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 simulation-based tools for calculating statistical power and sample size in interrupted time series (ITS) studies evaluating health policy and environmental interventions with count outcomes. The authors develop ready-to-use computer programmes for two commonly used count data models (Poisson and negative binomial), accounting for autocorrelation and various effect sizes. The work provides practical guidance for investigators designing ITS studies by demonstrating how power varies substantially depending on whether changes in level, trend, or both are being tested.

UK applicability

The statistical methodology presented is directly applicable to UK health policy evaluation and implementation science research. UK researchers evaluating National Health Service interventions, public health policies, or health system changes could use these tools to design adequately powered ITS studies with count outcomes.

Key measures

Statistical power, sample size, segmented autoregressive error models, autocorrelation coefficients (-0.9 to 0.9), level change and trend change parameters

Outcomes reported

The study presents simulation-based methods to calculate statistical power and sample size for interrupted time series (ITS) analyses using count outcomes. It demonstrates application to two statistical models (Poisson and negative binomial) with varying levels of autocorrelation and effect sizes.

Topic tags