Conventional screening measure does not accurately capture screening status in a minority of patients with colorectal cancer

Abstract

Colorectal cancer (CRC) screening uptake in the Veterans Health Administration (VA) has been reported to be higher than the US general population, but CRC remains a prevalent cancer within the VA system. To examine CRC predictors and the extent to which the conventional definition of up-to-date screening applies to the population, we conducted a case-control study using VA data from 2012 to 2018. We classified patients into 5 categories: up-to-date or not up-to-date average-risk patients aged 50 to 75 (Categories 1 and 2), up-to-date or not up-to-date average-risk patients aged <50 or >75 (Categories 3 and 4), and high-risk patients (Category 5). Each CRC case was matched by age, sex, and facility with 4 controls. We performed multivariable conditional logistic regression, adjusting for race and ethnicity, diabetes, obesity, and alcohol use. Among 3714 CRC cases identified, Category 4 (odds ratio [OR] 1.40, 95% CI 1.11-1.78) and Category 5 (OR 6.23, 95% CI 5.06-7.66) patients had a higher risk of CRC compared to Category 1 patients. Compared with White patients, Black patients had a higher risk (OR 1.54, 95% CI 1.37-1.73). Diabetes (OR 1.65, 95% CI 1.51-1.81) and alcohol use disorder (OR 1.53, 95% CI 1.35-1.73) were also associated with CRC. Most CRC cases occurred in individuals aged 50 to 75, but 12.5% occurred in persons who were outside of this age range or had high-risk personal or family history. The conventional measure of CRC screening, focused on average-risk individuals aged 50 to 75, does not reflect screening status in an important minority of CRC patients.

Keywords: Veteran; colorectal cancer; screening.