Colorectal cancer (CRC) is the 3rd most common cause of cancer death. Colonoscopy is the ‘gold standard’ for CRC screening and early diagnosis but still sometimes misses lesions. Our aim was to identify potential factors that might influence CRC miss rates on colonoscopy.Methods
Data on all new CRC diagnoses during a 3 year period (1 st September 2014 – 31 st August 2017) were collected retrospectively from the electronic records of a UK teaching hospital. A CRC miss was defined as a case where CRC was missed on ‘initial’ colonoscopy and identified within 3 years upon subsequent ‘diagnostic’ colonoscopy. We analysed demographic data, and features of the ‘initial’ and ‘diagnostic’ colonoscopies.Results
691 new CRC cases were identified during the 3 year period. 12 of these were CRC misses (1.74%). Missed CRC cases had an average age of 65.25 years (29–80) and M:F ratio of 1:3. All 12 cases were adenocarcinomas with various histopathological and endoscopic features (table 1). ‘Initial’ colonoscopy identified polyps in 8 cases. In 4/8 (50%) cases the polyp location was identical to the cancer later identified on ‘diagnostic’ colonoscopy.Results
Interestingly, adjusting for volume of colonoscopies performed by each endoscopist type, trainees and surgeons were 2.5 and 5 times respectively, more likely to miss CRC than physician endoscopists (table 2). Finally, 4 out of 12 (33.33%) cases had the ‘initial’ colonoscopy that missed CRC on a Saturday morning. This translates to a 4-fold increased risk of missing CRC on a Saturday morning service list compared to any weekday morning (n=4) or afternoon (n=4) session.Conclusions
Despite a small missed CRC cohort, our results suggest that risk of missing CRC on colonoscopy might be higher in female patients and the miss rate might be linked to scope operator with trainee and surgical endoscopists perhaps more likely to miss CRC than physician endoscopists. It was common for ‘initial’ colonoscopy to identify a polyp which in ½ of cases was at the same anatomical site as the cancer identified later. The risk of colonoscopy missing CRC was higher on a Saturday service list than a normal working day session. Overall, our findings suggest discrepancies that warrant further investigation in larger more cohorts with more statistical power in order to improve CRC screening.