Colonoscopic Perforations: A Retrospective Review

• Published in: Journal of gastrointestinal surgery 2005-12, Vol.9 (9), p.1229-1236
• Main Authors: Iqbal, Corey W., Chun, Yun Shin, Farley, David R.
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Colon; colon injury; colon trauma; colonoscopic perforation; colonoscopy; Colonoscopy - adverse effects; endoscopy; Female; Humans; Intestinal Perforation - diagnosis; Intestinal Perforation - epidemiology; Intestinal Perforation - etiology; Intestinal Perforation - therapy; Male; Middle Aged; Mortality; Retrospective Studies
• ISSN: 1091-255X; 1873-4626
• DOI: 10.1016/j.gassur.2005.06.023

Colonic perforation is no longer a rare complication of colonoscopy. Our previous report identified 45 such iatrogenic injuries from 1980 through 1994 (3082 colonoscopies per year). This follow-up of the ensuing 7 years examines changing trends of endoscopic usage in addition to management and prognosis of patients with colonoscopic perforations. Retrospective analysis of 78,702 colonoscopies (1994 through 2000, 11,243 colonoscopies per year) allowed assessment of medical records in all patients treated at our institution for colonic perforation. Sixty-six patients from our institution (perforation rate, 0.084%; 1 per 1192 procedures) and six patients from outside institutions were treated for colonic perforation following colonoscopy (41 women, 31 men; ages, 30–92 years; median, 73 years). Sixty-two patients underwent laparotomy, while 10 were managed nonoperatively. All 10 patients managed nonoperatively were void of peritoneal irritation by physical examination; eight patients did well (median hospital stay, 5.5 days; range, 0–12), but one death (family declined operative intervention) and one pelvic abscess requiring percutaneous drainage were noted. Peritoneal irritation by physical examination was evident in 57 of 62 patients undergoing laparotomy. Perforations occurred throughout the colon: right, 22 (31%); transverse, 5 (7%); left, 44 (61%); and unknown, 1 (1%). Thirty-eight patients (61%) underwent primary repair or resection with anastomosis. Fecal diversion was used in 100% of patients with extensive peritoneal contamination (n = 12) and 40% of patients with moderate contamination (12 of 30). Perioperative morbidity (39%) and mortality (8%) were significant. Factors predicting a poor outcome included delayed diagnosis, extensive peritoneal contamination, and patients using anticoagulants ( P < .05). Compared with our prior study, the present review highlights a higher prevalence of injury based on more frequent use of colonoscopy. Perforation rates remain around 0.08%. While nonoperative management is viable in patients void of peritonitis, expedient surgical intervention seems to facilitate patient recovery.