MATHEMATICAL MODEL FOR PREDICTING THE OCCURRENCE OF PRETERM BIRTH IN WOMEN AT RISK BASED ON THE STUDY OF ESTRADIOL

Preterm birth is one of the most pressing problems in modern obstetric practice, which not only endangers the life and health of newborns but also has serious consequences for women's health. They are defined as labour occurring before the 37th week of pregnancy and can be caused by a variety of factors, including biological and endocrine disorders. One of the important hormones that plays a critical role in maintaining a normal pregnancy is estradiol. Low or abnormal levels of this hormone may indicate the presence of disorders that increase the likelihood of preterm birth. The aim of the study is to develop a mathematical model to determine the probability and percentage of prediction of preterm birth in women at risk based on the study of estradiol. Materials and methods: The study analysed the level of estradiol in 60 women (including 30 at risk of preterm birth and 30 absolutely healthy women) who were in the Chernivtsi Regional Perinatal Centre. Estradiol levels were assessed both before and after treatment. The logistic regression method was used to develop a mathematical model for predicting preterm birth, which allows to estimate the probability of preterm birth based on changes in estradiol levels. The Welch's T-test was used to test the statistical significance of the results, which allows comparing the mean values of hormone levels before and after treatment and assessing the effectiveness of treatment measures. Results: A one-unit increase in estradiol increased the odds of preterm birth by 0.985 times (decreased the odds of preterm birth by 1.5%). Conclusions: The mathematical logistic regression model based on estradiol levels demonstrated an inverse relationship between estradiol levels and the probability of preterm birth, which allows predicting risk based on changes in the concentration of this hormone. An increase in estradiol levels by one unit reduces the likelihood of preterm birth by 1.5%, which opens up new opportunities to improve the prediction and management of pregnancy in women at risk.