Department of Infection Control, School of Medicine, The First Affiliated Hospital of Xiamen University, Xiamen University, Xiamen, Fujian, 361003, PR China.
Xiamen Hospital Infection Management Quality Control Center, Xiamen, Fujian, 361003, PR China.
BMC Infect Dis. 2024 Sep 11;24(1):955. doi: 10.1186/s12879-024-09795-y.
This study aimed to develop and validate a nomogram for assessing the risk of nosocomial infections among obstetric inpatients, providing a valuable reference for predicting and mitigating the risk of postpartum infections.
A retrospective observational study was performed on a cohort of 28,608 obstetric patients admitted for childbirth between 2017 and 2022. Data from the year 2022, comprising 4,153 inpatients, were utilized for model validation. Univariable and multivariable stepwise logistic regression analyses were employed to identify the factors influencing nosocomial infections among obstetric inpatients. A nomogram was subsequently developed based on the final predictive model. The receiver operating characteristic (ROC) curve was utilized to calculate the area under the curve (AUC) to evaluate the predictive accuracy of the nomogram in both the training and validation datasets.
The gestational weeks > = 37, prenatal anemia, prenatal hypoproteinemia, premature rupture of membranes (PROM), cesarean sction, operative delivery, adverse birth outcomes, length of hospitalization (days) > 5, CVC use and catheterization of ureter were included in the ultimate prediction model. The AUC of the nomogram was 0.828 (0.823, 0.833) in the training dataset and 0.855 (0.844, 0.865) in the validation dataset.
Through a large-scale retrospective study conducted in China, we developed and independently validated a nomogram to enable personalized postpartum infections risk estimates for obstetric inpatients. Its clinical application can facilitate early identification of high-risk groups, enabling timely infection prevention and control measures.
本研究旨在开发和验证一种用于评估产科住院患者医院感染风险的列线图，为预测和减轻产后感染风险提供有价值的参考。
对 2017 年至 2022 年间分娩的 28608 例产科患者进行回顾性观察性研究。使用 2022 年的数据（包括 4153 名住院患者）进行模型验证。采用单变量和多变量逐步逻辑回归分析确定影响产科住院患者医院感染的因素。基于最终预测模型，构建列线图。使用接收者操作特征（ROC）曲线计算曲线下面积（AUC），以评估列线图在训练和验证数据集的预测准确性。
纳入最终预测模型的因素包括孕周≥37 周、产前贫血、产前低蛋白血症、胎膜早破（PROM）、剖宫产、手术分娩、不良分娩结局、住院时间（天）>5 天、CVC 使用和输尿管置管。该列线图在训练数据集和验证数据集中的 AUC 分别为 0.828（0.823，0.833）和 0.855（0.844，0.865）。
通过在中国进行的大规模回顾性研究，我们开发并独立验证了一种列线图，以实现对产科住院患者个性化的产后感染风险评估。其临床应用可以帮助早期识别高风险人群，从而及时采取感染预防和控制措施。
