ABSTRACT
This chapter begins by discussing some of the key choices that need to be made when designing a self-controlled case series (SCCS) study. These include the primary time line, risk periods, observation periods, and age groups. The chapter discusses a special SCCS design sometimes called the self-controlled risk interval design. It devotes to sample size and power calculations. The R package SCCS may be used to obtain nominal sample sizes with and without age effects, and to undertake simulations in more realistic scenarios. The chapter also discusses the asymptotic efficiency and identifiability of the SCCS design. It comparees the efficiencies of the SCCS method, the cohort method and the 1:1 matched case-control method. The chapter describes how SCCS design features may impact upon its efficiency. It shows how to use the R package SCCS to obtain event counts and person-time, and discusses the presentation of results tables.
