Artificial Intelligence Rabies Post-Exposure Prophylaxis Forecasting Under Variable Vaccine Cold-Chain Break Records: A Conformal Prediction Evaluation of Healthcare Stockout Risk

Rabies post-exposure prophylaxis (PEP) forecasting models are typically evaluated by point-prediction accuracy, yet the operational decision they inform is a stockout alert whose cost depends on whether the forecast interval covers the true demand. Cold-chain break records introduce a structured, record-dependent distortion into the vaccine consumption series that standard accuracy metrics do not penalise and that conformal calibration can absorb. The framework is evaluated analytically rather than empirically: no observational PEP series is used or claimed. A design-based power analysis of the before-after-control-impact estimator that would be used to validate the framework in a field setting was conducted under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64), with 2000 replications per design cell across 8, 16, 24 and 32 clusters and effects of 3, 5 and 8 percentage points. The analysis indicates that 54 clusters are required to detect a 5-percentage-point effect at 80% power. The contribution is a calibration-aware formulation of PEP stockout forecasting that treats cold-chain break records as a conditioning variable rather than as noise, together with a design-level statement of the evidence that a field validation would need to produce.