Short- and long-term outcome evaluation of a university-based opioid overdose education and naloxone distribution (OEND) program

Background The United States opioid overdose epidemic's reach is expanding. Opioid overdose education and naloxone distribution (OEND) programs need to be scaled nationally. College students are a key target for training, as they can respond to overdoses within their campuses, social networks, and communities in which they live, work, and travel, during and after college. University infrastructure could be utilized to scale and sustain OEND programming nationally. This paper addresses the need for more rigorous outcome evaluation to build this programming's evidence base. Methods This paper reports results of a long-term outcome evaluation among undergraduate students trained through a university-based OEND program, including baseline, 2-week, and 6-month follow-up survey data. Broadly informed by the Theory of Planned Behavior, we examined changes over time in seven outcomes related to knowledge, behavioral beliefs, attitudes, perceived behavioral control, and intention. We built a linear mixed-effects regression model for each outcome, analyzing main effects of time and training context (pre-COVID/during-COVID) and their interaction, adjusting for covariates. Results Baseline-to-2-week improvement was maintained at 6 months for five outcomes. We present baseline-to-6-month changes in adjusted marginal means of outcomes scores among pre-COVID and during-COVID trainees, including possible score ranges: opioid/overdose knowledge (time, p < 0.001; pre-COVID: 5.50 → 8.73, during-COVID: 5.68 → 9.10, possible: 0–11); naloxone distribution beliefs (time, p < 0.001; pre-COVID: 4.19 → 4.29, during-COVID: 4.28 → 4.42, possible: 1–5); naloxone distribution policy support (time, p < 0.001; training context, p = 0.028; pre-COVID: 4.60 → 4.81, during-COVID: 4.69 → 4.87, possible: 1–5); comfort distributing naloxone (time, p < 0.001; training context, p = 0.046; pre-COVID: 3.01 → 3.79, during-COVID: 3.17 → 3.86, possible: 1–5); confidence administering naloxone (time/training context interaction, p < 0.001; pre-COVID: 21.61 → 25.16, during-COVID: 22.63 → 24.97, possible: 1–35). Reduction in liability concerns was only maintained at 6 months among pre-COVID trainees (time/training context interaction, p = 0.001; pre-COVID: 2.67 → 2.14, during-COVID: 2.63 → 2.37, possible: 1–5). Intent to engage in key behaviors (measured at 2 weeks and 6 months only) decreased (time, p = 0.006; pre-COVID: 4.33 → 4.22, during-COVID 4.41 → 4.27, possible: 1–5). Conclusion Our findings demonstrate long-term improvement in potential key determinants of carrying and administering naloxone among undergraduates trained through a university-based OEND program. Significant time/training context interactions for two outcomes indicate need for more hands-on training and education on laws/policies protecting those assisting in overdose situations.