Inadequate critical appraisal of studies in systematic reviews of time to diagnosis.

Abstract Objective To analyze tools used to critically appraise primary studies included in systematic reviews (SRs) of time to diagnosis (TTD). Study Design and Setting We systematically searched MEDLINE via PubMed and Web of Science for SRs of TTD published up to the end of February 2015; we identified and characterized tools used for critical appraisal and classified their items. Results From 1,936 articles identified, we included 45 SRs that aimed to summarize the available information on the length (n = 16), determinants (n = 31), and/or consequences (n = 14) of TTD. For the 23 SRs (51%) reporting a critical appraisal process, 21 different tools were used, with 232 items assessing quality of reporting (64%), risk of bias or threats to generalizability (43%), statistical issues (5%), and/or an unclear domain (0.5%); 11% were specific to TTD issues. Overall, 36% of the 45 SRs assessed risk of bias and/or threats to generalizability. Conclusion Assessment of risk of bias and threats to generalizability in primary studies included in SRs of TTD is infrequent, nonstandardized and rarely concerns TTD study specificities. These findings highlight the need for guidance on critical appraisal of studies of TTD.