Download Adherence to Lipid-Lowering Drug Therapy among Members of the

MILITARY
MEDICINE,
173, 7:666, 2008
Adherence to Lipid-Lowering Drug Therapy among Members of
the Canadian Forces
Janice Ma, PharmD*; LCol Regis Vaillancourt, Canadian Forces Health Servicest;
CarolBennett,MSct
,
ABSTRACT Objective: This study was performed to quantify adherence rates to lipid-lowering drug therapy among
members of the Canadian Forces (CF) and to identify factors associated with nonadherence. Methods: Pharmacy claims
were reviewed for all CF members who received a lipid-lowering drug between April I and June 1, 2003. Subjects were
categorized as adherent if records indicated consumption of at least 80% of prescribed doses. Logistic regression was
performed to assess the impact of patient and drug characteristics upon adherence. Results: Overall adherence rate at I
year was 38.5% among all users of lipid-lowering medications. Adherence did not vary among the different classes of
lipid-lowering drugs. Duration of service was the only independent predict9r of adherence. Conclusions: Despite a
relative lack of treatment barriers and the presence of established treatment programs in the CF health care system,
long-term adherence with lipid-lowering medications remains suboptimal in this population.
I'
I'
INTRODUCTION
Lipid-lowering therapy, particularly
with 3-hydroxy-J=methyl"',
glutaryl-coenzyme A reductase inhibitors (statins), has been
proven to reduce cardiovascular morbidity and mortality.1-3
However, sustained use of such medications is required to
achieve these benefits. Two major studies have indicated that
long-term adherence to lipid-lowering drug therapy is low
among elderly patients, with less than a third of prescriptions
continuing to be taken 1 year after initiation of therapy.4,5
Adherence is similarly poor even among high-risk patients
with established cardiovascular disease.5,6This gap in treatment represents a lost opportunity to reduce morbidity and
mortality due to cardiovascular disease.
In the Canadian Forces (CF), significant expenditures are
associated with lipid-lowering medications, which are continuously among the most common and most costly classes of
medication prescribed for our members. A previous study
performed in our population demonstrated that screening for
*Directorate of Medical Policy, Pharmacy Policies and Standards, National Defence Medical Centre Building, 2nd Floor, 1745 Alta Vista Drive,
Ottawa, Ontario, KIA OK6, Canada.
tChildren's
Hospital of Eastern
Ontario, KIH 8Ll, Canada.
:j:Ottawa Health Research
Ontario,
Institute,
401 Smyth
Administrative
Road, Ottawa,
Services
Building,
1053 Carling Avenue, Ottawa, Ontario, KIY 4E9, Canada.
Presented at the Joint Therapeutics Congress 2005 (Canadian Association
for Population Therapeutics, Canadian Society of Clinical Pharmacology,
and the Canadian College of Clinical Pharmacy), April 17-19, 2005, Vancouver, British Columbia, Canada; the Western Branch Seminar 2006, Canadian Society of Hospital Pharmacists, March 10-12, 2006, Banff, Alberta,
Canada; and the World Congress of Pharmacy and Pharmaceutical Sciences
and the 66th International Congress of the International Pharmaceutical
Federation, International Pharmaceutical Federation and Federal Council of
Pharmacy of Brazil, July 25-August 31, 2006, Salvador Bahia, Brazil.
This manuscript was received for review in August 2007. The revised
manuscript was accepted for publication in April 2008.
Reprint & Copyright @ by Association of Military
2008.
666
Surgeons
of U.S.,
dyslipidemia is very comprehensive, with almost all serving
members screened irrespective of their baseline risk for cardiovascular disease.? However, this study also documented
that, for a significant proportion of our members, control of
their dyslipidemiawas suboptimaI.7
r
Our department (Directorate of Medical Policy, Pharmacy
Policies and Standards) has also evaluated the impact of two
different initiatives upon dyslipidemia control, both of which
relied primarily upon pharmacist interventions. In the first
initiative, selection of three preferred statins for inclusion on
our benefit list provided an opportunity for our pharmacists to
discuss optimal doses of statins during therapeutic substitution.8In the second initiative, pharmacists at three specific CF
practice sites provided more intensive counseling and follow-up to patients referred by their primary care physician.9
Both of these initiatives demonstrated improvement in dyslipidemia profiles for patients who were subject to pharmacists' interventions. Our organization has thus invested and
continues to dedicate a significant amount of resources to the
identification and management of dyslipidemia.
Despite the success demonstrated with these efforts, however, we recognized that not all CF members would be able
to access or otherwise receive intensive counseling and follow-up from pharmacists or other health care providers, due
to the constraints inherent in oui health care system. Because
our younger, relatively disease-free population stands to benefit substantially from lipid-lowering therapy, we were interested to know whether our previous initiatives, coupled with
increasing awareness of the importance of dyslipidemia management, were associated with greater adherence to lipidlowering therapy overall in CF members.
OBJECTIVES
The primaryobjectivefor this study was to quantify adherence rates to lipid-lowering drug therapy among CF members. Secondary objectives involved comparison of adherence
MILITARY MEDICINE, Vol. 173, July 2008
Clinical Report
rates for different classes of lipid-lowering drugs and identification of factors that were independently associated with
nonadherence.
METHODS
Study Design
A cohort study was perfonned using infonnation from two
phannacy databases: one containing prescriptions dispensed
from military phannacies and another containing prescriptions. obtained by CF members from civilian phannacies.
These two databases were merged into a single file which was
subsequently blinded for analysis. The study protocol was
submitted to and received approval from the Ottawa Hospital
Research Ethics Board before its execution.
Study Population
Male and female Regular Forces CF members who received
a lipid-lowering agent between the dates of April I, 2003 and
June 1, 2003 were eligible for inclusion in this study. Patients
who had <3 months of follow-up infonnation related to
prescription drug use were excluded from the study.
Data Collection
For each eligible patient, infonnation was collected on patient
demographics-age, sex, rank Uunior or senior noncommissioned member [NCM] vs. junior or senior officer), service
element (air, land, sea), duration of service, and geographic
location by military base/wing-and medication use (drug,
dose, date(s) of prescriptions, and quantities dispensed).
Data Analysis
Adherence rates were calculated based on the method previously described by Jackevicius.5 Prescriptions for lipid-lowering medications were assessed for 1 year following the
index prescription. Patients were deemed adherent if their use
of lipid-lowering medication(s) was sustained during 12 consecutive months. Sustained use was defined based on the
pattern of refills obtained: for each prescription, actual refill
dates were compared to expected refill date~ that were identified based on the dose prescribed and quantity dispensed,
allowing a 20% grace period for late refills. Patients were
required to refill all prescriptions for specific lipid-lowering
medications within the expected "windows" during a 12month period to be deemed adherent. The X2test statistic was
then used to examine differences between adherent and nonadherent members for different drug classes (with combination therapy excluded from this analysis.). Stepwise multivariate analysis was used to identify factors independently
associated with nonadherence.
RESULTS
A total of 1,360 members was included in the cohort. As
expected in a military population, the members were predominantly male (97% of study subjects); this proportion was
MILITARY MEDICINE, Vol. 173, July 2008
---7
higher than among the overall CF population, where ~85% of
CF members are male. Consistent with the disease state being
evaluated, the mean age in this cohort was higher than the
general CF population (44.6 years, range 24-61 years, com:"
pared to 36 yearsin the CF overall).Durationof servicereflected
the age of subjects, averaging 23.4 years (range, 1-38 years).
The majority of cohort members were senior NCMs (49%~.
The overall I-year adherence rate for all lipid-lowering
medications was 38.5%. Nonadherent members were younger,
had fewer years of service, and were less likely to be senior
NCMs (Table I). There were no significant differences in
adherence between members taking different classes of lipidlowering medications (Fig. 1). Due to variability in the numbers of eligible members located at different military sites
(ranging from 1 to 257 at 30 distinct sites), no significant
differences were found in adherence rates among the CF
bases (range, 14.3-100%, data not shown).
Medication class was not significantly associated with
nonadherence (p < 0.15) in univariate analysis, and was thus
excluded from the multivariate analysis. Univariate analysis
identified age and duration of employment to be significant
predictors of adherence (Table II), however, duration of service was the only remaining variable which was independently associated with nonadherence in multivariate analysis
(data not shown).
DISCUSSION
Our results, although disappointing, are not entirely unexpected. Medication taking is a complex behavior, with many
TABLE I.
Demographic
Characteristics
Nonadherent
Adherers
Characteristic
Age (categorical)
24-34 years
35-40 years
41-44 years
45-48 years
49-61 years
Sex
Male
Female
Rank
Junior NCM
Senior NCM
Junior officer
Senior officer
Service element
Air
Land
Sea
Duration of service
1-18
19-23
24-28
29-38
(n = 524)
of Adherent and
Members
Nonadherers
(n = 836)
p
<0.0001
16
73
139
132
164
(34.8%)
(28.5%)
(34.9%)
(42.6%)
(46.9%)
30
183
259
178
186
(65.2%)
(71.5%)
(65.1 %)
(57.4%)
(53.1 %)
0.09
514 (38.9%)
10 (25.6%)
807 (61.1 %)
29 (74.4%)
137
281
46
60
264
385
87
100
0.05
(34.2%)
(42.2%)
(34.6%)
(37.5%)
(65.8%)
(57.8%)
(65.4%)
(62.5%)
0.07
245 (42.0%)
186 (36.4%)
93 (35.0%)
,338 (58.0%)
325 (63.6%)
173 (65.0%)
<0.0001
100
117
152
155
(31.2%)
(33.9%)
(40.6%)
(48.4%)
221
228
222
165
(68.%)
(66.1%)
(59.4%)
(51.6%)
667
Clinical Report
Adherence Rates by Drug Class
classes of lipid-lowering medications. We had hypothesized
that, given greater awareness and concerns regarding the side
effects of statins-with one agent (cerivastatin) recently with80%
drawn from the market due to an increased risk of rhabdomyolysisl6-other lipid-lowering agents may have been used
60%
preferentially in managing dyslipidemia among our members.
Our results indicate, however, that statins remain by far the
40%
most commonly prescribed lipid-lowering agents, although a
20%
notable number of our members are managed with other
agents either alone or in combination with a statin. Adherence
0%
rates were not shown to vary with drug class, although our
i::P
.i
~q,
statistical
power to detect such a difference was likely limited
0,1>
~1>
~
«~
iJ-~
due
to
the
small numbers of individuals taking non-statin
o~
C3
lipid-lowering agents alone.
Therapeutic Class
Our study, like those of other authors, is also limited by
:IGURE 1. Adherence rates by lipid-lowering drug class. All comparithe quality of information in our pharmacy claims data,ons not significant (;f > 0.2). 'Not included in Jf analysis.
base. Although such databases are an invaluable source of
information on large groups, such information may not be
TABLE II. Factors Associated with Nonadherence
complete. Most notably, we cannot estimate the number of
(Univariate Analysis)
prescriptions for lipid-lowering medications which were
never filled at the outset. We also cannot determine
Factor
Odds Ratio
95% CI
p Value
whether all doses dispensed were indeed taken and taken
<0.0001
Age group (years)
0.87-3.14
24-34
1.65
appropriately, a point of importance given that certain
2.21
1.57-3.11
35-40
statins are more effective when administered in the evening.
41-44
1.64
1.22-2.21
Our
results may thus underestimate the true rates of nonad1.19
0.87-1.62
45-48
herence in our population. However, these limitations are
1.00
49-61
Male sex
0.54
0.26-1.12
0.10
also present in other studies of this design and thus allow
Rank
0.05
for a comparable analysis of our performance to other
0.79-1.69
Junior NCM
1.16
populations.
0.58-1.17
Senior NCM
0.82
Despite its limitations, our study does suggest some speJunior officer
1.14
0.70--1.83
Senior officer
1.00
cific opportunities for improving management of dyslipide<0.0001
Duration of employment
mia in CF members. Most importantly, despite the positive
2.08
1.50--2.87
1-18
impact associated with previous pharmacy-centered interven1.83
1.34-2.50
19-23
tions, such efforts alone are obviously not sufficient to
1.37
1.02-1.85
24-28
1.00
29-38
ensure consistent and sustained uptake of lipid-lowering
therapy over the longer term on a population-wide basis.
CI, Confidence interval.
This is particularly true in the present age, when extreme
shortages of pharma~ists and other primary health care
opportunities for mishap. Many contributing factors to non- providers-and competing demands on such personneladherence have been identified previously in diverse patient may not allow them to devote sufficient resources to manpopulations, both with regard to medications and other ther- age complex disease states such as dyslipidemia. Additionapeutic interventions.1OFactors influencing adherence include ally, the management of dyslipidemia has been evolving at
both patient-specific elements, such as attitudes toward ill- a rapid pace, with new guidelines and more aggressive
ness and medication taking, !l.12as well as characteristics treatment goals introduced on a frequent basis. This further
unique to the diseasel3-15and medication regimensl3 (includ- challenges practitioners attempting to remain up-to-date on
ing complexity of dosing, number of medications, and side the most appropriate management strategies, let alone
effects). In our population, many factors which have been communicate the importance of these to their patients in a
,
shown or hypothesized to contribute to nonadherence (e.g., consistent manner.
illiteracy, prohibitive drug acquisition costs) are lower or
The disappointing results of our current study, when
absent. Nevertheless, our study has found that levels of ad- viewed in light of our extensive work thus far to promote
herence to lipid-lowering medication ~e equally poor in judicious use of lipid-lowering medications, strongly suggest
Canadian military personnel as among other populations that future efforts to address this issue should perhaps be
studied to date.
expanded to involve other health professionals in a structured
In contrast to other studies, our research also sought to manner. Our previous review of activities undertaken by
determine whether adherence was similar among different pharmacists at lipid clinics revealed that many of the profes100%
668
MILITARY MEDICINE, Vol. 173, July 2008
--
~
~
~
sional interventions included referrals to other health care
providers (predominantly dieticians), and general recommendations to increase physical activity levels.9 Given that diet
and exercise also have positive influences upon lipid levels
and cardiovascular health outcomes, health care professionals
who are trained to address these elements should also be
involved in the standard management of members with dyslipidemia. This approach is also in accordance with many
recent Canadian guidelines for management of diabetes and
other conditions associated with dyslipidemia.17,18
Our results alsoLsuggestthat our lipid clinic programs may
do well to embrace a more comprehensive focus on cardiovascular health, or overall health in general, rather than
focusing primarily upon control of dyslipidemia. Presently,
entry into the lipid clinic program requires initial referral
from the primary care physician; the program will thus selectively target those members with established dyslipidemia,
as well as those who have failed to respond to usual treatment. However, many members may not perceive adequate
benefit from the multiple labor-intensive interventions (involving diet, exercise, drug therapy, laboratory monitoring,
and physician follow-up) required to adequately manage lipid
abnormalities alone, thus contributing to noncompliance in
the long-term. A relatively isolated focus upon dyslipidemia
management may also not be warranted, given the increasing
body of evidence implicating dyslipidemia as a factor in more
complex disease states such as metabolic syndrome.19Instead, programs which focus on improving overall cardiovascular health may be better received by military members, as
these would focus on both short-term and long-term benefits
together. Immediate improvements in exercise tolerance and
general feelings of well-being can often result from established cardiovascular risk reduction strategies, such as smoking cessation, weight loss, and structured exercise regimens,
and may also be associated with greater long-term adherence
to treatment.2O:21
By focusing on these interventions in place
of dyslipidemia-specific targets, we may achieve better longterm treatment adherence and improved health outcomes in
the future.
CONCLUSIONS
An extensive review of our pharmacy claims database revealed that adherence to lipid-lowering therapy was suboptimal among members of the Canadian Forces. Consistent with
reports in civilian populations, adherence rates at 1 year
following initiation of lipid-lowering therapy were -40%
overall. Adherence was independently associated with duration of military service, perhaps reflecting the greater perceived importance of reducing cardiovasc,ular risk through
medication-taking behavior among older members. Future
efforts to manage dyslipidemia on a more widespread basis
should involve multidisciplinary health care teams, and could
perhaps be tailored to address overall cardiovascular health
MILITARY MEDICINE, Vol. 173, July 2008
and associated short-term benefits, such as improved exercise
tolerance and weight loss, as opposed to focusing upon lipid
levels alone.
ACKNOWLEDGMENTS
This research was funded by the Canadian Forces Health Services Quality of
Life Research Programme through the Office of the Surgeon GeneraL We
also thank Ms. Julie Lanouette, who was involved in collecting data for this
study.
REFERENCES
I. Scandinavian Simvastatin Survival Study Group: Randomised lrial of
cholesterol lowering in 4444 patients with coronary heart disease: the
Scandinavian
Simvastatin Survival Study (4S). Lancet 1994; 344:
1383-9.
2. Shepherd J, Cobbe SM, Ford I, et al: For the West of Scotland Coronary
Prevention Study Group: prevention of coronary heart disease in men
with hypercholesterolemia.
N Engl J Med 1995; 333: 1301-7.
3, Downs JR, Clearfield M, Weis S, et al: For the AF-CAPSDrrexCAPS
Research Group: primary prevention of acute coronary events with
lovastatin in men and women with average cholesterol levels: results of
AFCAPSrrexCAPS.
JAMA 1998; 279: 1615-22.
4. Benner JS, Glynn RJ, Mogun H, Neumann PJ, Weinstein MC, Avorn J:
Long-term persistence in use of statin therapy in elderly patients. JAMA
2002; 288: 455-61.
5. Jackevicius CA, Mamdami M, Tu JV: Adherence with statin therapy in
elderly patients with and without acute coronary syndromes. JAMA
2002; 288: 462-7.
6. Tsuyuki RT, Johnson JA, Teo KK, et al: A randomized trial of the effect
of community pharmacist intervention on cholesterol risk management:
the Study of Cardiovascular Risk Intervention by Pharmacists (SCRIP).
Arch Intern Med 2002; 162: 1149-55.
7. Spaans IN, Coyle D, Fodor G, et al: Application of the 1998 Canadian
cholesterol guidelines to a military population: health benefits and cost
effectiveness of improved cholesterol management. Can J Cardiol 2003;
19: 790-6.
8. Moisan J, Vaillancourt
R, Gregoire JP, Gaudet M, Cote I, Leach A:
Preferred hydroxymethylglutaryl-coenzyme
A reductase
inhibitors:
treatment-modification
program and outcomes, Am J Health Syst Pharm
1999; 56: 1437-41.
9. Vaillancourt R, Gutschi LM, Ma J, Sinclair S, Beechinor D: Pharmacistmanaged lipid clinics: development and implementation in the Canadian
Forces. Can J Hosp Pharm 2003; 56: 24-31.
10. Krueger KP, Berger BA, Felkey B: Medication adherence and persistence: a comprehensive review. Adv Ther 2005; 22: 313-56.
II. Jones I, Britten N: Why do some patients not cash their prescriptions?
Br J Gen Pract 1998; 48: 903-5.
12. McElnay JC, McCallion CR, al-Deagi F, Scott M: Self-reported medication non-compliance in the elderly. Eur J Clin Pharmacol 1997; 53:
171-8.
13, Agarwal MR, Sharma VK, Kishore Kumar KV, Lowe D: Non-compliance with treatment in patients suffering from schizophrenia: a study to
evaluate possible contributing factors. Int J Sac Psychiatry 1998; 44:
92-106.
14, Yuan Y, L'italien G, Mukherjee J, Iloeje U: Determinants of discontinuation of initial highly active antiretrovlral therapy regimens in a US
mV-infected patient cohort. HIV Med 2006; 7: 156-62.
15. Chisholm MA, Lance CE, Mulloy LL: Patient factors associated with
adherence to immunosuppressant
therapy in renal transplant recipients.
Am J Health Syst Pharm 2005; 62: 1775-81.
16. Bayer, Inc.: Market withdrawal of Baycol (cerivastatin). Available at
http://www.hc-sc.gc.caldhp-mps/medeff/advisories-avis/profl2oo
.2_hpc-cps_e.html; accessed March 10, 2006.
Ilba ycol-
669
Clinical Report
17. Canadian Diabetes Association
Committee:
Canadian Diabetes
Clinical Practice Guidelines Expert
Association 2003 Clinical Practice
Guidelines for the Prevention and Management of Diabetes in Canada.
Can J Diabetes 2003; 27(Suppl 2): SI-152.
18. Khan NA, McAlister FA, Lewanczuk RZ, et al: Canadian Hypertension
Education Program: the 2005 Canadian Hypertension Education Program recommendations
for the man~gement of hypertension. Part II.
Therapy. Can J Cardiol 2005; 21: 657-72.
19. International Diabetes Federation: The International Diabetes Federation
consensus worldwide definition of the metabolic syndrome, 2005. Available at http://www.idf.org/home/index.cfm?unode=
1120071E-AACE41D2-9FAO-BAB6E25BA072;
accessed March 14, 2006.
20. Stotland SC, Larocque
M: Early treatment
response as a predictor of
ongoing weight loss in obesity treatment. Br J Health Psychol 2005; 10:
601-14.
21. Diabetes Prevention Program Research Group: Achieving weight and
activity goals among Diabetes Prevention Program lifestyle participants.
Obes Res 2004; 12: 1426~34.