Methodological quality assessment tools of non-experimental studies: a systematic review

Abstract

The evaluation of the methodological quality of primary studies in systematic reviews is of great importance in order to guarantee the validity and reliability of their results, but there is no agreement on which tool should be used. Our aim is to analyze the tools proposed so far for the assessment of cohort, case-control, and cross-sectional studies in psychology and health sciences. A systematic review was performed using 5 electronic databases and Google^®. In order to analyze the tools’ content, 6 domains of quality were defined based on reporting guidelines, the established bibliography, and previous similar studies. 74 tools were identified and analyzed. Few reported their reliability (20%) or validity (14%). The most frequently addressed content domains were Data collection (71.6%), Selection (67.6%), Statistics and data analysis (67.6%), and Measurement (58.1%); only 35.1% addressed Representativeness, and 6.8% addressed Funding. Despite the strengths we found scattered among the tools, there is no single obvious choice if we had to make any recommendation. Methodological quality assessment tools of non-experimental studies should meet standardized development criteria, but previously it is necessary to reach an agreement on which content domains they should take into account.