The Cochrane 1998 Albumin Review–not all it was cracked up to be

Ref ID	752
First Author	P.J.Horsey
Journal	EUROPEAN JOURNAL OF ANAESTHESIOLOGY
Year Of Publishing	2002
URL	https://pubmed.ncbi.nlm.nih.gov/12463381/
Keywords	Cochrane Harms Author Stakeholder Inference Influence Emergency medicine Team
Problem(s)	Incorrect interpretation or statistical inference error from meta-analysis Lack of clinical expert/ stakeholder/ user perspective
Number of systematic reviews included	1
Summary of Findings	The discussion piece regarding the Cochrane review of albumin administration in critically ill patients published in 1998 reports that none of the authors had the medical expertise or experience necessary to allow appropriate inferences from the included trial data regarding mortality outcomes and definitions. As a result the author argues that the Cochrane review erroneously concluded that albumin administration killed more patients than it saves. "In the seven highest quality trials in the hypovolaemia category the pooled relative risk was 0.87 (CI 0.67–1.14) implying a tendency to benefit from albumin in contrast with the Cochrane review figures of 1.46 (0.97–2.2) implying the opposite."
Did the article find that the problem(s) led to qualitative changes in interpretation of the results?	Not Applicable
Are the methods of the article described in enough detail to replicate the study?