*2.7. Quality Assessment*

A quality assessment was conducted to determine the rigor of study methods and relevance of the effect measures reported, in the context of the objectives of our study. For the quality assessment, five criteria were evaluated by the two authors (DB and LID) who completed the data extraction. These criteria were adapted from the Cochrane Collaboration Risk of Bias Tool, usually applied to evaluations of quantitative evidence from systematic reviews [18]. The authors of this tool recommended the assessment of five specific categories of bias and one category for other types of bias as needed in accordance with study objectives. Detail on bias categories is provided below; other categories were not

considered given the sufficient coverage provided by those applied. The authors evaluated the criteria on these categories independently by maintaining separate data sheets that were shared only when evaluations were complete. Discordance was discussed and resolved by a third reviewer (JS). Each criterion is defined below and was assigned a designation of, low, high, or unclear quality with point values of −1, 1, and 0, respectively. Criteria were weighted equally because each category of bias was deemed of equal importance to study quality. An overall assessment of study quality was ascertained by the sum of the point values, termed the "cumulative quality assessment", which was used to categorize studies as of high, low, or neutral quality. The unclear designation was assigned when information regarding a criterion were not provided in the article itself and was not possible to ascertain from supplemental materials or earlier publications describing study methods. Quality assessment criteria were as follows:

