Endoscopic perforation rates at a Canadian university teaching hospital.

BACKGROUND: Despite advances in training, operative techniques and endoscopic technology, upper and lower endoscopic procedures continue to have potential for intestinal perforation. Perforation rates provided to patients at the time of consent have frequently been derived from historical cohorts and survey datasets. OBJECTIVE: This study examined the perforation rates of upper and lower endoscopic procedures at a major Canadian tertiary care centre. METHODS: Inpatient and outpatient gastroscopies and colonoscopies performed during a three year period were evaluated. Endoscopies with perforations occurring within 14 days of procedure were retrospectively isolated using the International Classification of Diseases - 9th Revision code descriptions, then retrieved and hand searched to confirm a procedure-related perforation. Data were extracted to identify risk factors and patient outcomes. RESULTS: A total of 21,217 endoscopies (13,792 gastroscopies and 7425 colonoscopies) were reviewed. Of these, 359 were identified, isolated and hand searched for confirmation of a perforation event. Eighteen were found to have an endoscopy-associated perforation. Ten perforations occurred with colonoscopy (0.13%) (incidence, 1.3/1000 procedures), resulting in one death (0.013%) (incidence, 0.13/1000 procedures). Eight perforations occurred with gastroscopy (0.06%) (incidence, 0.6/1000 procedures), resulting in zero mortality. Of colonoscopy procedures the rate of perforation with diagnostic colonoscopy was 0.13% (incidence, 1.3/1000 procedures) and with therapeutic colonoscopy was 0.14% (incidence, 1.4/1000 procedures). Of gastroscopy procedures the rate with therapeutic gastroscopy was 0.15% (incidence, 1.5/1000 procedures). No perforations occurred with diagnostic gastroscopy. CONCLUSION: Gastroscopy and colonoscopy procedures, especially those with therapeutic maneuvers, continue to carry morbidity and mortality risks associated with perforation.