| Objective To investigate the diagnostic value of interleukin-6(il-6) as well ascombined with other diagnostic biomarkers of neonatal sepsis, and then to assess thediagnostic value of serum amyoid A (SAA) for neonatal sepsis.Materials and Methods The literatures about the il-6test for the diagnosis ofneonatal sepsis were searched from Pubmed, Embase and the Cochrane Library and othersources by two reviewers independently. The inclusion and exclusion criteria meet thestandards of diagnostic test. Each selected literature was evaluated according to QUADAS(quality assessment of diagnostic accuracy studies) by two reviewers independently, andthen scored it. The dispute will be resolved through consultation with the third partner. Theliterature data were collected by fourfold table. Meta-disc1.4, RevMan5and Stata11wereused for statistical analysis. Firstly, the sensitivity, specificity and diagnostic odds ratio(DOR) about the accuracy of biomarkers were pooled, and then constructed the receiveroperating characteristic curve (SROC) to acquire the area under the SROC curve(AUC)and Q*. At the same time, the P-value indicating whether the heterogeneity has statisticalsignificance was acquired by chi-square test, and I2was also calculated to quantitativelyassess heterogeneity. Regression analysis was applied to explore the main factors causingthe heterogeneity, In addition, linear regression drawing Deek figures were plotted todetect publication bias for inclusive literatures.Results The total of44relevant literatures about the il-6test for the diagnosis ofneonatal sepsis including4,758neonates was included. In the diagnosis of early-onsetneonatal sepsis (ENOS), the pooled sensitivity and specificity respectively are0.76(95%CI,0.72-0.80) and0.81(95%CI,0.79-0.83), and AUC is0.90(Q*=0.83). Especially, we foundthat Preterm neonates in Europe with preterm rupture of membranes(PPROM) and cordblood sample using the il-6test for the diagnosis of neonatal sepsis have better diagnosticindicators (0.87,0.88,0.96,0.86) in ENOS. In the diagnosis of early-onset neonatal sepsis(LNOS), the pooled sensitivity and specificity respectively are0.80(95%CI,0.76-0.84) and0.83(95%CI,0.81-0.86), and the area under the SROC curve (AUC) is0.88(Q*=0.81). Meanwhile, we also found that the pooled indicators of the il-6test for diagnosing theEuropean sample are0.84,0.89,0.93,0.87. In other categories, they show differentperformance for the il-6test. About joint diagnostic analysis, the summary results of thecombination of il-6and CRP in the diagnosis of neonatal sepsis are0.72（0.68-0.76）forpooled sensitivity,0.77(0.74-0.80) for pooled specificity and0.93for the AUC (Q*=0.86).the pooled results of the combination of il-6and TNF-α in the diagnosis of neonatal sepsisare0.95(95%CI0.90-0.98) for pooled sensitivity,0.83(95%CI0.78-0.86) for pooledspecificity and0.97for the AUC (Q*=0.92).About the SAA test for the diagnosis of neonatal sepsis, a total of9relevant literatureswith823neonates meet inclusion criteria. In the first suspicion of neonatal sepsis, thesummary results of the SAA test of the diagnosis of neonatal sepsis are respectively0.84(95%CI,0.80-0.87) for sensitivity and0.89(95%CI,0.86-0.92) for specificity, and theAUC is0.96(Q*=0.91); In8-96hours after the first suspicion of neonatal sepsis, thepooled sensitivity and specificity is0.78(95%CI,0.72-0.84) and0.84(95%CI,0.79-0.88),respectively.Conclusion The il-6test as a biomarker for NS has shown its moderate accuracy.Especially in the diagnosis of ENOS using the sample of the European neonates with Cordblood samples and PPROM and only using the sample of the European or prematureneonates in LONS, it shows better diagnostic accuracy. The combination of il-6and CRP,especially in LONS and the combination of il-6and TNF-α have better diagnostic accuracythan use il-6alone; SAA at the first suspicion of neonatal sepsis shows moderate diagnosticadvantages, and keeps its usefulness at8-96hours after the first suspicion of neonatalsepsis. |
PDF Full Text Request