Patient and Physician Preferences for Antibiotics in Acute Uncomplicated Diverticulitis: A Delphi Consensus Process to Generate Noninferiority Margins.

ABSTRACT

BACKGROUND: Despite the existing evidence, the omission of antibiotics in the management of acute uncomplicated diverticulitis has not gained widespread acceptance. OBJECTIVE: This study aims to incorporate the input of both patients and physicians on the omission of antibiotics in uncomplicated diverticulitis to generate noninferiority margins for 3 outcomes. DESIGN: This was a mixed-methods study, including in-person interviews with patients and a Delphi process with physicians. SETTINGS North American patients and physicians participated. PATIENTS Consecutive patients undergoing colonoscopy, 40% of whom had a previous history of diverticulitis, were selected. INTERVENTIONS: Informational video (for patients) and evidence summaries (for physicians) regarding antibiotics in diverticulitis were reviewed. MAIN OUTCOMES MEASURES: Noninferiority margins were generated for time to reach full recovery, persistent diverticulitis, and progression to complicated diverticulitis in the context of a nonantibiotic strategy. Consensus was defined as an interquartile range <2.5. RESULTS: Fifty patients participated in this study. To avoid antibiotics, patients were willing to accept up to 5.0 (3.0-7.0) days longer to reach full recovery, up to an absolute increase of 4.0% (4.0-6.0) in the risk of developing persistent diverticulitis, and up to an absolute increase of 2.0% (0-3.8) in the risk of progressing to complicated diverticulitis. A total of 55 physicians participated in the Delphi (round 1 response rate = 94.8%; round 2 response rate = 100%). Consensus noninferiority margins were generated for persistent diverticulitis (4.0%, 4.0-5.0) and progression to complicated diverticulitis (3.0%, 2.0-3.0), but could not be generated for time to reach full recovery (5.0 days, 3.5-7.0). LIMITATIONS: Patients were recruited from a single institution, and Delphi participants were invited and not randomly selected. CONCLUSION: Noninferiority margins were generated for 3 important outcomes after the treatment of acute uncomplicated diverticulitis in the context of a nonantibiotic strategy.