Arash Azin, MD, Fady Saleh, MD, MPH, Andrew Yuen, MD, Michelle Cleghorn, MSc, Timothy Jackson, MD, MPH, Allan Okrainec, MDCM, MHPE, Fayez A Quereshy, MD, MBA. University of Toronto
Introduction: Colonoscopy for colorectal cancer (CRC) has been shown to have a localization error rate as high as 21%. Such errors can have substantial clinical consequences especially in laparoscopic surgery. The primary objective of this study is to determine if initial diagnostic colonoscopic examination performed by the operating surgeon compared to non-operating referring endoscopists are associated with lower endoscopic localization errors.
Methods: A retrospective chart review of all patients who underwent elective surgical resection for CRC at a large tertiary academic hospital between January 2006 and August 2014 was performed. The primary exposure of interest was the operator of the initial endoscopy: 1) patients who had their initial endoscopy by the same surgeon who conducted the definitive operation (operating surgeon group), and 2) patients that had their initial endoscopy by a referring gastroenterologist or general surgeon (non-operating endoscopist group). The primary outcome was localization error; categorized as a difference in at least one anatomic segment between initial endoscopy and final operative location. Multivariable logistic regression was used to control for predictors of localization error and to provide an adjusted comparison of odds ratio (OR) of localization error rate between the operating surgeon group and non-operating endoscopist group.
Results: 557 patients were identified, 81 patients had their initial endoscopy completed by the operating surgeon, and 476 patients by a non-operating endoscopist. Initial diagnostic colonoscopy performed by the operating surgeon compared to non-operating endoscopist demonstrated statistically significant lower operative localization error rate (1.2% vs. 9.0%, p=0.016), shorter mean time from endoscopy to OR (52.3 days vs. 76.4 days, p=0.015), higher tattoo localization rate (32.1% vs. 21.0%, p=0.027), and lower repeat endoscopy rate (8.6% vs. 40.8%, p<0.001). Initial endoscopy performed by the operating surgeon was protective against error in both univariate, OR=7.94 [95% Confidence Interval (CI): 1.08-58.52; p=0.016] and multivariate analysis, OR=8.31 (95% CI: 1.12-63.37; p=0.040).
Conclusion: This study demonstrates that diagnostic colonoscopies performed by an operating surgeon are independently associated with a lower localization error rate. Further research exploring the factors influencing localization accuracy and explaining why operating surgeons have lower error rates relative to non-operating endoscopists is necessary to understand differences in care.