Download Evidence for Appropriate Pain Treatment Guidelines

Evidence For Appropriate Pain
Treatment Guidelines
Ro g e r C h o u , M D
Professor of Medicine
Oregon Health & Science University
D i r e c t o r, Pa c i f i c N o r t h w e s t E v i d e n c e - b a s e d P r a c t i c e C e n t e r
Conflict of interest disclosure
Dr. Chou has received researc h funding
from the Agency for Healthcare Researc h
and Quality
Purpose
• Understand opioid pharmacology
• Understand opioid prescribing patterns
• Understand risks and benefits of long term opioid therapy for c hronic pain
• Understand implications for clinical
practice
Nothing is intrinsically good or evil,
but its manner of usage may make it
so.
-St. Thomas Aquinas
5
Case
• 53-year old female transferring care because her PCP is
leaving practice
• Shoulder and hip pain 2˚ avascular necrosis, s/p shoulder replacements, hip
decompression, hip replacement
• Fibromyalgia, non-radicular LBP, chronic headache
• Depression, fatigue
• Gastroparesis, irritable bowel syndrome
• Morphine IR 30 mg 5 T (150 mg) q 8 hrs + oxycodone 5 mg 8 T (40 mg)
q 6 hrs
•
•
•
•
MED/day: 690 mg
Modafinil 20 mg po qD
Pain 6/10 on average, with day to day fluctuation
Can carry out ADLs with pain, limited exercise, no aberrant behaviors
Background
• C h ro n i c
•
n o n c a n c e r p a i n h i g h l y p r e va l e n t , w i t h
substantial burdens
• Estimates vary, up to 1/3 of adults report some CNCP
O p i o i d s h a ve b e c o m e c o m m o n l y p r e s c r i b e d fo r c h ro n i c
noncancer pain
• About 5% of adults report use of LOTa
• The U.S. is ~5% of the world’s population, but accounts for 80% of the
world’s supply of opioids (99% of hydrocodone)
• Opioids
a r e a s s o c i a t e d w i t h p o t e n t i a l h a r m s, b o t h t o
patients and to society
aBoudreau
et al Pharmacoepidemiol Drug Saf 2009
US opioid sales
quadrupled
2000-2010
Since 2008, 15,000
deaths per year.
This exceeds MVA
deaths in 30 states.
Slide courtesy Mark Sullivan
Trends in opioid prescribing and rates of abuse
Dart RC 2015;372:241-8
10
Rate Of Death Association with Heroin and
Prescription Opioids
Dart RC et al. N Engl J Med 2015;372:241-248.
11
Opioid pharmacology
• Opioid mu-receptors mediate analgesic effects and AE’s
•
•
•
Agonists, partial agonists, antagonists
Natural, semisynthetic, synthetic
Half-life 2-4 hours for most opioids; 15-30 hours for methadone
• Ongoing exposure causes tolerance
•
•
•
Larger dose required to maintain original effects (analgesic and AE’s)
Interindividual variability in development of tolerance
“There appears to be no limit to the development of tolerance, and
with appropriate dose adjustments, patients can continue to obtain pain
relief.”—Inturrisi C. Clin J Pain 2002;18:S3-13
• No theoretical dose ceiling
12
Dose-response relationship for respiratory depression
Dahan A. Br J Anaesth 2005;94:825-34
13
Prescribing trends
•
•
1980 vs. 2000: Opioid use for chronic pain increased from 8% to
16%
• Use of more potent opioids increased from 2 to 9%
1997 vs. 2005: Across age groups and in men and women, increase
in long-term opioid use ranged from 61% to 135%
• Increased use of schedule II opioids; greatest increase in daily doses occurred
in prescriptions of schedule II opioids
•
2000 vs. 2005: Increases in opioid use in commercial and Medicaid
populations
• Greatest increase in schedule II opioids; yearly dose increase greatest for
schedule II opioids
Caudill-Slosberg MA. Pain 2004;109:514; Boudreau D. Pharmacoepidemiol Drug Safety
2009;18:1166; Sulivan MD 2008;138:440; Campbell CI Am J Pub Health 2010;100:2541
14
How did we get here?
• Perceived undertreatment of chronic pain
• Laws or regulations passed in >20 states to allow use of opioids for chronic pain
• Low risk of abuse observed with use of opioids in palliative care settings
• “…patients rarely demonstrate euphoric responses to opioid drugs, and neither analgesic
tolerance nor physical dependence is a significant clinical problem.” –Portenoy RK. J Law
Medicine Ethics 1996;24:296
• Case series describing benefits of long-term opioid therapy for chronic pain, with
low rates of abuse, addiction, or other serious AE’s
•
Most prescribed low doses (<20 mg MED/day)
• No ceiling dose used in palliative care settings
• “Escalation of the opioid dose until either adequate analgesia occurs or intolerable and
unmanageable side effects supervene is standard practice in cancer pain management.”—
Portenoy RK. J Pain Symptom Management 1996;11:203
• Emphasis on round-the-clock dosing using sustained-release formulations
Portenoy RK 1986;25:171; Haythornthwaite JA 1998;15:185
15
Randomized Trial Evidence for Commonly
Used Medications from Recent Meta-Analyses
Medication class
Number
of trials
Number
Personof patients years
(est.)
Number US
adults using longterm
Antihypertensivesa
147
~464,000
~1,857,000
48 million
Statinsb
26
~169,000
~753,000
34 million
NSAIDsc
31
~116,000
~117,000
6 million
Opioidsd
62
~12,000
~1,500
7-9 million
a.
b.
c.
d.
Law et al., BMJ 2009.
CTT Collaboration, Lancet 2010.
Trelle et al., BMJ 2011.
Furlan et al. Pain Res Manage 2011.
Courtesy Michael Von Korff
• Short-term efficacy
•
Evidence on
effectiveness of
LOT for chronic
non-cancer pain
•
•
•
62 RCT’s in one recent meta-analysis, duration <16
weeks in 61a
Opioids more effective than placebo for nociceptive
and neuropathic pain (effect sizes 0.55-0.60)
Maximum dose ≤180 mg MED/day in all trials except
for 3
Long-term effectiveness
• Cochrane review included 26 studies >6 monthsb
• 25 studies were case series or uncontrolled long-term
•
trial continuations
Many discontinuations due to adverse effects (23%) or
insufficient pain relief (10%), but some evidence that
patients who continue on opioids experience long-term
pain relief
aFurlan
bNoble
et al. Pain Res Manag 2011
et al. Cochrane Database Syst Rev 2010
•
Other limitations
of the evidence
on effectiveness
of LOT
•
•
•
•
Effects on function generally smaller than
effects on pain, with some trials showing no
or minimal benefit
High loss to follow-up
Trials typically excluded patients at higher
risk for abuse or misuse, psychological
comorbidities, and serious medical
comorbidities
Limited evidence on commonly treated
conditions
• Low back pain, fibromyalgia, headache, others
No trials compared LOT vs. CBT-based
exercise therapy or interdisciplinary
rehabilitation
•
Overdose
1 fair-quality study of chronic pain
patients in a US healthcare system who
had received opioids at some point
• Recent opioid use associated with
increased risk of any overdose event
(adjusted HR 5.2, 95% CI 2.1 to 2.5)
and serious overdose events (adjusted
HR 8.4, 95% CI 2.5 to 2.8)
Factors
associated with
increased risk of
overdose, or
observed in high
proportions of
overdoses
•
•
•
•
•
•
•
Aberrant behaviors
Recent initiation of opioids
Methadone
Concomitant use of benzodiazepines
Obtaining opioid prescriptions from
multiple providers
Substance abuse and other
psychological comorbidities
Higher dose
• Tools for risk stratification are available,
Patient selection
and risk
stratification
but require further evaluation
• Opioid Risk Tool
• SOAPP version 1 and SOAPP-R
• Accuracy for predicting future aberrant
behaviors tend to be high in initial studies
but poor in subsequent (limited) studies
Administration
Opioid Risk Tool (ORT)
• On initial visit
• Prior to opioid therapy
Scoring
• 0-3: low risk (6%)
• 4-7: moderate risk
(28%)
• > 8: high risk (> 90%)
Webster & Webster. Pain Med. 2005;6:432.
22
High-dose opioid therapy prescribing patterns
• Small proportion of patients account for the majority of opioids prescribed
•
In one study, 5% of opioid users accounted for 48% to 70% of total use
• Factors associated with high dose therapy include:
•
•
•
•
•
Presence of substance use disorders
Presence of mental health disorders
Use of sedative-hypnotics and multiple opioids
Higher health service utilization
Multiple pain problems and high levels of medical and psychiatric comorbidity
• “Adverse selection”
• Trends show increases in opioid prescribing in patients with substance use and
mental health disorders
Weisner CM. Pain 2009;145:287; Edlund MJ. J Pain Symptom Manage 2010;40:279;
Morasco BJ 2010;151:625; Kobus AM. J Pain 2012;13:1131; Braden JB 2009;31:564
•
Overdose: Doseresponse
relationship
•
•
•
Observational studies consistently show an
association between opioid dose and risk of
overdose or death in patients with chronic
pain
Risk starts to increase at relatively low doses
and continues to increase
Studies matched or adjusted for potential
confounders available in administrative
databases
• Potential for residual confounding by indication
Difficult to determine whether patients had
chronic pain and duration of therapy

Cohort study (n=9940, 51 opioid overdoses, 6 fatal)

Association
between opioid
dose and
overdose risk

Case-control study (VA, 750 cases)


Risk of opioid overdose (vs. 1to <20 mg/day)
 >=100 mg/d:
HR 8.9 (4.0-20)
 50 -<100 mg/d: HR 3.7 (1.5-9.5)
 20-<50 mg/d:
HR 1.4 (0.57-3.6)
Risk of opioid overdose-related death (vs. 1 to <20
mg/day)
 >=100 mg/d:
HR 7.2 (4.8-11)
 50-<100 mg/d: HR 4.6 (3.2-6.7)
 20-<50 mg/d:
HR 1.9 (1.3-2.7)
Nested case-control study (Ontario, 498 cases)
 Risk of opioid-related mortality (vs. 1 to <20
mg/day)
 >=200 mg/d:
 100-199 mg/d:
 50-99 mg/d:
 20-49 mg/d:
OR 2.9 (1.8-4.6)
OR 2.0 (1.3-3.2)
OR 1.9 (1.3-2.8)
OR 1.3 (0.94-1.8)
Ref: Dunn et al. Ann Intern Med 2010;152:85-92; Bohnert et al. JAMA
2011;305:1315-21; Gomes et al. Arch Intern Med 2011;171:686-91
25
Dose-related risk of opioid overdose
Risk of adverse event
10
9
8
7
6
Dunn 2010
Risk Ratio 5
Bohnert 2011
Gomes 2011
4
Zedler 2014
3
2
1
0
<20 mg/day
20-49 mg/day
50-99 mg/day
Dose in mg MED
>=100 mg/day
Courtesy
Gary Franklin
•
Estimates vary from 4% to 26%, or higher
• One study (n=801) based on standardized
interviewsa
Abuse, addiction,
misuse
•
•
•
•
•
•
•
•
26% purposeful oversedation
39% increased dose without prescription
8% obtained extra opioids from other doctors
18% used for purposes other than pain
12% hoarded pain medications
Definitions inconsistent across studies and
behaviors evaluated vary in seriousness
Poorly standardized methods to detect these
outcomes
Data from efficacy trials underestimate risks
aFleming
et al. J Pain 2007
•
Abuse, addiction,
misuse: Dosedependent
effects
One fair-quality study of a large claims
database found long-term prescribed
opioid use (>90 days’ supply) associated
with increased risk of an opioid abuse or
dependence diagnosis vs. no opioid
treatment
• Low dose (1-36 mg MED/day): OR 15 (95% CI
•
•
10 to 21)
Moderate dose (36-120 mg MED/day): OR 29
(95% CI 20 to 41)
High dose (≥120 mg MED/day): 122 (95% CI 73
to 206)
Edlund MJ. Clin J Pain 2014;30:557
•
Opioid use associated with increased risk of
fracture in 1 fair-quality cohort study
(adjusted HR 1.28, 95% CI 0.99 to 1.64)
• 1 to <20 mg MED/day: HR 1.2 (95% CI 0.92 to 1.56)
• >=50 mg MED/day: HR 2.0 (95% CI 1.24 to 3.24)
• Overall test for dose-response did not reach statistical
Fracture
significance
•
Similar increased risk of fracture in 1 goodquality case-control study (OR 1.27, 95% CI
1.21 to 1.33)
• Risk highest with one prescription and no longer present
with >20 cumulative prescriptions
Saunders KW. J Gen Intern Med 2010;25:310 Li L. Am J
Epidemiol 2013;178:559.
•
Cardiovascular
events
•
Cohort study: adjusted IRR 2.66 (95% CI
2.30 to 3.08) for chronic opioid therapy
user vs. general population
•
•
•
•
•
Versus cumulative dose of 0 to 1350 mg MED over 90 days
1350 to <2700 mg: IRR 1.21 (95% CI 1.02 to 1.45)
2700 to <8100 mg: IRR 1.42 (95% CI 1.21 to 1.67)
8100 to <18,000 mg: IRR 1.89 (95 5CI 1.54 to 2.33)
>18,000 mg: IRR 1.73 (95% CI 1.32 to 2.26)
Case-control study: adjusted OR 1.28 (95%
CI 1.19 to 1.37) for MI with current opioid
therapy versus nonuse
• Highest risk with 11 to 50 cumulative prescriptions
• OR 1.09 to 1.29 with 1-2, 3-10, or >50 cumulative
prescriptions
Carman WJ. Pharmacoepidemiol Drug Saf 2011;20:754; Li L.
J Intern Med 2013;273:511
• Cross-sectional study
• Men with back pain
• Outcome: Use of medications for erectile
Endocrinological
events
•
•
•
dysfunction or testosterone replacement
Opioid use versus non-use: adjusted OR 1.5, 95%
CI 1.1 to 1.9
>=120 mg MED/day vs. 0 to <20 mg day: OR
1.6, 95% CI 1.03 to 2.4
No increased risk at doses of 20 to <120 mg
MED/day
Deyo RA. Spine 2013;38:909
31
FDA Public Health Advisory, November 2006
“Methadone Use for Pain Control May
Result in Death and Life-Threatening
Changes in Breathing and Heart Beat”
http://www.fda.gov/Drugs/DrugSafety/PublicHealthAdvisories/ucm124346.htm
•
Increased methadone deaths nationwide

Methadone

•
Half-life 15 to 60 hours, up to 120 hours



•
•
Methadone accounted for 1.7% of opioid rx’s in 2009
and 9.0% of morphine equivalents in 2010a
Involved in 31% of opioid-related deaths, and 40%
of single-drug deathsa
60 hour half-life=12 days to steady-state
Prolongation of QT intervals, sudden death
Start at 2.5 mg q8 hrs, increase slowly
Higher doses of methadone associated with
greater QTc interval prolongation
• High proportion of reported cases of torsades de
pointes occurred in patients prescribed >200 mg/day
Morphine to methadone conversion ratio
increases at higher doses
aMMWR
2012;61:493-7
Methadone
•
Little evidence on use of methadone for
CNCP




One small, poorly designed trial of methadone
vs. placebo
A VA cohort study found methadone associated
with lower mortality risk compared to morphine
(HR 0.38 to 0.66)a
A cohort study based on Tennessee Medicaid
data found methadone associated with higher
mortlaity risk compared to methadone (HR 1.46,
95% CI 1.17 to 1.83)b
No evidence on effects of ECG monitoring on
risks associated with methadone, but some
guidelines recommend routine ECG monitoring
Krebs EE et al. Pain 2011;152:1789-1795; Ray WA et al. JAMA Int Med
2015;175:420-7
a
34
Time to reach steady state
35
Prolonged QTc and torsades de pointes
36
Morphine to methadone conversion
Oral morphine to methadone
24 hour total oral morphine
conversion ratio
<30 mg
2:1
31-99 mg
4:1
100-299 mg
8:1
300-499 mg
12:1
500-999 mg
15:1
>1000 mg
20:1
Managing Cancer Pain in Skeel ed. Handbook of Cancer Chemotherapy. 6th
ed., Phil, Lippincott, 2003, p 663
•
Selection of
opioids
•
No clear differences in benefits and harms of
other long-acting opioids
Limited evidence on benefits and safety of
short- vs. long-acting opioids, round-the-clock
versus PRN



Practice of long-acting, round-the-clock dosing
based on cancer guidelines and expert opinion,
potential benefits not proven
Potential harms of long-acting, round-the-clock
opioids include development of hyperalgesia,
tolerance, endocrinologic adverse effects
VA cohort study found long-acting opioid associated
with higher risk of overdose than short-acting
opioids (HR 2.33, 95% CI 1.26 to 4.32)

Risk highest in first 2 weeks after initiation (HR 5.25,
95% CI 1.88 to 14.72)
Miller M. JAMA Intern Med 2015;175:608-15
 One RCT (n=140) of more liberal dose
Dose escalation
versus dose
maintenance




escalation versus maintenance of current doses
VA patients with musculoskeletal pain
No differences in pain, function, or use of
nonopioid medications or physical therapy after
12 months
No difference in withdrawal due to opioid
misuse (24% vs. 30%, RR 0.79, 95% CI 0.46 to
1.4)
Limited separation between groups in opioid
doses at end of trial (mean 52 vs. 40 mg
MED/day)
Naliboff BD. J Pain 2011;12:288
• 2007: Washington Agency Medical
Dose threshold
policies
Directors’ Opioid Dosing Guidelines
• 120 mg MED/day threshold dose
• Re-evaluation and pain management consultation if
needed
• 2009: APS/AAPM guideline
• 200 mg MED/day “watchful” dose
• Based on doses evaluated in trials and observed in
•
observational studies
Recommended re-evaluation for appropriateness
of therapy, enhanced monitoring, consider
consultation
• Subsequent policies have generally
recommended dosing thresholds of 80120 mg/day MED; as low as 50 mg/day
MED (ACOEM)
Effects of dose
limitation
strategies and
policies on
opioid-related
deaths
• In 2007, WA state implemented dosing
policy of <120 mg/day morphine
equivalents in workers’ compensationa
• After 2007, proportion prescribed >120
•
•
•
mg/day decreased by 35%
50% decrease from 2009 to 2010 in
number of opioid-related deaths
Data observational using pre-post
analysis, subject to confounding and
attribution bias
No information about effects of policy on
pain or function
aFranklin
et al, Am J Industrial Med 2011
41
*
 Monitoring, including urine drug testing
 Access prescription drug monitoring data
Mitigating risks
associated with
higher doses of
opioids
 Avoid sedative-hypnotics (particularly





benzodiazepines)
More frequent follow-up
Addiction, pain, or psychiatric consultation
More frequent refills with smaller quantities
Abuse-deterrent formulations?
Naloxone co-prescription?
 Available now in many states
Prescription drug
monitoring
programs
 Studies show that use of PDMPs can identify
cases of diversion and doctor shopping


Recent study found decreased inappropriate drug
prescribing with use of a centralized prescribing
system in Canadaa
Effects on clinical outcomes (e.g., overdose) not
known
 Use variable and generally suboptimal
 PDMPs vary in who can access, information
not available across states
aDormuth
et al. CMAJ 2012
 Optimal frequency and usefulness of
Urine drug tests
individualized vs. routine testing uncertain
 Urine drug tests can be difficult to interpret






Diagnostic accuracy for abuse/addiction difficult
to study
Need to understand metabolic pathways of
different opioids
Differential diagnosis for abnormal results
includes poorly controlled pain, drug abuse,
diversion
Potential for false reassurance
No evidence that urine drug testing improves
patient outcomes; potential for harms
Cost-effectiveness unclear
Opioid-deterrant
formulations
 Opioid-deterrant formulations have recently
been approved by FDA or undergoing FDA
approval process




Designed to be tamper-resistant or co-formulated
with medications that reverse opioid effects or
produce noxious side effects when tampered with
Effectiveness for reducing misuse/substance abuse
and improving clinical outcomes yet to be
established
Likely to be primarily effective in patients who crush
or inject opioids
One study found patients placed on a new tamperresistant formulation of long-acting oxycodone
frequently switched to an alternative opioid or
heroina
aCicero
et al. NEJM 2012
Unintended effects?
Larochelle MR. JAMA Intern Med 2015
48
Case
• Patient initially transferred care in 2008
• Unclear if benefitting from very high doses of long-term opioid therapy,
•
?worsening of GI symptoms; no signs of aberrant behaviors
Slow taper initiated
• Morphine 450 mg/day120 mg/day
• Oxycodone 160 mg/day5 mg po bid prn
•
• 690 mg MED/day135 mg/day
Added non-opioid medications
• Duloxetine 20 mg qD
• Buspirone 30 mg bid
•
Pain and function no worse than when on high doses, no serious
withdrawal
• Goal is to get down to <100-120 mg MED/day
•
Conclusions:
What is the
evidence?
•
•
•
•
•
•
•
Very limited data on long-term benefits of
high dose opioid therapy
Accumulating evidence on serious harms of
long-term opioid therapy that appear to be
dose-dependent
Titrating to achieve pain relief is inconsistent
with data on opioid benefits
Unclear whether tolerance to analgesic effects
of long-term opioid therapy occurs in clinical
practice
Special caution with methadone
No clear advantage to long-acting, RTC
prescribing
Patients on high doses warrant re-evaluation,
additional monitoring, and follow-up
Decrease dose or discontinue in patients who
are not improving
 Initiate at low doses and titrate slowly
Evidenceinformed
approach to
opioids
 View initial treatment as a therapeutic trial
 Routine monitoring
 Titration should be based on responsiveness of
patients to low doses


Patients who do not respond to low doses
probably will not respond to higher doses—
“opioid non-responders”
Taper in patients not responding or experiencing
adverse effects
 Caution when reaching threshold doses


Optimal dose threshold uncertain
Much easier to titrate up doses than to titrate
down
 Routinely integrate non-opioid therapies
 Opioid alone do not address the psychosocial
contributors to chronic pain