Should patients with anemia and low normal or normal serum ferritin undergo colonoscopy?

BACKGROUND: Patients with unexplained iron deficiency anemia have a greater prevalence of colonic neoplasia, and should be evaluated for a colonoscopy. The approach to patients with anemia without iron deficiency remains unclear. OBJECTIVE: To compare the prevalence of colonic neoplasia in anemic patients with normal ferritin (>50 ng/mL), to those with ferritin ≤50 ng/mL, and nonanemic individuals. METHODS: Patients referred for colonoscopy for anemia evaluation were stratified into 3 groups: ferritin ≤50 ng/mL, 51-100 ng/mL, and >100 ng/mL. We compared these groups to each other, and to asymptomatic nonanemic individuals undergoing screening colonoscopy. The prevalence of advanced colonic neoplasia was determined for each group using existing records. RESULTS: During the study period, 414 patients who underwent colonoscopy for anemia evaluation and 323 nonanemic individuals who underwent colonoscopy for cancer screening met inclusion criteria. Study subjects were mostly men. The prevalence of advanced colonic neoplasia in subjects with ferritin 51-100 ng/mL was 7.2% (95% CI 2.4-17.9%), similar to 7.9% (95% CI 5.1-11.9%) in those with ferritin ≤50 ng/mL. The incidence of advanced colonic neoplasia in subjects with ferritin >100 ng/mL was 1.7% (95% CI 0.1-6.6%), similar to 1.2% (95% CI 0.4-3.3%) in the asymptomatic nonanemic group. After adjusting for age, patients with ferritin ≤50 ng/mL and 51-100 ng/mL were almost 5 times more likely to harbor advanced colonic neoplasia than the other groups. The addition of other laboratory parameters did not improve the predictive value of ferritin. CONCLUSION: A ferritin cutoff of 100 ng/mL can be used to determine the need for colonoscopy in men with anemia.