Screening for Colorectal Cancer: Which Test Is Best?

To the Editor. —We commend Drs Lang and Ransohoff1for their effort to estimate the unknown, independent contribution of colonoscopy to fecal occult blood test (FOBT) screening. From their mathematical model, which used selected data from our FOBT trial2and from a case-control study of rigid sigmoidoscopy,3they concluded that FOBT screening with rehydrated slides is a haphazard method for selecting persons for colonoscopy. Their conclusion depends on several assumptions that may be incorrect. They assume a constant FOBT positivity rate of 10% throughout our study. It was not. Initially, with nonrehydrated slides, the positivity rate was 2.0%. Rehydration was phased in, and during the first screening phase (1976 to 1982) the overall positivity rate was 7.4%. During this time, two thirds of the FOBT slides were rehydrated. Screening was suspended for 3 years, not 4 years as assumed for the Lang-Ransohoff model. By February 1986, when