Increased odds of interval left-sided colorectal cancer after flexible sigmoidoscopy compared with colonoscopy in older patients in the United States: A population-based analysis of the SEER-medicare linked database, 2001-2005

Objectives: To compare the proportion of interval left-sided colorectal cancers (CRCs) after flexible sigmoidoscopy vs colonoscopy in older patients and to identify factors associated with interval CRC. Patients and Methods: Using the Surveillance, Epidemiology, and End ResultseMedicareelinked database, we studied patients 67 years or older with left-sided CRC who had at least one lower endoscopy performed within the previous 36 months between July 1, 2001, and December 31, 2005. The CRCs diagnosed within 6 months of lower endoscopy were defined as detected CRCs; CRCs diagnosed 6 to 36 months after lower endoscopy were defined as interval CRCs. The proportion of interval CRCs was calculated as number of interval CRCs divided by number of detected and interval CRCs. The χ² test and a multivariate logistic regression model were used in the statistical analysis. Results: Of 15,484 older patients with left-sided CRC, the proportion of interval CRCs after flexible sigmoidoscopy was 8.8%compared with 2.5%after colonoscopy (P<.001). This difference was similar across left colon locations and largest in the descending colon (17.1%vs 3.5%; P<.001). In multivariate logistic regression, the odds of interval CRC after flexible sigmoidoscopy was 3 times as high as that after colonoscopy (odds ratio, 3.52; 95%CI, 2.66-4.65). Conclusion: In older patients with left-sided CRC, the odds of interval CRC after flexible sigmoidoscopy was 3 times as high as that after colonoscopy. Whether this finding reflects differences in bowel preparation quality, sedation use, or depth of insertion warrants future research.