A Meta-analysis Of Perventricular Device Closure Of Perimembranous Ventricular Septal Defect

ObjectiveTo investigate the safety and efficacy of perventricular device closure of perimembranous VSD(pm VSD).MethodsPubMed(MEDLINE)、EMBASE and Cochrane were searched for studies in English focusing on perventricular device closure of pm VSD published up to the end of March2019.The random effect model was applied to the combined estimation of the success rate of plugging and the incidence of complications,and the corresponding forest map was made.Funnel figure was made to evaluate publication bias based on sample size and the success rate.If the funnel figure showed obvious asymmetry,Egger’s and Begg’s test were further done to verify the existence of publication bias.Trim-fill test were further done to evaluate the effect of publication bias.The inconsistent statistics(I~2)was used to assess heterogeneity.If I~2value greater than 50%suggested that the existance of heterogeneity,sensitivity analysis were done to find causes of heterogeneity for further analysis.Statistically significance is defined as 5%level on both sides.Stata 15 software was used in this study.ResultsA total of 15 publications comprising 1368 patients with pm VSD were included.The median follow-up duration was 2 months to 5 years,with a mean patient age ranging from 2 months to 56 years.The pooled success rate was 0.95(I2=86.2%,P=0.000).The pooled rate of postoperative residual shunting was 0.02(95%CI:0.01-0.03,I2=87.3%,P<0.001).The pooled rate of residual shunting in the follow-up period was 0.001(95%CI:-0.001-0.002,I2=30.5%,P=0.126).The pooled estimated rate of severe complications was 0.074(95%CI:0.046-0.102,I2=30.5%,P=0.126).The pooled incidence of complete atrioventricular block(c AVB)was 0.002(95%CI:0.000-0.005,I2=0.0%,p=0.577).ConclusionPerventricular device closure may be an alternative to conventional surgical repair in selected patients with pm VSD.The success rate was stable regarding the publication year and sample size and suggested both the short learning curve of this technology and its potential for wide application.The incidence of severe arrhythmia,especially c AVB,was low.These good results may be limited by the number of enrolled patients,and a more detailed and larger sample is required for further analysis.