Risk of Opioid Overdose Death in North
Carolina by Type of Opioid and Other
Controlled Substances Prescribed
Nabarun Dasgupta, PhD
Epidemico, Inc.
University of North Carolina at
Chapel Hill
CPDD – Phoenix
June 15, 2015
Slides Twitter @epidemico
1
Disclosures
Nabarun is a full-time employee of Epidemico, Inc., a
public health informatics company, which is a subsidiary
of Booz Allen Hamilton.
No opioid manufacturers were involved in this study.
This research was conducted in fulfillment of a doctoral
dissertation requirement and funded in part by CDC.
2
Research Questions
1. Does 100 daily MME represent a threshold for
increase in overdose mortality risk?
2. What proportion of clinicians prescribe (ER) opioids?
3. Are there dose-dependent differences in overdose
mortality between ER vs IR opioid analgesics, or
benzodiazepine exposure?
3
Literature Review
Dasgupta 2013 n=2,182,374 Single year; +/-codeine [NC]
Paulozzi 2012 n=730,381 Includes illicit drug OD [NM]
Gomes 2011a n=607,156
Gov’t drug benefit programme; 10
years [ON]
Gomes 2011b n=154,411
Non-cancer pain; -hydrocodone
[ON]
Bohnert 2011 n=143,684 Military veterans; -fentanyl [US]
Dunn 2010 n=9,940
Non-cancer pain; private insurance;
+propoxyphene, +tramadol [WA]
Coplan 2013 n=74,991
General practice EHR; non-fatal
OD [UK]
Bohnert 2015 n=420,386
Military veterans; intentional and
unintentional OD; -methadone [US]
4
Range of Highest and Reference Groups
Paulozzi 2012
OR 11.3 (8.1, 15.8) for >120 mg/day
Ref: 0-40 mg/day
Gomes 2011a
aOR 2.9 (1.8, 4.6) for 200+ mg/day
Ref: 1-19 mg/day
Gomes 2011b
IRR 2.2 for 201-399 mg/day
IRR 2.3 for 400+ mg/day
Ref: >0-200 mg/day
Bohnert 2011
Cancer HR 12.0 (4.4, 32.5) for100+ mg/day
Non-Ca HR 7.2 (4.8, 10.6) for 100+ mg/day
Ref: 1-20 mg/day
Dunn 2010
HR 8.9 (4.0, 19.7) for 100+ mg/day
Ref: 1-<20 mg/day
Coplan 2013
aRR 1.15 (0.87, 1.53) for 121+ mg/day
RD 0.30 (0.09, 0.52)
Ref: 1-30 mg/day
Bohnert 2015
No benzodiazepine aHR 3.3 (2.6, 4.2) for 100+ mg/day
Former benzodiazepine aHR 3.0 (2.2, 4.1) for 100+
mg/day
Current benzodiazepine aHR 3.9 (3.2, 4.8) for 100+
5
Study Design
Prospective cohort All NC residents, 2010
Outcome
Overdose death involving opioid analgesics
- primary or contributing role
- solid oral and transdermal: codeine, fentanyl,
hydrocodone, hydromorphone, morphine,
methadone, oxycodone, oxymorphone
Exposure
Average Daily MME
Approximately 20 mg increments
Models Poisson, GEE (independent)
Offset
Person-days
- Following intent-to-treat principle
Assumptions
- Uniform risk over interval
- Prescription status in one period does not
influence prescription status in subsequent
period
- Risk not influenced by: prior exposure, other
substances, number of prescriptions or
pharmacies, opioid formulation
6
Morphine Equivalents & Overlapping Scripts
Used by epidemiologists to standardize potency across
opioids
(Do or do not use as clinical guide during opioid rotation?)
Pharmacological potency relative to morphine (1.0)
Conversion factors suggested by CDC
MME = milligrams per unit x quantity of units x conversion
factor
Days supply = Days the prescription is intended to be used
Average daily MME =
7
MME
days supply
8
Study Flow
Numbers of participants
All North Carolina
Residents
(n=9,560,234)
EXPOSED
to opioid analgesics
(n=2,182,374)
UNEXPOSED
(n=7,377,860)
Immediate-release
opioid analgesics only
(n=2,042,645)
Extended-release
and immediate-release
opioid analgesics
(n=139,726)
Excluded (n=3)
• Formulation type unknown
Overdose deaths
(n=270)
Censored
(n=2,042,375)
Overdose deaths
(n=208)
Censored
(n=139,518)
Overdose deaths
(n=151)
Censored
(n=7,377,709)
OUTCOME
STRATIFIER
EXPOSURE
INCLUSION
Data Processing
9
Numbers of dispensed prescription
records
All prescriptions dispensed 2009 to 2011 and r ecorded in
North CarolinaControlled Substances Repor tingSystem
(n=54,825,930)
Excluded (n=1,094,717)
• Non-North Carolinacounty of residence (n=964,678)
• Unknown or missingdruginformation (n=67,064)
• Non-controlled substances (n=62,975)
Missingdataassessment
(n=53,731,213)
• Imputed missingdays supply (n=5,367,120)
Classification by active ingredient and formulation
to identify solid oral and transdermal opioid analgesics
(n=53,712,910)
Classification by year dispensed
to identify prescriptions only intended for use in 2010
(n=21,448,986)
Opioid analgesic prescriptions
intended for use in 2010
(n=7,393,375)
Excluded (n=18,303)
• Missingdays supply could not be imputed (n=3,364)
• Quantity dispensed could not be determined (n=14,939)
Excluded (n=32,263,924)
• Not opioid analgesics (n=31,881,052)
• Liquids (n=382,872)
Excluded (n=14,055,611)
• Dispensed in 2011 (n=7,522,050)
• Intended for use only in 2009 (n=6,533,561)
Opioid Analgesic Patients (n=2,182,374)
22.8% of NC residents received opioid analgesics
IR: 22.5% versus ER: 1.4%
89.6% of all licensed clinicians prescribed opioid analgesics
IR: 88.5% versus ER: 40.0%
80% of opioid analgesic patients prescribed benzo in previous
year
Overdose Decedents (n=629)
76% received opioid analgesic Rx in 365 days prior to death
43% of these opioid analgesic decedents had received ER
51% has active opioid analgesic prescription at time of death
61% of opioid analgesic overdose deaths also implicated benzos
10
11
Opioid Utilization in Context & Access to Care
12
Deaths Person-years n
Rate per 10,000
Person-Years
95% Confidence
Interval
Unexposed 151 4,700,647 7,377,860 0.3 0.27, 0.38
>0 to 39.9 mg/day 98 259,735 1,305,969 3.8 3.1, 4.6
40 to 59.9 mg/day 90 457,223 457,322 2.0 1.6, 2.4
60 to 79.9 mg/day 47 213,813 213,868 2.2 1.6, 2.9
80 to 99.9 mg/day 34 72,447 72,483 4.7 3.2, 6.5
100 to 119.9 mg/day 23 45,536 45,559 5.0 3.2, 7.6
120 to 139.9 mg/day 22 20,699 20,721 10.6 6.7, 16.1
140 to 159.9 mg/day 14 14,585 14,599 9.6 5.2, 16.1
160 to 179.9 mg/day 15 6,769 6,784 22.1 12.4, 36.5
180 to 199.9 mg/day 11 9,604 9,615 11.4 5.7, 20.5
200 to 249.9 mg/day 24 11,654 11,678 20.6 13.2, 30.6
250 to 299.9 mg/day 20 7,405 7,425 27.0 16.5, 41.7
300 to 349.9 mg/day 17 4,495 4,512 37.8 22.0, 60.5
350 to 399.9 mg/day 17 3,563 3,580 47.7 27.8, 76.4
400 to 499.9 mg/day 14 3,527 3,541 39.7 21.7, 66.6
500 to 5,000 mg/day 32 2,892 4,718 110.6 75.7, 156.2
Total 629 5,834,594 9,560,234 1.1 1.0, 1.2
13
14
Opioid analgesic + benzo: 7.0 per 10,000 person-years (95% CI: 6.3, 7.8
Only opioid analgesics: 0.7 per 10,000 person years (95% CI: 0.6, 0.9)
15
IR only: 1.3 per 10,000 person-years (95% CI: 1.2, 1.5)
ER + IR: 14.9 per 10,000 person years (95% CI: 12.9, 17.1)
Of “higher dose” (150 daily MME) IR-only patients, 14.1% were on opioid
analgesic therapy for more than 6 months.
Limitations
Exchangeability - why do patients receive higher doses?
Assumption of continuous risk during time exposed
Previous therapy and opioid tolerance not assessed
External factors that influence opioid mortality (e.g.,
ADFs)
Identification of overdose deaths and causality
assessment
Adherence to therapy and exposure to diverted opioids
16
Source: Dasgupta 2014, Drug & Alcohol Dependen
17
Deaths Person-years n
Rate per 10,000
Person-Years
95% Confidence
Interval
Unexposed 151 4,700,647 7,377,860 0.3 0.27, 0.38
>0 to 39.9 mg/day 98 259,735 1,305,969 3.8 3.1, 4.6
40 to 59.9 mg/day 90 457,223 457,322 2.0 1.6, 2.4
60 to 79.9 mg/day 47 213,813 213,868 2.2 1.6, 2.9
80 to 99.9 mg/day 34 72,447 72,483 4.7 3.2, 6.5
100 to 119.9 mg/day 23 45,536 45,559 5.0 3.2, 7.6
120 to 139.9 mg/day 22 20,699 20,721 10.6 6.7, 16.1
140 to 159.9 mg/day 14 14,585 14,599 9.6 5.2, 16.1
160 to 179.9 mg/day 15 6,769 6,784 22.1 12.4, 36.5
180 to 199.9 mg/day 11 9,604 9,615 11.4 5.7, 20.5
200 to 249.9 mg/day 24 11,654 11,678 20.6 13.2, 30.6
250 to 299.9 mg/day 20 7,405 7,425 27.0 16.5, 41.7
300 to 349.9 mg/day 17 4,495 4,512 37.8 22.0, 60.5
350 to 399.9 mg/day 17 3,563 3,580 47.7 27.8, 76.4
400 to 499.9 mg/day 14 3,527 3,541 39.7 21.7, 66.6
500 to 5,000 mg/day 32 2,892 4,718 110.6 75.7, 156.2
Total 629 5,834,594 9,560,234 1.1 1.0, 1.2
18
19
Linkage of Prescription and Mortality Data
Conducted by A. Hirsch at NC Division of Public Health
Identify deaths by vital statistics [X40-44, Y10-14] (n=896)
Toxicology results available (n=824)
Deaths involving opioid analgesics in primary/additive roles
(n=629)
Identify CSRS records using first 5 letters of last name and
DOB
Extract prescription dispensing history for 365 days prior to
death
20

Risk of Opioid Overdose Death in North Carolina by Type of Opioid and Other Controlled Substances Prescribed

  • 1.
    Risk of OpioidOverdose Death in North Carolina by Type of Opioid and Other Controlled Substances Prescribed Nabarun Dasgupta, PhD Epidemico, Inc. University of North Carolina at Chapel Hill CPDD – Phoenix June 15, 2015 Slides Twitter @epidemico 1
  • 2.
    Disclosures Nabarun is afull-time employee of Epidemico, Inc., a public health informatics company, which is a subsidiary of Booz Allen Hamilton. No opioid manufacturers were involved in this study. This research was conducted in fulfillment of a doctoral dissertation requirement and funded in part by CDC. 2
  • 3.
    Research Questions 1. Does100 daily MME represent a threshold for increase in overdose mortality risk? 2. What proportion of clinicians prescribe (ER) opioids? 3. Are there dose-dependent differences in overdose mortality between ER vs IR opioid analgesics, or benzodiazepine exposure? 3
  • 4.
    Literature Review Dasgupta 2013n=2,182,374 Single year; +/-codeine [NC] Paulozzi 2012 n=730,381 Includes illicit drug OD [NM] Gomes 2011a n=607,156 Gov’t drug benefit programme; 10 years [ON] Gomes 2011b n=154,411 Non-cancer pain; -hydrocodone [ON] Bohnert 2011 n=143,684 Military veterans; -fentanyl [US] Dunn 2010 n=9,940 Non-cancer pain; private insurance; +propoxyphene, +tramadol [WA] Coplan 2013 n=74,991 General practice EHR; non-fatal OD [UK] Bohnert 2015 n=420,386 Military veterans; intentional and unintentional OD; -methadone [US] 4
  • 5.
    Range of Highestand Reference Groups Paulozzi 2012 OR 11.3 (8.1, 15.8) for >120 mg/day Ref: 0-40 mg/day Gomes 2011a aOR 2.9 (1.8, 4.6) for 200+ mg/day Ref: 1-19 mg/day Gomes 2011b IRR 2.2 for 201-399 mg/day IRR 2.3 for 400+ mg/day Ref: >0-200 mg/day Bohnert 2011 Cancer HR 12.0 (4.4, 32.5) for100+ mg/day Non-Ca HR 7.2 (4.8, 10.6) for 100+ mg/day Ref: 1-20 mg/day Dunn 2010 HR 8.9 (4.0, 19.7) for 100+ mg/day Ref: 1-<20 mg/day Coplan 2013 aRR 1.15 (0.87, 1.53) for 121+ mg/day RD 0.30 (0.09, 0.52) Ref: 1-30 mg/day Bohnert 2015 No benzodiazepine aHR 3.3 (2.6, 4.2) for 100+ mg/day Former benzodiazepine aHR 3.0 (2.2, 4.1) for 100+ mg/day Current benzodiazepine aHR 3.9 (3.2, 4.8) for 100+ 5
  • 6.
    Study Design Prospective cohortAll NC residents, 2010 Outcome Overdose death involving opioid analgesics - primary or contributing role - solid oral and transdermal: codeine, fentanyl, hydrocodone, hydromorphone, morphine, methadone, oxycodone, oxymorphone Exposure Average Daily MME Approximately 20 mg increments Models Poisson, GEE (independent) Offset Person-days - Following intent-to-treat principle Assumptions - Uniform risk over interval - Prescription status in one period does not influence prescription status in subsequent period - Risk not influenced by: prior exposure, other substances, number of prescriptions or pharmacies, opioid formulation 6
  • 7.
    Morphine Equivalents &Overlapping Scripts Used by epidemiologists to standardize potency across opioids (Do or do not use as clinical guide during opioid rotation?) Pharmacological potency relative to morphine (1.0) Conversion factors suggested by CDC MME = milligrams per unit x quantity of units x conversion factor Days supply = Days the prescription is intended to be used Average daily MME = 7 MME days supply
  • 8.
    8 Study Flow Numbers ofparticipants All North Carolina Residents (n=9,560,234) EXPOSED to opioid analgesics (n=2,182,374) UNEXPOSED (n=7,377,860) Immediate-release opioid analgesics only (n=2,042,645) Extended-release and immediate-release opioid analgesics (n=139,726) Excluded (n=3) • Formulation type unknown Overdose deaths (n=270) Censored (n=2,042,375) Overdose deaths (n=208) Censored (n=139,518) Overdose deaths (n=151) Censored (n=7,377,709) OUTCOME STRATIFIER EXPOSURE INCLUSION
  • 9.
    Data Processing 9 Numbers ofdispensed prescription records All prescriptions dispensed 2009 to 2011 and r ecorded in North CarolinaControlled Substances Repor tingSystem (n=54,825,930) Excluded (n=1,094,717) • Non-North Carolinacounty of residence (n=964,678) • Unknown or missingdruginformation (n=67,064) • Non-controlled substances (n=62,975) Missingdataassessment (n=53,731,213) • Imputed missingdays supply (n=5,367,120) Classification by active ingredient and formulation to identify solid oral and transdermal opioid analgesics (n=53,712,910) Classification by year dispensed to identify prescriptions only intended for use in 2010 (n=21,448,986) Opioid analgesic prescriptions intended for use in 2010 (n=7,393,375) Excluded (n=18,303) • Missingdays supply could not be imputed (n=3,364) • Quantity dispensed could not be determined (n=14,939) Excluded (n=32,263,924) • Not opioid analgesics (n=31,881,052) • Liquids (n=382,872) Excluded (n=14,055,611) • Dispensed in 2011 (n=7,522,050) • Intended for use only in 2009 (n=6,533,561)
  • 10.
    Opioid Analgesic Patients(n=2,182,374) 22.8% of NC residents received opioid analgesics IR: 22.5% versus ER: 1.4% 89.6% of all licensed clinicians prescribed opioid analgesics IR: 88.5% versus ER: 40.0% 80% of opioid analgesic patients prescribed benzo in previous year Overdose Decedents (n=629) 76% received opioid analgesic Rx in 365 days prior to death 43% of these opioid analgesic decedents had received ER 51% has active opioid analgesic prescription at time of death 61% of opioid analgesic overdose deaths also implicated benzos 10
  • 11.
    11 Opioid Utilization inContext & Access to Care
  • 12.
  • 13.
    Deaths Person-years n Rateper 10,000 Person-Years 95% Confidence Interval Unexposed 151 4,700,647 7,377,860 0.3 0.27, 0.38 >0 to 39.9 mg/day 98 259,735 1,305,969 3.8 3.1, 4.6 40 to 59.9 mg/day 90 457,223 457,322 2.0 1.6, 2.4 60 to 79.9 mg/day 47 213,813 213,868 2.2 1.6, 2.9 80 to 99.9 mg/day 34 72,447 72,483 4.7 3.2, 6.5 100 to 119.9 mg/day 23 45,536 45,559 5.0 3.2, 7.6 120 to 139.9 mg/day 22 20,699 20,721 10.6 6.7, 16.1 140 to 159.9 mg/day 14 14,585 14,599 9.6 5.2, 16.1 160 to 179.9 mg/day 15 6,769 6,784 22.1 12.4, 36.5 180 to 199.9 mg/day 11 9,604 9,615 11.4 5.7, 20.5 200 to 249.9 mg/day 24 11,654 11,678 20.6 13.2, 30.6 250 to 299.9 mg/day 20 7,405 7,425 27.0 16.5, 41.7 300 to 349.9 mg/day 17 4,495 4,512 37.8 22.0, 60.5 350 to 399.9 mg/day 17 3,563 3,580 47.7 27.8, 76.4 400 to 499.9 mg/day 14 3,527 3,541 39.7 21.7, 66.6 500 to 5,000 mg/day 32 2,892 4,718 110.6 75.7, 156.2 Total 629 5,834,594 9,560,234 1.1 1.0, 1.2 13
  • 14.
    14 Opioid analgesic +benzo: 7.0 per 10,000 person-years (95% CI: 6.3, 7.8 Only opioid analgesics: 0.7 per 10,000 person years (95% CI: 0.6, 0.9)
  • 15.
    15 IR only: 1.3per 10,000 person-years (95% CI: 1.2, 1.5) ER + IR: 14.9 per 10,000 person years (95% CI: 12.9, 17.1) Of “higher dose” (150 daily MME) IR-only patients, 14.1% were on opioid analgesic therapy for more than 6 months.
  • 16.
    Limitations Exchangeability - whydo patients receive higher doses? Assumption of continuous risk during time exposed Previous therapy and opioid tolerance not assessed External factors that influence opioid mortality (e.g., ADFs) Identification of overdose deaths and causality assessment Adherence to therapy and exposure to diverted opioids 16
  • 17.
    Source: Dasgupta 2014,Drug & Alcohol Dependen 17
  • 18.
    Deaths Person-years n Rateper 10,000 Person-Years 95% Confidence Interval Unexposed 151 4,700,647 7,377,860 0.3 0.27, 0.38 >0 to 39.9 mg/day 98 259,735 1,305,969 3.8 3.1, 4.6 40 to 59.9 mg/day 90 457,223 457,322 2.0 1.6, 2.4 60 to 79.9 mg/day 47 213,813 213,868 2.2 1.6, 2.9 80 to 99.9 mg/day 34 72,447 72,483 4.7 3.2, 6.5 100 to 119.9 mg/day 23 45,536 45,559 5.0 3.2, 7.6 120 to 139.9 mg/day 22 20,699 20,721 10.6 6.7, 16.1 140 to 159.9 mg/day 14 14,585 14,599 9.6 5.2, 16.1 160 to 179.9 mg/day 15 6,769 6,784 22.1 12.4, 36.5 180 to 199.9 mg/day 11 9,604 9,615 11.4 5.7, 20.5 200 to 249.9 mg/day 24 11,654 11,678 20.6 13.2, 30.6 250 to 299.9 mg/day 20 7,405 7,425 27.0 16.5, 41.7 300 to 349.9 mg/day 17 4,495 4,512 37.8 22.0, 60.5 350 to 399.9 mg/day 17 3,563 3,580 47.7 27.8, 76.4 400 to 499.9 mg/day 14 3,527 3,541 39.7 21.7, 66.6 500 to 5,000 mg/day 32 2,892 4,718 110.6 75.7, 156.2 Total 629 5,834,594 9,560,234 1.1 1.0, 1.2 18
  • 19.
  • 20.
    Linkage of Prescriptionand Mortality Data Conducted by A. Hirsch at NC Division of Public Health Identify deaths by vital statistics [X40-44, Y10-14] (n=896) Toxicology results available (n=824) Deaths involving opioid analgesics in primary/additive roles (n=629) Identify CSRS records using first 5 letters of last name and DOB Extract prescription dispensing history for 365 days prior to death 20

Editor's Notes

  • #8 primary (the drug was at a concentration sufficient to have caused the death alone regardless of other drugs detected) additive (the drug was at a concentration not sufficient to have caused the death alone but acted in an additive manner with other drugs to have caused the death
  • #10 excluded for missing data: 148,342 = 0.3% of all prescriptions 1.8% of opioid analgesic prescriptions for liquids (not including cough syrups)
  • #11 43.1% (n=208) had received at least one extended-release formulation 83.3% (n=26,953) of licensed clinicians prescribed benzodiazepines, 57.2% (n=18,518) sleep aids, and 44.8% (n=14,487) stimulants. Ethanol was involved in 12.2% (n=77) of overdoses involving opioid analgesics. Heroin was present in only 1.3% (n=8) of opioid analgesic overdoses, whereas cocaine was present in 8.4% (n=53).
  • #20 PERCENT HIGH DOSE
  • #21 primary (the drug was at a concentration sufficient to have caused the death alone regardless of other drugs detected) additive (the drug was at a concentration not sufficient to have caused the death alone but acted in an additive manner with other drugs to have caused the death