摘要
目的分析常用的几种巨大儿产前临床预测方法的应用价值。方法将2000年1月1日至2005年3月31日在北京大学第三医院妇产科进行产前检查并分娩巨大儿的273例孕妇临床资料与随机抽取的同期单胎分娩正常体重儿的135例孕妇临床资料作对比分析。采用统计学方法对临床常用巨大儿预测方法包括B超综合指标估算胎儿体重(estimated fetal weight,EFW)法、B超双指标估算法、胎儿双顶径(biparietal diameter,BPD)与股骨长(femur length,FL)之和]、B超单一指标估算法(胎儿腹围、股骨长)、临床宫高与腹围之和法以及经验性临床评估法进行评价。结果(1)巨大儿组与非巨大儿组孕妇年龄无统计学差异,但身高和终止妊娠孕周均有统计学意义[(163.0±5.47)cm和(160.0±4.7)cm;(39.7±1.2)周和(39.1±1.1)周,P<0.01]。(2)临床常用的巨大儿预测指标及不同界值的预测价值,EFW法灵敏度为38.1%,特异度为96.4%。B超双指标估算法(BPD+FL≥17 cm)灵敏度为54.0%,特异度为83.8%;当界值取16.5 cm时,灵敏度为88.8%,特异度为55.9%;B超双指标估算法的AC+FL≥42.5 cm时,灵敏度为90.8%,特异度为82.9%。B超单一指标法,以AC≥35 cm为界定值时其灵敏度为95.0%,特异度为71.0%;以38 cm为界定值时灵敏度为31.0%,特异度为99.1%。B超单一采用FL指标,以7.5 cm为界定值时灵敏度为49.8%和,特异度为82.0%。临床宫高与腹围之和法的界定值为≥140 cm时,其灵敏度为77.9%特异度为77.5%。经验性临床评估法灵敏度为46.3%特异度为96.6%。(3)各预测指标的ROC曲线下面积:EFW法0.906;B超双指标估算法中,BPD+FL之和0.795;AC+FL之和0.914;B超单一指标估算法中,AC 0.904,FL 0.752;临床宫高腹围之和0.862。(4)对于各个独立参数采用Logistic多因素回归分析,求得概率(P)方程及P的ROC曲线下面积为0.938。结论本研究显示了在某些病例中临床上预测巨大儿方面仍然有不可预测性。但各项指标单独应用时均有较大局限性至今尚无一种方法可以准确预测巨大儿。胎儿腹围与股骨长之和可能成为一个临床价值较好的预测指标。综合孕妇身高、终止孕周、胎儿腹围、宫高腹围和多参数分析可提高预测的准确性,降低漏诊率。
Objective To evaluate the clinical values of some predictive methods for macrosomia in clinical practice. Methods Totally, 273 pregnant women with fetal macrosomia were enrolled in this study at Peking University Third Hospital from Jan. 2000 to Mar. 2005 and 135 singleton pregnant women with normal neonatal birth-weight were randomly chosen as a control group. Different measurements of fetal weight used in clinical practice by ultrasonography and clinical examination were compared and analyzed statistically. Cross-table analysis and Logistic linear regression were adopted to estimate their clinical value. Results （1） No statistical difference in maternal age was found between the two groups （P〉0.05）, but the maternal height and gestations at delivery showed significant difference [（163.0±5.47） cm vs （160.0±4.7） cm;（39.7±1. 2） wks vs （39. 1±1. 1）wks,P〈 0.01]. （2） Clinical value of these measures for predicting macrosomia was estimated. Estimating fetal weight （EFW） by Sonographer had a sensitivity of 38.1% with a specificity of 96.4%. The combination of fetal biparietal diameter （BPD） with femur length （FL） showed a sensitivity of 54.0% at the cut-off level of 17 cm, with a specificity of 83.8%. At the cut-off level of 16.5 cm, the sensitivity and specificity were 88.8% and 55.9%, respectively. The combination of fetal abdominal circumference （AC） and FL had the sensitivity of 90.8% and a specificity of 82.9% at the cut-off level of 42.5 cm. AC alone had a sensitivity of 95.0% at the cut-off level of 35 cm with a specificity of 71.0%. While at the cut-off level of 38 cm, the sensitivity and specificity were 31.0% and 99.1%, respectively. FL alone had a sensitivity of 49.8% at the cut-off level of 7.5 cm with a specificity of 82.0%. The measure using the sum of uterus fundal height and maternal AC showed a sensitivity of 77.9% at the cut-off level of 140 cm and with a specificity of 77.5 %. Finally, the sensitivity and specificity of experienced Leopoldrs maneuvers were 46.3% and 96.6%, respectively. （3） The areas under receiver operating characteristic curves （ROC） of the predictors were made. The results were： EFW 0. 906, BPD＋FL 0. 795, AC＋ FL 0. 914, AC 0. 904, FL 0. 752, and sum of uterus fundal height and maternal AC 0. 862. （4） The equation of probability of macrosomia can be obtained by Logistic linear regression, and the area under ROC of the probability was 0. 938. Conclusions Our study showed that although some means for predicting macrosomia play a role in clinical application, the limitation in either combination methods or single method alone remains. Misdiagnosis and underdiagnosis could not avoid in some cases. The sum of AC and FL may serve as better means for the clinical estimation of fetal weight. Multivariate analysis may improve the veracity of prediction.
出处
《中华围产医学杂志》
CAS
2006年第4期251-254,I0005,共5页
Chinese Journal of Perinatal Medicine
关键词
巨大胎儿
超声检查
产前
预测
Macrosomia
Ultrasonography, prenatal
Forecasting
