孕早期胎儿颈项透明层在产前超声筛查胎儿畸形中价值的Meta分析

Meta-analysis of the value of nuchal translucency as a screening test for fetal abnormalities in the first trimester ultrasound examination

目的探讨孕早期胎儿颈项透明层(NT)在产前超声筛查胎儿畸形(特别是心脏畸形)中的价值。方法检索Cochrane图书馆、PubMed、OVID、Springer数据库、中国期刊全文数据库和中国生物医学文献数据库(1990年1月至2008年8月)中孕早期NT产前超声筛查胎儿畸形(特别是心脏畸形)的中英文文献,按照诊断试验的纳入和排除标准筛选文献,提取纳入研究的特征信息。数据分析采用Meta-DiSc1.4软件,检验异质性,并根据异质性选择相应的效应模型。对所有研究予以加权定量合并分析,计算汇总敏感度和特异度及其95%CI。绘制汇总受试者工作特征曲线(SROC),并计算曲线下面积(AUC),最后剔除异质性文献和孕14周后检测NT厚度的文献进行敏感度分析。结果共纳入文献12篇,检查胎儿共计112099例。孕早期胎儿NT≥P95和P99对心脏畸形产前诊断的各研究间存在异质性,Meta分析合并效应量时采用随机效应模型;孕早期胎儿NT≥P95对胎儿畸形产前诊断的各研究间无异质性,Meta分析合并效应量时采用固定效应模型。9篇文献对孕早期胎儿NT≥P95对心脏畸形的诊断价值进行了分析,汇总灵敏度、特异度和SROCAUC分别为29.8%、96.9%和0.8047;7篇文献对孕早期胎儿NT≥P99对心脏畸形的诊断价值进行了分析,汇总灵敏度、特异度和SROCAUC分别为18.9%、99.3%和0.9712;5篇文献对孕早期胎儿NT≥P95对胎儿畸形的诊断价值进行了分析,汇总灵敏度、特异度和SROCAUC分别为25.0%、98.3%和0.1830。孕早期胎儿NT≥P95对心脏畸形产前诊断的汇总敏感度高于NT≥P99(χ2=6.58,P<0.05),两者的汇总特异度差异无统计学意义(P>0.05)。剔除异质性文献和14周后检测NT厚度的文献对汇总敏感度和特异度影响不大。结论孕早期胎儿NT增厚对笼统的胎儿畸形产前筛查的准确性低,产前筛查的价值不大;NT对不同种类胎儿畸形筛查的敏感度和特异度,目前还不完全清楚。本文对孕早期胎儿NT增厚对胎儿畸形中的心脏畸形产前筛查的特异度和准确性较高,可用于孕早期胎儿畸形中的心脏畸形的产前筛查,NT对其他不同种类胎儿畸形的筛查价值,需进一步的系统评价。中国孕早期胎儿NT厚度的判断标准和测量者间的技术水平差异较大,急需大规模、前瞻性和多中心的临床研究对中国孕早期胎儿NT厚度的标准和孕早期胎儿NT增厚对胎儿畸形(特别是心脏畸形)的筛查和诊断价值进行评估。

Objective To evaluate the value of nuchal translucency （NT） as a screening test for fetal abnormalities in the first trimester ultrasound examination. Methods A search in Cochrane Library, PubMed, OVID, Springer, China National Knowledge Infrastructure （CNKI） and Chinese Bio-medicine Database （CBM）was performed to identify relevant English and Chinese articles between Jan 1990 and August 2008. Criteria for inclusion were established based on validity criteria for diagnostic research. Subsequently, the characteristics of the included articles were extracted. Statistical analysis was performed with Meta- DiSc 1.4. Heterogeneity of the included articles was tested to select proper effect model to calculate pooled weighted sensitivity and specificity. Summary receiver operating characteristic curve （SROC） was made and the area under the curve （AUC） was calculated. Finally, sensitivity analysis was performed. Results Twelve articles were included, with a total of 112 099 fetuses. Nine articles meeting inclusion criteria were analyzed for the value of screening for cardiac abnormalities by nuchal translucency thickness above the 95th percentile for gestational age. The pooled sensitivity and specificity was 29.8%, 96. 9%, respectively, AUC =0. 804 7. Seven articles meeting inclusion criteria were analyzed for the value of screening for cardiac abnormalities by nuchal translucency thickness above the 99th percentile for gestational age. The pooled sensitivity and specificity was 18.9%, 99. 3% ,respectively, AUC = 0. 971 2. Five articles meeting inclusion criteria were analyzed for the value of screening for fetal abnormalities by nuchal translucency thickness above the 95th percentile for gestational age. The pooled sensitivity and specificity was 25.0%, 98. 3% respectively, AUC = 0. 183 0. The sensitivity of screening for cardiac abnormalities by nuchal translucency thickness above the 95th percentile for gestational age was significantly higher than the nuchal translucency thickness above the 99th percentile for gestational age（ X2 = 6. 58, P 〈0. 05 ） ,no difference was found in the specificity. The random-effect model was used in the analysis of screening for cardiac abnormalities by nuchal translucency thickness above the 95th or 99th percentile for gestational age because of the heterogeneity. The fixed-effect model was used in the analysis of screening for fetal abnormalities by nuchal translucency thickness above the 95th percentile for gestational age because of the homogenicity. Conclusions The results suggest that screening for fetal abnormalities by NT thickness above the 95th percentile for gestational age has a low accuracy and is unsuitable for prenatal screen. At present it is incompletely clear on the sensitivity and specificity of screening for different kinds of fetal abnormalities by NT thickness. Screening for cardiac abnormalities by first trimester NT thickness has a high accuracy and specificity and can be reliably used in prenatal screen. We need the further systemic reviews to evaluate the value of NT as a screening test for different kinds of fetal abnormalities. Further multi-centre and prospective studies are badly needed to be performed to optimize the standard of nuchal translucency thickness, evaluate the value of NT as a screening test for fetal abnormalities （especially cardiac abnormalities） in the first trimester ultrasound examination by a cost-effectiveness analysis, and develop an operation guideline on first trimester nuchal translucency thickness measurement for the prenatal screening of cardiac abnormalities in China.