[Detection of cancer, sensitivity of the test and sensitivity of the screening program]

In assessment of screening for cancer, no distinction is usually made between the sensitivity of the screening test (St) and the sensitivity of the screening program (Sp). This paper was aimed to distinguish meaning, method for assessment and interest for each of them, and to determine their relationship. Sensitivity of the screening program can be directly assessed with data from on-going trials whilst assessment of sensitivity of screening test requires modelisation techniques, especially for assessing the mean duration of the preclinical phase of cancer. Assuming an exponential distribution of this duration, lambda as the time parameter, a mathematical relation between St and Sp is suggested as follows: [formula: see text] with r being the interval between two screening tests. The implementation of this equation with data from a mass-screening program for colorectal cancer in the department of Calvados allowed us to investigate the influence of the mean preclinical phase and the interval between two screening tests on the value of the sensitivity of the screening procedure. Such a modelisation could be useful in the development of a rational screening policy.