ABSTRACT
Chapters 20 and 21 describe sample size calculations for the number of events that are independent of the anticipated event rates in the trial. If these results were applied, then the study would recruit until a specified number of events have been observed. However, for planning purposes, budgets and timescales, an estimate of the total sample size would also be required. This chapter discusses methods to estimate the total sample size to ensure there are a sufficient number of events in the trial.
The chapter highlights how the calculations are not immediately straightforward and are quite sensitive to the assumptions such that even a rounding error – three decimals places compared to four decimals places – can impact the results. However, it is shown how simple tables and figures can enhance the sample size estimation and assist in the interpretation of calculations as well as facilitate decisions as to the duration of follow-up to optimise sample size.
