Mathematical modeling of pathological changes in the body is the means of obtaining information for making decisions about the method of treatment. Numerous studies have shown that the exponential model describes the tumor cells growth, and the time of antigen doubling determines the aggression of cancer cells growth. The present work investigates inaccuracies in determining the antigen doubling time as a function of measurement errors. The study showed that the decision on the method of treatment could be changed by taking into account errors in the prognosis of patient’s condition. For patient’s stratification in groups of high, medium and low risks, various threshold values corresponding to the antigen level are proposed. The results are presented in the form of a table and graphs.