Zdravniški Vestnik (Jan 2005)
Meta-analysis in medicine
Abstract
Background: Meta-analysis is a process of using statistical methods to review and combine the results of different, independent clinical studies. Glass first used the term meta-analysis in 1976, when he and his co-worker Mary Lee Smith statistically combined the results of 375 studies that evaluated the efficacy of psychotherapy. Meta-analysis is of particular importance in the assessment of therapeutic efficacy as individual studies do not provide an overview over a topic in its entirety. As their samples are too small, individual studies cannot provide a quantitative evaluation of the effect of treatment, nor can they test the null hypothesis. Prior to meta-analysis, the traditional method was a narrative discourse on previous findings, which, however, could be misleading and subjective. In the past few years, meta-analysis has been increasingly used in all fields of science. This is particularly evident in the medical science, where two other terms are used as well – the systematic review and evidence based medicine. Other methods that have advanced markedly are the decision analysis and cost-effectiveness analysis. All these methods are connected, and the latter two are an upgrade of the first two. Systematic reviews are exact summaries of the best evidence related to exactly specified clinical dilemmas. Special centres, like the Cochrane Collaboration have been organised in different places around the world, where systematic reviews of scientific literature and their own findings are published in databases collecting data on most appropriate therapies of individual illnesses. These reviews support the synthesis of best evidence for treatment or establishment of best medical practices. In this case, meta-analysis has a broader impact and is not just a statistical method for collecting study results.Conclusions: In addition to reviewing and combining results, meta-analysis also offers an upgrade option for discovering and exploring differences in the existing scientific literature in the research field of interest and may provide the most plausible explanations or even result in the discovery of new knowledge. If well performed, meta-analysis can give practical answers to contraversal clinical issues and save costs of additional clinical experiments.
