The path from delay to dropout: A mediation analysis of how treatment settings explain the impact of wait time on treatment completion in substance use disorder programs.

Abstract

To examine whether treatment setting mediates the effect of wait time prior to admission on treatment completion in substance use disorder programs. We analyzed 2022 Treatment Episode Data Set-Discharges (TEDS-D) data from publicly funded substance use disorder treatment facilities across the United States. The analytic sample included 472,476 adult, non-transfer treatment episodes. We employed a mediation analysis using multinomial logistic regression for treatment setting (mediator) and logistic regression for treatment completion (outcome), adjusting for demographic, socioeconomic, referral, and clinical covariates. Natural Direct Effects (NDE) and Natural Indirect Effects (NIE) were estimated following the potential-outcomes framework of Imai, Keele, and Tingley (2010), with 500 bootstrap replications for confidence intervals. State-clustered standard errors accounted for geographic heterogeneity. Longer wait times were associated with significantly lower probabilities of treatment completion. Completion rates declined from 50.5% for same-day admissions to 43.4% for waits exceeding 30 days. Clients who experienced longer delays were disproportionately placed into regular outpatient treatment rather than detoxification or residential settings, which exhibited higher completion rates. The Natural Indirect Effect through treatment setting accounted for approximately 48-52% of the total association between extended waits (≥8 days) and reduced completion. For waits of 1-7 days, mediation through treatment setting was negligible. Administrative delays are associated with reduced treatment completion partly through shifts in treatment-setting allocation. These findings identify an operational pathway-treatment-setting placement-through which system-level delays may translate into differences in completion outcomes. Reducing intake delays and ensuring balanced access across treatment modalities may improve completion rates. These associational findings warrant further investigation using longitudinal or quasi-experimental designs.