Methodological quality of systematic reviews on treatments for osteoporosis: A cross-sectional study

Purpose: Systematic reviews (SRs) provide the best evidence on the effectiveness of treatment strategies for osteoporosis. Carefully conducted SRs provide high-quality evidence for supporting decision-making, but the trustworthiness of conclusions can be hampered by limitation in rigor. We aimed to appraise the methodological quality of a representative sample of SRs on osteoporosis treatments in a cross-sectional study. Methods: Cochrane Database of Systematic Reviews, EMBASE, MEDLINE, and PsycINFO were searched for SRs on osteoporotic treatments. AMSTAR (A MeaSurement Tool to Assess systematic Reviews) 2 was used to evaluate methodological quality of SRs. Associations between bibliographical characteristics and methodological quality ratings were explored using multivariate regression analyses. Results: A total of 101 SRs were appraised. Overall, one (1.0%) was rated "high quality", three (3.0%) were rated "moderate quality", eleven (10.9%) were rated "low quality", and eighty-six (85.1%) were rated "critically low quality". Ninety-nine (98.0%) did not explain study design selection, eighty-five (84.2%) did not provide a list of excluded studies (84.2%), and eighty-five (84.2%) did not report funding sources of included studies. SRs published in 2018 or after were associated with higher overall quality [adjusted odds ratio (AOR): 5.48; 95% confidence interval (CI): 1.12-26.89], while SRs focused on pharmacological interventions were associated with lower overall quality [AOR: 0.24; 95% CI: 0.06-0.96]. Conclusion: The methodological quality of the included SRs is far from satisfactory. Future reviewers must strengthen rigor by improving literature search comprehensiveness, registering and publishing a priori protocols, and optimising study selection and data extraction. Better transparency in reporting conflicts of interest among reviewers, as well as sources of funding among included primary studies, are also needed.