Introduction: There is little evidence-based medicine for the therapeutic approach to iatrogenic colonic perforation during colonoscopy. We report a retrospective cohort study comparing minimally invasive and open surgery in the management of colonoscopic perforation.
Methods: Data were extracted from the medical records of patients treated at a single institution from 1990 to 2006. Patients who underwent flexible colonoscopy and were found, by radiological or clinical presentation, to have a large bowel perforation were included. Median and ranges are calculated for the variables. Comparisons are made using Chi-squared and Fischer's Exact tests.
Results: Thirty patients were treated for iatrogenic colonic perforation. Twenty patients had undergone colonoscopy at our hospital for a perforation rate of 0.038%. Ten patients had delayed or no operation. Four patients presented in a delayed fashion to the emergency room 1-2 days after colonscopy. Twenty-five patients underwent operative treatment of their perforation, including resection (n=7), suture repair (n=10), or diversion with or without resection/repair (n=8). Of those 12 were attempted laparoscopically (lap). There were 2 conversions to open. The sites of perforation found at surgery were the cecum (6), ascending (3), descending (1), sigmoid (14), and rectum (1). In the operative cohort, overall morbidity and 30-day mortality was 68 and 8%. There were no deaths in the delayed or no operation cohorts. Morbidity and mortality was 50 and 8%, respectively, for lap repairs. There was no difference between the lap and open cohorts with regard to patient and perforation characteristics. The lap cohort had fewer SSIs and shorter median length of stay compared to the open cohort. Post-operative prophylactic antibiotics were used for a median duration of 7 days (1-15) in the operative cohort. SSIs occurred in 9% of operative patients (1/11) whose initial antibiotics course was less than 7 days compared to 29% (4/14) whose initial course was 7 days or more (p=0.34).
Conclusions: Colon perforation during colonoscopy is associated with high morbidity. Minimally invasive approaches and limited, perioperative antibiotics are appropriate for selected patients after iatrogenic colon perforation.

	Lap n=12	Open n=13	p-value
Age (years)	74.7 (53-84)	72.9 (40-89)	ns
ASA	3 (2-4)	3 (2-5)	ns
Albumin (g/dL)	3.5 (2.5-4.6)	3.9 (2.4-4.4)	ns
Perforated Sigmoid	8	6	ns
Gross spillage/pus	2	4	ns
Length of stay (days)	10.4 (4-21)	23 (8-60)	0.02
Surgical site infections (SSI)	0	5	0.02

ns=not significant