Abstract
Purpose
The American Academy of Pediatrics recently recommended that pediatricians consider medication-assisted treatment (MAT) for adolescents with severe opioid use disorders. Little is known about adolescents’ current use of MAT.
Methods
We use data on episodes of specialty treatment for heroin or opioid use (n = 139,092) from a database of publicly funded treatment programs in the U.S. We compare the proportions of adolescents and adults who received MAT, first using unadjusted comparison of proportions, then using logistic regression to adjust for potential confounders.
Results
Only 2.4% (95% confidence interval [CI], 1.4%–3.7%) of adolescents in treatment for heroin received MAT, as compared to 26.3% (95% CI, 26.0%–26.6%) of adults. Only .4% (95% CI, .2%–.7%) of adolescents in treatment for prescription opioids received MAT, as compared to 12.0% (95% CI, 11.7%–12.2%) of adults. Regression-adjusted results were qualitatively similar.
Conclusions
Regulatory changes and expansions of Medicaid/CHIP coverage for MAT may be needed to improve MAT access.
Keywords: Opioids, Heroin, Buprenorphine, Methadone, Medication-assisted treatment, Substance use
Prescribing rates for opioids among adolescents and young adults nearly doubled from 1994 to 2009 [1]. Between 1992 and 2012, the prevalence of nonmedical prescription opioid misuse and opioid use disorders among adolescents both doubled [2] (although prescription opioid misuse is beginning to decline again [3]). Only about 1 out of every 12 adolescents in need of treatment for opioid use received any care in 2014 [4].
The American Academy of Pediatrics recently recommended that pediatricians consider offering medication-assisted treatment (MAT) to adolescents with severe opioid use disorders (OUD) or discuss referrals to other providers for this treatment [2]. MAT reduces adolescents’ opioid misuse and injection drug use [5]. Increased availability of MAT for adults has also been associated with substantial population-level reductions in overdoses [6]. MAT by a pediatrician without referral would consist primarily of buprenorphine treatment (or long-acting injectable naltrexone, although there is limited evidence for the latter), but few primary care pediatricians have buprenorphine training. This likely leaves the provision of MAT to specialty substance use treatment centers like methadone clinics. However, methadone access is severely restricted for adolescents age 16 and 17 years. Federal regulations require methadone clinics to receive a special waiver to treat adolescents. Furthermore, adolescents must demonstrate two “failed” attempts without pharmacotherapy to be eligible for methadone treatment [7]. Little information is available regarding the extent of MAT use among adolescents treated for OUD.
Methods
We used data from the 2013 Treatment Episode Data Set (TEDS), a federal database of state administrative records on substance use treatment episodes that occur in publicly funded facilities (Table 1) [8]. TEDS has been estimated to cover more than 67% of substance use treatment admissions (public or private) in the U.S. [9].
Table 1.
Selected characteristics of adolescents and adults in treatment for heroin and opioid use by age, 2013
| Heroin | Other opioids | |||
|---|---|---|---|---|
|
|
|
|||
| Adults (N = 66,074) | Adolescents (N = 761) | Adults (N = 69,932) | Adolescents (N = 2,325) | |
|
|
|
|
|
|
| N (%) | N (%) | N (%) | N (%) | |
| No MAT | 48,694 (73.7%) | 743 (97.6%) | 61,569 (88%) | 2,316 (99.6%) |
| MAT | 17,380 (26.3%) | 18 (2.4%) | 8,363 (12%) | 9 (.4%) |
| Sex | ||||
| Male | 41,881 (63.4%) | 386 (50.7%) | 36,603 (52.3%) | 1,589 (68.3%) |
| Female | 24,193 (36.6%) | 375 (49.3%) | 33,329 (47.7%) | 736 (31.7%) |
| Race/ethnicity | ||||
| White | 46,653 (70.6%) | 558 (73.3%) | 58,764 (84%) | 1,627 (70%) |
| Black | 7,495 (11.3%) | 29 (3.8%) | 3,900 (5.6%) | 232 (10%) |
| Hispanic | 9,369 (14.2%) | 128 (16.8%) | 4,455 (6.4%) | 314 (13.5%) |
| Am Indian/Alaska Native | 707 (1.1%) | 16 (2.1%) | 1,288 (1.8%) | 50 (2.2%) |
| Asian/Hawaiian/Pac Islander | 507 (.8%) | 5 (.7%) | 456 (.7%) | 29 (1.2%) |
| Multiracial | 530 (.8%) | 9 (1.2%) | 546 (.8%) | 42 (1.8%) |
| Other | 813 (1.2%) | 16 (2.1%) | 523 (.7%) | 31 (1.3%) |
| Substances reported at admission | ||||
| 1 | 21,749 (32.9%) | 87 (11.4%) | 24,208 (34.6%) | 142 (6.1%) |
| 2 | 24,564 (37.2%) | 300 (39.4%) | 22,848 (32.7%) | 675 (29%) |
| 3 | 19,761 (29.9%) | 374 (49.1%) | 22,876 (32.7%) | 1,508 (64.9%) |
| Referral source | ||||
| Individual | 39,107 (59.2%) | 271 (35.6%) | 34,797 (49.8%) | 601 (25.8%) |
| Substance use provider | 4,701 (7.1%) | 65 (8.5%) | 3,938 (5.6%) | 109 (4.7%) |
| Other health provider | 3,918 (5.9%) | 68 (8.9%) | 7,479 (10.7%) | 236 (10.2%) |
| School/work | 136 (.2%) | 25 (3.3%) | 441 (.6%) | 170 (7.3%) |
| Other community | 4,872 (7.4%) | 100 (13.1%) | 7,376 (10.5%) | 261 (11.2%) |
| Criminal justice | 13,340 (20.2%) | 232 (30.5%) | 15,901 (22.7%) | 948 (40.8%) |
| Housing status | ||||
| Not homeless | 56,520 (85.5%) | 754 (99.1%) | 64,594 (92.4%) | 2,296 (98.8%) |
| Homeless | 9,554 (14.5%) | 7 (.9%) | 5,338 (7.6%) | 29 (1.2%) |
Note: Chi-square tests indicate that all between age-group differences are statistically significantly different at the p < .001 level.
MAT = medication-assisted treatment.
Persons treated primarily for “heroin,” “nonprescription use of methadone,” or “other opiates or synthetics” were included in our analysis. We restricted our sample to first treatment episodes. Pennsylvania, Georgia, West Virginia, Wisconsin, and Wyoming did not provide necessary data and were excluded. An additional 5.5% of remaining records were excluded because they were missing information on covariates included in the analysis. The final sample included 139,092 first treatment episodes. Adolescents comprised 2.2% of the sample of episodes, including 1.1% of episodes of treatment for heroin use and 3.2% of episodes of treatment for all other opioids.
Receipt of MAT was defined by whether methadone or buprenorphine (but not naltrexone) was part of a client’s treatment plan. We calculated the proportion of adolescent episodes (ages 15–17 years) and adult episodes (≥18) of opioid treatment that included MAT, stratified by whether the individual in treatment used heroin versus other opioids.
Adolescents and adults in drug treatment may differ on other characteristics that influence the receipt of MAT. We used logistic regression to model the association of MAT with age, adjusting for sex, race/ethnicity, referral source, homelessness status, and number of substances reported at admission and stratified by whether the client used heroin versus only other prescription opioids. The regression model was used to estimate the odds ratio of MAT comparing adolescents to adults and to estimate adjusted rates of MAT use where all covariates were set to the population average.
This research is exempt from the human subject’s ethical approval research requirements because it involves secondary analysis of existing data and subjects cannot be identified.
Results
In total, 761 adolescents had a first treatment admission for heroin use and 2,325 for other opioid use. Only 2.4% (95% confidence interval [CI], 1.4%–3.7%) of adolescent treatment admissions included MAT, as compared to 26.3% (95% CI, 26.0%–26.6%) of adult admissions. Only .4% (95% CI, .2%–.7%) of adolescent treatment admissions for other opioids included MAT, as compared to 12.0% (95% CI, 11.7%–12.2%) of adult admissions.
Adolescents were less likely than adults to be homeless, were more likely to report using more substances at admission, and were more likely to be referred by the criminal justice system. However, adjusting for these variables still indicated substantial differences in MAT use of heroin users, OR = .09 (95% CI, .05–.14) and other opioid users, OR = .05 (95% CI, .03–.10). (Unadjusted and adjusted rates of MAT are shown in Figure 1.)
Figure 1.
Use of medication-assisted treatment (MAT) among persons in treatment for heroin and opioid use by age, 2013. Authors’ analysis of the 2013 Treatment Episode Data Set (TEDS). Each observation represents a first episode of treatment for opioid use disorder. Adjusted rate represents a predictive margin derived from a logistic regression model that controls for sex, race/ethnicity, primary referral source, homelessness, and number of substances used in the adult group. Error bars show 95% confidence interval.
Discussion
Opioid misuse in adolescence is associated with numerous long-term health risks [10]. MAT may be effective in the treatment of OUD for adolescents [2,6,7]. We found that adolescents in specialty treatment for either heroin or other opioid misuse rarely receive MAT. This may reflect patient, parent, or provider preferences and concerns about the appropriateness of MAT for adolescents. However, Medicaid and CHIP, which cover the majority of all adolescents, could also help facilitate better access to MAT treatment for adolescents with opioid use disorder. For example, state Medicaid programs could include MAT within the early periodic screening, detection, and treatment benefit [7], and adolescents could be incorporated into ongoing Medicaid demonstrations to expand MAT.
This study has several limitations. First, we considered only specialty treatment programs. Office-based physicians were not included in TEDS but may provide buprenorphine or naltrexone to adolescents. Physicians provide a minority of all opioid use disorder treatment [4]. Second, we could not consider health insurance status in this analysis because this variable is not reported in many states. Third, TEDS also does not contain a strong measure of disease severity, which may influence the decision to prescribe MAT. Fourth, the MAT variable does not distinguish between methadone and buprenorphine treatment and may miss youth who were prescribed long-acting naltrexone. Finally, people with MAT in their treatment plan may not actually receive MAT if the treatment plan is not followed.
In spite of these limitations, the data presented here suggest that adolescents with opioid use disorders are much less likely than adults to receive MAT, and policy and practice changes are needed to expand access as recommended by the American Academy of Pediatrics.
IMPLICATIONS AND CONTRIBUTION.
The American Academy of Pediatrics recently recommended that pediatricians consider medication-assisted treatment (MAT) for adolescents with severe opioid use disorders. This study shows that adolescents in specialty substance treatment for opioid use disorders rarely receive MAT. Regulatory changes and expansions of Medicaid/CHIP coverage for MAT could improve MAT access.
Acknowledgments
Funding Sources
N.K. was supported by a NIDA T32 for Drug Dependence Epidemiology and Training (grant number DA007292).
Footnotes
Conflicts of Interest: The authors have no conflicts of interest to disclose.
References
- 1.Fortuna RJ, Robbins BW, Caiola E, et al. Prescribing of controlled medications to adolescents and young adults in the United States. Pediatrics. 2010;126:1108–16. doi: 10.1542/peds.2010-0791. [DOI] [PubMed] [Google Scholar]
- 2.Levy S, Ryan SA, Gonzalez PK, et al. Medication-assisted treatment of adolescents with opioid use disorders. Pediatrics. 2016;138:e20161893. doi: 10.1542/peds.2016-1893. [DOI] [PubMed] [Google Scholar]
- 3.Results from the 2014 National Survey on Drug Use and Health: Summary of National Findings. Rockville Md: Substance Abuse and Mental Health Services Administration; [PubMed] [Google Scholar]
- 4.Han B, Hedden SL, Lipari R, et al. Receipt of services for behavioral health problems: Results from the 2014 National Survey on Drug Use and Health. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2015. [PubMed] [Google Scholar]
- 5.Woody GE, Poole SA, Subramaniam G, et al. Extended vs short-term buprenorphine-naloxone for treatment of opioid-addicted youth: A randomized trial. [Accessed June 20, 2014];JAMA. 2008 300:2003–11. doi: 10.1001/jama.2008.574. Available at: www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2610690&tool=pmcentrez&rendertype=abstract. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Schwartz RP, Gryczynski J, O’Grady KE, et al. Opioid agonist treatments and heroin overdose deaths in Baltimore, Maryland, 1995–2009. Am J Public Health. 2013;103:917–22. doi: 10.2105/AJPH.2012.301049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wachino V, Hyd PS. Coverage of behavioral health services for youth with substance use disorders. [Accessed August 20, 2016];CMCS Informational Bulletin. 2015 Available at: https://www.medicaid.gov/federal-policy-guidance/downloads/cib-01-26-2015.pdf.
- 8.Substance Abuse and Mental Health Services Administration (SAMHSA) Treatment Episode Data Set-Admissions (TEDS-A) 2013. Rockville, MD: SAMHSA; 2012. [Accessed August 20, 2016]. Available at: https://wwwdasis.samhsa.gov/dasis2/teds.htm. [Google Scholar]
- 9.Saloner B, Lê Cook B. Blacks and Hispanics are less likely than whites to complete addiction treatment, largely due to socioeconomic factors. Health Aff. 2013;32:135–45. doi: 10.1377/hlthaff.2011.0983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fogger S, McGuinness TM. Adolescents at risk: Pain pills to heroin: Part II. J psychosocial Nurs Ment Health Serv. 2015;53:27–30. doi: 10.3928/02793695-20150106-01. [DOI] [PubMed] [Google Scholar]