Colon Cancer Screening in the Ambulatory Setting

Background. Despite evidence of decreased mortality, recommendations for colon cancer screening have not been widely implemented by physicians. The objective of this study was to determine patient and clinician factors associated with screening for colon cancer. Methods. A retrospective review of computerized medical records from primary care practices in an academic medical center was performed. Patients comprised men and women aged 50–74 with at least one visit between July 1, 1995 and June 30, 1997. Measurements included Fecal occult blood testing in the past 1 or 2 years, sigmoidoscopy in the previous 5 or 10 years, or colonoscopy in the past 10 years. Results. A total of 6,039 patients were included in the analysis. Fecal occult blood testing had been performed in 44% of patients and sigmoidoscopy in 26%. Fifty-three percent of patients had undergone some type of colon cancer screening. In multivariate analysis, patient factors predictive of fecal occult blood testing included age (odds ratio (OR) per 5 years 1.05; 95% confidence interval 1.04, 1.06), Asian ethnicity (OR 1.23; 1.08, 1.41), number of visits during the study period (OR 1.05; 1.04, 1.06), recency of the last visit (OR 1.03; 1.02, 1.04), and having private insurance (OR 1.65; 1.04, 2.62). Predictors of sigmoidoscopy were similar except that patients with a family history of colon cancer and those with managed care insurance were also more likely to undergo sigmoidoscopy. Patients of nurse practitioners were less likely to receive fecal occult blood testing than were patients of physicians (OR 0.78; 0.65, 0.93). Patients of residents were less likely to undergo sigmoidoscopy than were patients of faculty (OR 0.79; 0.66, 0.94). Conclusions. Among patients seen in primary care practices, rates of colon cancer screening remain low, especially in patients who are younger, who have been seen less frequently, who are uninsured, or who do not have managed care insurance. Future research should explore these differences and should focus on increasing screening in all patient groups, especially in those who are underscreened.