Simulation as a practical and versatile tool for sample size calculation

Richard Hooper

doi:10.1186/1745-6215-14-s1-o99

Abstract

Background The ethical significance of sample size is enshrined in principles of Good Clinical Practice and in quality standards for reporting trial results, yet sample size is often one of the most ill-considered aspects of a grant application. Researchers are also battling with an increasing complexity in study design. With some adaptive trial designs, for example, analytical solutions to sample size calculation problems may simply be unattainable. Monte Carlo simulation offers a general alternative. Methods A versatile tool using simulation to calculate sample size for given statistical power was developed as an add-on to a general statistics software package. This tool allows determination of sample size for any analysis under any statistical model that can be programmed in the host package, and uses a novel algorithm that is more efficient than previously published alternatives. Results

Highlights

The ethical significance of sample size is enshrined in principles of Good Clinical Practice and in quality standards for reporting trial results, yet sample size is often one of the most ill-considered aspects of a grant application
A versatile tool using simulation to calculate sample size for given statistical power was developed as an add-on to a general statistics software package
The software can cope with non-parametric tests, multilevel models, multiple testing, group sequential methods, adaptive trial designs, and more

Summary

Introduction

The ethical significance of sample size is enshrined in principles of Good Clinical Practice and in quality standards for reporting trial results, yet sample size is often one of the most ill-considered aspects of a grant application. Simulation as a practical and versatile tool for sample size calculation Richard Hooper From 2nd Clinical Trials Methodology Conference: Methodology Matters Edinburgh, UK. Background The ethical significance of sample size is enshrined in principles of Good Clinical Practice and in quality standards for reporting trial results, yet sample size is often one of the most ill-considered aspects of a grant application.

Results

Conclusion