Objective: The aim: Identification of new, non-trivial knowledge on the prediction of thyroid recurrence on the basis of follow-up data from medical histories.
Patients and methods: Materials and methods: The development of long-term oncological effects was studied on the catamnestic data of 157 patients diagnosed with thyroid cancer who were treated according to a standard scheme, including radical surgery, radioiodine therapy and hormone therapy.
Results: Results: It is shown that the specificity of thyroglobulin as a cancer marker for thyroid cancer is not an unambiguous question and the probability of obtaining false-positive results on its basis is quite significant. It is shown that violation of the recommended terms for special treatment (surgical and radioiodine therapy) can be used as a factor in the prognosis of relapse, and patients who received special treatment with violation of the terms for various reasons require careful attention and more careful examination. The dose of thyroxine that should be used to achieve suppression can be used as a marker of thyroid relapse: an excess of thyroxine levels of 2.8 μg / kg is an indicator of the risk of relapse in the future. Statistically there was no significant effect on the prevention of long-term oncological complications by prolonging the duration of suppressive hormone therapy as a component of thyroid cancer treatment, but there are grounds to believe that prolonged suppression leads to increased cardiovascular and female genital complications.
Conclusion: Conclusions: the use of modern information technologies in relation to the arrays of catamnestic data of medical histories allowed to obtain additional knowledge to prevent the development of distant oncological complications resulting from thyroid cancer.
Keywords: Data Mining technology; Differentiated thyroid cancer; long-term effects.