应用倾向性评分匹配构建母亲产时发热伴新生儿感染的临床预测模型

Construction of a clinical prediction model for neonatal infection associated with intrapartum maternal fever based on propensity score matching

摘要

目的 探讨母亲产时发热伴新生儿感染的危险因素,并建立临床预测模型,以提高新生儿感染风险的早期评估和预测能力。方法 回顾性分析2020年1月至2022年12月在首都医科大学附属北京妇产医院分娩且出现产时发热产妇及活产新生儿的临床资料.以新生儿生后是否出现感染为观察结局,分为感染组和非感染组.采用倾向性评分匹配(PSM)方法,对胎龄和出生体重进行匹配,建立母亲产时发热伴新生儿感染的临床预测模型.通过Logistic回归分析确定影响因素,构建列线图模型,并通过受试者工作特征(ROC)曲线、校准曲线和临床决策曲线(DCA)进行模型评价与验证。结果 共纳入3269名产妇及3212例活产新生儿,其中591例新生儿生后因感染住院治疗.通过PSM法匹配208对新生儿后,两组新生儿胎龄和出生体重的差异无统计学意义.多因素Logistic回归分析结果显示,胎膜早破超过18小时(OR=1.885,95%CI:1.123~3.166)、B族链球菌(GBS)阳性(OR=4.443,95%CI:2.156~9.156)和胎粪污染(OR=3.137,95%CI:1.104~8.910)是母亲产时发热伴新生儿感染的独立影响因素(P<0.05).模型的训练集和验证集的ROC曲线下面积(AUC)分别为0.720和0.711,校准曲线中,训练集的拟合优度检验(χ²=4.917,P=0.766)和验证集的拟合优度检验(χ²=0.130,P=1.000),显示出较好的预测能力和一致性。结论 建立基于PSM方法的母亲产时发热伴新生儿感染的临床预测模型,通过收集易获取的临床指标进行早期判断,适合临床应用,减少不必要的干预和并发症,提高新生儿的健康和生存质量.

Abstract

Objective To investigate the risk factors associated with neonatal infection in cases of intrapartum maternal fever and to develop a clinical prediction model to improve early assessment and prediction of neonatal infection risk. Methods A retrospective analysis was conducted on clinical data from mothers who experienced intrapartum fever and their live-born neonates delivered at Beijing Obstetrics and Gynecology Hospital, Capital Medical University, between January 2020 and December 2022. Neonates were divided into an infection group and a non-infection group based on postnatal infection status. Propensity score matching (PSM) was applied to match gestational age and birth weight, and a clinical prediction model for neonatal infection associated with maternal intrapartum fever was constructed. Multivariate logistic regression was used to identify significant risk factors, and a nomogram model was developed. The model's performance was evaluated using the receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). Results A total of 3269 mothers and 3212 live-born neonates were included, among whom 591 neonates were hospitalized due to infection. After PSM, 208 pairs of neonates were matched, and no statistically significant differences were found in gestational age and birth weight between the two groups. Multivariate logistic regression identified prolonged rupture of membranes over 18 hours (OR=1.885, 95%CI: 1.123-3.166), group B Streptococcus (GBS) positivity (OR=4.443, 95%CI: 2.156-9.156), and meconium-stained amniotic fluid (OR=3.137, 95%CI: 1.104-8.910) as independent risk factors (P<0.05). The areas under the ROC curve (AUC) for the training and validation sets were 0.720 and 0.711, respectively. Good calibration was observed in both the training set (χ²=4.917, P=0.766) and the validation set (χ²=0.130, P=1.000), indicating satisfactory predictive performance and consistency. Conclusion A clinical prediction model for neonatal infection associated with intrapartum maternal fever was successfully developed using the PSM method. By incorporating easily accessible clinical indicators, the model facilitates early risk assessment and is suitable for clinical application, potentially reducing unnecessary interventions and complications while improving neonatal health outcomes and quality of life.

关键词

倾向性评分匹配 / 母亲产时发热 / 新生儿感染 / 预测模型

Key words

propensity score matching / intrapartum maternal fever / neonatal infection / prediction model

参考文献

[1] Laikemariam M, Aklilu A, Waltengus F, et al. Adverse neonatal outcomes and associated factors among mothers who gave birth through induced and spontaneous labor in public hospitals of Awizone, Northwest Ethiopia: a comparative cross-sectional study[J]. BMC Pregnancy Childbirth, 2023, 23(1): 307.

[2] Mersha A, Assefa N, Teji K, et al. Essential newborn care practice and its predictors among mother who delivered within the past six months in Chencha District[J]. PLoS One, 2018, 13(12): e0208984.

[3] Jansen S, Lopriore E, Naaktgeboren C, et al. Epidural-related fever and maternal and neonatal morbidity: a systematic review and meta-analysis[J]. Neonatology, 2020, 117(3): 259-270.

[4] Austin PC, Stuart EA. The effect of a constraint on the maximum number of controls matched to each treated subject on the performance of full matching on the propensity score when estimating risk differences[J]. Stat Med, 2021, 40(1): 101-118.

[5] 中国妇幼保健协会新生儿保健专业委员会, 中国医师协会新生儿科医师分会. 母婴同室早发感染高危新生儿临床管理专家共识[J]. 中华围产医学杂志, 2021, 24(8): 567-575.

[6] 中华医学会儿科学分会新生儿学组, 中华儿科杂志编辑委员会. 新生儿败血症诊断与治疗专家共识(2024)[J]. 中华儿科杂志, 2024, 62(10): 931-940.

[7] Segal S. Labor epidural analgesia and maternal fever[J]. Anesth Analg, 2010, 111(6): 1467-1475.

[8] Pantell R H, Roberts K B, Adams W G, et al. Clinical practice guideline: evaluation and management of well-appearing febrile infants 8 to 60 days old[J]. Pediatrics, 2021, 148(2): e2021052228.

[9] Puopolo K M, Benitz W E, Zaoutis T E, et al. Management of neonates born at ≥35 0/7 weeks' gestation with suspected or proven early-onset bacterial sepsis[J]. Pediatrics, 2018, 142(6): e20182894.

[10] Abebe H, Adane D, Shitu S. Essential newborn care practice and its associated factors in Southwest Ethiopia[J]. Arch Public Health, 2021, 79(1): 42.

[11] Linehan L A, Walsh J, Morris A, et al. Neonatal and maternal outcomes following midtrimester preterm premature rupture of the membranes: a retrospective cohort study[J]. BMC Pregnancy Childbirth, 2016, 16(1): 25.

[12] Stoll K, Titoria R, Turner M, et al. Perinatal outcomes of midwife-led care, stratified by medical risk: a retrospective cohort study from British Columbia (2008-2018)[J]. CMAJ, 2023, 195(8): E292-E299.

[13] Muglia L J, Benhalima K, Tong S, et al. Maternal factors during pregnancy influencing maternal, fetal, and childhood outcomes[J]. BMC Med, 2022, 20(1): 418.

[14] Mohr S, Amylidi-Mohr S K, Stadelmann P, et al. Systemic inflammation in pregnant women with periodontitis and preterm prelabor rupture of membranes: a prospective case-control study[J]. Front Immunol, 2019, 10: 2624.

[15] Zamstein O, Wainstock T, Sheiner E. Intrapartum maternal fever and long-term infectious morbidity of the offspring[J]. Clin Med, 2023, 12(9): 3329.

[16] Madan I, Jackson F I, Figueroa R, et al. Preterm prelabor rupture of membranes in singletons: maternal and neonatal outcomes[J]. Perinat Med, 2023, 51(6): 787-791.

[17] Cassidy-Bushrow A E, Sitarik A, Levin A M, et al. Maternal group B streptococcus and the infant gut microbiota[J]. Dev Orig Health Dis, 2016, 7(1): 45-53.

[18] Berardi A, Lugli L, Baronciani D, et al. Group B Streptococcus early-onset disease in Emilia-romagna: review after introduction of a screening-based approach[J]. Pediatr Infect Dis J, 2010, 29(2): 115-121.

[19] Filkins L, Hauser J R, Robinson-Dunn B, et al. American society for microbiology provides 2020 guidelines for detection and identification of Group B Streptococcus[J]. J Clin Microbiol, 2020, 59(1): e01230-20.

[20] Klingner M C, Kruse J. Meconium aspiration syndrome: pathophysiology and prevention[J]. J Am Board Fam Pract, 1999, 12(6): 450-466.

[21] Carr B L, Copnell B, McIntyre M. Differences in meconium stained amniotic fluid in an Australian population: a retrospective study[J]. Women Birth, 2019, 32(2): e259-e263.

[22] Hochler H, Lipschuetz M, Guedalia J, et al. The impact of peak and duration of maternal intrapartum fever on perinatal outcomes[J]. Am J Obstet Gynecol MFM, 2021, 3(4): 100390.