Letter commenting on: Outcomes of multivessel vs culprit lesion‐only percutaneous coronary intervention in patients with acute myocardial infarction complicated by cardiogenic shock: Evidence from an updated meta‐analysis (Kundu et al. Catheter Cardiovasc Interv. 2018 Dec 28. [epub ahead of print])

Abstract

To the Editor: We read with interest the recently published meta-analysis by Kundu et al which addresses outcomes of multivessel vs culprit lesion-only percutaneous coronary intervention in patients with infarct-related cardiogenic shock. We want to stress a serious design flaw of the meta-analysis not mentioned in the manuscript. The analysis combines evidence from one randomized study (CULPRIT-SHOCK) and several non-randomized studies. The Cochrane Collaboration explicitly recommends against such practice as the robustness of data from the two designs cannot be considered equal. Only in case randomized trials are not available should a metaanalysis of observational studies be considered. This is especially true in clinical fields such as cardiogenic shock where a high level of selection and reporting bias in non-randomized trials is likely.We believe that adherence to these scientific standards is a “conditio sine qua non.” Therefore, clinical conclusions drawn from the presentedmeta-analysis warrant caution.