Evidence based decision and meta-analysis with applications in cancer research studies

Abstract

In the age of evidence based decision making through a systematic review of the literature, statistical meta-analysis has been extensively used to synthesise published summary data on a particular topic of interest from a number of independent studies in order to make credible and scientifically valid conclusions. The main objective is to estimate the common effect size as a pooled statistic for any selected outcome variables from the relevant data. There are several issues concerning the quality and type of the published summary statistics and inherent heterogeneity among the estimates of the effect size across the studies. This paper covers several estimators of the common effect size and some of their major impacts in meta-analysis through redistribution of weights to the individual studies. Some examples from recent literature on cancer research studies are used to illustrate the alternative estimators and discuss their usefulness in analysing data from randomised controlled trials in medicine.