Surgical Decision-Making in Colonoscopic Perforation: Predictors and Optimal Timing of Intervention in a 20-Year...

Colonoscopic perforation is a rare but potentially life-threatening complication, and optimal selection of patients for surgical intervention remains challenging. This study aimed to identify independent predictors of the need for surgery and to evaluate the association of endoscopic closure attempts and surgical timing on clinical outcomes in patients with colonoscopic perforation.

We conducted a single-center retrospective cohort study of patients with colonoscopic perforation managed between June 2005 and September 2025 at a tertiary referral center, including patients referred after perforation at outside institutions. A total of 68 patients were included. Clinical and procedural variables were compared between the non-surgery (n = 39) and surgery (n = 29) groups. Technical success was defined as complete endoscopic closure of the perforation defect during the index procedure. Clinical success was defined as absence of persistent leakage, intra-abdominal abscess, generalized peritonitis, or clinical deterioration requiring additional invasive intervention during hospitalization. Multivariable logistic regression was performed to identify independent predictors of the need for surgery. Clinical outcomes were further analyzed according to endoscopic closure attempts, timing of perforation recognition (≤ 6 h vs. > 6 h), and interval from recognition to surgery (≤ 24 h vs. > 24 h).

Of the 68 patients, 29 (42.6%) required surgical intervention. In multivariable analysis, perforation at an outside institution (odds ratio [OR] 11.35; 95% confidence interval [CI] 2.51-74.14) and intraperitoneal perforation (OR 6.73; 95% CI 1.78-34.95) were independently associated with the need for surgery, although confidence intervals were wide, reflecting limited sample size. Diagnostic colonoscopy-related perforations (OR 3.51; 95% CI 0.95-14.01) and calendar era (2005-2015 vs. 2016-2025; OR 4.68; 95% CI 0.90-31.03) were not independently associated with surgical intervention after adjustment. Endoscopic closure was attempted in 46 patients, with technical success achieved in 35 cases (76.1%). The technical success rate was lower for diagnostic perforations (54%) than for therapeutic perforations (85%). The timing of perforation recognition was not associated with surgical requirement or postoperative outcomes. In contrast, among the 29 patients who underwent surgery, primary closure was feasible in 11 of 20 (55.0%) operated within 24 h of recognition, whereas all 9 patients operated after 24 h required bowel resection (p = 0.005).

These findings underscore the importance of appropriate patient selection, timely endoscopic closure attempts when feasible, and prompt surgical intervention without delays when surgery is indicated in colonoscopic perforation.