Trauma-Focused Therapy Fidelity Drift and Mental Health Relapse in Malawian District Clinics: A Stepped-Wedge Analysis of Supervision Interval Variation

Supervision interval variation in trauma-focused therapy delivery is widely assumed to influence fidelity and, through fidelity, relapse. The assumption is rarely examined as a design problem: whether a stepped-wedge evaluation of supervision interval variation can recover the fidelity-drift-relapse pathway at the cluster counts available to district health systems is unknown. Power was estimated across 2000 design cells spanning cluster counts of 8, 16, 24, 32 and effect sizes of 3, 5, 8 percentage points. The analysis was designed to characterise the operating characteristics of the design rather than to estimate a treatment effect from observational data. At 80% power, 54 clusters were required to detect a 5 percentage point effect. The intraclass correlation of 0.64 dominated the power profile: at 8 and 16 clusters, power remained below conventional thresholds even at the largest effect size examined, and only at 32 clusters did power approach acceptable levels for the 8 percentage point effect. Minimum detectable effect declined steeply between 8 and 24 clusters and flattened thereafter. The results imply that district-level evaluations of supervision interval variation are underpowered at the cluster counts typically available, and that the fidelity-drift-relapse pathway cannot be reliably identified without either larger cluster counts or a design that reduces the between-cluster variance component.