Start Over

Reporting and methodological qualities of published surgical meta-analyses.

Authors :: Zhang H
Han J
Zhu YB
Lau WY
Schwartz ME
Xie GQ
Dai SY
Shen YN
Wu MC
Shen F
Yang T
Source :: Journal of clinical epidemiology [J Clin Epidemiol] 2016 Feb; Vol. 70, pp. 4-16. Date of Electronic Publication: 2015 Jun 24.
Publication Year :: 2016
Abstract: Objectives: To assess the overall qualities of published surgical meta-analysis and predictive factors for high qualities.<br />Study Design and Setting: All meta-analyses pertinent to surgical procedures published in year 2013 were selected from PubMed and EMBASE. The characteristics of the included meta-analyses were collected, and their reporting and methodologic qualities were assessed by the PRISMA (27 items) and AMSTAR (11 items) checklists, respectively. Independent predictive factors associated with these two qualities were evaluated by univariate and multivariate analyses.<br />Results: Hundred ninety-seven meta-analyses representing 10 surgical subspecialties were included. The mean PRISMA and AMSTAR adherences (by items) were 22.2 ± 2.4 and 7.8 ± 1.2, respectively, and a positive linear correlation was found between them with an R(2) of 0.793. Those meta-analyses conducted by the first authors having meta-analysis publication previously had significantly higher reporting and methodologic qualities than those who did not (P = 0.002 and P = 0.001). Meanwhile, there were also significant differences in these two qualities between studies published in Q1-ranked and (Q2 + Q3)-ranked journals as rated by the SCImago indicator (P < 0.001 and P < 0.001). On multivariate analyses, region of origin (non-Asia vs. Asia), publishing experience of first authors (ever vs. never), rank of publishing journals (Q1 vs. Q2 + Q3), and preregistration (presence vs. absence) were independently associated with superior reporting and methodologic qualities.<br />Conclusions: The reporting and methodologic qualities of current surgical meta-analyses remained suboptimal, and first authors' experience and ranking of publishing journals were independently associated with both qualities. Preregistration might be an effective measure to improve the quality of meta-analyses, which deserves more attention from future study conductors.<br /> (Copyright © 2016 Elsevier Inc. All rights reserved.)