Self-controlled case series analyses: Small-sample performance

Second-order expressions for the asymptotic bias and variance of the log relative incidence estimator are derived for the self-controlled case series model in a simplified scenario. The dependence of the bias and variance on factors such as the relative incidence and ratio of risk to observation period are studied. Small-sample performance of the estimator in realistic scenarios is investigated using simulations. It is found that, in scenarios likely to arise in practice, asymptotic methods are valid for numbers of cases in excess of 20-50 depending on the ratio of the risk period to the observation period and on the relative incidence. The application of Monte Carlo methods to self-controlled case series analyses is also discussed.