Background: Previous research on the quality of systematic reviews (SR) and meta-analyses (MA) showed it was unsatisfactory, mainly due to poor methodological reporting. Thomson Reuters Master Journal List (TRMJL) contains journals assigned Journal Citation Reports Impact Factor. No research on the quality of SR and MA published in Polish journals figuring on this list has yet been reported.
Objectives: To assess the quality of studies published as SR or MA in Polish journals listed on 2016 TRMJL within the field of internal medicine and surgery and to compare it to the quality of such studies published in journals listed in 2011.
Methods: Following a protocol published in PROSPERO we identified Polish journals in the field of surgery and internal medicine. We searched for the studies in those journals using terms 'meta-analysis' and 'systematic review' in an Ovid Medline and Embase and in every journal's official site. The quality of each SR and MA was scored using the AMSTAR checklist (total score from 0 to 11). Two authors assessed the quality of each article independently, any discrepancies were resolved by discussion among all authors.
Results: We identified 101 papers (24-2011; 77-2016). After de-duplication and applying of exclusion criteria, we assessed the quality of 35 included articles (9-2011; 26-2016). Mean AMSTAR score was approx. 25% higher in 2016 than in 2011 (3.73 vs. 3.0). The item most frequently scoring 0 was 'conflict of interest (COI) statements' in 2016, while in 2011 - items related to: 'using the quality of primary studies in formulating conclusions' and 'COI statements'. The largest difference in favour of 2016 vs. 2011 was observed in the item 'combining the findings of studies' (MD=0.4; Ratio of mean quality=2.2). The most similar results were in the item 'inclusion of studies on the basis of publication status' (MD=0.01; Ratio of mean quality=1.04).
Conclusions: Although the quality of studies published as SR and MA in 2016 improved as compared to those published in 2011, it is still unsatisfactory. We highly recommend that journal editors and peer reviewers pay more attention to the methodological quality of accepted SR and MA.