Analysis of lesion localisation at colonoscopy: outcomes from a multi-centre U.K. study

Abstract

Background: Colonoscopy is currently the gold standard for detection of colorectal lesions, but may be limited in anatomically localising lesions. This audit aimed to determine the accuracy of colonoscopy lesion localisation, any subsequent changes in surgical management and any potentially influencing factors.

Methods: Patients undergoing colonoscopy prior to elective curative surgery for colorectal lesion/s were included from 8 registered U.K. sites (2012-2014). Three sets of data were recorded: patient factors (age, sex, BMI, screener vs. symptomatic, previous abdominal surgery); colonoscopy factors (caecal intubation, scope guide used, colonoscopist accreditation) and imaging modality. Lesion localisation was standardised with intra-operative location taken as the gold standard. Changes to surgical management were recorded.

Results: 364 cases were included; majority of lesions were colonic, solitary, malignant and in symptomatic referrals. 82% patients had their lesion/s correctly located at colonoscopy. Pre-operative CT visualised lesion/s in only 73% of cases with a reduction in screening patients (64 vs. 77%; p = 0.008). 5.2% incorrectly located cases at colonoscopy underwent altered surgical management, including conversion to open. Univariate analysis found colonoscopy accreditation, scope guide use, incomplete colonoscopy and previous abdominal surgery significantly influenced lesion localisation. On multi-variate analysis, caecal intubation and scope guide use remained significant (HR 0.35, 0.20-0.60 95% CI and 0.47; 0.25-0.88, respectively).

Conclusion: Lesion localisation at colonoscopy is incorrect in 18% of cases leading to potentially significant surgical management alterations. As part of accreditation, colonoscopists need lesion localisation training and awareness of when inaccuracies can occur.

Keywords: Colonoscopy; Lesion localisation; Multi-centred audit.