Designing the Transition Window: Continuity, Modality, and Sequencing in Opioid Treatment Initiation

When patients experience an opioid use disorder (OUD)-related hospitalization or emergency visit, the transition back to ambulatory care represents a critical window for initiating medication-assisted treatment (MAT). Yet only about one in five patients initiates MAT within 30 days. The authors examine whether timely primary care follow-up facilitates treatment initiation and identify the care delivery mechanisms through which this e!ect operates. Using claims data covering more than 111,000 patients, the authors find that follow-up within seven days increases MAT initiation (IV-adjusted OR = 1.11) and is associated with improved long-term adherence (baseline-adjusted OR = 1.17) and lower relapse risk (baseline-adjusted OR = 0.82), with convergent estimates from propensity score matching and Heckman selection sensitivity. An instrumental variable approach reveals that addressing selection roughly cuts the estimated MAT-initiation lift from 33% to 11% (OR = 1.33 vs. 1.11), underscoring the importance of addressing selection into follow-up care. The authors examine two mechanisms: telehealth moderates the relationship through a substitution effect, functioning as an alternative pathway when in-person access is limited; non-pharmacologic treatments mediate approximately 18% of the follow-up-to-initiation relationship, serving as enablers of pharmacologic uptake. Together, these findings reframe post-acute transitions as a care delivery design problem and identify three operational levers for discharge protocol design: stratifying patients by accessibility profile, routing them to a single care modality rather than layering modalities, and sequencing supportive services as on-ramps to pharmacologic treatment. The asymmetric magnitudes of selection and causal effect imply that current discharge protocols disproportionately reach patients who would initiate treatment regardless of system design, a finding with direct implications for which transition-design interventions can plausibly expand treatment uptake.