The reporting completeness of observational systematic reviews and meta-analysis in social science could be improved: a cross-sectional survey

Liping Guo,Xin Xing,Junjie Ren,Xinyu Huang,Sarah Miller,Howard White,Kehu Yang
DOI: https://doi.org/10.1016/j.jclinepi.2024.111548
2024-10-09
Abstract:Objectives: To evaluate and compare the reporting quality of observational systematic reviews and meta-analyses (OSRMA) published in Chinese- and English-language social science journals using the Meta-analyses Of Observational Studies in Epidemiology guideline. Study design and setting: We searched the Social Science Core Index via Web of Science and the Chinese Social Science Core Index via China National Knowledge Infrastructure databases for OSRMAs published from 2012 to 2022 (before December 2022). The process of literature search, selection, and data extraction were performed double blind. Results: We identified 634 OSRMAs, of which 349 were in English-language journals and 285 in Chinese-language journals. The mean reporting completeness rate was 54.6%, and reviews in English had higher reporting completeness than those in Chinese (55.2% vs 53.9%). All the reviews reported well on background, method, and conclusion, with average reporting rates of 68.1%, 65.2%, and 88.8%, respectively. However, the reporting of search strategy was ignored, with an average rate of 42.9% for reviews in English and 25.8% for reviews in Chinese, especially on the qualifications of searchers, effort to include all available studies, search software used, handling unpublished studies, and contact with authors. The reporting transparency of these reviews was statistically significantly influenced by multiple review factors, including the language of journals, year of publication, the number of authors, the reporting guideline followed, and the declaration of funding source. Conclusion: OSRMAs in social science demonstrate low reporting quality, especially in Chinese-language journals. We suggest adapting the Meta-analyses Of Observational Studies in Epidemiology guideline to the social science context and promoting its use among researchers and reviewers in this field.
What problem does this paper attempt to address?