新生儿危重症先天性心脏病适宜筛查窗口期和筛查地点的系统评价和Meta分析

Appropriate screening timing and locations for neonatal major congenital heart disease:A systematic review and meta-analysis

背景针对新生儿先天性心脏病(CHD)筛查的组配方案的系统评价报道较多,而筛查的最佳窗口期和适宜的筛查机构也同样值得关注。目的评估在不同筛查窗口期和筛查机构中进行新生儿危重症CHD(mCHD)筛查的诊断准确性和假阳性率。设计诊断准确性研究的Meta分析。方法本文文献检索和筛选是在新生儿CHD筛查(NCHDS)指南中实现的。筛查金标准:筛查阳性婴儿行心脏超声检查确诊为mCHD,或随访出现CHD症状和体征后经心脏超声或手术或心脏导管介入术证实为mCHD。以QUADAS-2量表对纳入文献行偏倚风险及临床适用性评价,提取文献数据,应用随机效应模型汇总敏感度、特异度、假阳性率和假阴性率等诊断准确性参数。主要结局指标不同筛查窗口期和筛查机构的筛查准确性。结果①关于筛查窗口期的16篇文献中,筛查窗口期在生后≤24 h 9篇,~48 h 7篇,~72 h 3篇,6~72 h 3篇,筛查新生儿mCHD的敏感度分别为0.788(95%CI:0.600~0.921)、0.579(95%CI:0.378~0.757)、0.586(95%CI:0.369~0.775)和0.897(95%CI:0.836~0.937),特异度分别为0.985(95%CI:0.946~0.996)、0.998(95%CI:0.994~0.999)、0.996(95%CI:0.987~0.999)和0.994(95%CI:0.983~0.998),假阳性率分别为0.015(95%CI:0.004~0.057)、0.002(95%CI:0.001~0.006)、0.004(95%CI:0.001~0.016)和0.006(95%CI:0.002~0.019);生后≤24 h、~48 h、~72 h和6~72 h筛查新生儿mCHD的AUC分别为0.916、0.973、0.939和0.973。②关于筛查机构的20篇文献中,产院15篇、NICU 2篇、社区2篇、产院+NICU 1篇,筛查新生儿mCHD的敏感度分别为0.701(95%CI:0.576~0.802)、0.797(95%CI:0.675~0.881)、0.571(95%CI:0.230~0.856)和0.821(95%CI:0.555~0.944);特异度分别为0.995(95%CI:0.990~0.998)、0.885(95%CI:0.778~0.944)、0.993(95%CI:0.988~0.996)和0.916(95%CI:0.189~0.998);产院和NICU筛查新生儿mCHD的假阳性率分别为0.006(95%CI:0.003~0.011)和0.030(95%CI:0.001~0.636),AUC分别为0.960和0.757;产院生后≤24 h亚组敏感度和特异度分别为0.817和0.996,生后>24 h亚组敏感度和特异度分别为0.680和0.996。结论追求筛查的高敏感度、接受高假阳性率时可以选择生后6~24 h内进行筛查;追求低假阳性率时可选择生后24~72 h之后筛查,应选择在产院开展筛查mCHD,准确性高、假阳性率低。

Background There are many reports on the systematic review of the screening protocols for neonatal congenital heart disease,and the appropriate screening timing and locations for screening are also worthy of attention.Objective To evaluate the diagnostic accuracy and false positive rate of neonatal major congenital heart disease(mCHD)screening at different screening timing and locations.Design Meta-analysis of diagnostic accuracy studies.Methods The literature search and selection have been completed in the development of neonatal CHD screening(NCHDS)guideline.The positive infants in screening were diagnosed as mCHD by cardiac ultrasound(gold standard),or infants with symptoms and signs were confirmed as mCHD in the follow-up by cardiac ultrasound,surgery,or cardiac catheterization(gold standard).The risk of bias and clinical applicability of the included literature were evaluated with the QUADAS-2 scale.The data were extracted and the diagnostic accuracy parameters such as sensitivity,specificity and false positive rate were summarized with the random effect model.Main outcome measures Accuracy of neonatal mCHD screening at different timing and locations.Results Among the 16 studies on the screening timing,9 were in the postnatal period of≤24 h,7 were in the period of-48 h,3 were in the period of-72 h and 3 were in the period of 6-72 h.The sensitivity of neonatal mCHD screening at different timing were 0.788(95%CI:0.600-0.921),0.579(95%CI:0.378-0.757),0.586(95%CI:0.369-0.775)and 0.897(95%CI:0.836-0.937),respectively.And the specificity were 0.985(95%CI:0.946-0.996),0.998(95%CI:0.994-0.999),0.996(95%CI:0.987-0.999)and 0.994(95%CI:0.983-0.998),respectively.The false positive rates were 0.015(95%CI:0.004-0.057),0.002(95%CI:0.001-0.006),0.004(95%CI:0.001-0.016)and 0.006(95%CI:0.002-0.019),respectively.The AUC of neonatal mCHD screening at≤24 h,-48 h,-72h and 6-72 h after birth was 0.916,0.973,0.939 and 0.973,respectively.Among the 20 literatures on screening locations,there were 15 in maternity hospitals,2 in NICU,2 in out of hospital settings,and 1 in maternity hospitals+NICU.The sensitivity of neonatal mCHD screening was 0.701(95%CI:0.576-0.802),0.797(95%CI:0.675-0.881),0.571(95%CI:0.230-0.856)and 0.821(95%CI:0.555-0.944),respectively.The specificity was 0.995(95%CI:0.990-0.998),0.885(95%CI:0.778-0.944),0.993(95%CI:0.988-0.996)and 0.916(95%CI:0.189-0.998),respectively.The false positive rates of the neonatal mCHD screening in maternity hospital and NICU were 0.006(95%CI:0.003-0.011)and 0.030(95%CI:0.001-0.636)respectively,and the AUC were 0.960 and 0.757 respectively.The sensitivity and specificity were 0.817 and 0.996 in the subgroup within 24 h after birth,and 0.680 and 0.996 in the subgroup>24 h after birth.Conclusion When pursuing high sensitivity of screening and accepting high false positive rate,screening within 24 hours after birth can be selected.For low false-positive rate,screening after 24 hours after birth can be selected.Neonatal mCHD screening should be implemented in the maternity hospital for its good accuracy and low false-positive rate.