Simulation-based power and sample size calculation for designing interrupted time series analyses of count outcomes in evaluation of health policy interventions

Summary

This methodological paper presents a simulation-based approach for calculating power and sample size in interrupted time series studies evaluating health policy and environmental interventions using count outcomes. The authors developed ready-to-use computational tools for two common count models (Poisson and negative binomial), demonstrating that power varies substantially depending on whether the intervention effect manifests as a level change, trend change, or both. The work provides practical guidance for investigators designing robust ITS studies in health services and policy evaluation.

UK applicability

The statistical methods and software tools described are broadly applicable to UK health policy evaluation studies, particularly those assessing implementation outcomes, service utilisation, or count-based endpoints in the NHS and public health domains. The methodology is agnostic to geography and could support study planning for UK-based interrupted time series evaluations.

Key measures

Statistical power, sample size, autoregressive error models, segmented regression parameters (level change, trend change), Poisson and negative binomial model performance

Outcomes reported

The study developed and validated simulation-based methods to calculate statistical power and sample size for interrupted time series (ITS) analyses of count outcomes in health policy evaluations. Power calculations were estimated for both Poisson and negative binomial models across varying autocorrelation structures and effect sizes.

Topic tags