Alʹmanah Kliničeskoj Mediciny (Feb 2023)
Chronic glomerulonephritis and pregnancy: predictors of preterm birth
Abstract
Background: Chronic kidney disease (CKD) in pregnant women, with one of its most important causes being chronic glomerulonephritis (CGN), increases the incidence of adverse perinatal outcomes and gestational complications, including preterm birth (PB). Being the main cause of infant morbidity and mortality, PB has serious medical and social significance. Aim: To identify clinical predictors and develop a predictive model of PB in pregnant women with CGN. Materials and methods: A retrospective/prospective study included 122 CGN patients, whose 128 pregnancies resulted in childbirth from January 2009 to November 2022. Eighty-eight pregnancies were in the patients with CKD stage 1, 15 in stage 2, 21 in stage 3a, 3 in stage 3b, and one in stage 4. One hundred and nine (109) patients (115 pregnancies) delivered on term (at least 37 weeks of gestation) and were included into the group of term deliveries, whereas 13 women with 13 pregnancies had PB within the range of 22 weeks to 36 weeks 6 days. In the patients of both groups, we assessed nephrological and obstetric history, proteinuria and arterial hypertension at baseline and during pregnancy, complications of the index pregnancy, such as preeclampsia (PE) and severe PE, anemia, urinary tract infections, acute kidney injury, placental insufficiency, and cervical insufficiency. Binary logistic regression was used for prediction modeling of PB in women with CGN. Results: The proportion of PB in total cohort of the CGN patients was 10.2%. PB was spontaneous only in 2/13 (15.4%) cases, while in the rest of 11 pregnancies (84.6%) the delivery was induced due to maternal and fetal indications. Six independent predictors of PB were identified: body mass index, CKD stage, history of non-developing pregnancies, proteinuria during pregnancy 1 g/day, PE and placental insufficiency. The predictive model had sensitivity of 76.9%, specificity 99.1%, diagnostic efficiency 96.9%, positive predictive value 90.9%, and negative predictive value 97.4%. Conclusion: Predicting PB and targeting modifiable factors associated with PB may improve pregnancy outcomes in patients with CGN.
Keywords