Abstract

Objective: to evaluate the quality of diagnostic accuracy studies related to Physiotherapy with the STARD statement and QUADAS-2 tool. Material and method: a research was performed with Pubmed/MEDLINE data base following guidelines by the authors Haynes and Leeflang, and using specific terms of diagnostic accuracy (sensitivity, specificity, diagnostic, predictive value, likelihood, ROC, Physical Therapy) without using filters and its specificity was improved by adding terms used in this kind of study. Finally quality of works was evaluated with the Quality Assessment Tool for Diagnostic Accuracy Studies (QUADAS-2; 2010) and also with the STARD initiative (Standards for Reporting of Diagnostic Accuracy). Results and discussion: a total of 282 articles were located and 257 of them were excluded. The final sample consisted of 25 works. The majority of examined articles, erroneously follow the recommendations proposed in QUADAS-2. In general, observing obtained data, it appears that STARD have been better applied to QUADAS-2. However the lack of a proper methodological application in the results section is evidenced. Conclusions: the items in the STARD guide are used correctly in most of the sections but not in methodology and results exposition items. The analysis with QUADAS-2 tool shows that there is a high risk of bias in most of the articles.