Reasons for Failure to Diagnose Colorectal Carcinoma at Colonoscopy

BACKGROUND AND STUDY AIMS: Colonoscopy can produce false-negative results, and the reasons for this remain obscure. The aim of this study was to examine why cancers are missed at colonoscopy. PATIENTS AND METHODS: All colonoscopies carried out at Christchurch Hospital, New Zealand, over a 43-month period (1 October 1997 - 30 April 2001) were retrospectively analyzed (the data having been prospectively collected). All cases of colorectal carcinoma during the period 1 October 1997 - 30 July 2001 (3 months longer, to capture delayed diagnoses) were also identified. The two databases were then compared, and all cases in which a colonoscopy had been performed more than 6 weeks before a colorectal carcinoma specimen being received by the pathology department were identified and analyzed. RESULTS: A total of 5055 colonoscopies were undertaken in 4598 patients. Over this period, 630 colorectal carcinomas were identified in the pathology database; 286 of the patients affected were in the colonoscopy and pathology database. Sixty-six patients had had a colonoscopy performed more than 6 weeks before the diagnosis of colorectal carcinoma. Carcinoma was identified in 48 of these 66 patients, and management was being provided. Seventeen cancers (5.9 %) were missed at colonoscopy, and the patients had had an incomplete colonoscopy in nine of these cases. In seven of the 17, an alternative benign cause was recorded. In four patients, a lesion was seen and thought to be benign, although subsequently proven not to be. In another four cases, the cancer was not diagnosed despite adequate bowel preparation and what was thought by the colonoscopist to be an adequate colonoscopy. CONCLUSIONS: Colonoscopy missed 17 of 286 cancers (5.9 %). The reasons why cancers were missed relate to incomplete colonoscopy, poor bowel preparation, misinterpretation of what was seen, failure to carry out adequate biopsy (and follow-up) of lesions seen, and systems failures related to follow-up investigations in patients who had an incomplete colonoscopy. The fact that colonoscopy and barium enema investigations may fail to diagnose cancers has important medicolegal implications. The recognition that colonoscopy may miss a cancer should encourage doctors to reinvestigate patients when there is a lack of correlation between the clinical and investigative findings.