Impact of Comorbidity on Colorectal Cancer Screening in the Veterans Healthcare System

Background & Aims: The quality assessment measure of colorectal cancer screening in the veteran's health system reports the proportion of patients aged 52-80 years who were tested. This approach does little to assess for comorbid illnesses, which might limit the utility of screening. Our aim was to determine the relationship between patient comorbidity and screening by fecal occult blood test in a national sample of veterans. Methods: We examined the Veterans Health Administration's national databases (October 2003-February 2005) for a random sample of primary care patients, aged ≥50 years. The Charlson score, a validated measure of comorbidity burden, was calculated from diagnosis codes by the Deyo method. The association between Charlson score and colorectal cancer screening was assessed with logistic regression. Results: The sample of 77,268 was 97% men; mean age was 67 years. Charlson score distribution was 0, 45%; 1, 24%; 2, 14%; 3, 7%; 4, 4%; 5, 2%; 6, 1%; 7, 0.8%; 8, 0.6%; 9, 0.4%; ≥10, 1%. Overall there was no consistent significant association between Charlson score and use of fecal occult blood testing except in the sickest 1%. There was a strong and incremental relationship between Charlson score and 1-year mortality. Conclusions: Although there was a strong relationship in the veteran population between the Charlson score and survival, colorectal cancer screening utilization was not impacted by Charlson score. Instead, resources were expended evenly throughout the population, rather than directed toward screening the patients with the most life-years at stake. The quality measure for colorectal cancer screening should be modified to account for patient comorbidity.