%0 Journal Article
%T Construction of a nomogram model for predicting infectious intrapartum fever.
%A Lu B
%A Hong L
%A Dai Q
%A Cai H
%A Lu Z
%A Chen A
%J Zhejiang Da Xue Xue Bao Yi Xue Ban
%V 52
%N 1
%D 2023 Feb 25
%M 37283118
暂无%R 10.3724/zdxbyxb-2022-0479
%X <B>OBJECTIVE: </B>To investigate influencing factors of intrapartum fever during vaginal delivery and to construct a prediction model for infectious intrapartum fever.<BR><B>METHODS: </B>A total of 444 patients with intrapartum fever admitted in Ningbo Women and Children's Hospital from January 2020 to December 2021 were enrolled. The clinical data and laboratory findings were compared between patients with infectious intrapartum fever and non-infectious intrapartum fever, and the factors associated with intrapartum fever were analyzed with a multivariate logistic regression model. A prediction nomogram model was constructed based on the factors of intrapartum fever and its predictive efficiency was evaluated by correction curve and receiver operator characteristic curve.<BR><B>RESULTS: </B>In the 444 cases, 182 (41.0%) had definite intrauterine infection and 262 (59.0%) had no infectious intrapartum fever. Univariate analysis showed that the length of hospital stay before induced labor, the time of induced abortion, misoprostol administration, autoimmune diseases, white blood cell count (WBC) and hypersensitive C-reactive protein (hs-CRP) levels were significantly different between the two groups (all P<0.05). Multivariate analysis showed that misoprostol administration and autoimmune diseases were protective factors (OR=0.31 and 0.36, both P<0.05) for infectious intrapartum fever, while high WBC and hs-CRP were risk factors (OR=1.20 and 1.09, both P<0.05). The area under the curve of nomogram model for predicting infectious intrapartum fever was 0.823, and the calibration curve validation showed that the predicted and measured values were in general agreement.<BR><B>CONCLUSIONS: </B>Multiple factors cause intrapartum fever. The nomogram model constructed in this study has good predictive accuracy for infectious intrapartum fever.<BR>目的: 研究阴道试产过程中产时发热的影响因素，并构建宫内感染所致产时发热的列线图预测模型。方法: 收集2020年1年—2021年12月宁波市妇女儿童医院收治的阴道试产过程中出现产时发热的444例产妇的资料。回顾性分析患者的临床资料、症状、体征及实验室检查结果，比较宫内感染所致产时发热与非宫内感染所致产时发热的临床特点及实验室指标差异，采用多因素logistic回归模型分析产时发热的相关因素。应用R4.1.0软件及rms和regplot程序包建立列线图模型，应用Bootstrap重采样法进行模型验证，绘制预测模型校准曲线及受试者操作特征曲线。结果: 444例产时发热产妇中，182例（41.0%）产后明确为宫内感染，262例（59.0%）为非宫内感染所致。单因素分析结果显示，引产前住院时间、引产时间、使用米索前列醇、妊娠合并免疫系统疾病、白细胞计数（WBC）及超敏C反应蛋白（hs-CRP）在宫内感染所致产时发热与非宫内感染所致产时发热组间差异有统计学意义（均P<0.05）；多因素logistic回归分析结果显示，妊娠合并免疫系统疾病和使用米索前列醇是宫内感染所致产时发热的保护因素（OR=0.31和0.36，均P<0.05），而WBC和hs-CRP增加是危险因素（OR=1.20和1.09，均P<0.05）。基于上述四项独立影响因素构建的列线图模型预测宫内感染所致产时发热的曲线下面积为0.823，校准曲线显示预测值与实测值基本一致。结论: 产时发热影响因素众多，本研究构建的列线图模型对于宫内感染所致产时发热具有较好的预测价值。.