted a cost
25
effective analysis based on a decision analytic model that compared long
26
acting risperidone with a range of other antipsychotic agents, including oral
27
risperidone and depot haloperidol. The population was drawn from patients
28
with schizophrenia in community dwellings who have previously suffered
29
relapse requiring hospitalisation.
30
Compliance was assumed to be improved by the long acting injectable
31
risperidone formula. It was estimated that a 20% point difference in
32
compliance would predict a 3.1 point improvement in the PANSS (Positive
33
and Negative Syndrome Scale for Schizophrenia). Such improvement in turn
Medicines concordance: full guideline DRAFT (July 2008)
page 342 of 373
DRAFT FOR CONSULTATION
1
stabilised patients so that a further 6.1 point in PANSS was achieved by
2
further improved medicine taking behaviour, and aversion of relapse. The
3
model predicts that patients receiving long acting risperidone will have the
4
best clinical outcomes in terms of the frequency and duration of relapses over
5
the one year duration. For example, on long acting risperidone 26% of
6
patients will experience relapse requiring hospitalisation and 24% relapse not
7
requiring hospitalisation. On haloperidol nearly two thirds of patients are
8
predicted to have relapses requiring hospitalisation and over 60% not
9
requiring hospitalisation. This analysis predicts dominance of long acting
10
risperidone over the comparators, with providing a health outcome
11
improvement in terms of days of relapse averted whilst costing less over the
12
time horizon of one year. Univariate sensitivity analysis was reported to have
13
been robust. However, at the upper bound of the 95%CI for relapse rates
14
requiring hospitalisation there was an incremental cost for long acting
15
risperidone with an ICER of US$821per days of hospitalisation averted
16
compared to oral risperidone. The model seems also sensitive to the cost of
17
hospitalisation as well as frequency rates of relapse.
18
Overall, the analysis seems of interest; however, there are issues with its
19
robustness. Adherence and clinical outcomes were estimated and not
20
quantified in a single measure, such as the QALY. The outcome of cost per
21
days of hospitalisation averted poses a challenge on the interpretation of the
22
findings in the context of this guideline. Values used in the sensitivity analysis
23
seem relatively conservative. The short time horizon could be an issue and
24
has not been thoroughly discussed. Quantifying treatment effect and quality of
25
life losses in one measurement such as the QALY could considerably help
26
interpret the findings from the analysis.
27
28
The cost utility analysis conducted by Munakata (2006) (Munakata, J.,
29
Benner, J. S., Becker, S. et al , 2006) was based on a decision analytic
30
model. The aim was to quantify the clinical and economic effects of
31
nonadherence and estimate the cost effectiveness of improving adherence in
32
treatment naïve patients. For this, HAART treatment with an assumed good
33
adherence was compared with HAART on ‘typical’ adherence. The authors
Medicines concordance: full guideline DRAFT (July 2008)
page 343 of 373
DRAFT FOR CONSULTATION
1
drew on data from randomised controlled trials and observational data for the
2
comparators, respectively. The model population was HIV positive, with a
3
mean age of 33. The assumed portion of drugs consumed of 0.98 (0.95-1.0)
4
was defined as adherent and 0.55 (0-0.95) as nonadherent. The proportion of
5
adherent patients in the typical comparator arm was imputed as 0.52 (0.3-
6
0.88). Lifetime discounted costs in the typical and ideal scenarios were $308
7
000 and $341 000, respectively. This gives an incremental cost of $33 000.
8
People in the ideal scenario generated 10.2 QALYs per patient compared to
9
9.0 QALYs per patient in the typical scenario. This gives an incremental effect
10
of 1.2 QALYs. The incremental cost effectiveness ratio (ICER) resulted in $29
11
400 per QALY. This means that there is scope for an intervention to increase
12
adherence. The authors calculated a willingness to pay (WTP) ceiling value
13
for an intervention to increase adherence. They conclude that $1 600 could be
14
spent per patient to increase adherence to ideal levels, giving 15-33%
15
reductions in treatment failure. Univariate sensitivity analysis was conducted
16
for all parameters, as well as multivariate SA for selected values. The analysis
17
was described as robust in sensitivity analysis. In severe diseases where
18
adherence and related comorbidities are a big issue, adherence improving
19
interventions may be cost effective. Given that there are interventions that are
20
effective in increasing adherence, this analysis found that $1 600 per patient
21
could be spent.
22
23
Please refer to the discussion section 10.6 for a summary of the evidence.
Medicines concordance: full guideline DRAFT (July 2008)
page 344 of 373
DRAFT FOR CONSULTATION
1
2
10.6
3
It has been shown that the relationship between cost effectiveness and
4
adherence is not a simple a priori relationship. An increase in adherence does
5
not necessarily lead to an improvement in the cost effectiveness profile, even
6
for highly cost-effective interventions.
7
In order to make any predictions on the cost effectiveness of an intervention to
8
increase adherence (or indeed patient involvement as discussed) it will be
9
necessary to determine outcomes from direct evidence, comprising both
10
Discussion
harms and benefits.
11
12
10.6.1 Theoretical discussion of review findings, caveats and
13
opportunities for future research
14
The presented findings from the studies selected for this review update
15
underpins the discussion from the previous evidence narrative of the three
16
systematic reviews (Cleemput, I, Kesteloot, K, and DeGeest, S, 2002) (Elliott,
17
Rachel A., Barber, Nick, and Horne, Rob, 2005) (Hughes, D. A., Bagust, A.,
18
Haycox, A. et al , 2001). This states that, where nonadherence is prevalent
19
and leads to reductions in quality of life and even survival, there is scope for
20
an intervention that effectively raises adherence. However, it could not be
21
concluded which intervention could increase adherence appropriately to be
22
cost effective. This update review gives examples such as HAART and renal
23
transplant patients where there may be scope for such interventions, provided
24
they are effective in increasing adherence.
25
26
The reviews have revealed areas where there are particular methodological
27
difficulties/caveats in this area of research. Firstly, it became evident that
28
there is an important distinction between efficacy and effectiveness. Clinical
29
trials may demonstrate higher adherence rates than would be found in clinical
30
practice. This may be due to bias or length of follow up, particularly for
31
medicines where medication is indicated for the long term (adherence to
32
statins, for example, may be significantly lower after four years rather than
Medicines concordance: full guideline DRAFT (July 2008)
page 345 of 373
DRAFT FOR CONSULTATION
1
three months after initiation, as often measured in clinical trials (Avorn, J.,
2
Monette, J., Lacour, A. et al , 1998). Limitations to follow up result from a
3
number of factors, including funding. There have been suggestions for
4
research to overcome the problem of follow up in form of persistence
5
measures (Elliott, R. A., Shinogle, J. A., Peele, P. et al , 2008) (Hughes,
6
Dyfrig, Cowell, W, Koncz, T et al , 2007). This will be discussed in more detail
7
below.
8
9
Secondly, the review of three systematic reviews presented in the previous
10
section found ‘forgiveness’ in connection with severity and pathophysiology of
11
the disease an important factor in determining the cost effectiveness of an
12
intervention to increase adherence. As described in section 10.3.2, a drug is
13
‘forgiving’ if it is still efficacious despite missed doses. In such situations the
14
scope for an intervention is reduced. For example, statins are considered to
15
be relatively forgiving. From the update review, the study of monitoring
16
devices for the use of statins in hypertensive patients confirmed a limited use
17
of adherence enhancing measures in terms of QALY gain and cost
18
effectiveness (Brunenberg-Danielle, E. M., Wetzels-Gwenn, E. C., Nelemans,
19
Patricia J. et al , 2007). Measuring and reporting adherence as well as health
20
related outcomes is paramount for research to be interpreted meaningfully.
21
22
Thirdly, process utility was not measured in any of the studies reviewed, but
23
there is likely to be an effect on the quality of life by the process of taking the
24
medication. Please refer to Section 10.3 for some further detail on process
25
utility.
26
27
Despite all problems and caveats, the area of medicine taking is an interesting
28
and important one to investigate. As some studies in the review have shown,
29
there is scope to provide cost effective interventions to increase adherence.
30
Firstly, the capacity to benefit is evident to be of importance here. Where not
31
taking medication has likely severe effects on quality of life perhaps via effect
32
of the illness (e.g. in antidepressants) or comorbidities (e.g. in HIV AIDS
33
defining illnesses) and even survival (e.g. HIV, Renal transplants), effective
Medicines concordance: full guideline DRAFT (July 2008)
page 346 of 373
DRAFT FOR CONSULTATION
1
interventions to enhance adherence are more likely to be cost effective. The
2
papers by Cleemput et al. (Cleemput, Irina, Kesteloot, Katrien,
3
Vanrenterghem, Yves et al , 2004) and Munakata et al. (Munakata, J.,
4
Benner, J. S., Becker, S. et al , 2006) provide examples for this.
5
6
Secondly, an effective intervention to increase adherence is more likely to be
7
cost effective where even only few missed doses of an ‘unforgiving’ have
8
more immediate consequences. Munakata et al. (Munakata, J., Benner, J. S.,
9
Becker, S. et al , 2006) suggested that for people taking HAART medication,
10
there is scope to provide effective programmes to help patients take their
11
medication. Other examples could be antiepileptic or immunosuppressant
12
drugs.
13
Thirdly, externalities may widen the role of interventions to increase
14
adherence to a societal level. In economics, externalities (or spill over effects)
15
occur when firms or people impose costs or benefits on others outside the
16
market place (Samuelson, Paul A. and William D Nordhaus, 2004). An
17
example of this occurs in communicable diseases such as for example
18
tuberculosis. The social value curve for high adherence to active TB treatment
19
may deviate from the individual value curve for some patients as they may
20
experience a situation under treatment they would not personally have
21
chosen. An example of negative externality is antibiotic resistance which
22
occurs with poor adherence to TB treatment, and we see the devastating
23
effect it can have in some Baltic states and Latin American countries (Pablos-
24
Mendez, Ariel, Raviglione, Mario C., Laszlo, Adalbert et al , 1998).
25
26
27
10.6.2 Recommendations for minimum reporting requirements
28
The review of interventions to increase adherence has shown limitations of
29
the reporting style from studies. Often, reporting was inconsistent, incomplete
30
and/or not validated. In accordance with the NICE technical manual for
31
guideline development, we recommend the following data to be reported for
32
health economic analyses:
33
Medicines concordance: full guideline DRAFT (July 2008)
page 347 of 373
DRAFT FOR CONSULTATION
1
2
3
1. Adequate time horizons and length of follow up need to be considered and
reported.
2. Perspective of the evaluation needs to be reported. Where perspective not
4
healthcare, all relevant healthcare and social costs need to be presented
5
separately
6
3. Instead of presenting total costs per patient, microlevel costing and
7
providing mean resource use and unit costs is preferred in order to be able
8
to adjust for the prevailing setting.
9
4. Incremental analysis is used in health economics to reveal the value of
10
adopting a decision to provide a programme over and above an
11
alternative.
12
5. To report health related outcome, including all related side effects and co-
13
morbidities. The QALY is the preferred choice as it combines all health
14
effects into one measure, thereby allowing comparisons across
15
programmes.
16
6. Essentially, a definition of adherence needs to be stated, justified and used
17
consistently. It would be useful to develop standardised definitions for
18
specific formulations. For long term follow up, a binary measure such as
19
persistence may be most useful, particularly in more forgiving treatment
20
options (Stason, W. B., 1999) (Hughes, Dyfrig, Cowell, W, Koncz, T et al ,
21
2007). Beware of censoring when using persistence measures, this
22
phenomenon can often be found with other methods of survival analysis
23
(Lamiraud, Karine and Moatti, Jean Paul, 2006)
24
7. A modelling approach may be of value in order to reflect the complexity of
25
concordance and adherence behaviour. Where decision trees oversimplify,
26
cohort based Markov modelling and patient-level discrete event simulation
27
may be of more valuable methodologies (Hughes, Dyfrig, Cowell, W,
28
Koncz, T et al , 2007)
Medicines concordance: full guideline DRAFT (July 2008)
page 348 of 373
DRAFT FOR CONSULTATION
1
8. Probabilistic sensitivity analysis should be carried out in order to estimate
2
uncertainty around incremental estimates and robustness of assumptions.
3
(Claxton, K., Sculpher, M., McCabe, C. et al , 2005).
4
9. Heterogeneity of intervention needs to be addressed appropriately by
5
providing an in-depth description of all components of the intervention.
6
External validity and replicability issues need to be adequately considered.
7
10. Heterogeneity of the patient population needs to be addressed by
8
providing an in-depth description of the study population. This is of great
9
importance for considering validity and replicability issues, especially since
10
medicine taking is an individual behaviour and qualitative research
11
approaches could be advocated.
12
Recommendations for future research
13
10.6.3
14
Further from the minimum reporting requirements set out in the previous
15
section, there are methodological options that the research community could
16
utilise to address the specific complexities described in this document more
17
adequately.
18
19

Minimum reporting recommendations would give scope for doing
20
economic modelling in the specific area. A decision model could
21
directly analyse the effect of changing intervention adherence levels on
22
outcome and cost effectiveness. A price ceiling for the intervention
23
could be established, and the uncertainty and robustness of single
24
parameters, such as rate of relapse, age etc evaluated by means of
25
formal sensitivity analysis.
26
27

Interventions to increase adherence involve a complex set of
28
components that may be difficult to incorporate into standard cost
29
effectiveness models. More research is needed to identify and assess
30
their components and how they can lead to improvements in quality of
31
life and/or a reduction in costs. It has been suggested that in the field of
Medicines concordance: full guideline DRAFT (July 2008)
page 349 of 373
DRAFT FOR CONSULTATION
1
adherence research endogeneity and interdependence play an
2
important role that require careful consideration for sound statistical
3
and econometric analysis (Lamiraud, Karine and Moatti, Jean Paul,
4
2006). They also say that using a microeconomic perspective could
5
help to better understand the complex trade offs that patients have to
6
face when deciding about heir level of adherence with medical
7
prescriptions. Further from describing the potential implications from
8
variable endogeneity and complexity to understand the dimensions of
9
behaviour change in both HCP and patient, pioneering research to
10
assess GP behaviour change has drawn upon psychometric tests to
11
assess GP change, as well as testing if actual behaviour has changed
12
(Bonetti, Debbie, Eccles, Martin, Johnston, Marie et al , 2005) (Hrisos,
13
S., Eccles, M., Johnston, M. et al , 2008a) (Hrisos, S., Eccles, M.,
14
Johnston, M. et al , 2008b). Incorporated into a model, such data would
15
allow retrospective analysis in order to understand why an intervention
16
worked or did not work.
17
18

The binary measure of treatment persistence could enable long term
19
follow up after clinical trials have stopped, and thereby help determine
20
long term cost effectiveness of an intervention to enhance adherence.
21
This would be of particular interest in treatment situations where
22
treatments affect outcomes over long periods of time and/or where
23
medicines have to be taken for long periods of time and adherence
24
rates tend to drop over time in practice. Please see also section 10.3
25
where the concepts of efficiency and effectiveness were contrasted in
26
this context.
27
28

The evaluation of process utility may be important for evaluating the
29
benefits of shared decision making as well as adherence behaviour.
30
Therefore, further research in this area could play an important role as
31
insight into the individual gain other than health related from a health
32
care programme may change the outcome in terms of cost
33
effectiveness. This has not yet been considered by NICE as robust
Medicines concordance: full guideline DRAFT (July 2008)
page 350 of 373
DRAFT FOR CONSULTATION
1
methodology is absent, however, as a field of research changes in
2
process utility could be of particular importance for evaluating the
3
benefits and costs of shared decision making.
4
5
6
7
8
9
10
11
12
Medicines concordance: full guideline DRAFT (July 2008)
page 351 of 373
DRAFT FOR CONSULTATION
Reference List
(1) Adam BD, Maticka TE, Cohen JJ. Adherence practices among people living
with HIV. AIDS Care: Psychological Socio Medical Aspects of AIDS/HIV.
2003; 15 (2):263-274.
(2) Adler DA, Bungay KM, Wilson IB, Pei Y et al. The impact of a pharmacist
intervention on 6-month outcomes in depressed primary care patients. Gen
Hosp Psychiatry. 2004; 26 (3):199-209.
(3) Agård A, Hermerén G, Herlitz J. When is a patient with heart failure
adequately informed? A study of patients' knowledge of and attitudes toward
medical information. Heart & lung : the journal of critical care. 2004; 33
(4):219-226.
(4) Aikens JE, Nease-Donald-E-Jr, Klinkman MS. Explaining patients' beliefs
about the necessity and harmfulness of antidepressants. Annals of family
medicine. 2008; 6 (1):23-29.
(5) Albus-Christian S-NS-BF. Preferences regarding medical and psychosocial
support in HIV-infected patients. Patient Educ Couns. 2005; 56 (1):16-20.
(6) Alfonso V, Bermbach N, Geller J, Montaner J-SG. Individual variability in
barriers affecting people's decision to take HAART: A qualitative study
identifying barriers to being on HAART. Aids Patient Care STDS. 2006; 20
(12):848-857.
(7) Antoni MH, Carrico AW, Duran RE, Spitzer S et al. Randomized clinical trial
of cognitive behavioral stress management on human immunodeficiency virus
viral load in gay men treated with highly active antiretroviral therapy.
Psychosom Med. 2006; 68 (1):143-151.
(8) Aristides M, Weston AR, FitzGerald P, Le RC et al. Patient preference and
willingness-to-pay for Humalog Mix25 relative to Humulin 30/70: A
multicountry application of a discrete choice experiment. Value Health. 2004;
7 (4):442-454.
(9) Aronson B. Medication management behaviors of adherent short-term
antibiotic users. Clinical Excellence for Nurse Practitioners. 2005; 9 (1):23-30.
(10) Astrom K, Carlsson J, Bates I, Webb DG et al. Desire for information about
drugs. A multi-method study in general medical inpatients. Pharm World Sci.
2000; 22 (4):159-164.
Medicines concordance: full guideline DRAFT (July 2008) page 352 of 373
DRAFT FOR CONSULTATION
(11) Atella V, Schafheutle E, Noyce P, Hassell K. Affordability of medicines and
patients' cost-reducing behaviour: empirical evidence based on SUR estimates
from Italy and the UK. Appl Health Econ Health Policy. 2005; 4 (1):23-35.
(12) Atherton-Naji A, Hamilton R, Riddle W, Naji S. Improving adherence to
antidepressant drug treatment in primary care: A feasibility study for a
randomized controlled trial of education intervention. Primary Care
Psychiatry. 2001; 7 (2):61-67.
(13) Avorn J, Monette J, Lacour A, Bohn RL et al. Persistence of use of lipidlowering medications: a cross-national study. JAMA. 1998; 279 (18):14581462.
(14) Bailey WC, Richards-Jr JM, Brooks CM, Soong S et al. A randomized trial to
improve self-management practices of adults with asthma. Arch Intern Med.
1990; 150 (8):1664-1668.
(15) Bain D. Doctor-patient communication in general practice consultations.
Medical Education 10, 125-131. 1976.
(16) Baird MG, Bentley-Taylor MM, Carruthers SG, Dawson KG et al. A study of
efficacy, tolerance and compliance of once-daily versus twice-daily
metoprolol (Betaloc) in hypertension. Betaloc Compliance Canadian
Cooperative Study Group. Clin Invest Med. 1984; 7 (2):95-102.
(17) Bambra C, Egan M, Thomas S, Petticrew M et al. The psychosocial and health
effects of workplace reorganisation. 2. A systematic review of task
restructuring interventions. J Epidemiol Community Health. 2007; 61
(12):1028-1037.
(18) Bane C, Hughes CM, Cupples ME, McElnay JC. The journey to concordance
for patients with hypertension: a qualitative study in primary care. Pharmacy
world & science : PWS. 2007; 29 (5):534-540.
(19) Bangalore S, Kamalakkannan G, Parkar S, Messerli FH. Fixed-dose
combinations improve medication compliance: a meta- analysis. Am J Med.
2007; 120 :713-719.
(20) Beaucage K, Lachance-Demers H, Ngo TT, Vachon C et al. Telephone
follow-up of patients receiving antibiotic prescriptions from community
pharmacies. American Journal of Health-System Pharmacy. 2006; 63 (6):557563.
(21) Bechdolf A, Knost B, Kuntermann C, Schiller S et al. A randomized
comparison of group cognitive-behavioural therapy and group
psychoeducation in patients with schizophrenia. Acta Psychiatr Scand. 2004;
110 (1):21-28.
(22) Bechdolf A, Köhn D, Knost B, Pukrop R et al. A randomized comparison of
group cognitive-behavioural therapy and group psychoeducation in acute
patients with schizophrenia: Outcome at 24 months. Acta Psychiatr Scand.
2005; 112 (3):173-179.
Medicines concordance: full guideline DRAFT (July 2008) page 353 of 373
DRAFT FOR CONSULTATION
(23) Becker LA, Glanz K, Sobel E. A randomized trial of special packaging of
antihypertensive medications. J Fam Pract. 1986; 22 (4):357-361.
(24) Begley S, Livingstone C, Hodges N, Williamson V. Impact of domiciliary
pharmacy visits on medication management in an elderly population.
International Journal of Pharmacy Practice. 1997; 5 (3):111-121.
(25) Bender B, Milgrom H, Rand C. Nonadherence in asthmatic patients: is there a
solution to the problem? Annals of allergy asthma & immunology : official
publication of the American College of Allergy Asthma & Immunology. 1997;
79 (3):177-186.
(26) Bernsten C, Bjorkman I, Caramona M, Crealey G et al. Improving the wellbeing of elderly patients via community pharmacy-based provision of
pharmaceutical care. Drugs Aging. 2001; 18 (1):63-77.
(27) Beusterien KM, Dziekan K, Flood E, Harding G et al. Understanding patient
preferences for HIV medications using adaptive conjoint analysis: feasibility
assessment. Value Health. 2005; 8 (4):453-461.
(28) Bikowski RM, Ripsin CM, Lorraine VL. Physician-patient congruence
regarding medication regimens. J Am Geriatr Soc. 2001; 49 (10):1353-1357.
(29) Bollini P, Tibaldi G, Testa C, Munizza C. Understanding treatment adherence
in affective disorders: a qualitative study. J Psychiatr Ment Health Nurs. 2004;
11 (6):668-674.
(30) Bonetti D, Eccles M, Johnston M, Steen N et al. Guiding the design and
selection of interventions to influence the implementation of evidence-based
practice: an experimental simulation of a complex intervention trial. Social
Science & Medicine. 2005; 60 (9):2135-2147.
(31) Bosch C, X, Garner P. Contracts between patients and healthcare practitioners
for improving patients' adherence to treatment, prevention and health
promotion activities. 2006.
(32) Bosch-Capblanch, X., Abba, K., Prictor, M., and Garner, P. Contracts
between patients and healthcare practitioners for improving patients'
adherence to treatment, prevention and health promotion activities.
2007
(33) Bosmans JE, Brook OH, Van-Hout H-PJ, De-Bruijne MC et al. Cost
effectiveness of a pharmacy-based coaching programme to improve adherence
to antidepressants. Pharmacoeconomics. 2007; 25 (1):25-37.
(34) Braman AC. Patient personality predicts preference for relationships with
doctors. Personality and Individual Differences. 2004; 37 (4):815-826.
(35) Brouwer WBF, Exel NJ, Berg Bvd, Bos GAM et al. Process utility from
providing informal care: the benefit of caring. Health Policy. 2005; 74 (1):8599.
Medicines concordance: full guideline DRAFT (July 2008) page 354 of 373
DRAFT FOR CONSULTATION
(36) Brown BG, Bardsley J, Poulin D, Hillger LA et al. Moderate dose, three-drug
therapy with niacin, lovastatin, and colestipol to reduce low-density
lipoprotein cholesterol <100 mg/dl in patients with hyperlipidemia and
coronary artery disease. Am J Cardiol. 1997; 80 (2):111-115.
(37) Brown C, Battista DR, Bruehlman R, Sereika SS et al. Beliefs about
antidepressant medications in primary care patients: relationship to selfreported adherence. Med Care. 2005; 43 (12):1203-1207.
(38) Brunenberg-Danielle EM, Wetzels-Gwenn EC, Nelemans PJ, Dirksen CD et
al. Cost effectiveness of an adherence-improving programme in hypertensive
patients. Pharmacoeconomics. 2007; 25 (3):239-251.
(39) Campero L, Herrera C, Kendall T, Caballero M. Bridging the gap between
antiretroviral access and adherence in Mexico. Qual Health Res. 2007; 17
:599-611.
(40) Canto-De-Cetina TE, Canto P, Ordoñez LM. Effect of counseling to improve
compliance in Mexican women receiving depot-medroxyprogesterone acetate.
Contraception. 2001; 63 (3):143-146.
(41) Caress A. Patient roles in decision-making. Nurs Times. 1997; 93 (31):45-48.
(42) Carrick R, Mitchell A, Powell RA, Lloyd K. The quest for well-being: a
qualitative study of the experience of taking antipsychotic medication.
Psychology and psychotherapy. 2004; 77 (1):19-33.
(43) Chaplin R, Kent A. Informing patients about tardive dyskinesia. Controlled
trial of patient education. Br J Psychiatry. 1998; 172 (Jan):78-81.
(44) Charles C, Gafni A, Whelan T. Decision-making in the physician-patient
encounter: revisiting the shared treatment decision-making model. Soc Sci
Med. 1999; 49 (5):651-661.
(45) Charles C, Gafni A, Whelan T. Shared decision-making in the medical
encounter: what does it mean? (or it takes at least two to tango). Social
Science & Medicine. 1997; 44 (5):681-692.
(46) Chisholm MA, Mulloy LL, Jagadeesan M, DiPiro JT. Impact of clinical
pharmacy services on renal transplant patients' compliance with
immunosuppressive medications. Clin Transplant. 2001; 15 (5):330-336.
(47) Claxton AJ, Cramer J, Pierce C. A systematic review of the associations
between dose regimens and medication compliance. Clin Ther. 2001; 23
(8):1296-1310.
(48) Claxton A, Cramer J, Pierce C. A systematic Review of the Associations
Between Dose Regimens and Medication Compliance. Clinical Therapeutics
23[8], 1296. 2001.
Medicines concordance: full guideline DRAFT (July 2008) page 355 of 373
DRAFT FOR CONSULTATION
(49) Claxton K, Sculpher M, McCabe C, Briggs A et al. Probabilistic sensitivity
analysis for NICE technology assessment: not an optional extra. Health Econ.
2005; 14 (4):339-347.
(50) Cleemput I, Kesteloot K, DeGeest S. A review of the literature on the
economics of noncompliance. Room for methodological improvement. Health
Policy. 2002; 59 (1):65-94.
(51) Cleemput I, Kesteloot K, Vanrenterghem Y, De GS. The economic
implications of non-adherence after renal transplantation.
Pharmacoeconomics. 2004; 22 (18):1217-1234.
(52) Cohen D, Longo MF, Hood K, Edwards A et al. Resource effects of training
general practitioners in risk communication skills and shared decision making
competences. Journal of Evaluation in Clinical Practice. 2004; 10 (3):439-445.
(53) Collier AC, Ribaudo H, Mukherjee AL, Feinberg J et al. A randomized study
of serial telephone call support to increase adherence and thereby improve
virologic outcome in persons initiating antiretroviral therapy. J Infect Dis.
2002; 192 (8):1398-1406.
(54) Connor J, Rafter N, Rodgers A. Do fixed-dose combination pills or unit-of-use
packaging improve adherence? A systematic review. [Review] [26 refs].
Bulletin of the World Health Organization. 2004; 82 (12):935-939.
(55) Cooper V, Buick D, Horne R, Lambert N et al. Perceptions of HAART among
gay men who declined a treatment offer: preliminary results from an
interview-based study. AIDS Care. 2002; 14 (3):319-328.
(56) Cote J, Bowie DM, Robichaud P, Parent JG et al. Evaluation of two different
educational interventions for adult patients consulting with an acute asthma
exacerbation. Am J Respir Crit Care Med. 2001; 163 (6):1415-1419.
(57) Côté J, Bowie DM, Robichaud P, Parent JG et al. Evaluation of two different
educational interventions for adult patients consulting with an acute asthma
exacerbation. Am J Respir Crit Care Med. 2001; 163 (6):1415-1419.
(58) Coté J, Cartier A, Robichaud P, Boutin H et al. Influence on asthma morbidity
of asthma education programs based on self-management plans following
treatment optimization. Am J Respir Crit Care Med. 1997; 155 (5):1509-1514.
(59) Cox, F, Stevenson, N, Brtten, N, and Dundar, Y. A systematic review of
communication between patients and health care professionals about
medicine-taking and prescribing. Medicines partnership. 2004
(60) Cox K, Britten N, Hooper R, White P. Patients' involvement in decisions about
medicines: GPs' perceptions of their preferences. Br J Gen Pract. 2007; 57
(543):777-784.
Medicines concordance: full guideline DRAFT (July 2008) page 356 of 373
DRAFT FOR CONSULTATION
(61) Cramer JA. A systematic review of adherence with medications for diabetes.
Diabetes Care. 2004; 27 (5):1218-1224.
(62) De G, Schafer-Keller P, Denhaerynck K, Thannberger N et al. Supporting
medication adherence in renal transplantation (SMART): a pilot RCT to
improve adherence to immunosuppressive regimens. Clinical Transplantation.
2003; #2006 May-Jun. (3).
(63) Deegan PE. The importance of personal medicine: A qualitative study of
resilience in people with psychiatric disabilities. Scandinavian Journal of
Public Health. 2005; 33 (SUPPL. 66):29-35.
(64) Degner L. Decision making during serious illness: What role do patients really
want to play?. Sloan J, editor. Journal of Clinical Epidemiology 45, 941-950.
2008.
(65) Doherty C, Doherty W. Patients' preferences for involvement in clinical
decision-making within secondary care and the factors that influence their
preferences. J Nurs Manag. 2005; 13 (2):119-127.
(66) Dowell J, Jones A, Snadden D. Exploring medication use to seek concordance
with 'non-adherent' patients: a qualitative study. The British journal of general
practice : the journal of the Royal College of General Practitioners. 2002; 52
(474):24-32.
(67) Duggan C, Bates I. Development and evaluation of a survey tool to explore
patients' perceptions of their prescribed drugs and their need for drug
information. International Journal of Pharmacy Practice. 2000; 8 (1):42-52.
(68) Duggan PS, Geller G, Cooper LA, Beach MC. The moral nature of patientcenteredness: is it "just the right thing to do"?. [Review] [36 refs]. Patient
Education & Counseling. 2006; 62 (2):271-276.
(69) Dunbar J, Dunning EJ, Dwyer K. Compliance measurement with arthritis
regimen. Arthritis care and research : the official journal of the Arthritis
Health Professions Association. 1989;(3):S8-16.
(70) Edwards A, Elwyn G. Inside the black box of shared decision making:
distinguishing between the process of involvement and who makes the
decision. Health Expect. 2006; 9 (4):307-320.
(71) Edwards A, Elwyn G, Hood K, Atwell C et al. Patient-based outcome results
from a cluster randomized trial of shared decision making skill development
and use of risk communication aids in general practice. Fam Pract. 2004; 21
(4):347-354.
(72) Edwards G, Anderson I. Systematic review and guide to selection of selective
serotonin reuptake inhibitors. Drugs. 1999; 57 (4):507-533.
(73) Elliott RA, Shinogle JA, Peele P, Bhosle M et al. Understanding Medication
Compliance and Persistence from an Economics Perspective. Value Health.
2008.
Medicines concordance: full guideline DRAFT (July 2008) page 357 of 373
DRAFT FOR CONSULTATION
(74) Elliott RA, Barber N, Horne R. Cost-effectiveness of adherence-enhancing
interventions: a quality assessment of the evidence. The Annals of
pharmacotherapy. 2005; 39 (3):508-515.
(75) Elliott RA, Ross DD, Adams AS, Safran DG et al. Strategies for coping in a
complex world: adherence behavior among older adults with chronic illness. J
Gen Intern Med. 2007; 22 (6):805-810.
(76) Elwyn G, Edwards A, Dwyn R, Grol R. Towards a feasible model for shared
decision-making: focus group study wuth general practice registrars. British
Medical Journal 319, 753-756. 1999.
(77) Elwyn G, Edwards A, hutchings A, Rapport F et al. The OPTION
scale:measuring the extent that clinicians involve patients in decision-making
tasks. Health Expectations 8[1], 34-42. 2005.
(78) Elwyn G, Edwards A, Mowle S, Wensing M et al. Measuring the involvement
of patients in shared decision-making: a systematic review of instruments.
[Review] [88 refs]. Patient Education & Counseling. 2001; 43 (1):5-22.
(79) Elwyn G, Edwards A, Wensing M, Hood K et al. Shared decision making:
developing the OPTION scale for measuring patient involvement. Qual Saf
Health Care. 2003; 12 (2):93-99.
(80) Ende J, Kazis L, Ash A, Moskowitz MA. Measuring patients' desire for
autonomy: decision making and information-seeking preferences among
medical patients. Journal of general internal medicine : official journal of the
Society for Research and Education in Primary Care Internal Medicine. 1989;
4 (1):23-30.
(81) Erwin J, Peters B. Treatment issues for HIV+ Africans in London. Soc Sci
Med. 1999; 49 (11):1519-1528.
(82) F Légaré, S Ratté D Stacey J Kryworuchko K Gravel L Turcot ID
Graham. Interventions for improving the adoption of shared decision
making by healthcare professionals. Cochrane Collaboration. 2008
(83) Farmer KC. Methods for measuring and monitoring medication regimen
adherence in clinical trials and clinical practice. Clin Ther. 1999; 21 (6):-90.
(84) Finley PR, Rens HR, Pont JT, Gess SL et al. Impact of a collaborative care
model on depression in a primary care setting: a randomized controlled trial.
Pharmacotherapy. 2003; 23 (9):1175-1185.
(85) Fraenkel L, Rabidou N, Wittink D, Fried T. Improving informed decisionmaking for patients with knee pain. The Journal of rheumatology. 2007; 34
(9):1894-1898.
(86) Friedman RH, Kazis LE, Jette A, Smith MB et al. A telecommunications
system for monitoring and counseling patients with hypertension. Impact on
medication adherence and blood pressure control. Am J Hypertens. 1996; 9
(4):285-292.
Medicines concordance: full guideline DRAFT (July 2008) page 358 of 373
DRAFT FOR CONSULTATION
(87) Frileux-Stéphanie S-M-T-MM-ES. The impact of the preventive medical
message on intention to change behavior. Patient Educ Couns. 2004; 52
(1):79-88.
(88) Gagné-Camille GG. Improving self-report measures of non-adherence to HIV
medications. Psychology & Health. 2005; 20 (6):803-816.
(89) Garber MC, Nau DP, Erickson SR, Aikens JE et al. The concordance of selfreport with other measures of medication adherence: a summary of the
literature. Med Care. 2004; 42 (7):649-652.
(90) Garcia-Popa-Lisseanu MG, Greisinger A, Richardson M, Malley KJ et al.
Determinants of treatment adherence in ethnically diverse, economically
disadvantaged patients with rheumatic disease. J Rheumatol. 2005; 32 (5):913919.
(91) George J, Kong-David CM, Stewart K. Adherence to disease management
programs in patients with COPD. International journal of chronic obstructive
pulmonary disease. 2007; 2 (3):253-262.
(92) Girvin B, McDermott BJ, Johnston GD. A comparison of enalapril 20 mg once
daily versus 10 mg twice daily in terms of blood pressure lowering and patient
compliance. J Hypertens. 1999; 17 (11):1627-1631.
(93) Girvin BG, Johnston GD. Comparison of the effects of a 7-day period of noncompliance on blood pressure control using three different antihypertensive
agents. J Hypertens. 2004; 22 (7):1409-1414.
(94) Goldring AB, Taylor SE, Kemeny ME, Anton PA. Impact of health beliefs,
quality of life, and the physician-patient relationship on the treatment
intentions of inflammatory bowel disease patients. Health Psychology. 2002;
21 (3):219-228.
(95) Gravel K, Legare F, Graham ID. barriers and facilitators to implementing
shared decision-making in clinical practice: a systematic review of health care
professionals perceptions. implementation science. 2006; 1 (16):1-12.
(96) Gray R, Leese M, Bindman J, Becker T et al. Adherence therapy for people
with schizophrenia: European multicentre randomised controlled trial. Br J
Psychiatry. 2006; 189 (DEC.):508-514.
(97) Grymonpre RE, Williamson DA, Montgomery PR. Impact of a pharmaceutical
care model for non-institutionalised elderly: Results of a randomised,
controlled trial. International Journal of Pharmacy Practice. 2001; 9 (4):235241.
(98) Guthrie RM. The effects of postal and telephone reminders on compliance
with pravastatin therapy in a national registry: results of the first myocardial
infarction risk reduction program. Clin Ther. 2001; 23 (6):970-980.
(99) Hamann J, Langer B, Winkler V, Busch R et al. Shared decision making for
in-patients with schizophrenia. Acta Psychiatr Scand. 2006; 114 (4):265-273.
Medicines concordance: full guideline DRAFT (July 2008) page 359 of 373
DRAFT FOR CONSULTATION
(100) Hamann J, Cohen R, Leucht S, Busch R et al. Shared decision making and
long-term outcome in schizophrenia treatment. The Journal of clinical
psychiatry. 2007; 68 (7):992-997.
(101) Hamet P, Campbell N, Curnew G, Eastwood C et al. Avapromise: a
randomized clinical trial for increasing adherence through behavioural
modification in essential hypertension. Exp Clin Cardiol. 2003; 7 :165-172.
(102) Hamilton W, Russell D, Stabb C, Seamark D et al. The effect of patient selfcompletion agenda forms on prescribing and adherence in general practice: a
randomized controlled trial. Fam Pract. 2007; 24 (1):77-83.
(103) Hanlon JT, Weinberger M, Samsa GP, Schmader KE et al. A randomized,
controlled trial of a clinical pharmacist intervention to improve inappropriate
prescribing in elderly outpatients with polypharmacy. The American journal of
medicine. 1996; 100 (4):428-437.
(104) Harrington-Jane NL. Improving patients' communication with doctors: A
systematic review of intervention studies. Patient Educ Couns. 2004; 52 (1):716.
(105) Hawkshead J, Krousel-Wood MA. Techniques for measuring medication
adherence in hypertensive patients in outpatient settings: Advantages and
limitations. Disease Management and Health Outcomes. 2007; 15 (2):109-118.
(106) Hayes RP, Bowman L, Monahan PO, Marrero DG et al. Understanding
diabetes medications from the perspective of patients with type 2 diabetes:
prerequisite to medication concordance. The Diabetes educator. 2006; 32
(3):404-414.
(107) Haynes RB, Ackloo E, Sahota N, McDonald HP et al. Interventions for
enhancing medication adherence. Cochrane Database of Systematic Reviews.
2008;(CD000011).
(108) Haynes RB, Sackett DL, Gibson ES. Improvement of medication compliance
in uncontrolled hypertension. Lancet. 1976; 1 (7972):1265-1268.
(109) Haynes RB, Yao X, Degani A, Kripalani S et al. Interventions to enhance
medication adherence. Cochrane Database Syst Rev. 2005;(4):CD000011.
(110) Haynes TB, Taylor DW, Sackett DL. Behavioural Strategies for improving
compliance. In: Haynes TB, Taylor DW, Sackett DL. Compliance in Health
Care. 174-190. 1979. John Hopkins University Press.
(111) Hecht FM. Measuring HIV treatment adherence in clinical practice. AIDS
clinical care. 1998; 10 (8):57-59.
(112) Heneghan CJ, Glasziou P, Perera R. Reminder packaging for improving
adherence to self-administered long-term medications. [Review] [86 refs].
Cochrane Database of Systematic Reviews (1):CD005025. 2006;(1).
Medicines concordance: full guideline DRAFT (July 2008) page 360 of 373
DRAFT FOR CONSULTATION
(113) Henry A, Batey RG. Enhancing compliance not a prerequisite for effective
eradication of Helicobacter pylori: the HelP Study. Am J Gastroenterol. 1999;
94 (3):811-815.
(114) Hill SA, Laugharne R. Decision making and information seeking preferences
among psychiatric patients. J Ment Health. 2006; 15 (1):75-84.
(115) Hirth RA, Greer SL, Albert JM, Young EW et al. Out-of-pocket spending and
medication adherence among dialysis patients in twelve countries. Health Aff.
2008; 27 (1):89-102.
(116) Honish A, Westerfield W, Ashby A, Momin S et al. Health-related quality of
life and treatment compliance with diabetes care. DIS MANAGE. 2006; 9
(4):195-200.
(117) Horne R, Weinman J. Concordance, A Partnership in Medicine Taking. 2004.
The Pharmaceutical Press.
(118) Horne R, Weinman J, Barber N, Elliott R et al. Concordance, adherence and
compliance in medicine taking. Report for the National Co-ordinatingCentre
for NHS Service Delivery andOrganisation R & D (NCCSDO . 2005.
(119) Horne R. Beliefs about medicines questionnaire: the development and
evaluation of a new method for assessing the cognitive representation of
mediciation. Psychology and Health. 2008; 37 (2):113-124.
(120) Horne R, Cooper V, Gellaitry G, Date HL et al. Patients' perceptions of highly
active antiretroviral therapy in relation to treatment uptake and adherence: The
utility of the necessity-concerns framework. J Acquir Immune Defic Syndr.
2007; 45 (3):334-341.
(121) Howland JS, Baker MG, Poe T. Does patient education cause side effects? A
controlled trial. J Fam Pract. 1990; 31 (1):62-64.
(122) Hrisos S, Eccles M, Johnston M, Francis J et al. An intervention modelling
experiment to change GPs' intentions to implement evidence-based practice:
using theory-based interventions to promote GP management of upper
respiratory tract infection without prescribing antibiotics #2. BMC Health Serv
Res. 2008a; 8 :10.
(123) Hrisos S, Eccles M, Johnston M, Francis J et al. Developing the content of two
behavioural interventions: using theory-based interventions to promote GP
management of upper respiratory tract infection without prescribing antibiotics
#1. BMC Health Serv Res. 2008b; 8 :11.
(124) Hughes DA, Bagust A, Haycox A, Walley T. The impact of non-compliance
on the cost-effectiveness of pharmaceuticals: a review of the literature.
[Review] [52 refs]. Health Economics. 2001; 10 (7):601-615.
(125) Hughes D, Cowell W, Koncz T, Cramer J. Methods for integrating medication
compliance and persistence in pharmacoeconomic evaluations. Value Health.
2007.
Medicines concordance: full guideline DRAFT (July 2008) page 361 of 373
DRAFT FOR CONSULTATION
(126) Iskedjian M, Einarson TR, MacKeigan LD, Shear N et al. Relationship
between daily dose frequency and adherence to antihypertensive
pharmacotherapy: Evidence from a meta-analysis. Clin Ther. 2002; 24
(2):302-316.
(127) Jenkins L, Britten N, Stevenson F, Barber N et al. Developing and using
quantitative instruments for measuring doctor- patient communication about
drugs. Patient Educ Couns. 2003; 50 (3):273-278.
(128) Johnson AL, Taylor DW, Sackett DL. Self-recording of blood pressure in the
management of hypertension. Can Med Assoc J. 1978; 119 (9):1034-1039.
(129) Johnson SB. Methodological issues in diabetes research. Measuring
adherence. Diabetes Care. 1992; 15 (11):1658-1667.
(130) Katon W, Russo J, Von KM, Lin E et al. Long-term effects of a collaborative
care intervention in persistently depressed primary care patients. J Gen Intern
Med. 2002; 17 (10):741-748.
(131) Kellett N, West F, Finlay AY. Conjoint analysis: a novel, rigorous tool for
determining patient preferences for topical antibiotic treatment for acne. A
randomised controlled trial. The British journal of dermatology. 2006; 154
(3):524-532.
(132) Kemp R, Kirov G, Everitt B, Hayward P et al. Randomized controlled trial of
compliance therapy. 18 month follow up. Br J Psychiatry. 1998; 172 :413-419.
(133) Kemp R, Hayward P, Applewhaite G, Everitt B et al. Compliance therapy in
psychotic patients: Randomised controlled trial. Br Med J. 1996; 312
(7027):345-349.
(134) Kemp-Steven F-MH-AW-HC-C. Psychological factors and use of antiepileptic
drugs: Pilot work using an objective measure of adherence. Psychology. 2007;
12 (1):107-113.
(135) Khanderia U, Townsend KA, Erickson SR, Vlasnik J et al. Medication
adherence following coronary artery bypass graft surgery: Assessment of
beliefs and attitudes. Ann Pharmacother. 2008; 42 (2):192-199.
(136) Kikkert MJ, Schene AH, Koeter-Maarten WJ, Robson D et al. Medication
adherence in schizophrenia: exploring patients', carers' and professionals'
views. Schizophr Bull. 2006; 32 (4):786-794.
(137) King MT, Hall J, Lancsar E, Fiebig D et al. Patient preferences for managing
asthma: results from a discrete choice experiment. Health Economics. 2007;
16 :703-717.
(138) Kinnersley P, Edwards A, Hood K, Cadbury N et al. Interventions before
consultations for helping patients address their information needs. Cochrane
database of systematic reviews (Online). 2007;(3):CD004565, 156CD004565,493X.
Medicines concordance: full guideline DRAFT (July 2008) page 362 of 373
DRAFT FOR CONSULTATION
(139) Kumar K, Gordon C, Toescu V, Buckley CD et al. Beliefs about medicines in
patients with rheumatoid arthritis and systemic lupus erythematosus: a
comparison between patients of South Asian and White British origin.
Rheumatology (Oxford England). 2008; 47 (5):690-697.
(140) La FJ. Methods to Measure Patient Compliance with Medication Regimens. JPain-Pall-Care-Pharmacother. 2004; 18 (3):81-87.
(141) Lam DH, Watkins ER, Hayward P, Bright J et al. A randomized controlled
study of cognitive therapy for relapse prevention for bipolar affective disorder:
outcome of the first year. Arch Gen Psychiatry. 2003; 60 (2):145-152.
(142) Lamiraud K, Moatti JP. Economic implications of nonadherence to highly
active antiretroviral treatment in HIV patients. Expert opinion on
pharmacotherapy. 2006; 7 (2):135-143.
(143) Lancsar EJ, Hall JP, King M, Kenny P et al. Using discrete choice experiments
to investigate subject preferences for preventive asthma medication.
Respirology (Carlton, Vic ). 2007; 12 (1):127-136.
(144) Langewitz W, Nübling M, Weber H. Hospital patients' preferences for
involvement in decision-making: A questionnaire survey of 1040 patients from
a Swiss university hospital. Swiss Medical Weekly. 2006; 136 (3-4):59-64.
(145) Lawton J, Ahmad N, Hallowell N, Hanna L et al. Perceptions and experiences
of taking oral hypoglycaemic agents among people of Pakistani and Indian
origin: Qualitative study. Br Med J. 2005; 330 (7502):1247-1249.
(146) Lee JK, Grace KA, Taylor AJ. Effect of a pharmacy care program on
medication adherence and persistence, blood pressure, and low-density
lipoprotein cholesterol: a randomized controlled trial.[see comment]. JAMA.
2006; 296 (21):2563-2571.
(147) Levy ML, Robb M, Allen J, Doherty C et al. A randomized controlled
evaluation of specialist nurse education following accident and emergency
department attendance for acute asthma. Respir Med. 2000; 94 :900-908.
(148) Lewin, S. A., Skea, Z. C., Entwistle, V., Zwarenstein, M. et al.
Interventions for providers to promote a patient-centred approach in
clinical consultations. 2001
(149) Lewis MP, Colbert A, Erlen J, Meyers M. A qualitative study of persons who
are 100% adherent to antiretroviral therapy. AIDS Care - Psychological and
Socio-Medical Aspects of AIDS/HIV. 2006; 18 (2):140-148.
(150) Lipton HL, Bird JA. The impact of clinical pharmacists' consultations on
geriatric patients' compliance and medical care use: a randomized controlled
trial. The Gerontologist. 1994; 34 (3):307-315.
(151) Little P, Dorward M, Warner G, Moore M et al. Randomised controlled trial of
effect of leaflets to empower patients in consultations in primary care. Br Med
J. 2004; 328 (7437):441-444.
Medicines concordance: full guideline DRAFT (July 2008) page 363 of 373
DRAFT FOR CONSULTATION
(152) Loh A, Simon D, Wills CE, Kriston L et al. The effects of a shared decisionmaking intervention in primary care of depression: a cluster-randomized
controlled trial. Patient Educ Couns. 2007; 67 (3):324-332.
(153) Lowe CJ, Raynor DK, Purvis J, Farrin A et al. Effects of a medicine review
and education programme for older people in general practice. British journal
of clinical pharmacology. 2000; 50 (2):172-175.
(154) Lukoschek P. African Americans' beliefs and attitudes regarding hypertension
and its treatment: a qualitative study. J Health Care Poor Underserved. 2003;
14 (4):566-587.
(155) Mahadevia PJ, Shah S, Leibman C, Kleinman L et al. Patient preferences for
sensory attributes of intranasal corticosteroids and willingness to adhere to
prescribed therapy for allergic rhinitis: a conjoint analysis. Annals of allergy
asthma & immunology : official publication of the American College of
Allergy Asthma & Immunology. 2004; 93 (4):345-350.
(156) Makoul G, Clayman ML. An integrative model of shared decision making in
medical encounters. Patient Educ Couns. 2006; 60 (3):301-312.
(157) Mannheimer SB, Morse E, Matts JP, Andrews L et al. Sustained benefit from
a long-term antiretroviral adherence intervention: Results of a large
randomized clinical trial. J Acquir Immune Defic Syndr. 2006; 43 (SUPPL.
1):S41-S47.
(158) Mark TL, Swait J. Using stated preference and revealed preference modeling
to evaluate prescribing decisions. Health Economics. 2004; 13 (6):563-573.
(159) Marquez-Contreras E, Martell-Claros N, Gil-Guillen V, de lV et al. Efficacy
of a home blood pressure monitoring programme on therapeutic compliance in
hypertension: the EAPACUM-HTA study. Journal of Hypertension. 2006; 24
(1):169-175.
(160) McCann-Simon WJ. Encouraging patient participation in general practice
consultations: Effect on consultation length and content, patient satisfaction
and health. Psychology & Health. 1996; 11 (6):857-869.
(161) McKinstry, B., Ashcroft, R. E., Car, J., Freeman, G. K. et al.
Interventions for improving patients' trust in doctors and groups of
doctors. 2006
(162) McLean M, Armstrong D. Eliciting patients' concerns: a randomised
controlled trial of different approaches by the doctor. British Journal of
General Practice. 2004; 54 (506):663-666.
(163) Mead N, Power P. Patient-centredness: a conceptual framework and review of
the emprirical literature. Social Science & Medicine 51[7], 1087-1110. 2000.
(164) Medical Research Council. A Framework for development and evaluation of
RCTs for Complex Interventions to Improve Health. 2000.
Medicines concordance: full guideline DRAFT (July 2008) page 364 of 373
DRAFT FOR CONSULTATION
(165) Menckeberg TT, Bouvy ML, Bracke M, Kaptein AA et al. Beliefs about
medicines predict refill adherence to inhaled corticosteroids. J Psychosom Res.
2008; 64 (1):47-54.
(166) Middleton JF, McKinley RK, Gillies CL. Effect of patient completed agenda
forms and doctors' education about the agenda on the outcome of
consultations: randomised controlled trial. Br Med J. 2006; 332 (7552):12381242.
(167) Miklowitz DJ, George EL, Richards JA, Simoneau TL et al. A randomized
study of family-focused psychoeducation and pharmacotherapy in the
outpatient management of bipolar disorder. Arch Gen Psychiatry. 2003; 60
(9):904-912.
(168) Miller LG, Hays RD. Measuring adherence to antiretroviral medications in
clinical trials. HIV Clin Trials. 2000; 1 (1):36-46.
(169) Mills EJ. Adherence to Antiretroviral Therapy in Sub-Saharan Africa and
North America: A Meta-Analysis. JAMA. 2006; 296 (6):679-690.
(170) Mills EJ, Nachega JB, Bangsberg DR, Singh S et al. Adherence to HAART: a
systematic review of developed and developing nation patient-reported
barriers and facilitators. PLoS medicine. 2006; 3 (11):2039-2064.
(171) Molina JM, Podsadecki TJ, Johnson MA, Wilkin A et al. A
lopinavir/ritonavir-based once-daily regimen results in better compliance and
is non-inferior to a twice-daily regimen through 96 weeks. AIDS Res Hum
Retroviruses. 2007; 23 (12):1505-1514.
(172) Montgomery AA, Fahey T, Peters TJ. A factorial randomised controlled trial
of decision analysis and an information video plus leaflet for newly diagnosed
hypertensive patients. The British journal of general practice : the journal of
the Royal College of General Practitioners. 2003; 53 (491):446-453.
(173) Morgan-Myfanwy F-M-J. Barriers to Uptake and Adherence with Malaria
Prophylaxis by the African Community in London, England: Focus Group
Study. Ethnicity & Health. 2005; 10 (4):355-372.
(174) Morice AH, Wrench C. The role of the asthma nurse in treatment compliance
and self- management following hospital admission. Respir Med. 2001; 95
(11):851-856.
(175) Munakata J, Benner JS, Becker S, Dezii CM et al. Clinical and economic
outcomes of nonadherence to highly active antiretroviral therapy in patients
with human immunodeficiency virus. Med Care. 2006; 44 (10):893-899.
(176) Munro SA, Lewin SA, Smith HJ, Engel ME et al. Patient adherence to
tuberculosis treatment: a systematic review of qualitative research. PLoS
medicine. 2007; 4 (7):1230-1245.
Medicines concordance: full guideline DRAFT (July 2008) page 365 of 373
DRAFT FOR CONSULTATION
(177) Murray E, Charles C, Gafni A. Shared decision-making in primary care:
tailoring the Charles et al. model to fit the context of general practice.
[Review] [32 refs]. Patient Education & Counseling. 2006; 62 (2):205-211.
(178) Mutchler JE, Bacigalupe G, Coppin A, Gottlieb A. Language barriers
surrounding medication use among older Latinos. Journal of cross cultural
gerontology. 2007; 22 (1):101-114.
(179) Nazareth I, Burton A, Shulman S, Smith P et al. A pharmacy discharge plan
for hospitalized elderly patients--a randomized controlled trial. Age Ageing.
2001; 30 (1):33-40.
(180) Neame R, Hammond A, Deighton C. Need for information and for
involvement in decision making among patients with rheumatoid arthritis: a
questionnaire survey. Arthritis Rheum. 2005; 53 :249-255.
(181) O'Connor, A. M., Stacey, D., Entwistle, V., Llewellyn-Thomas, H. et al.
Decision aids for people facing health treatment or screening decisions.
2003
(182) O'Donnell C, Donohoe G, Sharkey L, Owens N et al. Compliance therapy: a
randomised controlled trial in schizophrenia. Br Med J. 2003; 327 (7419):834837.
(183) Oakley S, Walley T. A pilot study assessing the effectiveness of a decision aid
on patient adherence with oral bisphosphonate medication. Pharmaceutical
Journal. 2006; 276 (7399):536-538.
(184) Ogedegbe G, Harrison M, Robbins L, Mancuso CA et al. Barriers and
facilitators of medication adherence in hypertensive African Americans: a
qualitative study. Ethnicity & disease. 2004; 14 (1):3-12.
(185) Orton L, Barnish G. Unit-dose packaged drugs for treating malaria. [Review]
[40 refs]. Cochrane Database of Systematic Reviews. 2005;(CD004614).
(186) Pablos-Mendez A, Raviglione MC, Laszlo A, Binkin N et al. Global
Surveillance for Antituberculosis-Drug Resistance, 1994-1997. The New
England journal of medicine. 1998; 338 (23):1641-1649.
(187) Parienti JJ, Massari V, Reliquet V, Chaillot F et al. Effect of twice-daily
nevirapine on adherence in HIV-1-infected patients: a randomized controlled
study. AIDS. 2007; 21 (16):2217-2222.
(188) Paterson DL, Potoski B, Capitano B. Measurement of adherence to
antiretroviral medications. J Acquir Immune Defic Syndr. 2002; 31 (Suppl
3):S103-S106.
(189) Peterson GM, Fitzmaurice KD, Naunton M, Vial JH et al. Impact of
pharmacist-conducted home visits on the outcomes of lipid-lowering drug
therapy. J Clin Pharm Ther. 2004; 29 (1):23-30.
Medicines concordance: full guideline DRAFT (July 2008) page 366 of 373
DRAFT FOR CONSULTATION
(190) Peveler R, George C, Kinmonth AL, Campbell M et al. Effect of
antidepressant drug counselling and information leaflets on adherence to drug
treatment in primary care: randomised controlled trial. Br Med J. 1999; 319
(7210):612-615.
(191) Piette JD, Weinberger M, McPhee SJ, Mah CA et al. Do automated calls with
nurse follow-up improve self-care and glycemic control among vulnerable
patients with diabetes? Am J Med. 2000; 108 (1):20-27.
(192) Portsmouth SD, Osorio J, McCormick K, Gazzard BG et al. Better maintained
adherence on switching from twice-daily to once-daily therapy for HIV: a 24week randomized trial of treatment simplification using stavudine prolongedrelease capsules. HIV Medicine. 2005; 6 (3):185-190.
(193) Pound P, Britten N, Morgan M, Yardley L et al. Resisting medicines: a
synthesis of qualitative studies of medicine taking. Soc Sci Med. 2005; 61
(1):133-155.
(194) Pradier C, Bentz L, Spire B, Tourette TC et al. Efficacy of an educational and
counseling intervention on adherence to highly active antiretroviral therapy:
French prospective controlled study. HIV Clin Trials. 2003; 4 (2):121-131.
(195) Pyne JM, McSweeney J, Kane HS, Harvey S et al. Agreement between
patients with schizophrenia and providers on factors of antipsychotic
medication adherence. Psychiatr Serv. 2006; 57 (8):1170-1178.
(196) Rand CS, Wise RA. Measuring adherence to asthma medication regimens. Am
J Respir Crit Care Med. 1994; 149 (2 Pt 2):S69-S76.
(197) Rao JK. Communication interventions make a difference in conversations
between physicians and patients: A systematic review of the evidence. Med
Care. 2007; 45 (4):340-349.
(198) Rathbun RC, Farmer KC, Stephens JR, Lockhart SM. Impact of an adherence
clinic on behavioral outcomes and virologic response in treatment of HIV
infection: a prospective, randomized, controlled pilot study. Clinical
Therapeutics. 2005; 27 (2):199-209.
(199) Raynor DK, Blenkinsopp A, Knapp P, Grime J et al. A system review of
quantitative and qualitative research on the role and effectiveness of written
information available to patients about individual medicines. Health Technol
Assess. 2007; 11 (5):iii-98.
(200) Razali SM, Hasanah C, I, Khan UA, Subramaniam M. Psychosocial
interventions for schizophrenia. J Ment Health. 2000; 9 (3):283-289.
(201) Reginster JY, Rabenda V, Neuprez A, Reginster JY et al. Adherence, patient
preference and dosing frequency: understanding the relationship. [Review] [30
refs]. Bone. 2006; 38 (4 Suppl 1):S2-S6.
Medicines concordance: full guideline DRAFT (July 2008) page 367 of 373
DRAFT FOR CONSULTATION
(202) Reid M, Clark A, Murdoch DL, Morrison C et al. Patients strategies for
managing medication for chronic heart failure. Int J Cardiol. 2006; 109 (1):6673.
(203) Remien RH, Stirratt MJ, Dolezal C, Dognin JS et al. Couple-focused support
to improve HIV medication adherence: a randomized controlled trial. AIDS.
2005; 19 (8):807-814.
(204) Richards C, Coulter A. Is the NHS becoming more patient centred: trends
from the national surveys of
NHS patients in England 2002-07. 2007. Picker Institute.
(205) Richter J, Eisemann M, Zgonnikova E. Doctors' authoritarianism in end-of-life
treatment decisions. A comparison between Russia, Sweden and Germany.
Journal of Medical Ethics. 2001; 27 (3):186-191.
(206) Rickles NM, Svarstad BL, Statz-Paynter JL, Taylor LV et al. Pharmacist
telemonitoring of antidepressant use: effects on pharmacist-patient
collaboration. J Am Pharm Assoc. 2005; 45 (3):344-353.
(207) Ross SE, Moore LA, Earnest MA, Wittevrongel L et al. Providing a webbased online medical record with electronic communication capabilities to
patients with congestive heart failure: randomized trial.[see comment]. Journal
of Medical Internet Research. 2004; 6 (2):e12.
(208) Rudd P, Miller NH, Kaufman J, Kraemer HC et al. Nurse management for
hypertension. A systems approach. Am J Hypertens. 2004; 17 (10):921-927.
(209) Ruskin PE, Silver-Aylaian M, Kling MA, Reed SA et al. Treatment outcomes
in depression: comparison of remote treatment through telepsychiatry to inperson treatment. Am J Psychiatry. 2004; 161 (8):1471-1476.
(210) Ryan M, Farrar S. Using conjoint analysis to elicit preferences for health care.
BMJ. 2000; 320 (7248):1530-1533.
(211) Sackett DL, Gibson ES, Taylor DW. Randomised clinical trial of strategies for
improving medication compliance in primary hypertension. Lancet. 1975; 1
(7918):1205-1207.
(212) Sadik A, Yousif M, McElnay JC. Pharmaceutical care of patients with heart
failure. Br J Clin Pharmacol. 2005; 60 (2):183-193.
(213) Samuelson PA, William D Nordhaus. Economics. 18th edition. ed. McGrawHill; 2004.
(214) Savage R, Armstrong D. Effect of a general practitioner's consulting style on
patients' satisfaction: a controlled study. Br Med J. 1990; 301 (6758):968-970.
(215) Schaffer SD, Tian L. Promoting adherence: effects of theory-based asthma
education. Clin Nurs Res. 2004; 13 (1):69-89.
Medicines concordance: full guideline DRAFT (July 2008) page 368 of 373
DRAFT FOR CONSULTATION
(216) Schneider A, Wensing M, Quinzler R, Bieber C et al. Higher preference for
participation in treatment decisions is associated with lower medication
adherence in asthma patients. Patient Educ Couns. 2007; 67 (1-2):57-62.
(217) Schneider PJ, Murphy JE, Pedersen CA. Impact of medication packaging on
adherence and treatment outcomes in older ambulatory patients. J Am Pharm
Assoc. 2008; 48 (1):58-63.
(218) Scotto CJ. The lived experience of adherence for patients with heart failure. J
Cardpulm Rehabil. 2005; 25 (3):158-163.
(219) Segador J, Gil-Guillen VF, Orozco D, Quirce F et al. The effect of written
information on adherence to antibiotic treatment in acute sore throat.
International Journal of Antimicrobial Agents. 2005; 26 (1):56-61.
(220) Shi L, Hudges M, Yurgin N, Boye KS. Impact of dose frequency on
compliance and health outcomes: A literature review (1966-2006). Expert
Review of Pharmacoeconomics and Outcomes Research. 2007; 7 (2):187-202.
(221) Shields CG, Epstein RM, Fiscella K, Franks P et al. Influence of accompanied
encounters on patient-centeredness with older patients. The Journal of the
American Board of Family Practice / American Board of Family Practice.
2005; 18 (5):344-354.
(222) Sidat M, Fairley C, Grierson J. Experiences and perceptions of patients with
100% adherence to highly active antiretroviral therapy: a qualitative study.
Aids Patient Care STDS. 2007; 21 (7):509-520.
(223) Sookaneknun P, Richards RM, Sanguansermsri J, Teerasut C. Pharmacist
involvement in primary care improves hypertensive patient clinical outcomes.
Annals of Pharmacotherapy. 2004; 38 (12):2023-2028.
(224) Stason WB. Compliance, quality of life, and cost effectiveness. Curr
Hypertens Rep. 1999; 1 (6):471-474.
(225) Stevenson FA. General practitioners' views on shared decision-making: a
qualitative analysis. Patient Educ Couns. 50, 291-293. 2003.
(226) Stewart A, Noakes T, Eales C, Shepard K et al. Adherence to cardiovascular
risk factor modification in patients with hypertension. Cardiovascular Journal
of Southern Africa. 2005; 16 (2):102-107.
(227) Stone VE, Jordan J, Tolson J, Miller R et al. Perspectives on adherence and
simplicity for HIV-infected patients on antiretroviral therapy: self-report of the
relative importance of multiple attributes of highly active antiretroviral therapy
(HAART) regimens in predicting adherence. J Acquir Immune Defic Syndr.
2004; 36 (3):808-816.
(228) Strang JS, Falloon I-RH, Moss HB. The effects of family therapy on treatment
compliance in schizophrenia. Psychopharmacol Bull. 1981; 17 :87-88.
Medicines concordance: full guideline DRAFT (July 2008) page 369 of 373
DRAFT FOR CONSULTATION
(229) Strydom A, Forster M, Wilkie BM, Edwards C et al. Patient information
leaflets for people with learning disabilities who take psychiatric medication.
British Journal of Learning Disabilities. 2001; 29 (2):72-76.
(230) Sturgess IK, McElnay JC, Hughes CM, Crealey G. Community pharmacy
based provision of pharmaceutical care to older patients. Pharm World Sci.
2003; 25 (5):218-226.
(231) Taylor CT, Byrd DC, Krueger K. Improving primary care in rural Alabama
with a pharmacy initiative. Am J Health Syst Pharm. 2003; 60 (11):11231129.
(232) Theunissen-Nicolet CM. Manipulation of patient-provider interaction:
Discussing illness representations or action plans concerning adherence.
Patient Educ Couns. 2003; 51 (3):247-258.
(233) Thompson AG. The meaning of patient involvement and participation in
health care consultations: a taxonomy. Social Science & Medicine. 2007; 64
(6):1297-1310.
(234) Thomson RG, Eccles MP, Steen IN, Greenaway J et al. A patient decision aid
to support shared decision-making on anti-thrombotic treatment of patients
with atrial fibrillation: randomised controlled trial. Quality & Safety in Health
Care. 2007; 16 (3):216-223.
(235) Tortolero-Luna-Guillermo B-TG-J. Relationship between English Language
Use and Preferences for Involvement in Medical Care among Hispanic
Women. Journal of Women's Health. 2006; 15 (6):774-785.
(236) Towle A. Physician and patient communication skills: competencies for
informaed shared decision-making. 1997. Vancouver, Canada, University of
British Columbia.
(237) Trevena LJ, Davey HM, Barratt A, Butow P et al. A systematic review on
communicating with patients about evidence. J Eval Clin Pract. 2006; 12
(1):13-23.
(238) Tsuyuki RT, Fradette M, Johnson JA, Bungard TJ et al. A multicenter disease
management program for hospitalized patients with heart failure.[see
comment]. Journal of Cardiac Failure. 2004; 10 (6):473-480.
(239) Tuckett D, Boutlon M, Olson C, Williams A. Meetings between experts: an
appraoch to sharing ideas in medical consultation. 1985. Tavistock Publishers.
(240) Tuldrà A, Fumaz CR, Ferrer MJ, Bayés R et al. Prospective randomized twoArm controlled study to determine the efficacy of a specific intervention to
improve long-term adherence to highly active antiretroviral therapy. J Acquir
Immune Defic Syndr. 2000; 25 (3):221-228.
(241) Turner BJ. Adherence to antiretroviral therapy by human immunodeficiency
virus- infected patients. J Infect Dis. 2002; 185 (Suppl 2):S143-S151.
Medicines concordance: full guideline DRAFT (July 2008) page 370 of 373
DRAFT FOR CONSULTATION
(242) Urien AM, Guillen VF, Beltran DO, Pinzotas CL et al. Telephonic back-up
improves antibiotic compliance in acute tonsillitis/pharyngitis. International
Journal of Antimicrobial Agents. 2004; 23 (2):138-143.
(243) Urquhart J. Patient non-compliance with drug regimens: measurement, clinical
correlates, economic impact. European heart journal. 1996; 17 Suppl A :8-15.
(244) van S, Nyamathi A, Carpio F, Pearce D et al. Effects of a treatment adherence
enhancement program on health literacy, patient-provider relationships, and
adherence to HAART among low-income HIV-positive Spanish-speaking
Latinos. Aids Patient Care STDS. 2005; 19 (11):745-759.
(245) van-Dam HA. Provider-patient interaction in diabetes care: Effects on patient
self-care and outcomes A systematic review. Patient Educ Couns. 2003; 51
(1):17-28.
(246) Vick S, Scott A. Agency in health care. Examining patients' preferences for
attributes of the doctor-patient relationship. J Health Econ. 1998; 17 (5):587605.
(247) Vinter RN, Petricek G, Katic M. Obstacles which patients with type 2 diabetes
meet while adhering to the therapeutic regimen in everyday life: Qualitative
study. Croat Med J. 2004; 45 (5):630-636.
(248) Vivian EM. Improving blood pressure control in a pharmacist-managed
hypertension clinic. Pharmacotherapy. 2002; 22 (12):1533-1540.
(249) Vrijens B, Belmans A, Matthys K, de K et al. Effect of intervention through a
pharmaceutical care program on patient adherence with prescribed once-daily
atorvastatin. Pharmacoepidemiology & Drug Safety. 2006; 15 (2):115-121.
(250) Wagner GJ, Kanouse DE, Golinelli D, Miller LG et al. Cognitive-behavioral
intervention to enhance adherence to antiretroviral therapy: a randomized
controlled trial (CCTG 578). AIDS. 1909; 2006 Jun 12. (9).
(251) Weber R, Christen L, Christen S, Tschopp S et al. Effect of individual
cognitive behaviour intervention on adherence to antiretroviral therapy:
prospective randomized trial. Antiviral Therapy. 2004; 9 (1):85-95.
(252) Wetzels R, Harmsen M, van WC, Grol R et al. Interventions for improving
older patients' involvement in primary care episodes. Cochrane Database of
Systematic Reviews. 2007;(4 (CD004273) Source notes: Update Software
online or CD-ROM updated quarterly ISSN: 1469-493X.).
(253) Wetzels R, Wensing M, van WC, Grol R. A consultation leaflet to improve an
older patient's involvement in general practice care: A randomized trial.
Health Expect. 2005; 8 (4):286-294.
(254) Weymiller AJ, Montori VM, Jones LA, Gafni A et al. Helping patients with
type 2 diabetes mellitus make treatment decisions: statin choice randomized
trial. Arch Intern Med. 2007; 167 (10):1076-1082.
Medicines concordance: full guideline DRAFT (July 2008) page 371 of 373
DRAFT FOR CONSULTATION
(255) Wilkinson CR, Williams M. Strengthening patient-provider relationships.
Lippincott's case management : managing the process of patient care. 2002; 7
(3):86-99.
(256) Wills CE, Holmes RM. Patient comprehension of information for shared
treatment decision making: State of the art and future directions. Patient Educ
Couns. 2003; 50 (3):285-290.
(257) Wilson HS, Hutchinson SA, Holzemer WL. Reconciling incompatibilities: a
grounded theory of HIV medication adherence and symptom management.
Qual Health Res. 2002; 12 :1309-1322.
(258) World Health Organization. Adherence to long-term therapies;
evidence for action. WHO. 2003
(259) Wyatt GE, Longshore D, Chin D, Carmona JV et al. The efficacy of an
integrated risk reduction intervention for HIV-positive women with child
sexual abuse histories. AIDS & Behavior. 2004; 8 (4):453-462.
(260) Xiong W, Phillips MR, Hu X, Wang R et al. Family-based intervention for
schizophrenic patients in China. A randomised controlled trial. Br J
Psychiatry. 1994; 165 (2):239-247.
(261) Zermansky AG, Petty DR, Raynor DK, Lowe CJ et al. Clinical medication
review by a pharmacist of patients on repeat prescriptions in general practice:
A randomised controlled trial. Health Technol Assess. 2002; 6 (20).
(262) Zhang M, Wang M, Li J, Phillips MR. Randomised-control trial of family
intervention for 78 first-episode male schizophrenic patients. An 18-month
study in Suzhou, Jiangsu. Br J Psychiatry Suppl. 1994; 165 (Suppl 24):96-102.
Medicines concordance: full guideline DRAFT (July 2008) page 372 of 373
DRAFT FOR CONSULTATION
Appendices
Please note the appendices are available as separate files.
Medicines concordance: full guideline DRAFT (July 2008) page 373 of 373