Methodological Quality Analysis of Systematic Review/Meta-analysis Published in Chinese Science Citation Database Journals

  Objective  To evaluate the methodological quality of systematic review/meta analysis (SR/MA) published in Chinese Science Citation Database (CSCD) journals in 2020, with the aim of providing suggestions for researchers in China to develop SR/MA and reference for editors and reviewers to review SR/MA.

  Methods  The list of journals was formed with a sample of journals published by the Chinese Medical Association Publishing House and included in CSCD. Four researchers working in pairs conducted independently manual searches on the official websites of the journals to include SR/MA published in 2020, and one researcher conducted a supplementary search on China National Knowledge Infrastructure. After identifying the included literature, the researchers worked in pairs to extract information independently, evaluated methodological quality by using AMSTAR 2(A Measurement Tool to Assess Systematic Reviews 2) and performed statistical analysis of the results.

  Results  A total of 79 journals from the Chinese Medical Association Publishing House were included in CSCD, 46 of which published SR/MA in 2020, and 126 SR/MA were included for analysis after screening according to inclusion and exclusion criteria. AMSTAR 2 evaluation results showed that the average score of SR/MA was 10.38±1.91. The scored proportion of 9 items was more than 70%. SR/MA met partially/met a minimum of 5 AMSTAR 2 items and a maximum of 12 items. Before the key items were adjusted, over 90% of SR/MA's credibility rating was critically low. After adjusting the key items according to the actual Chinese situation, about 40% of SR/MA's credibility rating was medium and low, respectively. The results of the stratified analysis showed that the use of GRADE, adherence to the PRISMA statement, the number of authors greater than 5, the language of publication, the type of included studies including RCTs, and the reporting of funding were not factors that influenced the methodological quality of SR/MA.

  Conclusions  The methodological quality of SR/MA published in CSCD journals in 2020 need to improve. Attention should be paid to the registration, retrieval, inclusion and exclusion criteria of SR/MA as well as financial support, so as to effectively improve the quality of SR/MA.