Hanzha R. Statistical evaluation of the results of clinical trials in oncology.

The dissertation is devoted to the research and improvement of the theoretical-methodological and statistical bases of the assessment of oncological diseases and breast cancer as an object of statistical research. Theoretical and methodological principles of the process of diagnosis of malignant and benign tumors, their general histology, natural mechanics of appearance and development are studied. The main characteristics of breast cancer as a disease that occupies a very special place among other possible human malignancies are presented. The economic component of treatment and diagnosis of breast cancer has been clarified, as well as the key stages of the disease treatment are described. The direct medical costs are identified and further analyzed. The theoretical and methodological bases of statistical research of oncological diseases as a mass phenomenon and particular object of statistical study are investigated, the comparative statistical analysis of oncology morbidity in Ukraine and the world is carried out, as well as social and economic aspects of the treatment of patients with breast cancer are covered. Based on the preliminary analysis, the survival of women with breast cancer was modeled in order to figure out the most significant factors that are predicted to have an effect on the probability of survival of women with multiple malignant neoplasms. A series of survival models have been built – a general model, as well as separate models for synchronous and metachronous types of cancer-related processes. It has been shown that the course of the disease in patients with the synchronous type of cancer is more aggressive and prognostically unfavorable as compared to those patients with the metachronous type of disease. It is proved that the factors of formation of survival of patients aged less than and more than 30 years at the time of the Chernobyl accident are fundamentally different between these 2 age groups. The possibility of further use of the obtained models for predicting the survival of women in the study groups is indicated. In order to determine the criteria for the objective choice of the type of surgical treatment at the diagnostic stage, the factors that predict the choice of the type and extent of particular surgery for breast cancer were analyzed, along with modeling performed considering different case scenarios. Despite convincing evidence of the benefits of breast-conserving surgery, it has been shown that identified predictors of breast-conserving surgery often do not match those commonly known and used ones suggested in the protocols. Based on the results of analyses and modeling of data from patients with breast cancer who underwent mammography examination, the type and scope of surgery were planned, actions to further compare the types and scope of actual surgery with the planned were performed too, allowing one to further determine which of the mammographic features appear to be leading. A statistical model was constructed using mammographic examination data, as well as the coefficient of the lesion was derived and subsequently included into the statistical model as a continuous covariate. The constructed model provided a fairly decent opportunity to quite effectively choose the type of surgical intervention and further determine the potential scope of required actions. Its application took into account the characteristics of the tumor and the anatomical features of patients, which, in addition to providing real-time information, further allows predicting the type and extent of optimal surgery for breast cancer. In turn, this can not only increase the effectiveness of cancer treatment of women with breast cancer but also ensure their proper quality of life after the treatment in the future, which, in turn, is likely to help to reduce the risk of recurrence and to increase overall survival.