Oral Presentation
Context. Pre-eclampsia (PE) is one of the major problems of modern obstetrics, due to high frequency of this pathology during pregnancy (2-8%). One of the perspective approaches to the prediction of the disease is a comprehensive analysis of risk factors that have a different nature. Objective. To determine the probability of developing PE according on testing thrombophilia genes, level of antibodies to β2 glycoprotein 1, D-dimer and the coefficient of atherogenicity were made. Methods. With PCR genetic polymorphisms in PAI-1 675 5G / 4G and fibrinogen β (455 G → A) genes were detected, antibodies to β2 glycoprotein-1 with IFA (ELISA) and level of D-dimer with immunоturbodimetric assay were determine. For a mathematical model of prognosis the method of maximum likelihood was used. Patient(s). This prospective, cohort study included 177 patients: 133 with PE and 44 healthy pregnant women in the 3rd trimester of pregnancy. Result(s). Applying the maximum likelihood method, the risk function was calculated, as follows: y=-11,74+6,25xD+0,80xCA+1,08xPAI+1,25xFIB+0,49xAtβ.; were: xD is the concentration of D-dimer, xCA is the value of coefficient of atherogenicity, xAtβ2 is the amount of antibodies to β2 glycoprotein 1, xPAI is the value of PAI-1, and xFGB is the FGB value. The value of xPAI and xFGB is 1 if the gene is normal, equal to 2, if the allelic polymorphism is heterozygous, and is equal to 3 if the mutant (pathological) gene is homozygous. Proposed formula that determines the probability of developing PE: P(y)=P(xD, xCA, xPAI,xFIB,xAtβ2)=1/(1+exp{-y})==1/(1+exp{-(-11,74+6,25xD+0,80xCA+1,08xPAI+1,25xFIB+0,49xAtβ2)}). P value (0.765) = 0.683 is interpreted as critical, above which pregnant woman should be included to the high-risk group of preeclampsia development. Conclusions. The sensitivity of the proposed model is 82.5% (95% CI 74.2-88.9%), the specificity is 90.9% (95% CI 78.3-97.5%). The significance of Hosmer-Lemeshow statistics, which characterizes the degree of evasion between the predicted number and the observed, is 7.99 (p < 0.05), it indicate good agreement between the model and real data.