Presentation delivered at the University of Michigan Injury Center Opioid Overdose Summit, Ann Arbor, MI, December 1, 2015

1. EXPANDING THE UNDERSTANDING OF
RISKS ASSOCIATED WITH OPIOIDS AS
WELL AS STRATEGIES TO REDUCE
OPIOID OVERUSE
Mark Ilgen, PhD
VA Center for Clinical Management Research
Department of Psychiatry, University of Michigan

2. Acknowledgements and
Disclosures1
 This work was supported by the Department of Veterans
Affairs.
 Research - VA Quality Enhancement Research Initiative (RRP 13-
251)
 Evaluation - VA Serious Mental Illness Treatment Research and
Evaluation Center (SMITREC)
 These are my opinions and do not necessary represent those of VHA
 Collaborators
 Amy Bohnert, PhD
 Allison Lin, MD
 Dara Ganoczy, MPH

3. Overview
2
 Goal – utilize VHA data to inform an
understanding of risks associated with opioids as
well as the potential impact of strategies to
mitigate those risks.
 Two studies (both based in VHA):
 An examination of the relationship between
opioids and suicide risk
 An evaluation of changes in risky opioid
prescribing before, during, and after the
national roll-out of the opioid safety initiative
(OSI)

4. 3
Pain and suicide-related
outcomes
 Several studies have documented:
 The elevated prevalence of suicidal thoughts and
behaviors in pain clinic patients (Fishbain, Clin J Pain, 1991)
 The cross-sectional association between self-reported
pain and suicidal ideation and non-fatal attempts
(Breslau, Neurology, 1992; Ilgen et al., Gen Hosp Psych, 2008)
 The longitudinal relationship between
 Self-reported pain severity and suicide mortality
(Ilgen et al., SLTB, 2010)
 Pain conditions and suicide mortality (Ilgen et al., JAMA
Psychiatry, 2013)

5. 4
Treating pain with opioids: A
“lifeline” that could help reduce risk
of suicide
LYNN R. WEBSTER, MD: President, American Academy of Pain Medicine

6. Opioid Access and Suicide
Prevention5
 Limiting Access to Means of Suicide
 Most studies looking at access to means—
whether guns, pills, carbon monoxide, bridges, or
other suicide methods—have found that making
these methods less available reduces suicide
rates.
www.afsp.org/preventing-suicide

7. Methods
6
 Case-cohort design
 For each of the two study years, a 5% random
sample of patients was drawn, irrespective of
case status.
 Cases were all FY04-FY05 VHA patients who died by
suicide before the end of FY09.
 Both cases and controls were further restricted to all
individuals with a chronic pain condition who were
treated with an opioid.
 Individuals with indicators of palliative care
consultations or hospice care in their VHA medical
records were excluded (n=1926)
 The sample size was 123,946.

8. Methods
7
 Primary outcomes:
 Suicide mortality was based on deaths classified by
the International Classification of Diseases-10 (ICD-
10) codes X60-X84 and Y87.0 in the NDI.
 intentional overdose was identified by ICD-10 codes
X60-X69.
 Predictors: This study focused on maximum
prescribed morphine-equivalent daily opioid dose.
Morphine-equivalent doses were calculated for
codeine, morphine, oxycodone, hydrocodone,
oxymorphone, and hydromorphone using
established methods. The present analyses did
not examine synthetic opioids (which include
buprenorphine) or methadone.

9. Results: Cox Proportional Hazards
Models of Risk of Death by Suicide8
Suicide, Any
Mechanism
Intentional Overdose
HR (95% CI) HR (95% CI)
Prescribed Daily Opioid Dose
1 to < 20 mg/d 1.00 1.00
20 to < 50 mg/d 1.48 (1.25, 1.75) 1.59 (1.12, 2.27)
50 to < 100 mg/d 1.69 (1.33, 2.14) 1.74 (1.09, 2.76)
100+ mg/d 2.15 (1.64, 2.81) 2.09 (1.22, 3.56)
Adjusted for age, sex, race, Hispanic ethnicity, number of pain conditions,
number of psychiatric conditions, Charlson comorbidity Index, and opioid schedule.

10. Note: All comparisons for opioids significant at p < 0.05 compared to 1 to < 20; comparison
between 1300 to 2000 and less than 1300 for acetaminophen significant at p < 0.05
Examining acetaminophen for comparison……

11. 10
Bohnert, Valenstein, Bair, Ganoczy, McCarthy, Ilgen, Blow, JAMA. 2011.

12. Conclusion: Opioids and suicide
11
 Limitations: Is an observational study
 Patient: Is opioid dose a proxy for increased pain?
 Treatment: Is higher opioid dose a proxy for poor pain
care?
 No signal for potential protective effect of opioids
on risk of suicide
 Increases in opioid dose are associated with
increased risk of suicide
 This association is not limited to intentional overdose
 Chance that opioids impair judgment and could
increase the likelihood of engaging in suicidal
behaviors
 The magnitude of the observed association was
much lower than what has been described for
unintentional overdose.

13. GIVEN THE ASSOCIATION
BETWEEN OPIOIDS AND
ADVERSE OUTCOMES… WHAT
CAN BE DONE ABOUT IT

14. A national approach to address
high-risk opioid prescribing13
 The Opioid Safety Initiative (OSI) was designed to
increase safe opioid prescribing.
 The Under Secretary for Health charged a task
force to develop and deploy an opioid surveillance
system that provides detailed and specific
information about opioid prescribing to all VHA
facilities.
 This resulted in a Business Intelligence (BI) tool to
provide data on opioid prescribing to all facilities.
 Goals, to reduce
 Use of very high dosages of opioids
 Co-use of opioids and sedatives

15. Evaluation goals for this project
14
 To examine recent trends in opioid prescribing
in VHA and whether any notable changes
coincided with the national roll-out of the
Opioid Safety Initiative (OSI)

16. Methods
15
 Obtain facility-level data on opioid and
sedative prescribing during the year prior to,
and the year following, the national roll-out of
the OSI
 Conduct interrupted time series analyses to
describe overall change in opioid prescribing
during this time period and whether
prescribing differed from before to after the
OSI:
 >100 meq of opioids
 > 200 meq of opioid
 Opioid and benzodiazepine co-prescribing

17. 16
intercept P value β 1 –
original
slope
P value Β2 –
immediate
interventio
n effect
P
valu
e
β 3 –
change
in slope
P value
% > 100meq 9.82 <.0001 -.019 .0015 .15 .008
3
-.035 .0001

18. 17
intercept P value β 1 –
original
slope
P value Β2 –
immediate
intervention
effect
P value β 3 –
change
in
slope
P value
% > 200meq 3.53 <.0001 -.016 <.0001 .017 <.0001 -.022 <.0001

19. 18
0
5
10
15
20
25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
%ofpts
Oct 2012-Sept 2014
All facilities combined, % of pts w/concurrent opioid & benzo fills
by month
intercept P value β 1 –
original
slope
P value Β2 –
immediate
intervention
effect
P
value
β 3 –
change in
slope
P value
% w/ overlapping
benzo’s &
opioids
23.31 <.0001 -.185 <.0001 -.155 .5792 .158 .0137

20. 19
0
5
10
15
20
25
30
35
40
<-10 -10 to -5 -5 to 0 0 to 5 >5
%offacilities
% change
Change in prescription of high-dose opioid fills following
OSI rollout among VHA facilities
>100 mg
>200 mg

21. Discussion: OSI and change in
opioids20
 Use of higher dosages of opioids is decreasing in VHA
 The pace of this decrease appears to have accelerated post OSI
(especially for > 200 meq)
 Co-use of opioids and benzodiazepines is quite common. It
decreased over 2 years but no noticeable decrease
associated with the OSI.
 Possible that it was not emphasized, or
 Harder to address
 Caution – observational data on trends over time
 Significant variability in change pre-post OSI was observed
across VHA facilities. Highlighting:
 The difficulties in rolling-out national strategies to change care
 Opportunities to encourage use of strategies to improve the
impact of the OSI (e.g., the Minneapolis approach) throughout
VHA

22. General recommendations
21
 Understanding and quantifying the potential
implications of the expanded use of opioids for
chronic pain requires an examination of multiple
adverse outcomes (unintentional overdose,
drugged driving, crime, heroin initiation, …
suicide)
 Health systems have played a significant role in
the increased use of opioids and will need to
invest in strategies to decrease opioid use.
 Developing, evaluating and refining these
strategies will help to reverse the trend of overuse
of opioids and decrease the occurrence of opioid-
related adverse outcomes

23. Thank You!
Please feel free to contact me:
marki@med.umich.edu
22