Predictors of Retention in a Clinical Trial of a Fentanyl Overdose Prevention Intervention: Implications for...

Structural barriers can pose challenges to retaining people who use drugs in longitudinal research. This study assessed baseline characteristics associated with retention in a year-long clinical trial addressing whether fentanyl test strip provision decreased rates of nonfatal overdose among people who use drugs.

From September 2020 to February 2023, 505 participants enrolled in the year-long RAPIDS clinical trial. Outcome visits were completed at 6 and 12 months following enrollment. Bivariable and multivariable analyses were conducted to assess baseline characteristics associated with the completion of none, one, or both study outcome assessments. We used multinomial logistic regression to identify statistically significant baseline predictors of retention in the RAPIDS trial.

Among eligible participants, 41.6% (n = 210) only completed a baseline assessment, 23.4% (n = 118) completed one outcome visit, and 35.0% (n = 177) completed both. In the final multinomial logistic regression model, women were more likely to complete both outcome visits compared to men (adjusted odds ratio [AOR] = 1.69, 95% CI: 1.08-2.62), while participants with a history of drug selling were less likely to complete two visits (AOR = 0.62, 95% CI: 0.39-0.98).

Study results challenge stigmatizing beliefs about the impact of sociodemographic and drug use-related characteristics on retention in longitudinal research. Retention strategies that target men and people who sell drugs are needed. Investigators can collect multiple pieces of contact information, have monthly check-ins with participants, and partner with community organizations to make research more accessible and acceptable to structurally vulnerable populations.