State-level prescription drug monitoring program mandates and adolescent injection drug use in the United States, 1995-2017: A difference-in-differences analysis

Author summaryWhy was this study done? The misuse of prescription opioids is a critical step in the trajectory toward injection of opioids and other substances. Intervening early on in this trajectory may prevent transitions to injection drug use and related health harms. Prescription drug monitoring programs track prescriptions of controlled substances, including opioids, and are designed to limit excessive prescriptions of opioids, but many states do not require clinicians to use this tool. As of January 1, 2017, 18 states in our sample had implemented laws that require the use of a state prescription drug monitoring program. We evaluated whether state prescription drug monitoring program mandates were associated with changes in adolescent injection drug use. What did the researchers do and find? Using data from over 330,000 US high school students 17-18 years old, we compared changes in self-reported injection drug use over time in states that implemented prescription drug monitoring program mandates versus states that did not. Our results show that injection drug use among 17 and 18 year olds declined after states implemented prescription drug monitoring program mandates compared to states that did not. What do these findings mean? The reduction in adolescent injection drug use associated with prescription drug monitoring program mandates is equivalent to an estimated 81,754 (95% CI 32,620 to 125,357) fewer adolescents initiating injection drug use over a year if states without prescription drug monitoring program mandates implemented these mandates. This study finds that prescription drug monitoring program mandates may be an important and effective strategy to prevent adolescent injection drug use. Policymakers might consider implementing prescription drug monitoring program mandates along with other substance use policies such as expanded addiction treatment, naloxone distribution, safe prescribing interventions, and harm reduction services. Background Prescription opioid misuse is an ongoing crisis and a risk factor for injection drug use (IDU). Few studies have evaluated strategies for preventing opioid or IDU initiation among adolescents. We evaluated changes in the proportion of adolescents reporting IDU before and after prescription drug monitoring program (PDMP) mandates were implemented in 18 states compared to 29 states without such mandates. Methods and findings This difference-in-differences analysis used biannual Youth Risk Behavioral Surveillance System (YRBSS) data representative of adolescents 17 to 18 years old across 47 states from 1995 to 2017. We compared changes in adolescent IDU in 18 states with and 29 states without PDMP mandates. Among 331,025 adolescents, 51.7% identified as male, 62.1% as non-Hispanic white, 17.4% as non-Hispanic black, and 14.6% as Hispanic. Overall, 3.5% reported IDU during the 2 years prior to PDMP mandates. In the final multivariable difference-in-differences model, we included individual age, sex, and race/ethnicity, as well as state and year as covariates from the YRBSS. We also included state- and year-specific poverty rates based on US Census Bureau data. Additionally, we controlled for state implementation of (non-mandated) PDMPs before states subsequently implemented mandates and pill mill laws. We conducted several sensitivity analyses, including repeating our main analysis using a logistic, rather than linear, model, and with a lead indicator on PDMP mandate implementation, a lag indicator, and alternative policy implementation dates. PDMP mandates were associated with a 1.5 percentage point reduction (95% CI -2.3 to -0.6 percentage points;p =0.001) in adolescent IDU, on average over the years following mandate implementation, a relative reduction of 42.9% (95% CI -65.7% to -17.1%). The association of PDMP mandates with this reduction persisted at least 4 years beyond implementation. Sensitivity analyses were consistent with the main results. Limitations include the multi-stepped causal pathway from PDMP mandate implementation to changes in IDU and the potential for omitted state-level time-varying confounders. Conclusions Our analysis indicated that PDMP mandates were associated with a reduction in adolescent IDU, providing empirical evidence that such mandates may prevent adolescents from initiating IDU. Policymakers might consider PDMP mandates as a potential strategy for preventing adolescent IDU.