目的 探讨CRP诊断新生儿败血症的诊断价值.方法 计算机检索维普中文科技期刊数据库、中国期刊全文数据库、万方数据库、中国生物医学文献数据库、Medline、EMBASE、Cochrane图书馆、PubMed等数据库(检索时间为1989年1月至2011年7月),获得CRP诊断新生儿确诊败血症的文献.采用QUADAS工具对纳入文献进行质量评价.采用MetaDisc 1.4和RevMan 5.0软件检验各文献间的异质性,并根据异质性结果选择相应的效应模型进行加权定量合并,计算敏感度和特异度及其95%CI.绘制受试者工作特征(SROC)曲线并计算曲线下面积(AUC),并行敏感度分析.结果 20篇文献进入Meta分析.分析结果显示:CRP诊断新生儿败血症的汇总敏感度和特异度分别为0.69(95%CI:0.65～0.72)和0.87(95%CI:0.86～0.89),SROC AUC为0.88,Q*指数为0.81.CRP＞8 mg·L-1诊断新生儿败血症的汇总敏感度和特异度分别为0.88(95%CI:0.82～0.92)和0.86 (95%CI:0.81～0.90),SROC AUC为0.94,Q*指数为0.88;CRP≥8 mg·L-1诊断新生儿败血症的汇总敏感度和特异度分别为0.54(95%CI:0.47～0.61)和0.81(95%CI:0.76～0.85),SROC AUC为0.80,Q*指数为0.74;两者合并诊断新生儿败血症的汇总敏感度和特异度分别为0.70(95%CI:0.65～0.74)和0.83(95%CI:0.80～0.86),SROC AUC为0.88,Q*指数为0.82.在纳入文献中标本来源和检测方法的差异等非阈值效应因素是产生异质性的原因.结论 CRP对新生儿败血症诊断的敏感度和特异度较高,有助于早期诊断新生儿败血症.%Objective Despite major advances in neonatal care during the past few decades, neonatal septicemia has high mortality and morbidity. The usage of CRP has recently been proven to be useful for the diagnosis of neonatal septicemia. A meta-analysis was performed to assess the accuracy of the CRP test for diagnosing neonatal septicemia. Methods VIP, China National Knowledge Infrastructure ( CNKI ), Wanfang Chinese Periodical Database and Chinese Bio-medicine Database ( CBM ), Medline, EMBASE, Cochrane library and PubMed were searched to identify retrieve relevant studies on CRP in diagnosis of neonatal septicemia from January 1989 to July 2011. QUADAS tools were used for quality evaluation of literature by two reviewers. Meta-analysis was performed by employing MetaDisc 1.4 and Revman 5. 0 software. Heterogeneity of the included articles was tested to select proper efficacy model for calculate pooled weighted sensitivity, specificity and 95% CI. Summary receiver operating characteristic ( SROC ) curve was made and the area under the curve ( AUC ) and Q index were calculated. Finally, sensitivity analysis and comparison of sensitivity among different groups were performed. Descriptive analysis was performed if meta-analysis was inappropriate. Results A total of 739 literatures were reviewed, including 430 relevant English articles and 307 Chinese articles. Twenty studies including 9 English records and 11 Chinese records met the inclusion criteria. Sensitivity and specificity of CRP test for the diagnosis of neonatal septicemia was 0. 69( 95% CI:0. 65 -0. 72 )and 0. 87( 95% CI 0. 86 -0. 89 ). The area under SROC curve was 0. 88, Q index was 0. 81. The pooled sensitivity , specificity, AUC and Q index for CRP > 8 mg ? L~ were 0. 88 ( 95%CI:0. 82-0. 92 ),0. 86 ( 95%CI:0. 81 -0. 90 ),0. 94 andO. 88 respectively. The pooled sensitivity, specificity, AUC and Qindex for CRP&8 mg ? L"1 were 0. 54( 95% CI:0. 47 - 0. 61 ) ,0. 8l( 95% CI:0. 76 - 0. 85 ) ,0. 80 and 0. 74 respectively. The pooled sensitivity, specificity ,SROC curve and Q index for CRP 8 mg ? L"1 were 0. 70( 95% CI:0. 65 -0. 74 ),0. 83( 95% CI: 0. 80 - 0. 86 ), 0. 88 and 0. 82 respectively . CRP assay method and differences in sample sources may partially explain the heterogeneity. Conclusions CRP test as a non-invasive test is a highly accurate method for diagnosing neonatal septicemian. CRP should be combined with other diagnostic markers to further improve the sensitivity and accuracy in the diagnosis of septicemia.
展开▼
