Meta-Analysis Methods

Pairwise meta-analysis may compare TS-GBR with other bone augmentation techniques if applicable. The minimum clinically important difference was set at 10 percentage points, deliberately conservative relative to the 32 percentage point difference seen in a prior autonomous titration trial. A sample of 300 patients provides greater than 99% statistical power to detect a 10-point absolute difference in normoxemia proportion at α=0.05. Heterogeneity will be measured using the I² statistic, and publication bias will be assessed using funnel plots and Egger’s regression test. Risk factors examined in at least three eligible LMIC studies will be included in meta-analyses. Risk factors reported in fewer than three studies will be narratively described rather than meta-analysed because of limited statistical reliability. Evidence certainty for analyzed outcomes will be assessed using the GRADE framework. Risk of bias will be assessed independently by two reviewers, with disagreements resolved by a third reviewer. The review plans to use a random-effects model when heterogeneity is acceptable, such as I² below 75%. The tool for risk of bias assessment will depend on study design. Mean prim…