双胎妊娠剖宫产回收式自体血回输预测模型构建及临床应用研究

Construction and clinical application of prediction model for intra-operative cell salvage in gemellary pregnancy after cesarean section

目的 构建双胎妊娠剖宫产回收式自体血回输(IOCS)预测模型并探究其临床应用。方法 回顾性收集2021年10月—2023年6月金华市妇幼保健院自体血回收的150例双胎妊娠剖宫产贫血产妇的临床资料,按照4∶1随机抽样分为建模组(120例)与验证组(30例),基于建模组实际数据,根据是否采用IOCS技术分为IOCS组(63例)和未回输组(57例)。采用logistic多因素回归分析接受IOCS的危险因素,R语言建立列线图模型,采用ROC曲线评估模型的预测能力。结果 建模组和验证组贫血孕产妇基线资料比较,差异均无统计学意义(均P>0.05)。IOCS回输率52.50%,自体血浪费率47.50%。logistic多因素分析结果显示,产前阴道出血、胎盘植入或穿透、胎盘早剥、完全性前置胎盘及凶险性前置胎盘是发生IOCS的独立危险因素(P<0.05)。ROC结果显示,建模组的AUC为0.923(95%CI:0.868~0.979),灵敏度为85.87%,特异度为95.02%;验证组AUC为0.916(95%CI:0.840~0.992),灵敏度为89.45%,特异度为82.48%。结论 临床对于有产后大出血高风险人群应加强监管,基于危险因素建立的列线图模型具有较好的诊断效能,有利于临床决策。

Objective To construct the prediction model for intra-operative cell salvage(IOCS)and explore its clinical application in gemellary pregnancy after cesarean section.Methods A retrospective collection was performed on the clinical data of 150 puerperae with gemellary pregnancy after cesarean section in Jinhua Maternal and Child Health Hospital between October 2021 and June 2023.According to random sampling method,they were divided into modeling group(n=120)and verification group(n=30).According to presence or absence of IOCS,patients in modeling group were divided into IOCS group(n=63)and non-IOCS group(n=57).The risk factors of IOCS were analyzed by logistic multivariate regression analysis.The nomogram model was constructed by R language,and its predictive ability was evaluated by ROC curves.Results There was no significant difference in baseline data between modeling group and verification group(P>0.05).The incidence of IOCS and autologous blood waste was 52.50%and 47.50%,respectively.The results of logistic multivariate analysis showed that prenatal vaginal bleeding,placental implantation or penetration,placental abruption,complete placenta previa and dangerous placenta previa were independent risk factors of IOCS(P<0.05).The results of ROC curves analysis showed that AUC,sensitivity and specificity in modeling group and verification group were[0.923(95%CI:0.868-0.979),85.87%,95.02%]and[0.916(95%CI:0.840-0.992),89.45%,82.48%],respectively.Conclusion Clinical supervision should be strengthened for people with high risk of postpartum massive hemorrhage.The nomogram model constructed based on risk factors has good diagnostic efficiency,which is conducive to clinical decision-making.