Meta-analysis of non-randomized studies in interventional cardiology: a critical appraisal

Abstract

Utilisation of meta-analysis is becoming more and more common in interventional cardiology. The aim of this statistical approach is to collect a large number of patients from randomized clinical studies and nonrandomized registries in order to obtain a pooled estimate of the results. Nevertheless, simply pooling these results without a correct methodological approach can easily lead to biased conclusions. In this report we analyse the possible methodological drawbacks of such an approach and we suggest a simplified check-list of items to be considered in the effort of building-up a meta-analysis from non-randomized studies.