Matt Vassar edited figures/Table 3- 1 page/caption.tex  almost 9 years ago

Commit id: b3e62604f8c0880fadfc3c3cc7e537afc24c7cdc

deletions | additions      

       

We compared components of the commonly used appraisal instruments to one another in Table 3 to determine which domains were most often evaluated across instruments.  A total of 41 quality components were evaluated across all scales (see Table 3). Blinding of both the participants and the assessors was measured by all of the included MQ/ROB tools (100\%; 9/9). The assessment of withdrawals/dropouts was the second most common component addressed by the included tools (77.78\%; 7/9), followed by randomization, which was measured by 66.67\% (6/9) tools. There were multiple components that were evaluated by only 11.11\% (1/9) of MQ/ROB tools (non-response rate, cases and controls comparable on basis of design, benefits worth harm/cost, sufficient Power, were intervention groups recruited over same time period, were intervention groups compliant, data dredging, staff, places, facilities representative of treatment, provided estimates of random variability in data, confounders described, funding/declarations of conflicts of interest, and selection of controls). Tools ranged from evaluating 63.41\% (Downs and Black) to only 14.63\% (Cochrane Risk of Bias Tool) of the listed components. The average number of quality and risk of bias components assessed across all included tools was 32.56\%.