S.A. Vozianov, S.N. Shamrayev, V.P. Stus, V.N. Krasnov, V.D. Vasylieva, M.Yu. Polion, D.N. Shamrayeva

Prediction of early postoperative complications of radical cystectomy with different method of urinary derivation through methods of mathematical modeling

The aim of the study was to develop a mathematical model for predicting the outcome of the early postoperative period in patients with musculo-invasive bladder cancer.

Materials and Methods.Treatment of 120 patients that underwent open radical cystectomy (ORC) became the basis for the retrospective research with every 3rd patient having postoperative complications. To identify the predictors for the development of intra- and postoperative complications, all patients were divided into two groups according to the Clavien-Dindo classification: the first group comprised 76 (63.3%) patients with uncomplicated postoperative period; the second group included 44 (67.7%) patients who developed complications of various severity and required additional instrumental or operational intraoperative procedures. The value of the diagnostic coefficient (DK) was used to assess the objective parameters studied. The results are based on the method of the inhomogeneous sequential recognition procedure, which is based on the Bayes method. The calculation of the informativeness of both the individual ranges of characteristics and their total informativeness was carried out.

Results. 60 indicators were analyzed in both groups of patients, 50 (83.3%) of which were uninformative. It was discovered and statistically confirmed that several factors play a role in the development of postoperative complications. Among them: surgical intervention, where the risk factors are its traumatism and duration (SI = 0.5), the volume of intraoperative blood loss (SI = 0.8), the choice of the method of urine diversion, and the duration of abdominal drainage and the initial nutritional insufficiency (SI = 6.15). The revealed factors of the negative prognosis of the course of the early postoperative period require preoperative correction, which can increase the cohort of patients who will be able to undergo radical surgical treatment of muscle-invasive bladder cancer.

Conclusion. The mathematical model for predicting the outcome of the early postoperative period after ORC provides an objective quantitative assessment of the patient’s body parameters, allows for the mutual influence of different parameters on each other, and stratifies each individual patient in the appropriate group of postoperative course. It objectifies the tactics of surgical treatment based on the biological parameters of the patient’s organism on the development of complications of radical cystectomy. Further improvements of the proposed method for predicting postoperative complications of radical cystectomy in the form of computer software will improve the quality and accuracy of decision making in the choice of tactics for treating this category of patients.