Complication rates of colonoscopy in an Australian teaching hospital environment.

BACKGROUND: Quality assurance is an important part of health-care delivery. With the high level of awareness relating to adverse events from medical care, demonstration of a high standard of practice in gastroenterology is desirable. AIMS: To determine the incidence of significant complications or death within 30 days of an outpatient colonoscopy, and confirm that these are in keeping with international standards.
METHODS: A retrospective audit of linked endoscopy and other hospital databases and selected medical records was carried out, based on reports of 30,463 colonoscopies performed between 5 September 1989 and 31 December 1999 in the three Western Australian public teaching hospitals.
RESULTS: A total of 23,508 colonoscopies was performed on an outpatient basis between 5 September 1989 and 31 December 1999. Post-procedural complications identified (and incidence) were: bleeding episodes 49 (0.21%), colonic perforation 23 (0.1%), abdominal pain 22 (0.09%), and others 19 (0.08%). A total of 196 patients died within 30 days of undergoing colonoscopy (0.83%), although only three deaths were attributable to the procedure itself (incidence 0.01%). Two were inpatients at the time of the procedure (outpatient mortality rate 0.004%). The combined incidence of bleeding and perforation was not significantly different between consultant endoscopists and unassisted trainees (incidence 0.21% vs 0.20%, P=0.98).
CONCLUSIONS: The incidence of bleeding and perforation is similar to other reported series and reflects procedures performed by personnel with a wide range of endoscopic experience. The incidence of complications was not greater for trainees compared with consultant endoscopists. All bleeding episodes and the majority of perforations were associated with a therapeutic intervention. Diagnostic colonoscopy in particular is a very safe procedure.