Fragility of statistically significant outcomes in randomized trials comparing bariatric surgeries.

Bariatric surgery RCTs are fragile, with only a few patients required to change from non-events to events to reverse the statistical significance of most trials. Future research should examine the use of FI in trial design.