Community Mental Health Service Withdrawal After Involuntary Outpatient Commitment Expiry and Re-Engagement in South African Districts: A Regression-Discontinuity Estimation of Coercive-Leverage Exposure

Involuntary outpatient commitment (IOC) in South Africa creates a time-limited legal lever that compels community mental health service attendance, yet the moment of expiry is poorly understood as a discontinuity in care engagement. The design exploits the sharp statutory boundary at commitment expiry, comparing patients whose orders lapse just before versus just after a district-level administrative cut-off, with running variable defined as days from expiry to the next scheduled review. The paper specifies the estimand, the local randomisation assumption, the bandwidth and polynomial choices, the McCrary density test, covariate balance checks, and the handling of partial compliance and district-level clustering. It further proposes a companion analysis of the DHS Program indicator 'Believes fertile period is after period ended' (N = 154 country-survey observations, 38 African countries, 1986–2025; mean 27.756, SD 9.876) as a benchmark for the kind of population-level measurement error that district health information systems are likely to reproduce. The contribution is a pre-specified, falsifiable design that separates the legal status of coercion from the clinical content of care, and that can be executed with routine district data without inventing outcomes. The protocol does not report empirical estimates; it reports the conditions under which such estimates would be credible.