Principle of sample size calculation in gastroenterology research: a practical guide for clinicians
ABSTRACT
Introduction
A proper sample size calculation enables to conduct adequately powered randomized controlled trials (RCTs) and to provide a valid assessment of a specific clinical question.
Areas covered
In the current manuscript, we tried to provide the reader with an easy guide on the principles of sample size calculation in RCTs, tailored specifically to the context of gastroenterology and hepatology. The basics of sample size calculation were commented with a description of some of the main methods, including the calculation of the non-inferiority margin for non-inferiority RCTs and the calculation of the minimum clinically important difference (MCID). Some examples from the gastroenterology literature were also provided.
Expert opinion
Collaborating with a biostatistician can provide valuable insights into the nuances of sample size calculation and study design. However, it is crucial that the clinicians understand the basics of calculating sample size, so they could provide valuable input in designing the study from a clinical point of view.
Article highlights
-
A proper sample size calculation leads to adequately powered RCTs and to clinically meaningful results for the patients.
-
Generally, sample size calculation depends on several variables, including effect size, power and alpha level.
-
The MCID can determine which observed difference correlates with a clinically important improvement (or worsening) for the patients.
-
Non-inferiority RCTs should take into account a non-inferiority margin that could be based on the literature or calculated through statistical methods.
-
Meta-analyses or trial sequential analyses could represent valuable options to base the sample size calculation for future RCTs.
