论文部分内容阅读
目的建立新的预测足月胎儿出生体质量(FW)的回归方程;比较新建方程与现有19种方程(5种临床参数方程及14种超声参数方程)的准确性。方法选取2015年10月~2016年4月于产前0~5 d在华北理工大学附属医院进行超声检查并住院分娩的单胎孕足月孕妇224例,收集孕妇的临床及超声资料。根据胎儿实际体质量,分为非巨大儿组(FW<4 000 g,n=183)和巨大儿组(FW≥4 000 g,n=41)。使用绝对误差评价不同方程预测FW的准确性。采用Pearson相关性分析及多重线性回归法建立新的回归方程,方差分析、LSD检验比较不同方程准确性。结果新建立了3个方程:临床参数方程、超声参数方程、联合参数方程。在非巨大儿组、巨大儿组及整体中预测FW绝对误差最小的均为联合参数方程:EFW=40.39×宫高+84.60×AC+244.17×TCD+221.79×FSTT+141.98×LL+127.31×BPD-4 467.98。5种临床参数方程中,预测FW绝对误差最小的分别为卓晶如法(非巨大儿组、整体)和罗来敏法(巨大儿组);14种超声参数方程中,在非巨大儿组及整体中绝对误差最小的方程均为Hadlock FP(BPD,HC,AC,FL)法;在巨大儿组中为Merz E(BPD,AC)法。与已知方程相比,新建立的联合参数方程绝对误差最小(P<0.05)。结论联合参数方程可以较准确地预测FW,且与已知的19种方程比较,其预测FW的绝对误差最小。
Objective To establish a new regression equation to predict the birth weight (FW) of full-term fetuses and to compare the accuracy of the new equation with the existing 19 equations (five kinds of clinical parameter equations and 14 kinds of ultrasound parameter equations). Methods Totally 224 singleton pregnant term pregnant women were enrolled in the Affiliated Hospital of North China University of Science and Technology from October 2015 to April 2016 at 0 ~ 5 days. The clinical and ultrasound data of pregnant women were collected. According to the actual body weight of the fetus, they were divided into non-giant children group (FW <4 000 g, n = 183) and giant children group (FW≥4 000 g, n = 41). The absolute error is used to evaluate the accuracy of different FW equations. Pearson correlation analysis and multiple linear regression method were used to establish new regression equation, variance analysis and LSD test to compare the accuracy of different equations. Results Three new equations were established: clinical parameter equation, ultrasonic parameter equation and joint parameter equation. In the non-macrosomia group, the macrosomia group and the whole body, the predicted absolute error of the FW is the joint parameter equation EFW = 40.39 × H + 84.60 × AC + 244.17 × TCD + 221.79 × FSTT + 141.98 × LL + 127.31 × BPD -4 467.98.5 The clinical parameters of the equation, the smallest prediction of FW absolute error were Zhuo Jingru method (non-huge children group, the whole) and Luo Laimin method (giant children group); 14 kinds of ultrasonic parameter equation, in the non-huge The equations with the smallest absolute error in the group and in the population were all Hadlock FP (BPD, HC, AC, FL) and Merz E (BPD, AC) in the macrosomia. Compared with the known equations, the absolute error of the newly established joint parameter equation is the smallest (P <0.05). Conclusion The joint parametric equation can predict the FW more accurately, and compared with the known 19 kinds of equations, the absolute error of the predicted FW is the smallest.