Algorithm of prenatal prognostication of the intrauterine infection

Abstract

DOI: 10.52705/2788-6190-2026-01-1
УДК 618.33-022.7-037

The objective: to develop the algorithm of prenatal prognostication of the intrauterine infection.
Materials and methods. By us a complete database (clinical and laboratory indexes, anamnesis, motion of pregnancy, results of births, information for motion of neonatal period et al) was formed about 201 pregnant, transmitters of exciters perinatal meaningful infections (Herpes of simplex 1,2, Staphylococcus of aureus, Candida of albicans and Chlamydia of trachomatis). From them healthy children gave birth in 50,3 % (101) women, and in 49,7 % (100) – to put with the intrauterine infection. Age of women, in which children gave birth with intrauterine infections, made ± 25,5 (23–30) years, and for women which gave healthy children ± 25 (24–28) years, distributing of index in both groups differently from normal (DKS = 1,43; p = 0,033 and DKS = 1,94; p = 0,001 respectively) statistically meaningful differences it is not discovered (U = 0,233; p = 0,816). To the complex of the conducted researches were included clinical, microbiological, virologic, biochemical and statistical.
Results. By a model it is classified correctly 82 % (50/61) all new-born, such which gave birth healthy, among new-born with points after a scale Apgar from 0 to 7 this index is evened 75 %. On the whole by the built model on an instructive multiplicity it was correctly classified (coefficient of concordance) 78 % (105/134) all cases. Size of coefficient of connection of D-Зоммера (SOMERS’D), which represents connection of actual frequency and logistic regression, which is foreseen on equation, is 0,573, p < 0,001. By a model it is classified correctly 79 % (11/14) all healthy new-born, among new-born with points after a scale Apgar from 0 to 7 this index is evened 76 %. On the whole, by the built model on a test multiplicity it was correctly classified (coefficient of concordance) 77 % (23/30) all cases. Size of coefficient of connection of D-Sommer (SOMERS’D), which represents connection of actual frequency and logistic regression, which is foreseen on equation, is 0,464; p = 0,004. The analysis of clinical and laboratory data at the use of mathematical model of regression of logit allowed to form the aggregate of predictors risk of development of intrauterine infection with the high level of concordance over 90 %. It underlay development of the computer program of calculation of risk of birth of children with intrauterine infections.
Conclusions. We succeeded to develop the program of screening prenatal prognostication of intrauterine infections, which includes clinical predictors of the intrauterine infection and regulator-transport squirrel at blood of expectant mothers and amniotic fluid of puerpera. Created program of calculation of risk of the intrauterine infection in an early neonatal period it is recommended to apply from the moment of determination of signs of live birth, from 22 a week’s terms of pregnancy, term of premature births, after second ultrasonic screening.