Abstract
We previously reported the prevalence of colorectal neoplasm and the sensitivity of fecal immunochemical tests (FIT) based on the results of simultaneous screening colonoscopy and single-sample FIT. In the present study, we attempted to simulate two-sample FIT and estimate the positive predictive values (PPV) using one or two positive FIT results in two samples based on the same dataset collected at the initial screening. We hypothesized that the sensitivity of each individual sample from each lesion category was the same as that of the single-sample method used in previous studies, and that the first and second samples were completely independent. The PPV for colorectal cancer in men aged in their 40s was estimated to be 0.3% when only one of two samples was positive and 10% when both samples were positive. For women in their 40s, the respective PPV were 0.7% and 19%, compared with 1.7% and 26% for men in their 50s, 0.8% and 15% for women in their 50s, 4.1% and 49% for men in their 60s, and 2.9% and 44% for women in their 60s. The PPV increased with age, and subjects with positivity in both samples were estimated to be at much higher risk of colorectal cancer than those positive for only one sample.
