Predictive factors on CT imaging for progression of uncomplicated into complicated acute diverticulitis

PURPOSE: Since outpatient treatment and omitting antibiotics for uncomplicated acute colonic diverticulitis have been proven to be safe in the majority of patients, selection of patients that may not be suited for this treatment strategy becomes an important topic. The aim of this study is to identify computed tomography (CT) imaging predictors for a complicated disease course of initially uncomplicated acute diverticulitis. METHODS: CT imaging from a randomized controlled trial (DIABOLO study) of an observational vs. antibiotic treatment strategy of first-episode uncomplicated acute diverticulitis patients was re-evaluated. For each patient that developed complicated diverticulitis within 90 days after randomization, two patients with an uncomplicated disease course were randomly selected. Two abdominal radiologists, blinded for outcomes, independently re-evaluated all CTs. RESULTS: Of the 528 patients in the DIABOLO trial, 16 patients developed complications (abscess > 5 cm, perforation, bowel obstruction) within 90 days after randomization. In the group with a complicated course of initially uncomplicated diverticulitis, more patients with fluid collections (25 vs. 0%; p = 0.009) and a longer inflamed colon segment (86 ± 26 mm vs. 65 ± 21 mm; p = 0.007) were observed compared to an uncomplicated course of disease. Pericolic extraluminal air was no predictive factor. CONCLUSION: Fluid collections and to a lesser extent the length of the inflamed colon segment may serve as predictive factors on initial CT for a complicated disease course in patients with uncomplicated acute colonic diverticulitis. These findings may aid in the selection of patients not suitable for outpatient treatment and treatment without antibiotics.