Question special
Chief Resident

It is interesting to think about CODA vs APPAC, and as noted in the paper, CODA's population included patients with "more severe appendicitis" and also included those with appendicolith, who were excluded from APPAC. While the sicker population of CODA may explain the differences in overall outcome, at least at 90 days, I wonder what happens when one compares the population subsets, looking at more similar populations.
For example, incidence of appy in APPAC antibx group was 16% (at 90 days), vs 29% in CODA (at 90 days), or more specifically, 25% when looking at those without appendicolith, which is a better comparison to APPAC. Why do you think there existed this difference in crossover rate between the trials (16% vs. 25%) even when comparing participants w/o appendicoliths, which presumably gets a closer approximation of similar populations? Is that something that can be assessed-- whether that difference is something that can be evaluated and if so, is it statistically significant and what drives that difference?