Issues relating to confounding and meta-analysis when including non-randomized studies in systematic reviews on the effects of interventions.

摘要: Background Confounding caused by selection bias is often a key difference between non-randomized studies (NRS) and randomized controlled trials (RCTs) of interventions. Key methodological issues In this third paper the series, we consider issues relating to inclusion NRS in systematic reviews on effects interventions. We discuss whether potential biases from confounding can be accounted for, limitations current methods for attempting do so, different contexts RCTs, problems these create reviewers, research agenda future. Guidance Reviewers who are considering or not include meta-analyses must weigh number factors. Including may allow review address outcomes pragmatic implementations an intervention studied but it will also increase workload team, as well their required technical repertoire. Furthermore, results synthesis involving likely more difficult interpret, less certain, relative only studies. When both evidence available, favor strategy including RCTs same synthesizing separately. Conclusion Including make derived apparent, thereby guiding inferences about generalizability, help with design next generation RCTs. Copyright © 2012 John Wiley & Sons, Ltd.