Community Mental Health Service Same-Day Antipsychotic Initiation Policy and Thirty-Day Readmission in South African District Clinics: A Regression-Discontinuity Estimation of Post-Discharge Timing Exposure

Delayed antipsychotic initiation after psychiatric discharge is associated with elevated readmission risk, yet the causal effect of same-day initiation policies remains poorly identified in resource-constrained settings. The identification strategy exploits a policy discontinuity at a defined discharge time threshold, comparing patients discharged just before and just after the cutoff. The design addresses confounding by indication and unobserved severity through local randomisation near the threshold. The proposed analysis uses routine administrative data from district clinics, with readmission within thirty days as the primary outcome. The framework specifies power calculations, bandwidth selection, and robustness checks including donut-hole and covariate-balance tests. The contribution is methodological: it provides a template for evaluating timing-based mental health policies in low-resource settings where randomised trials are infeasible. The paper concludes that regression-discontinuity designs can yield credible estimates of post-discharge timing effects, informing community mental health service delivery in South Africa and comparable contexts.