Sample Size Calculation In Medical Research

Abstract

For the reliability and general ability of the results of any epidemiological or medical research study, adequate number of study units is extremely important. Study of both less than needed and unnecessarily large sample size is ethically and economically unfair. Considering the importance of sample size, its determination and justification in all kinds of epidemiological or medical research studies, we propose simple and non-technical explanation to the available sample size calculation methods. In this review we covered most frequently encountered study designs in medical research. Present review provides statistical formulas for the sample size calculation of the research studies with varying objectives such as estimation of population proportion(s) and hypothesis testing, estimation of population mean, estimation of diagnostic accuracy, estimation of association and reliability, sample size calculation for case control studies, and sample size calculation for cohort studies. Limited technical details and explanations of underlying theoretical assumption for each method have been included to ensure the adaptability of the method with minimal theoretical understating and statistical knowledge.