Background: It is routine practice to perform colonoscopy as a follow-up after an attack of diverticulitis, with the main aim to exclude any underlying malignancy.
Purpose: This study aimed to determine whether colonoscopy is necessary and what additional information is gained from this procedure.
Design: This is a study of a retrospective cohort.
Settings and patients: From January 2003 to June 2009, patients in whom left-sided diverticulitis was diagnosed on CT scan were matched with colonoscopy reports within 1 year from the date of CT by the use of radiology and endoscopy databases. Patients who had colonoscopy within 1 year before the CT scan were excluded. The Western Australian Cancer Registry was cross-referenced to identify patients who subsequently received diagnoses of cancers for whom colonoscopy reports were unavailable.
Main outcome measures: The main outcome measures were the number of patients in whom colorectal cancers were diagnosed and other incidental findings, eg, polyps, colitis, and stricture.
Results: Left-sided diverticulitis was diagnosed in 1088 patients on CT scan, whereas follow-up colonoscopy reports were available for 319 patients. Eighty-two (26%) patients had incidental findings of polyps (9 polyps >1 cm), and 9 patients (2.8%) received diagnoses of colorectal cancers on colonoscopy. After cross-referencing with the cancer registry, the overall prevalence of colorectal cancer among the cohort within 1 year of CT scan was 2.1% (23 cases). The odds of a diagnosis of colorectal cancer were 6.7 times (95% CI 2.4-18.7) in patients with an abscess reported on CT, 4 times (95% CI 1.1-14.9) in patients with local perforation, and 18 times (95% CI 5.1-63.7) in patients with fistula compared with patients with uncomplicated diverticulitis.
Limitations: This study was limited by the unavailability of data for private/interstate hospitals, and the relatively small number of cancer cases reduced the statistical power of the study.
Conclusions: We recommend routine colonoscopy after an attack of presumed left-sided diverticulitis in patients who have not had recent colonic luminal evaluation. The rate of occult carcinoma is substantial in this patient population, in particular, when abscess, local perforation, and fistula are observed.