Mathematical modeling in predicting the effectiveness of anti-tuberculosis treatment based on biochemical markers of tuberculosis

Abstract

Introduction. Today it is necessary to create new methods for studying and predicting the course of diseases. One of these tools is the construction of a mathematical model. The purpose of the study was to make a mathematical model for predicting tuberculosis treatment effectiveness based on determining the levels of Human-beta-defensin-1, ferritin and interleukin-6. Materials and methods. The study included 44 patients with pulmonary tuberculosis. Results. To predict the effectiveness of anti-tuberculosis treatment based on the level of HBD-1, the equation was obtained: A = 0.63877*Defl - 0.09831*Def2 - 9.18519 B = 0.3405*Defl + 0.4666*Def2- 23.5351 For ferritin, the equation was obtained: A = 0.63877*Ferl - 0.09831*Fer2 - 9.18519 B = 0.3405*Ferl + 0.4666*Fer2- 23.5351 For IL-6, the equation was obtained: A = 0.01538*IL6-1 + 0.13728*IL6-2 - 3.38667 B = -0.0563*IL6-1 + 0.7154*IL6-2- 50.9235 Conclusions. Based on a study of the relationships between the effectiveness of tuberculosis treatment and the levels of Human-beta-defensin-1, ferritin and interleukin-6, a mathematical model was built that allows predicting the effectiveness of anti-tuberculosis therapy based on determining the data of biochemical markers at the beginning of treatment and after 60 days of anti-tuberculosis therapy with specificity and sensitivity of at least 88%.