Public Health Insurance Enrollment Interruptions and Diabetes Medication Adherence in Kenyan County Health Systems: A Target-Trial Emulation of Continuous Coverage Effects

Kenyan county health systems face a persistent challenge: diabetes medication adherence is often interrupted when public health insurance enrollment lapses, yet the causal effect of continuous coverage remains poorly characterised because enrollment is dynamic and confounded by health status. The design treats continuous enrollment as a time-varying treatment and uses the before-after-control-impact estimator to recover the interaction between coverage status and calendar period across county clusters. A design-based power analysis, parameterised by an intraclass correlation of 0.64, between-cluster standard deviation of 6.0 and residual standard deviation of 4.5, evaluated 2000 replications per cell across cluster counts of 8, 16, 24 and 32 and effect sizes 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 protocol specifies identification assumptions, estimand definition, sensitivity analyses for unmeasured confounding and implementation requirements for county health records. The contribution is a replicable design template that connects insurance continuity to chronic-disease outcomes without requiring individual-level randomisation, and that makes the trade-off between cluster numbers and detectable effects explicit for county health managers.