Magnitude and Severity of Colonoscopy Complications

David A. Johnson, MD, FACG, FACP


January 31, 2007

Complications of Colonoscopy in an Integrated Health Care Delivery System

Levin TR, Zhao W, Conell C, et al
Ann Intern Med. 2006;145:880-886

Colonoscopy is the preferred and most effective strategy for colorectal cancer screening. However, the benefit of any procedure must be weighed against the potential risks. This is particularly important as it relates to screening exams, which for the most part are done in otherwise healthy individuals.

There are somewhat limited data on the risks of colonoscopy. Limitations to accurate data recovery may involve the time lag for development of some delayed complications, as well as perhaps reporting bias in terms of how accurately the follow-up is performed and recorded. Another potential bias is that most estimates of colonoscopy complications have come from referral centers or carefully monitored trials conducted by experts.

This retrospective cohort study was conducted via a review of electronic records involving approximately 36,000 colonoscopies performed at Kaiser Permanente, an integrated healthcare delivery system. Cases were excluded if they involved patients younger than 40 years of age, were conducted for clinical signs or symptoms, involved poor preparation, if there was prior colonic surgery, if they involved inpatients, or if interval colonoscopy was performed less than 6 months prior. Hard-copy review of all possible cases of colonoscopy-related complications was performed by 2 physicians, and cases were independently adjudicated by an independent process when necessary. Of the 16,318 eligible colonoscopies (96% performed by board-certified gastroenterologists), the incidence of serious complications was 5/1000 procedures (95% confidence interval [CI]: 4.0-6.2). The 82 cases of serious complications that occurred involved 15 perforations, 6 cases of postpolypectomy syndrome, 53 cases of bleeding requiring hospitalization (15 requiring surgery or transfusion), 38 cases of bleeding requiring inpatient observation, 6 cases of diverticulitis, and 2 unusual complications (1 snare caught in a large polyp requiring surgery and 1 case of diabetic ketoacidosis associated with the colon preparation). There were 10 deaths (0.6/1000) within 30 days of the procedure, but only 1 of these was directly related to colonoscopy (a patient with congestive heart failure and sepsis after a transfusion for postpolypectomy-related bleeding).

Serious complications were associated with colonoscopy with biopsy (P < .0001) with an incidence of 7/1000 colonoscopies (95% CI: 5.6-8.7) compared with an incidence of only 0.9/1000 colonoscopies done without biopsy. Bleeding either post biopsy or post polypectomy occurred in 4.8/1000 procedures (95% CI: 3.6-6.2). The serious complication rate occurred more for larger (> 1 cm) polyps (11.4/1000) than for smaller polyps (6.5/1000), but because most polyps removed were < 1 cm, overall 62% of the bleeding and 40% of all serious complications occurred with removal of the smaller polyps.

The perforation rate reported in this study was 1/1000 procedures, and "serious complications" occurred in 5/1000. Recognizably, these numbers are consistent with those reported in other studies[1,2] and emphasize the need for discussing with each patient the defined risks, potential benefits, and alternatives for each procedure performed. Additionally, it is extremely important that the procedures are performed by individuals who are appropriately trained, certified, and credentialed to provide a quality colonoscopy examination.

This being said, it is important to realize that the procedures in this study were not screening (< 1% of total procedures in this series), and therefore, these complication rates may not be applicable to screening examinations. It is also important to maintain perspective and understand that perforation also occurs with other technologies. In 2 large studies[3,4] involving a cumulative of approximately 12,000 and 17,000 computed tomographic colonography ("virtual colonoscopy") cases, the reported perforation rates were 0.059% (1/1696 studies) and 0.053% (1/1896 studies), respectively.

A perhaps more important take-home message from this study is the importance of recognizing the consequences of removing small lesions that otherwise might not have needed removal. It is recognized by experts that the prime risk in removal of many of these smaller lesions actually derives from the use of cautery. Therefore, many experts have switched to a "cold snare" guillotine-type removal approach for these smaller lesions. Additionally, optical enhancements for imaging smaller lesions with narrow-band imaging, chromoendoscopy, or confocal laser endoscopy should further enhance the ability of the endoscopist to remove only those lesions that are truly neoplastic. These advances in technology and imaging are certain to further enhance the value as well as the safety and efficacy of colorectal cancer screening via colonoscopy.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.