Endoscopy 2018; 50(09): 861-870
DOI: 10.1055/a-0584-7138
Original article
© Georg Thieme Verlag KG Stuttgart · New York

Colonoscopic perforations in the English National Health Service Bowel Cancer Screening Programme

Edmund Derbyshire
1  Department of Gastroenterology, Royal Liverpool University Hospital, Liverpool, United Kingdom
,
Pali Hungin
2  School of Medicine, Pharmacy and Health, Durham University, Stockton-on-Tees, United Kingdom
,
Claire Nickerson
3  Cancer Screening Programmes, Public Health England, Sheffield, United Kingdom
,
Matthew D. Rutter
4  Department of Gastroenterology, University Hospital of North Tees, Stockton-on-Tees, United Kingdom
› Author Affiliations
Further Information

Publication History

submitted: 25 July 2017

accepted after revision: 30 January 2018

Publication Date:
28 March 2018 (eFirst)

Abstract

Background Perforation is the most serious adverse event associated with colonoscopy. In this study of data from the English National Health Service Bowel Cancer Screening Programme, we aimed to describe the presentation and management of perforations, and to determine factors associated with poorer outcomes post-perforation.

Methods The medical records of patients with a perforation following the national screening colonoscopy were retrospectively examined. All colonoscopies performed from 02/08/2006 to 13/03/2014 were studied. Bowel Cancer Screening Centres across England were contacted and asked to complete a detailed dataset relating to perforation presentation, management, and outcome.

Results 263 129 colonoscopies were analyzed, and the rate of perforation was 0.06 %. Complete data were reviewed for 117 perforations: 70.1 % of perforations (82/117) occurred during therapeutic colonoscopies; 54.9 % (62/113) of patients with perforations who were admitted to hospital and in whom data were complete underwent surgery; 26.1 % (30/115) of hospitalized patients left the hospital with a stoma and 19.1 % (22/115) developed post-perforation morbidity. Perforations not detected during colonoscopy were significantly more likely to require surgery (P = 0.03). Diagnostic perforations were significantly more likely to require surgery (P = 0.002) and were associated with higher rates of post-perforation morbidity (P = 0.01). At presentation, the presence of abdominal pain (P = 0.01), a pulse rate > 100 beats per minute (P = 0.049), and a respiratory rate > 20 breaths per minute (P = 0.01) were significantly associated with the patient having surgery.

Conclusions This is the largest retrospective observational case series in Europe to describe post-perforation presentation, management, and outcomes. We have confirmed that perforation leads to surgical intervention, stoma formation, and post-perforation morbidity. Perforations not recognized during the colonoscopy were significantly more likely to require surgery. Diagnostic perforations were at greater risk of requiring surgery and developing post-perforation morbidity.