Determining the sample size

Abstract

Determining the adequate sample size for a clinical trial is crucial in the design of an epidemiological study. In fact the question about the number of subjects need to study is common for clinical investigators, because a correct sample size is fundamental to obtain reliable findings. The larger the sample size under study, the greater the chance of detecting, as statistically significant, a clinically important effect it exists. This issue is related to the precision and the power of a study in measuring the difference between treatments being studied, the validity and accuracy of a diagnostic test, the occurrence of a disease. However, conducting a study with an adequate sample size is fundamental not only in statistical terms, but also from an ethical point of view. It is unjustifiable to expose patients to the risks of a research if the study has not the necessary preconditions to obtain findings useful to substantial scientific progress. Calculating sample size depends on several issues, such as the type of sampling method, the type of the study, the desired power and level of confidence fixed for the study. The aim of this article is to summarize the criterions for defining the appropriate sample size and to present some examples of methods for its calculating.(Epidemiology_statistics)