Changes in Substance Abuse Treatment Use Among Individuals With Opioid Use Disorders in the United States, 2004-2013

During the last decade, nonmedical use of opioid analgesics and heroin increased substantially in the United States, contributing to an increased rate of overdose deaths.¹ Expanding access to substance use treatment among individuals with opioid use disorders (OUDs) may be an important strategy for reducing harmful use.² In the early 2000s, less than one-sixth of individuals with OUDs received any treatment, and use of office-based treatment was rare.³ It is unknown whether treatment patterns have changed in recent years.

Buprenorphine, approved in 2002, provides an office-based treatment, but it remains unclear whether buprenorphine availability has increased treatment rates.⁴ The 2008 mental health parity law eliminating some insurance-based barriers to addiction treatment could also expand treatment options; research indicates addiction treatment rates did not increase, but average expenditures did.⁵

We used the 2004-2013 rounds of the National Survey of Drug Use and Health (NSDUH), a nationally representative annual cross-sectional survey of individuals aged 12 years or older. Respondents were interviewed in person and overall response rates ranged from 60.2% to 70.0%. Using consistent measures, we identified individuals with opioid abuse or dependence symptoms based on Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) criteria⁶ and if they received treatment for OUD in the prior 12 months. Those indicating treatment receipt were asked whether they received treatment in the following settings (>1 setting could be indicated): hospital, outpatient, inpatient, mental health center, emergency department, physician’s office, jail, and self-help group.

We divided the sample into two 5-year periods (2004-2008 vs 2009-2013) to provide reliable estimates; the 2008 cut point also coincides with enactment of the federal parity law. We calculated the prevalence of individuals receiving treatment and, among the treated, the percentage receiving treatment separately for each setting. We calculated both unadjusted and regression-adjusted rates, accounting for sociodemographic factors that might influence treatment use (age, sex, race/ethnicity, self-reported health status, income category, insurance status, and types of substances used).

The 2-sided, pairwise t test was used to test for significant differences (P < .05) in means between the periods. We applied survey weights and adjusted standard errors to account for sampling design. The analysis was conducted using Stata version 13 (StataCorp). This study was determined exempt by the Johns Hopkins University institutional review board.

In the combined sample, we identified 6770 respondents with OUDs. Over time, this population became older, more likely to use heroin and other substances, and less likely to have private insurance (Table 1). In an unadjusted analysis (Table 2), the percentage of individuals with OUDs receiving treatment was 16.6% in 2004-2008 and 21.5% in 2009-2013, which is a statistically significant difference (5.0 percentage points [95% CI, 1.5 to 8.4 percentage points]; P = .005). The regression-adjusted rates were similar (18.8% in 2004-2008 vs 19.7% in 2009-2013); therefore, the difference was not statistically significant (0.8 percentage points [95% CI, −2.3 to 4.0 percentage points]; P = .59).

The mean number of settings visited increased from 2.8 in 2004-2008 to 3.3 in 2009-2013, which is a significant increase (difference, 0.5 [95% CI, 0.1-0.8]; P = .007). The most common setting in both periods was self-help groups, which did not significantly change. More than half of individuals receiving treatment during both periods also visited outpatient treatment. Use of inpatient treatment increased from 37.5% in 2004-2008 to 51.9% in 2009-2013 (difference, 14.4 percentage points [95% CI, 4.4-24.5 percentage points]; P = .005), and office-based treatment increased from 25.1% to 34.8% (difference, 9.7 percentage points [95% CI, 0.5%-18.8 percentage points]; P = .04). Adjusted treatment setting differences were consistent with unadjusted differences.

During the decade from 2004 to 2013, use of treatment remained low for individuals with OUDs and did not increase after accounting for changing population characteristics, underscoring substantial room for improvement.

Individuals in treatment received care in more settings, with the greatest increases in inpatient treatment and at physician’s offices. Although physician’s offices may provide access to buprenorphine, medication-assisted treatments are often unavailable in inpatient settings, which could hinder patient recovery.

A limitation of our study is that the NSDUH does not measure use of medication-assisted treatment or care quality. Measuring these outcomes is important for future research.

Corresponding Author: Brendan Saloner, PhD, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, 624 N Broadway, Room 344, Baltimore, MD 21205 (bsaloner@jhu.edu).

Author Contributions: Dr Saloner had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Saloner.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: Saloner.

Statistical analysis: All authors.

Study supervision: Saloner.

Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.