初产妇会阴裂伤风险列线图预测模型的建立与验证

Construction and verification of nomogram prediction model of risk for perineal laceration in primiparas

目的构建初产妇会阴裂伤风险列线图模型。方法选取2020年12月至2021年9月同济大学附属第一妇婴保健院的932名初产妇为研究对象。746名初产妇作为建模组构建模型,按照会阴是否发生裂伤将其分为裂伤组(506例)和未裂伤(240例)组;采用logistic回归建模,Hosmer-Lemeshow拟合优度、校准曲线和受试者操作特征曲线评价模型。结果裂伤组第一、第二和总产程时间长于未裂伤组,差异有统计学意义(P<0.05);裂伤组足月产、可疑巨大儿、硬膜外镇痛、就诊助产士门诊、胎儿窘迫和妊娠高血压占比高于未裂伤组,差异有统计学意义(P<0.05);两组会阴弹性、会阴体长度、产妇配合程度、助产士年资、引产方式、自由体位、分娩体位、分娩恐惧程度和孕期体力活动水平比较,差异有统计学意义(P<0.05)。胎龄、可疑巨大儿、第二产程时间、分娩恐惧程度、助产士年资、孕期体力活动水平是初产妇会阴裂伤的影响因素(P<0.05)。Hosmer-Lemeshow拟合优度和校准曲线提示模型拟合佳。建模组和验证组曲线下面积分别为0.822和0.799,模型预测性能良好。结论本研究构建的会阴裂伤风险列线图预测模型具有良好的区分度和校准度,可为临床提供高效便捷的评估工具。

Objective To construct the nomogram prediction model of risk for perineal laceration in primiparas.Methods A total of 932 primiparas from the First Maternity and Infant Hospital,Tongji University from December 2020 to September 2021 were selected as research objects.A total of 746 primiparas were used as the modeling group to construct the model,and they were divided into laceration group(506 cases)and non-laceration group(240 cases)according to whether perineal laceration occurred;186 primiparas were included in validation group to validate the model.Logistic regression model was used to construct the model,Hosmer-Lemeshow goodness of fit,calibration curve,and receiver operating characteristic curve were used to evaluate the model.Results The first labor time,second labor time,and total labor time in laceration group were longer than those in non-laceration group,and the differences were statistically significant(P<0.05).The proportion of full-term birth,suspected macrosomia,epidural analgesia,midwife outpatient service,fetal distress,and hypertension of pregnancy in laceration group was higher than that in non-laceration group,and the difference was statistically significant(P<0.05);there were significant differences in perineal elasticity,length of perineal body,degree of puerperal cooperation,seniority of midwife,mode of induced labor,free position,delivery position,degree of fear of childbirth,and physical activity level during pregnancy between two groups(P<0.05).Gestational age,suspected macrosomia,second labor time,degree of fear of childbirth,seniority of midwives,and physical activity level during pregnancy were the influencing factors of perineal lacerations in primiparas(P<0.05).The Hosmer-Lemeshow goodness of fit and calibration curve indicated that the model fit was good.The area under the curve of the modeling group and the verification group were 0.822 and 0.799,respectively,and the prediction performance of the model was good.Conclusion The prediction model of risk for perineal laceration in this study has good differentiation and calibration,which can provide an efficient and convenient evaluation tool for clinical practice.