Klin Onkol 2007; 20(Suppl 1 20): 176-181.

The role of stochastic modelling in the evaluation of inputs and outputs of population-based screening programme Sis discussed. Breast carcinoma was used as model diagnosis to document valuable contribution of stochastic models that is apparent even in comparison with randomized clinical trials. In evaluation of some specific time-to-event parameters like so called Lead time, the randomized
trials cannot be conducted at all and stochastic predictions remain the only accessible methodology. Paper evaluates advantages, disadvantages and assumptions of Markovian chains as well as analytical and simulation models (one-stage one-test breast cancer model, MISCAN model, MICROLIFE model, CISNET – The Cancer Intervention and Surveillance Network). The work is a pilot theoretical contribution to exact evaluation of Czech national breast cancer screening program.