Diagnosis Value of Cerebral Ultrasound on White Matter Damage: A Meta Analysis

Objectives A meta analysis was performed to determine the value and accuracy in diagnosis of white matter damage. Methods A comprehensive and systematic search was conducted in PubMed, EMBASE, CBM, Wanfang, Weipu and CJFD databases. A software called ‘Meta-disc’ was used to obtain pooled estimates. Results Ten articles were included. The included studies had heterogeneity（I² of sensitivity=0.753, I² of specificity=0.91）. Mean sensitivity and specificity was 97.0% （95.0%~98.0%） and 97.4% （96.4%~98.2%）, respectively. Summary positive likelihood ratio was 36.24（9.67~135.83）, summary negative likelihood ratio was 0.041 （0.02~0.10） and summary diagnostic odds ratio was 1 210.8（182.57~8 029.5）. The AUC of SROC was 0.994 2, SE=0.012 5. Sensitivity analysis demonstrated no disproportionate influences of individual studies. Deeks funnel plot suggested the publication bias was not significant（P=0.07）. Conclusion SROC shows that cerebral ultrasound is of value in diagnosis of white matter damage.