Quality Assessment of Cochrane and Non-Cochrane Systematic Reviews in Neonatal Medicine: A Meta-Analysis.

Abstract

To compare the methodological and reporting performance of Cochrane reviews (CR) and non-Cochrane reviews (NCR) of clinical trials evaluating neonatal interventions. PubMed, Embase, the Cochrane database, and Web of Science were searched from January 2015 to December 2023 for pairwise meta-analyses and systematic reviews. Review quality was assessed using A Measurement Tool to Assess Systematic Reviews-2 and compared across covariate-matched cohorts of CR and NCR. Trends in quality over time and across journal quartiles were examined using the Jonckheere Terpstra test. A total of 453 systematic reviews were included (202 CR and 251 NCR). A significantly greater number of NCR were rated as critically low quality compared with CR (71% vs 8%; P < .001); significantly fewer NCR were rated high or moderate quality (5.5% vs 13%; P = .007). The performance of CR was significantly higher than NCR for review protocol establishment (100% vs 49%), comprehensiveness of literature search (100% vs 57%), justification of exclusions (99% vs 38%), and reporting of trial funding (23% vs 5%). Reporting of review protocol, literature search, study exclusions, and sensitivity analyses improved across NCR categories ordered by increasing impact factor (P < .01 for trend). Over the 9-year period, a positive trend was observed in protocol reporting of NCR (from 23% to 71%) and reporting of trial funding in CR (from 9% to 72%). The mmethodological quality of neonatal CR was significantly better than NCR. NCR published in high-impact journals were superior to those in low-impact journals. Methodology and reporting of both review categories improved consistently over a 9-year time frame.