A Meta Analysis Of Status Epilepticus Severity Score For Its Clinical Application

Objective: 1.To investigate the diagnostic value of status epilepticus severity score(STESS)for judging the survival condition in status epilepticus(SE)patients.2.To investigate whether STESS at the cut-off value 3 or 4 could be a ruled-out tool.3.To investigate which could be the best cut-off value for STESS.Methods: PubMed,EMBASE,CBM,CNKI were searched to identify and retrieve relevant studies on STESS from Jan.2005 to Jan.2017.QUADAS-2 and Revman5.3 were applied to the quality evaluation of the publications.Meta-analysis was performed by Meta.Disc1.4 and STATA12.0 software.The pooled diagnostic characteristics and the corresponding 95%CI were obtained.We drew the summary receiver operating characteristic(SROC)curve and figured out the area under the curve(AUC).At last,subgrouping capacities,meta regression and sensitivity analysis were fulfilled.Results: Altogether 10 literatures including 17 studies and 1878 episodes were incorporated.1.All The 17 literatures had threshold effect caused heterogeneity which made us could only perform the SROC curve,The SROC-AUC was 0.7495.2.For STESS-3 group,diagnostic criterion caused the heterogeneity after subgroup analysis,meta regression and sensitivity analysis.We choosed the 5 minutes group as the sample,pooled sensitivity and specificity calculated by fixed effect model were 0.817(95% CI:0.752-0.870)and 0.527(95% CI:0.492-0.561).Pooled positive likelihood ratio negative likelihood ratio 1.688(95% CI:1.362-2.091)and 0.398(95% CI:0.283-0.560).The DOR was 4.822(95% CI:2.936-7.918)and the SROC-AUC was 0.740.3.For STESS-4 group,we did not find the source caused the heterogeneity.We choosed all the 7 studies as the sample,pooled sensitivity and specificity calculated by fixed effect model were 0.595(95% CI:0.500-0.685)and 0.770(95% CI:0.735-0.803).Pooled positive likelihood ratio and negative likelihood ratio were 2.556(95%CI:2.077-3.146)and 0.582(95% CI:0.404-0.837).The DOR was 5.249(95% CI:2.797-9.850)and the SROC-AUC was 0.7542.Conclusion: 1.STESS has a moderate efficiency inidentifying the survival condition in patients with SE.2.When STESS is greater than the cutoff value,it indicates patients will die.Otherwise,the patiens will survive.As the negative likelihood ratio are greater than 0.1,it could not be a a ruled-out tool.3.The STESS AUC at the cutoff value 4 are greater than it at the cutoff value 3,we may guess 4 may be the better one.However,the STESS-4 group is not as homogeneous as STESS-3,more prospective researchs are needed.