НОВАЯ КОРОНАВИРУСНАЯ ИНФЕКЦИЯ БЕРЕМЕННЫХ ЖЕНЩИН: ОСОБЕННОСТИ ТЕЧЕНИЯ БЕРЕМЕННОСТИ И ВОЗМОЖНОСТИ ПРОГНОЗИРОВАНИЯ ПНЕВМОНИИ НОВОРОЖДЕННЫХ

Abstract

Introduction. Pregnancy profoundly impacts the immune system, yet the influence of gestational age at COVID-19 contraction and its implications for perinatal outcomes remain uncertain. Aim. This study aimed to evaluate the effects of COVID-19 contracted at various gestational stages on pregnancy and perinatal outcomes. Materials and Methods. Retrospective analysis of medical records from 113 mother-newborn pairs, including COVID-19-positive pregnant women, was conducted at the clinical sites of the Department of Obstetrics and Gynecology of the Novosibirsk State Medical University from 2021 to 2022. The sample was categorized into three study groups based on gestational age at COVID-19 contraction, with a control group of COVID-19-negative pregnant women. Results. Findings revealed a higher incidence of mild COVID-19 in Group 1, while Groups 2 and 3 exhibited a moderate course (p=0.03). Additionally, Groups 2 and 3 demonstrated a higher prevalence of pneumonia (p=0.03) and antibiotic therapy (p=0.02). Complications such as preeclampsia (p=0.002), oligohydramnios (p<0.001), and fetal growth restriction (p=0.02) were more frequent in patients with COVID-19 during pregnancy. Furthermore, fetal distress was more common in study groups 1 and 2 (p=0.002). Neural network analysis allowed for the prediction of pneumonia in newborns with 98.2% accuracy based on factors such as gestational age at COVID-19 contraction, disease severity, presence of oligohydramnios and fetal growth retardation, meconium amniotic fluid, and delivery date. Conclusions. In conclusion, COVID-19 contraction before 34 weeks of gestation often complicates pregnancy, leading to placental insufficiency, preeclampsia, and gestational diabetes mellitus, thereby increasing the risk of early delivery and the birth of low-birth-weight children with reduced adaptive capabilities. The development of pneumonia in these newborns exacerbates their condition, often resulting in perinatal losses. Utilizing a neural network for pneumonia prediction may potentially mitigate perinatal losses.