Complex And Simple Appendicitis: REstrictive or Liberal postoperative Antibiotic eXposure (CASA RELAX) using Desirability of Outcome Ranking (DOOR) and Response Adjusted for Duration of Antibiotic Risk (RADAR): a multicenter Bayesian randomized controlled trial

Authors:
D. Dante Yeh, Gabrielle Hatton, Claudia Pedroza, Rafael Torres Fajardo, Sean Thomas Dieffenbaugher, Gerd Daniel Pust, Luciana Tito Bustillos, Diedra Turnacliff, Erin Fitzgerald, Jesse Victory, Lucy Z. Kornblith, Caitlin Collins, Genna Beattie, William G. Cheadle, Nicholas Caminiti, Alonso Andrade, Susan F. McLean, Andrew Bernard, Matthew Ray, Lillian Kao

Body of Abstract:
Background: The optimal duration of antibiotics (abx) after appendectomy for simple or complicated (gangrenous or perforated) appendicitis is unknown. We performed a randomized trial to compare a restrictive to a liberal antibiotic strategy on a composite, patient-centered outcome, the Desirability of Outcome Ranking (DOOR).

Methods: Adults with appendicitis undergoing appendectomy at 9 sites were randomized 1:1 to either a Restrictive (no postop abx for simple, up to 1 day of postop abx for complicated) or Liberal (up to 1 day of postop abx for simple, 4 days of postop abx for complicated) strategy. The primary endpoint was based on DOOR, an ordinal scale of mutually exclusive clinical complications with within-category rankings determined by duration of antibiotic exposure.

Randomization was stratified by age >65 and site. Bayesian cumulative logistic models using a neutral prior were used to assess the probability of benefit with a Restrictive strategy. Posterior probabilities of benefit of Restrictive strategy were generated. Type of appendicitis (simple vs complicated) was assessed as an interaction.

Results: A total of 346 subjects (182 Restrictive, 164 Liberal) enrolled from 9 sites were included in the final analysis. Baseline demographics between groups were similar. There were 265 simple and 81 complicated (11 gangrenous) appendicitis cases. The DOOR category outcomes were similar between groups (Figure). The majority experienced the best possible outcome, DOOR 1 (Cure; no adverse effects), and no patient experienced death (DOOR 7). The Restrictive strategy had a cumulative odds ratio of 0.84 (95% CrI 0.42-1.72) for each sequential DOOR category. This correlated with a 68% probability that Restrictive strategy reduces (improves) DOOR category. However, this finding was driven by patients with complicated appendicitis, resulting in a DOOR OR 0.69 (95% CrI 0.27-1.68), which correlates with a 79% probability that the Restrictive strategy reduces DOOR. In contrast, the Restrictive strategy with simple appendicitis resulted in a DOOR OR 1.17 (95% CrI 0.47-3.0), which correlates with a 37% probability that the Restrictive strategy reduces DOOR.

Conclusions: A Restrictive postoperative antibiotic strategy in acute appendicitis resulted in similar to mildly improved clinical outcomes, when compared to a Liberal strategy overall. The patients who benefitted most from a Restrictive strategy were patients with complicated appendicitis. Given the well-established risks of prolonged antibiotic exposure in the absence of clinical benefits, we suggest that a restrictive postoperative antibiotic strategy be adopted for patients with complicated appendicitis.