Artificial Intelligence Sepsis Alerting Under Variable Vital-Sign Monitor Probe Reuse: A Drift-Aware Calibration Evaluation of Healthcare Early Warning Reliability

Sepsis alerting algorithms that consume vital-sign streams inherit a measurement problem that is rarely modelled: reusable monitor probes degrade between calibrations, so the input distribution drifts while the alert threshold stays fixed. the present analysis treats probe-induced drift as a structured perturbation of the vital-sign signal and asks how much of the nominal alert reliability survives it. A drift-aware calibration framework is developed in which probe reuse is represented as a monotone attenuation of signal amplitude plus an additive bias, and the resulting shift in the alerting score is propagated analytically to the operating characteristics of a threshold rule. The framework yields a closed-form expression for the drift-induced change in sensitivity and a calibration condition under which the false-alarm rate remains bounded. A design-based evaluation of the before-after-control-impact estimator was conducted under a declared variance structure with between-cluster SD 6.0, residual SD 4.5 and intraclass correlation 0.64; power was estimated across 2000 design cells at cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points. The analysis indicates that 56 clusters are required for 80% power at a 5-percentage-point effect, and that the intraclass correlation of 0.64 dominates the power budget. The contribution is a calibration account that makes probe reuse an explicit parameter of alert reliability rather than an unmodelled nuisance.