Abstract
Background and Aims:
Some adolescents and young adults (termed ‘youth’) prescribed an opioid will develop opioid use disorder or experience overdose. This study aimed to identify patient and prescription characteristics associated with subsequent risk of opioid use disorder or overdose during the year after an opioid is first dispensed.
Design:
Retrospective cohort study.
Setting:
Commercial health insurance claims in a large US database from 2006 to 2016.
Participants:
Youth aged 11–25 years filling an initial opioid prescription.
Exposures:
Patient (sociodemographic information, and physical and mental health diagnoses) and prescription characteristics (opioid formulation, dose, and duration).
Measurement:
The primary outcome was development of an ‘opioid-related complication’ (a diagnosis of opioid use disorder or opioid-related overdose) during the subsequent 12 months. Exposures of interest were patient (sociodemographic information, and physical and mental health diagnoses) and prescription characteristics (opioid formulation, dose, and duration).
Findings:
Among 3,278,990 youth filling an initial opioid prescription, median age was 18 years (IQR, 16–21) and 56.1% were female. During the subsequent 12 months, 10,405 (0.3%) youth experienced an opioid-related complication. Conditions associated with increased risk included mood/anxiety disorders (aRR, 4.45 [95% CI, 4.25–4.66]) and substance use (aRR, 20.77 [95% CI, 19.74–21.84]). Comorbid substance use disorders were present among 72.8% of youth experiencing an opioid-related complication and included alcohol (33.4%), cannabis (33.0%), nicotine (43.2%), and other substance use disorders (75.5%). Long-acting opioids (aRR, 2.59 [95% CI, 2.18–3.09]) and longer durations were associated with increased risk (7–14 days: aRR, 1.15 [95% CI, 1.08–1.22]; ≥15 days: aRR, 1.96 [95% CI, 1.80–2.12]) compared to short-acting formulations and durations ≤3 days, respectively.
Conclusions:
Complications after an initial opioid prescription are relatively rare. Screening for mental health conditions and substance use before prescribing might identify youth at risk for OUD and overdose. Clinicians might mitigate the risk by prescribing short-acting opioids for short durations.
MeSH Keywords: substance-related disorders, opioid-related disorders, prescriptions, adolescent, young adult, mental health
INTRODUCTION
North America and numerous other regions worldwide continue to experience an epidemic of opioid overdose deaths.1–3 A critical strategy for reducing mortality internationally is to prevent the development of opioid use disorder (OUD) and opioid overdose.4 Adolescence and young adulthood are key periods during which to intervene in the development of OUD. Most individuals who develop OUD first start using opioids when they are an adolescent or young adult (termed “youth”).5 In the United States (US), which has among the highest overdose mortality rates globally, opioid overdose deaths have risen more than 250% among youth since 1999.6 Although a majority of overdose deaths among older adults in the US currently involve heroin and illicitly manufactured fentanyl,1 prescription opioids are involved in one-third of all overdose deaths among youth,7 and are commonly the first opioids that youth use before transitioning to heroin or fentanyl.8,9
Increasingly, there is attention to the risk for developing OUD or experiencing overdose after youth are prescribed an opioid. Although friends or relatives are the most common sources of prescription opioids that youth misuse, one-fifth of adolescents and one-quarter of young adults who misuse prescription opioids obtain them directly from a physician.10 Although recent data suggest that youth who receive an opioid following dental and surgical procedures may be at risk of developing OUD or experiencing overdose,11–14 the contribution of patient and prescription risk factors to subsequent risk for OUD and overdose remains poorly described for youth who receive an opioid prescription. Identifying such factors could inform screening practices and prescribing guidelines that clinicians can follow when they prescribe opioids to youth. Currently, opioid-prescribing guidelines are focused on older adult patients, largely due to a dearth of data on the rates of OUD and opioid overdose among youth and predisposing factors.15
Using a large, US claims database, we sought to identify risk factors for progression to OUD or opioid overdose during the year after youth fill their first observed opioid prescription. Whereas previous studies have identified a potentially higher incidence of opioid misuse following an initial prescription,11–14 we aimed instead to determine which risk factors predispose to this transition, with an emphasis on those that are clinically relevant and intervenable. We hypothesized that patient characteristics (such as comorbid pain conditions, mental health conditions, and other substance use disorders) and prescription characteristics (such as long-acting formulation, high opioid dose, and long duration) would be associated with subsequent opioid-related complications.
METHODS
Sample
We conducted a retrospective cohort study using the IBM MarketScan Commercial Database, which included all inpatient, outpatient, emergency department, behavioral health, and prescription drug claims from over 150 million unique US individuals with employer-provided health insurance between January 1, 2006 and December 31, 2016.
Eligible individuals were aged 11–25 years and filled their first observed opioid prescription between July 1, 2006 and January 1, 2016, allowing for ≥6 months before and ≥12 months after the prescription. Opioids were identified using National Drug Codes for codeine, fentanyl, hydrocodone, hydromorphone, levorphanol, meperidine, methadone, morphine, oxycodone, propoxyphene, tapendatol, and tramadol from the IBM Micromedex RED BOOK. Further inclusion criteria were: no known preexisting cancer diagnosis; no observed diagnosis of OUD or opioid-related overdose (defined below) during the 6 months before the opioid prescription was filled; no dispensing of medications suggestive of OUD treatment (i.e., pharmacy claims for buprenorphine or naltrexone, or procedure codes for administration of buprenorphine, naltrexone or methadone) during the 6 months before the opioid prescription was filled (Table S2); and ≥12 months of continuous enrollment after the opioid prescription was filled (i.e., last possible day of study follow-up was December 31, 2016). The final analytic sample included 3,278,990 individuals (Figure S1). The study was not considered human subjects research by the Boston University School of Medicine Institutional Review Board.
Variables
The primary outcome was development of ≥1 ‘opioid-related complication’ during the 12 months after the initial opioid prescription was filled, categorized as a binary variable. We defined opioid-related complications as receipt of a diagnosis of OUD or opioid overdose using International Classification of Diseases, Ninth Revision (ICD-9) and Tenth Revision (ICD-10) diagnosis codes (Table S2). Diagnoses could occur in any healthcare setting in which claims were documented. The 12-month window was selected to be consistent with other studies showing a relatively short time course between an opioid prescription and subsequent complications;11–14 we did not examine diagnoses after 12 months given that an observation association between the initial opioid prescription and later complications was less plausible beyond one year. Diagnoses documented in settings in which a health insurance claim was not filed with a youth’s commercial health insurer were not captured in the database. Similarly, information on overdose mortality was not available in the database.
Study covariates included patient and prescription factors selected a priori based on their previously identified or hypothesized associations with opioid-related complications.11,14,16–18 Patient factors included age at the time the opioid prescription was filled; sex; urban/rural residence based on living inside/outside a metropolitan statistical area; pregnancy;19 acute pain condition;20 chronic pain condition;20 intellectual disability, learning disorder, or developmental disorder; mood or anxiety disorder; psychosis; attention deficit hyperactivity disorder; trauma or stress-related condition; conduct disorder or antisocial personality disorder; and other substance use disorders based on ICD-9 and ICD-10 diagnosis codes. (Comorbid mental health conditions and substance use disorder diagnosis codes are listed in Tables S3 and S4; other diagnosis codes available from cited references.) Acute pain conditions were identified between 14 days before and 14 days after the opioid prescription was filled. The remaining clinical covariates (i.e., pregnancy, chronic pain condition, and all psychiatric diagnoses) are frequently missed diagnoses in youth and thus might be present but not yet documented at the time the opioid prescription was filled;21 to increase the sensitivity of detecting these commonly underdiagnosed conditions, they were identified during an 18-month observation period that included the 6 months before and 12 months after the opioid prescription was filled. Prescription factors included short- vs. long-acting formulation, daily dose in morphine milligram equivalents (MME)/day, and duration of the opioid prescription, all defined for the first opioid dispensed.
Statistical Analysis
We used descriptive statistics to characterize the sample with regard to patient and prescription factors. We used the chi-square test (or Fisher’s exact test when cell size was <5) to compare factors among individuals who received a diagnosis of OUD, opioid overdose, both, or neither. We used a modified Poisson regression approach using robust sandwich variance estimation to identify patient and prescription factors associated with development of ≥1 opioid-related complication in the 12 months after the initial opioid prescription was filled.22 The Poisson approach allowed us to express our findings in terms of relative risk, which is more intuitive for stakeholders than the odds ratio. After identifying bivariate associations, we conducted a final multivariable model including all study covariates. To better elucidate the timing of comorbid substance use disorders diagnoses relative to opioid-related complications, we conducted a subanalysis examining whether comorbid alcohol, cannabis, nicotine, or other substance use disorder diagnoses were documented during the 6 months before or the 12 months after the opioid prescription was filled. We then repeated this subanalysis for youth who did not experience any opioid-related complications.
Analyses were conducted using SAS Version 9.4 (SAS Institute, Inc.). All statistical tests were two-sided and considered significant at p<0.05. The analysis was not pre-registered and results should be considered exploratory.
RESULTS
Among 3,278,990 youth aged 11–25 years who filled an initial opioid prescription and met study eligibility criteria, median age was 18 years (interquartile range [IQR], 16–21 years) and 56.1% were female (Table 1). Overall, 19.7% of youth had an acute pain condition documented during the two weeks before or after their opioid prescription was filled, and 36.4% had a chronic pain condition documented during the 18 months they were observed in the study (i.e., six months before or 12 months after their opioid was prescribed; the range of chronic pain conditions documented are reported in Table S5). Most psychiatric diagnoses were relatively uncommon; mood and anxiety disorders were the most prevalent, with 13.6% of youth receiving a diagnosis during the 18-month observation period. Among the initial opioid prescriptions that youth filled, the median daily dose and duration were 33 MME (IQR, 23–49 MME) and 3 days (IQR, 3–5 days), respectively. Nearly all prescriptions (99.9%) were for short-acting formulations.
Table 1.
Development of opioid use disorder (OUD), opioid overdose, or both during the year following an initial opioid prescription among 3,278,990 youth.
|
Characteristic |
Total(N = 3,278,990) N (column %) |
OUD Diagnosis(N = 8,667) N (row %) |
OD Diagnosis(N = 1,270) N (row %) |
OUD and OD(N = 468) N (row %) |
P value |
|---|---|---|---|---|---|
| Patient factors | |||||
| Age at time of prescription, years | |||||
| 11–13 | 377,170 (11.5%) | 36 (<0.1%) | 40 (<0.1%) | 1 (<0.1%) | <0.001 |
| 14–16 | 707,657 (21.6%) | 673 (0.1%) | 274 (<0.1%) | 30 (<0.1%) | |
| 17–19 | 946,011 (28.9%) | 2,558 (0.3%) | 441 (<0.1%) | 154 (<0.1%) | |
| 20–22 | 639,141 (19.5%) | 3,031 (0.5%) | 311 (<0.1%) | 175 (<0.1%) | |
| 23–25 | 609,011 (18.6%) | 2,369 (0.4%) | 204 (<0.1%) | 108 (<0.1%) | |
| Sex | |||||
| Female | 1,839,377 (56.1%) | 3,006 (0.2%) | 734 (<0.1%) | 152 (<0.1%) | <0.001 |
| Male | 1,439,613 (43.9%) | 5,661 (0.4%) | 536 (<0.1%) | 316 (<0.1%) | |
| Location | |||||
| Rural | 501,400 (15.3%) | 1,133 (0.2%) | 183 (<0.1%) | 45 (<0.1%) | <0.001 |
| Urban | 2,777,590 (84.7%) | 7,534 (0.3%) | 1,087 (<0.1%) | 423 (<0.1%) | |
| Pregnancya | |||||
| No | 3,046,948 (92.9%) | 8,099 (0.3%) | 1,164 (<0.1%) | 442 (<0.1%) | 0.039 |
| Yes | 232,042 (7.1%) | 568 (0.2%) | 106 (<0.1%) | 26 (<0.1%) | |
| Acute pain conditionb | |||||
| No | 2,634,323 (80.3%) | 6,868 (0.3%) | 1,027 (<0.1%) | 365 (<0.1%) | 0.037 |
| Yes | 644,667 (19.7%) | 1,799 (0.3%) | 243 (<0.1%) | 103 (<0.1%) | |
| Chronic pain conditiona | |||||
| No | 2,084,847 (63.6%) | 2,025 (0.1%) | 261 (<0.1%) | 80 (<0.1%) | <0.001 |
| Yes | 1,194,143 (36.4%) | 6,642 (0.6%) | 1,009 (<0.1%) | 388 (<0.1%) | |
| Intellectual disability, learning disorder, or developmental disordera | |||||
| No | 3,235,236 (98.7%) | 8,511 (0.3%) | 1,234 (<0.1%) | 458 (<0.1%) | <0.001 |
| Yes | 43,754 (1.3%) | 156 (0.4%) | 36 (<0.1%) | 10 (<0.1%) | |
| Mood or anxiety disordera | |||||
| No | 2,833,410 (86.4%) | 3,242 (0.1%) | 391 (<0.1%) | 142 (<0.1%) | <0.001 |
| Yes | 445,580 (13.6%) | 5,425 (1.2%) | 879 (0.2%) | 326 (<0.1%) | |
| Psychosisa | |||||
| No | 3,261,752 (99.5%) | 8,107 (0.2%) | 1,138 (<0.1%) | 411 (<0.1%) | <0.001 |
| Yes | 17,238 (0.5%) | 560 (3.2%) | 132 (0.8%) | 57 (0.3%) | |
| Attention deficit hyperactivity disordera | |||||
| No | 3,054,324 (93.2%) | 7,185 (0.2%) | 1,130 (<0.1%) | 374 (<0.1%) | <0.001 |
| Yes | 224,666 (6.9%) | 1,482 (0.7%) | 140 (<0.1%) | 94 (<0.1%) | |
| Trauma or stress-related conditiona | |||||
| No | 3,127,063 (95.3%) | 7,258 (0.2%) | 999 (<0.1%) | 379 (<0.1%) | <0.001 |
| Yes | 151,927 (4.6%) | 1,409 (0.9%) | 271 (0.2%) | 89 (<0.1%) | |
| Conduct disorder or antisocial personality disordera | |||||
| No | 3,250,082 (99.1%) | 8,250 (0.3%) | 1,187 (<0.1%) | 434 (<0.1%) | <0.001 |
| Yes | 28,908 (0.9%) | 417 (1.4%) | 83 (0.3%) | 34 (0.1%) | |
| Other substance use disordera | |||||
| No | 3,092,972 (94.3%) | 2,020 (<0.1%) | 706 (<0.1%) | 61 (<0.1%) | <0.001 |
| Yes | 186,018 (5.7%) | 6,647 (3.6%) | 564 (0.3%) | 407 (0.2%) | |
| Prescription factors | |||||
| Long-acting opioid | |||||
| No | 3,274,997 (99.9%) | 8,551 (0.3%) | 1,268 (<0.1%) | 467 (<0.1%) | |
| Yes | 3,993 (0.1%) | 116 (2.9%) | 2 (<0.1%) | 1 (<0.1%) | |
| Daily dose, morphine milligram equivalents | |||||
| <30 | 1,174,465 (35.8%) | 3,210 (0.3%) | 471 (<0.1%) | 199 (<0.1%) | <0.001 |
| 30–59 | 1,627,234 (49.6%) | 4,089 (0.3%) | 595 (<0.1%) | 207 (<0.1%) | |
| 60–89 | 314,602 (9.6%) | 768 (0.2%) | 124 (<0.1%) | 34 (<0.1%) | |
| 90–119 | 109,629 (3.3%) | 285 (0.3%) | 50 (<0.1%) | 15 (<0.1%) | |
| ≥120 | 53,060 (1.6%) | 315 (0.6%) | 30 (<0.1%) | 13 (<0.1%) | |
| Duration of prescription, days | |||||
| 1–3 | 1,656,389 (50.5%) | 4,154 (0.6%) | 608 (<0.1%) | 227 (<0.1%) | <0.001 |
| 4–6 | 1,176,646 (35.9%) | 2,693 (0.2%) | 435 (<0.1%) | 163 (<0.1%) | |
| 7–14 | 360,728 (11.0%) | 1,128 (0.3%) | 169 (0.4%) | 56 (0.2%) | |
| ≥15 | 85,227 (2.6%) | 692 (0.8%) | 58 (<0.1%) | 22 (<0.1%) |
Diagnosed within the 6 months preceding or 12 months following the initial opioid prescription
Diagnosed within the 14 days preceding or following the initial opioid prescription
During the year after they filled their opioid prescription, 10,405 (0.3%) youth experienced an opioid-related complication (incidence rate, 3.18 per 1,000 person-years; 95% confidence interval [CI], 3.14–3.21). Overall, 8,667 (0.26%) youth received a diagnosis of OUD, 1,270 (0.04%) had a documented opioid-related overdose, 468 (0.014%) had both opioid-related complications documented, and 3,268,583 (99.7%) had no opioid-related complications documented. Among youth who received an OUD diagnosis, the median age at the time the initial opioid prescription was filled was 21 years (IQR, 18–23 years); among youth who had a documented opioid-related overdose, it was 19 years (IQR, 17–22 years). Youth experiencing opioid-related complications were more commonly male, from an urban location, and diagnosed with a chronic, rather than an acute, pain condition. Psychiatric comorbidities were relatively more common among youth with opioid-related complications than those without. Most diagnoses of OUD (71.8%) and overdose (61.3%) occurred >90 days after the initial opioid prescription (Table 2). The percentage of youth experiencing an opioid-related complication increased over the study period (Table S6).
Table 2.
Timing of first observed opioid use disorder (OUD) and opioid overdose diagnosesa relative to an initial opioid prescription among 10,405 youth experiencing opioid-related complications.
|
Time since initial prescription, days |
OUD diagnosis (N=8,935) N (%) |
Overdose (N=1,738) N (%) |
|---|---|---|
| 1–7 | 295 (3.2%) | 133 (7.7%) |
| 8–14 | 239 (2.6%) | 66 (3.8%) |
| 15–30 | 498 (5.4%) | 131 (7.5%) |
| 31–60 | 857 (9.4%) | 192 (11.1%) |
| 61–90 | 830 (9.1%) | 150 (8.6%) |
| 91–180 | 2,301 (25.2%) | 420 (24.2%) |
| 181–365 | 4,115 (45.1%) | 646 (37.2%) |
468 individuals received both an OUD and an overdose diagnosis and are included in both columns.
In multivariable models, patient factors associated with increased risk of experiencing an opioid-related complication were older age, male sex, and urban location (Table 3). Youth with diagnosed acute pain conditions were less likely to experience an opioid-related complication than those without (adjusted relative risk [aRR], 0.88 [95% CI, 0.84–0.93]); however, youth with a chronic pain condition were more likely to experience an opioid-related complication (aRR, 1.15 [95% CI, 1.09–1.20]). Numerous comorbid mental health conditions were associated with increased risk of an opioid-related complication, including mood and anxiety disorders (aRR, 4.45 [95% CI, 4.25–4.66]), psychosis (aRR, 1.35 [95% CI, 1.25–1.45]), attention deficit hyperactivity disorder (aRR, 1.10 [95% CI, 1.04–1.16]), trauma or stress-related conditions (aRR, 1.42 [95% CI, 1.35–1.50]), and conduct disorder or antisocial personality disorder (aRR, 1.18 [95% CI, 1.08–1.30]). Among patient factors, the largest effect size was observed for those with an additional comorbid substance use disorder (aRR, 20.77 [95% CI, 19.74–21.84]).
Table 3.
Associations of patient and opioid prescription factors with diagnosis of opioid use disorder or opioid overdose during the year after an opioid prescription among 3,278,990 youth.
|
Characteristic |
Unadjusted RR (95% CI) | Adjusteda RR (95% CI) |
|---|---|---|
| Patient factors | ||
| Age, years | ||
| 11–13 | Reference | Reference |
| 14–16 | 6.76 (5.36 – 8.53) | 3.54 (2.80 – 4.47) |
| 17–19 | 16.33 (13.02 – 20.47) | 5.65 (4.50 – 7.10) |
| 20–22 | 26.95 (21.51 – 33.78) | 7.31 (5.81 – 9.19) |
| 23–25 | 21.56 (17.19 – 27.05) | 6.17 (4.90 – 7.76) |
| Sex | ||
| Female | Reference | Reference |
| Male | 2.14 (2.06 – 2.22) | 2.00 (1.91 – 2.08) |
| Location | ||
| Rural | Reference | Reference |
| Urban | 1.20 (1.13 – 1.27) | 1.15 (1.09 – 1.20) |
| Pregnancyb | ||
| No | Reference | Reference |
| Yes | 0.95 (0.88 – 1.02) | 0.91 (0.84 – 0.99) |
| Acute pain conditionc | ||
| No | Reference | Reference |
| Yes | 1.06 (1.01 – 1.11) | 0.88 (0.84 – 0.93) |
| Chronic pain conditionb | ||
| No | Reference | Reference |
| Yes | 1.95 (1.86 – 2.04) | 1.15 (1.09 – 1.20) |
| Intellectual disability, learning disorder, or developmental disordera | ||
| No | Reference | Reference |
| Yes | 1.46 (1.27 – 1.68) | 0.84 (0.73 – 0.96) |
| Mood or anxiety disorderb | ||
| No | Reference | Reference |
| Yes | 11.17 (10.73 – 11.62) | 4.45 (4.25 – 4.66) |
| Psychosisb | ||
| No | Reference | Reference |
| Yes | 14.68 (13.65 – 15.79) | 1.35 (1.25 – 1.45) |
| Attention deficit hyperactivity disorderb | ||
| No | Reference | Reference |
| Yes | 2.68 (2.55 – 2.83) | 1.10 (1.04 – 1.16) |
| Trauma or stress-related conditionb | ||
| No | Reference | Reference |
| Yes | 4.22 (4.01 – 4.44) | 1.42 (1.35 – 1.50) |
| Conduct disorder or antisocial personality disorderb | ||
| No | Reference | Reference |
| Yes | 6.08 (5.58 – 6.63) | 1.18 (1.08 – 1.30) |
| Other substance use disorderb | ||
| No | Reference | Reference |
| Yes | 45.45 (43.53 – 47.45) | 20.77 (19.74 – 21.84) |
| Prescription factors | ||
| Long-acting opioid | ||
| No | Reference | Reference |
| Yes | 9.49 (7.94 – 11.34) | 2.59 (2.18 – 3.09) |
| Daily dose, morphine milligram equivalents | ||
| < 30 | Reference | Reference |
| 30–59 | 0.91 (0.87 – 0.95) | 0.95 (0.91 – 0.99) |
| 60–89 | 0.89 (0.83 – 0.96) | 0.90 (0.83 – 0.96) |
| 90–119 | 0.97 (0.87 – 1.08) | 0.92 (0.83 – 1.03) |
| ≥ 120 | 2.04 (1.83 – 2.28) | 1.23 (1.11 – 1.37) |
| Duration of prescription, days | ||
| 1–3 | Reference | Reference |
| 4–6 | 0.93 (0.89 – 0.97) | 0.96 (0.92 – 1.01) |
| 7–14 | 1.25 (1.17 – 1.32) | 1.15 (1.08 – 1.22) |
| ≥ 15 | 3.01 (2.79 – 3.24) | 1.96 (1.80 – 2.12) |
Adjusted for all other covariates listed in the table
Diagnosed within the 6 months preceding or 12 months following the initial opioid prescription
Diagnosed within the 14 days preceding or following the initial opioid prescription
Prescription factors associated with increased risk of experiencing an opioid-related complication included formulation, dose, and duration. Compared to short-acting opioid formulations, long-acting formulations were associated with a 159% increased risk of experiencing an opioid-related complication (aRR, 2.59 [95% CI, 2.18–3.09]). Daily dose was associated with decreased risk of opioid-related complications at moderate doses (30–59 MME: aRR, 0.95 [95% CI, 0.91–0.99]; and 60–89 MME: aRR, 0.90 [95% CI, 0.83–0.96]) and with increased risk at high doses (aRR, 1.23 [95% CI, 1.11–1.37]) compared to doses <30 MME daily. Longer prescription durations were also associated with increased risk of opioid-related complications (7–14 days: aRR, 1.15 [95% CI, 1.08–1.22]; and ≥15 days: aRR, 1.96 [95% CI, 1.80–2.12]) compared to prescriptions ≤3 days.
Among the 10,405 youth who experienced an opioid-related complication, 7,572 (72.8%) were identified as having an additional comorbid substance use disorder (Table 4). An alcohol use disorder was identified among 2,546 (33.4%) youth; cannabis use disorder, 2,514 (33.0%); nicotine use disorder, 3.289 (43.2%); and other substance use disorders, 5,751 (75.5%). In 29.0% of individuals who had an additional substance use disorder, the disorder had been identified before the initial opioid prescription, and in 71.0% it was not identified until after. Among the 3,278,990 youth who did not experience an opioid-related complication, 178,400 (5.4%) were identified as having a non-opioid substance use disorder (Table S7).
Table 4.
Timing of identification of other substance use disorders among 10,405 youth who experienced an opioid-related complication.
| Diagnosis | Totala N (% of all individuals with identified substance use) |
Identified at time of initial prescription or during 6 preceding monthsa N (row %) |
Identified during 12 months following initial prescriptiona N (row %) |
|---|---|---|---|
| Any substance use disorder (excluding opioid use disorder) | 7,618 (100.0%) | 2,208 (29.0%) | 5,410 (71.0%) |
| Alcohol use disorder | 2,546 (33.4%) | 448 (17.6%) | 2,098 (82.4%) |
| Cannabis use disorder | 2,514 (33.0%) | 447 (17.8%) | 2,067 (82.2%) |
| Nicotine use disorder | 3,289 (43.2%) | 1,145 (34.8%) | 2,144 (65.2%) |
| Other non-opioid substance use disorder | 5,751 (75.5%) | 904 (15.7%) | 4,847 (84.3%) |
Throughout, individuals could be counted in multiple rows if they had multiple substance use disorders (e.g., an alcohol use disorder and a cannabis use disorder).
DISCUSSION
In this US study of commercially insured youth, 0.3% of individuals filling an initial opioid prescription developed OUD or experienced an opioid overdose in the year that followed. These opioid-related complications, the majority of which were OUD diagnoses, most commonly occurred >90 days after the initial prescription. Patient risk factors included older age, male sex, urban location, comorbid chronic pain condition, comorbid mental health diagnoses, and other substance use disorders. Prescription risk factors included long-acting opioid formulations, daily doses ≥120 MME, and prescription duration ≥7 days. The percentage of youth experiencing an opioid-related complication was greatly elevated (4.1%) among youth who used other substances. Indeed, nearly three-quarters of the youth who experienced an opioid-related complication were identified as having another substance use disorder; however, this comorbid disorder was identified before the initial opioid prescription in only 29% of these youth.
Our findings build on three other studies suggesting that an initial opioid prescription is associated with subsequent diagnoses of OUD or persistent opioid use.11–14 Notably, the one-year risk of opioid-related complications observed in our study (0.3%) was much lower than that of these other studies. A large study of 16- to 25-year-olds receiving opioid prescriptions from dental clinicians found that approximately 6% of youth were diagnosed with OUD within the subsequent year.11 Since we examined opioid prescriptions across all indications and clinicians, it is possible that opioids prescribed for dental pain are associated with greater risk for the development of OUD than opioids prescribed for other indications. Two studies of opioid prescribing after surgical procedures found that 4–5% of youth progress to persistent opioid use, defined as continued receipt of opioid prescriptions 90–180 days following a procedure.12,13 However, persistent opioid use likely does not always reflect a true diagnosis of OUD. Regardless, it remains unclear why opioid-related complications occur months after an initial prescription. Two possible explanations include that OUD is a condition that develops over time, and that it often goes undiagnosed until serious problems result. Further studies should examine these possibilities.
Our findings suggest that opioid-related complications are rare following an initial prescription and support the judicious use of opioids for pain management. Consistent with US prescribing guidelines for adults,15 our results support the recommendations that when opioids are prescribed, clinicians should provide a short-acting formulation, use the lowest effective dose, and provide a short duration. Even still, the prescription factors that we identified had smaller effect sizes than those of patient factors. In particular, comorbid mental health conditions and other substance use disorders were associated with greatly elevated risk. The presence of a mood or anxiety disorder was associated with more than 4 times the risk for opioid-related complications, and other substance use disorders with more than 20 times the risk. These findings are consistent with that of another recent study of commercially insured youth in which individuals with comorbid mental health conditions or substance use disorders were more likely to experience an opioid overdose during the time that they were receiving a prescription.14 There are numerous time-efficient, developmentally appropriate, validated instruments to identify depression,23 anxiety,24–26 and substance use27,28 among youth in clinical settings, and our findings highlight the potentially critical role of screening for these conditions when clinicians prescribe opioids. We do not advocate for withholding opioid prescriptions from youth with comorbid mental health conditions or substance use; however, clinicians might consider ensuring close follow-up after the prescription for youth identified as having comorbid mental health conditions or substance use to minimize the risk for opioid-related complications.29 Such follow-up may need to continue >90 days after the prescription since risk appears to persist beyond this time.
Other patient risk factors deserve further consideration. Even after accounting for mental health and substance use disorder diagnoses that may increase in prevalence with age, we observed that older age was associated with increased risk for opioid-related complications. This risk peaked at age 20–22 years. We also found that male sex was associated with double the risk of opioid-related complications compared to female sex. These findings are consistent with national epidemiologic trends suggesting that the incidence of fatal opioid overdose increases with older age during young adulthood and is highest among males.1,17,30 However, the interplay between age and sex in the trajectory towards OUD and overdose remains poorly described; the development of opioid-related complications by age and sex among youth should therefore be further delineated in future studies. Additionally, we were unable to examine the role of family history, which should be the subject of further research.
The association of pain conditions with risk of opioid-related complications that we observed was complex. Our findings support the conclusion that opioids prescribed appropriately for the management of acute pain may be associated with lower risk.31 However, underlying chronic pain conditions, either present at the time of an initial opioid prescription or diagnosed in the following year, appear to be associated with increased risk. These findings are consistent with those from adult studies, which highlight that acute pain diagnoses are often absent at the time of an opioid prescription,32 and that risk for OUD is elevated among individuals with chronic pain.15,33,34 Taken together, these findings suggest that youth receiving opioids for chronic pain should receive close clinical follow-up to potentially reduce the risk for opioid-related complications, particularly if they have comorbid mental health conditions or substance use disorder.15
Consistent with other studies and with US opioid prescribing guidelines for adults, we found that long-acting opioid formulations and prescription durations of a week or more were associated with increased risk for opioid-related complications.14,35,36 However, unlike other studies, we did not observe a clear dose-dependent association with the daily MME prescribed and the risk for opioid-related complications. Although we did observe significantly increased risk for opioid prescriptions with very high doses of ≥120 MME/day (equivalent to, for example, 20 milligrams of oxycodone taken every 6 hours), moderate doses were associated with decreased risk compared to doses of <30 MME/day (equivalent to, for example, 5 milligrams of oxycodone taken every 6 hours). It is possible that the duration of an initial prescription may be more important for the risk of opioid-related complications than the dose, and that after adjustment for patient risk factors such as comorbid mental health conditions and substance use disorders, the risk of moderate opioid doses is mitigated. Nonetheless, this finding was unexpected and should be examined further in future studies.
This study has limitations. First, the techniques we used in this study do not establish a cause-and-effect relationship between opioid prescriptions and opioid-related complications. Nonetheless, we worked to ensure that opioid prescriptions preceded any opioid-related complications to support a temporal relationship. Additionally, our findings are consistent with the apparent increase in risk following an initial opioid prescription reported in other studies.11–14,37 Second, OUD, opioid overdose, other substance use disorders, and mental health conditions frequently go undiagnosed; thus, the true prevalence of these conditions may have been underestimated in our study. Also, youth who received diagnoses in situations in which no health insurance claim was made—including if, for example, they received care through private-pay treatment—would also not be identified in this study. Additionally, some of these variables may have been misclassified, which is known to occur in studies of substance use disorders and mental health conditions.38,39 Third, the measures that we used for opioid dose (i.e., MME) have not been validated for children and were not adjusted for weight. Fourth, our sample was restricted to commercially insured youth in the US context who met specific sample eligibility criteria (e.g., continuous insurance enrollment) and results may not generalize to youth who are uninsured or receive other forms of health insurance. Notably, the prevalence of opioid use disorder in commercially and publicly insured youth is similar; however, characteristics of youth with opioid use disorder differ, including by race/ethnicity, geographic location, and socioeconomic status, among other factors.16–18,40,41 Our findings should therefore be replicated in samples of publicly insured youth and individuals from outside the US, including those with lapses in insurance coverage.
The risk for OUD and opioid overdose following an initial opioid prescription is relatively rare, occurring in only approximately 1 in 315 US youth but often manifesting months after the prescription. Although our findings highlight that the risk for opioid-related complications may be minimized by prescribing short-acting opioid formulations for the shortest duration required, it may be even more important for clinicians to conduct a careful risk assessment of youth before prescribing. Our results suggest that such an assessment should include screening for mental health conditions such as depression and anxiety. Clinicians should also screen for substance use disorders—including problems with alcohol, cannabis, nicotine, and other substances—which may be associated with >20 times the risk for opioid-related complications and may already be present but as-yet unidentified at the time of an opioid prescription.
Supplementary Material
Funding Source:
National Institute on Drug Abuse (R01DA046527-02S1): Hadland, Bagley, Gai, Barocas
National Institute on Drug Abuse (K23DA045085): Hadland, Earlywine
National Institute on Drug Abuse (L40DA042434): Hadland
National Institute on Drug Abuse (K23DA044324): Bagley, Schoenberger
Charles A. King Trust: Barocas
Financial Disclosure:
The authors have no financial relationships relevant to this article to disclose.
Footnotes
Conflicts of Interest Declaration:
The authors have no conflicts of interest relevant to this article to disclose.
REFERENCES
- 1.Hedegaard H, Miniño AM, Warner M. Drug Overdose Deaths in the United States, 1999–2018. NCSH Data Brief, No. 356. National Center for Health Statistics; 2020. [PubMed] [Google Scholar]
- 2.GBD 2016 Alcohol and Drug Use Collaborators. The global burden of disease attributable to alcohol and drug use in 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Psychiatry. 2018;5(12):987–1012. doi: 10.1016/S2215-0366(18)30337-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Special Advisory Committee on the Epidemic of Opioid Overdoses. National Report: Apparent Opioid-Related Deaths in Canada (January 2016 to June 2017) Web-Based Report. Public Health Agency of Canada; https://www.canada.ca/en/public-health/services/publications/healthy-living/apparent-opioid-related-deaths-report-2016-2017-december.html [Google Scholar]
- 4.Levy S. Youth and the Opioid Epidemic. Pediatrics. 2019;143(2):e20182752. doi: 10.1542/peds.2018-2752 [DOI] [PubMed] [Google Scholar]
- 5.Substance Abuse and Mental Health Services Administration Center for Behavioral Health Statistics and Quality. Treatment Episode Data Set (TEDS): 2005–2015. National Admissions to Substance Abuse Treatment Services. BHSIS Series S-91, HHS Publication No. (SMA) 17–5037. Substance Abuse and Mental Health Services Administration; 2017. https://wwwdasis.samhsa.gov/dasis2/teds_pubs/2015_teds_rpt_natl.pdf [Google Scholar]
- 6.Gaither JR, Shabanova V, Leventhal JM. US National Trends in Pediatric Deaths From Prescription and Illicit Opioids, 1999–2016. JAMA Netw Open. 2018;1(8):e186558. doi: 10.1001/jamanetworkopen.2018.6558 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Seth P, Rudd RA, Noonan RK, Haegerich TM. Quantifying the Epidemic of Prescription Opioid Overdose Deaths. Am J Public Health. 2018;108(4):500–502. doi: 10.2105/AJPH.2017.304265 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cerdá M, Santaella J, Marshall BDL, Kim JH, Martins SS. Nonmedical Prescription Opioid Use in Childhood and Early Adolescence Predicts Transitions to Heroin Use in Young Adulthood: A National Study. J Pediatr. 2015;167(3):605—−612.e2. doi: 10.1016/j.jpeds.2015.04.071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Miech RA, Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE, Patrick ME. Monitoring the Future National Survey Results on Drug Use, 1975–2018: Volume I, Secondary School Students. Institute for Social Research, The University of Michigan; 2019. http://www.monitoringthefuture.org//pubs/monographs/mtf-vol1_2018.pdf [Google Scholar]
- 10.Hudgins JD, Porter JJ, Monuteaux MC, Bourgeois FT. Prescription opioid use and misuse among adolescents and young adults in the United States: A national survey study. PLoS Med. 2019;16(11):e1002922. doi: 10.1371/journal.pmed.1002922 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schroeder AR, Dehghan M, Newman TB, Bentley JP, Park KT. Association of Opioid Prescriptions From Dental Clinicians for US Adolescents and Young Adults With Subsequent Opioid Use and Abuse. JAMA Intern Med. Published onlineDecember2018. doi: 10.1001/jamainternmed.2018.5419 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harbaugh CM, Lee JS, Hu HM, et al. Persistent Opioid Use Among Pediatric Patients After Surgery. Pediatrics. 2018;141(1). doi: 10.1542/peds.2017-2439 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bennett KG, Harbaugh CM, Hu HM, et al. Persistent Opioid Use Among Children, Adolescents, and Young Adults After Common Cleft Operations. J Craniofac Surg. 2018;29(7):1697–1701. doi: 10.1097/SCS.0000000000004762 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chua K-P, Brummett CM, Conti RM, Bohnert A. Opioid Prescribing Patterns Associated with Prescription Opioid Overdose in Adolescents and Young Adults. JAMA Pediatr. Published online2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Dowell D, Haegerich TM, Chou R. CDC Guideline for Prescribing Opioids for Chronic Pain–United States, 2016. JAMA. 2016;315(15):1624–45. doi: 10.1001/jama.2016.1464 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Alinsky RH, Zima BT, Rodean J, et al. Receipt of Addiction Treatment After Opioid Overdose Among Medicaid-Enrolled Adolescents and Young Adults. JAMA Pediatr. Published onlineJanuary6, 2020:e195183. doi: 10.1001/jamapediatrics.2019.5183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hadland SE, Wharam JF, Schuster MA, Zhang F, Samet JH, Larochelle MR. Trends in Receipt of Buprenorphine and Naltrexone for Opioid Use Disorder Among Adolescents and Young Adults, 2001–2014. JAMA Pediatr. 2017;171(8):747–755. doi: 10.1001/jamapediatrics.2017.0745 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hadland SE, Bagley SM, Rodean J, et al. Receipt of Timely Addiction Treatment and Association of Early Medication Treatment With Retention in Care Among Youths With Opioid Use Disorder. JAMA Pediatr. 2018;172(11):1029–1037. doi: 10.1001/jamapediatrics.2018.2143 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.HCUP Tools and Software. Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality, Rockville, MD. PublishedFebruary19, 2020. AccessedMarch 11, 2020. https://www.hcup-us.ahrq.gov/tools_software.jsp [Google Scholar]
- 20.Mayhew M, DeBar LL, Deyo RA, et al. Development and Assessment of a Crosswalk Between ICD-9-CM and ICD-10-CM to Identify Patients with Common Pain Conditions. J Pain Off J Am Pain Soc. Published onlineMay2019. doi: 10.1016/j.jpain.2019.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Simon GE, Stewart C, Yarborough BJ, et al. Mortality Rates After the First Diagnosis of Psychotic Disorder in Adolescents and Young Adults. JAMA Psychiatry. 2018;75(3):254–260. doi: 10.1001/jamapsychiatry.2017.4437 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zou G A Modified Poisson Regression Approach to Prospective Studies with Binary Data. Am J Epidemiol. 2004;159(7):702–706. doi: 10.1093/aje/kwh090 [DOI] [PubMed] [Google Scholar]
- 23.Richardson LP, McCauley E, Grossman DC, et al. Evaluation of the Patient Health Questionnaire-9 Item for detecting major depression among adolescents. Pediatrics. 2010;126(6):1117–1123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mossman SA, Luft MJ, Schroeder HK, et al. The Generalized Anxiety Disorder 7-item scale in adolescents with generalized anxiety disorder: Signal detection and validation. Ann Clin Psychiatry Off J Am Acad Clin Psychiatr. 2017;29(4):227–234A. [PMC free article] [PubMed] [Google Scholar]
- 25.HALE WW III, Raaijmakers Q, Muris P, Meeus W. Psychometric properties of the Screen for Child Anxiety Related Emotional Disorders (SCARED) in the general adolescent population. J Am Acad Child Adolesc Psychiatry. 2005;44(3):283–290. [DOI] [PubMed] [Google Scholar]
- 26.Rappaport BI, Pagliaccio D, Pine DS, Klein DN, Jarcho JM. Discriminant validity, diagnostic utility, and parent-child agreement on the Screen for Child Anxiety Related Emotional Disorders (SCARED) in treatment- and non-treatment-seeking youth. J Anxiety Disord. 2017;51:22–31. doi: 10.1016/j.janxdis.2017.08.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Knight JR, Sherritt L, Shrier LA, Harris SK, Chang G. Validity of the CRAFFT substance abuse screening test among adolescent clinic patients. Arch Pediatr Adolesc Med. 2002;156(6):607–614. [DOI] [PubMed] [Google Scholar]
- 28.Levy S, Weiss R, Sherritt L, et al. An electronic screen for triaging adolescent substance use by risk levels. JAMA Pediatr. 2014;168(9):822–828. doi: 10.1001/jamapediatrics.2014.774 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hadland SE, Bagley SM. Opioid Prescribing Patterns and Subsequent Overdose. JAMA Pediatr. Published onlineDecember16, 2019. doi: 10.1001/jamapediatrics.2019.4885 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Scholl L, Seth P, Kariisa M, Wilson N, Baldwin G. Drug and Opioid-Involved Overdose Deaths - United States, 2013–2017. MMWR Morb Mortal Wkly Rep. 2018;67(5152):1419–1427. doi: 10.15585/mmwr.mm675152e1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hoppe JA, Kim H, Heard K. Association of emergency department opioid initiation with recurrent opioid use. Ann Emerg Med. 2015;65(5):493–499.e4. doi: 10.1016/j.annemergmed.2014.11.015 [DOI] [PubMed] [Google Scholar]
- 32.Sherry TB, Sabety A, Maestas N. Documented Pain Diagnoses in Adults Prescribed Opioids: Results From the National Ambulatory Medical Care Survey, 2006–2015. Ann Intern Med. 2018;169(12):892–894. doi: 10.7326/M18-0644 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Liebschutz JM, Saitz R, Weiss RD, et al. Clinical factors associated with prescription drug use disorder in urban primary care patients with chronic pain. J Pain Off J Am Pain Soc. 2010;11(11):1047–1055. doi: 10.1016/j.jpain.2009.10.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Dunn KM, Saunders KW, Rutter CM, et al. Opioid prescriptions for chronic pain and overdose: a cohort study. Ann Intern Med. 2010;152(2):85–92. doi: 10.7326/0003-4819-152-2-201001190-00006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Miller M, Barber CW, Leatherman S, et al. Prescription opioid duration of action and the risk of unintentional overdose among patients receiving opioid therapy. JAMA Intern Med. 2015;175(4):608–615. doi: 10.1001/jamainternmed.2014.8071 [DOI] [PubMed] [Google Scholar]
- 36.Ray WA, Chung CP, Murray KT, Hall K, Stein CM. Prescription of Long-Acting Opioids and Mortality in Patients With Chronic Noncancer Pain. JAMA. 2016;315(22):2415–2423. doi: 10.1001/jama.2016.7789 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Harbaugh CM, Nalliah RP, Hu HM, Englesbe MJ, Waljee JF, Brummett CM. Persistent Opioid Use After Wisdom Tooth Extraction. JAMA. 2018;320(5):504–506. doi: 10.1001/jama.2018.9023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Doktorchik C, Patten S, Eastwood C, et al. Validation of a case definition for depression in administrative data against primary chart data as a reference standard. BMC Psychiatry. 2019;19(1):9. doi: 10.1186/s12888-018-1990-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Murphy SM, Friesner DL, Rosenman R. Opioid misuse among adolescents: new evidence from a misclassification analysis. Appl Health Econ Health Policy. 2015;13(2):181–192. doi: 10.1007/s40258-015-0151-z [DOI] [PubMed] [Google Scholar]
- 40.Bagley SM, Larochelle MR, Xuan Z, et al. Characteristics and Receipt of Medication Treatment Among Young Adults Who Experience a Nonfatal Opioid-Related Overdose. Ann Emerg Med. 2020;75(1):29–38. doi: 10.1016/j.annemergmed.2019.07.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Chatterjee A, Larochelle MR, Xuan Z, et al. Non-fatal opioid-related overdoses among adolescents in Massachusetts 2012–2014. Drug Alcohol Depend. 2019;194:28–31. doi: 10.1016/j.drugalcdep.2018.09.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.