Medicaid expansion is not associated with prescription opioid and benzodiazepine misuse among people who inject drugs: A serial cross-sectional observational study using generalized difference-in-differences models

Abstract

Background

While evidence suggests Medicaid expansion can reduce overdose, some expressed concern expansion fueled the US opioid overdose crisis by increasing access to low-cost prescription opioids diverted for non-prescribed use. Ecologic studies find a protective relationship or no relationship between expansion and area-level opioid prescribing. Little is known about the relationship between expansion and opioid use among people experiencing poverty who inject drugs (PWID), a population at heightened risk of overdose likely to benefit from Medicaid expansion. We examined whether expansion was associated with prescription opioid and benzodiazepine misuse among PWID experiencing poverty and whether associations varied by race/ethnicity and HIV status.

Methods

This serial cross-sectional observational study used generalized difference-in-differences models to analyze data (2012, 2015, 2018) from 19,728 PWID aged 18–64 with income ≤138 % of federal poverty line from 13 states in the CDC's National HIV Behavioral Surveillance. Outcomes included past 12-month non-injection and injection prescription opioid misuse and benzodiazepine misuse.

Results

The sample (N = 19,728) was 40 % non-Latinx Black persons and 22 % Latinx persons. Past 12-month non-injection prescription opioid misuse was 33 %, injection prescription opioid misuse was 16 %, and benzodiazepine use was 40 %. Across all models, there was no association between expansion and prescription opioid misuse (confidence intervals included 0) or prescription benzodiazepine misuse (confidence intervals included 0). Associations did not vary by race/ethnicity or HIV status.

Conclusions

We found no association between Medicaid expansion and opioid or benzodiazepine misuse overall, by race/ethnicity, or HIV status among a large, geographically diverse sample of PWID. These findings provide empirical evidence that expansion is not associated with prescription opioid or benzodiazepine misuse in a population likely to benefit from expansion.