Development of risk prediction nomogram for neonatal sepsis in Group B Streptococcus-colonized mothers: a retrospective study

Abstract

AbstractNeonatal clinical sepsis is recognized as a significant health problem, This study sought to identify a predictive model of risk factors for clinical neonatal sepsis. A retrospective study was conducted from 1 October 2018 to 31 March 2023 in a large tertiary hospital in China. Neonates were divided into patients and controls based on the occurrence of neonatal sepsis. A multivariable model was used to determine risk factors and construct models.The utilization and assessment of model presentation were conducted using Norman charts and web calculators, with a focus on model differentiation, calibration, and clinical applicability (DCA). Furthermore, the hospital’s data from 1 April 2023 to 1 January 2024 was utilized for internal validation. In the modelling dataset, a total of 339 pairs of mothers and their newborns were included in the study and divided into two groups: patients (n = 84, 24.78%) and controls (n = 255, 75.22%). Logistic regression analysis was performed to examine the relationship between various factors and outcome. The results showed that maternal age < 26 years (odds ratio [OR] = 2.16, 95% confidence interval [CI] 1.06–4.42, p = 0.034), maternal gestational diabetes (OR = 2.17, 95% CI 1.11–4.27, p = 0.024), forceps assisted delivery (OR = 3.76, 95% CI 1.72–5.21, p = 0.032), umbilical cord winding (OR = 1.75, 95% CI 1.32–2.67, p = 0.041) and male neonatal sex (OR = 1.59, 95% CI 1.00–2.62, p = 0.050) were identified as independent factors influencing the outcome of neonatal clinical sepsis. A main effects model was developed incorporating these five significant factors, resulting in an area under the curve (AUC) value of 0.713 (95% CI 0.635–0.773) for predicting the occurrence of neonatal clinical sepsis. In the internal validation cohort, the AUC value of the model was 0.711, with a 95% CI of 0.592–0.808. A main effects model incorporating the five significant factors was constructed to help healthcare professionals make informed decisions and improve clinical outcomes.