Network Meta-analysis Of Interventional Therapy For Gastric Variceal Hemorrhage In Cirrhosis

Objective:To compare to each other non-surgical interventions for gastric variceal bleeding(GVB)associated with liver cirrhosis by network meta-analysis in order to provide evidence support for clinical emergency hemostasis and secondary prevention.Methods:According to Cochrane Library search strategy,the literatures collected from Database Day to March 2022 in major Chinese and English databases were searched,including Cochrane Library,Embase database,Pub Med,Science Direct,CBM,CNKI,WANFANG MED ONLINE,VIP Paper Check System,Baidu Academic,Google Academic,etc.According to the established inclusion and exclusion criteria,clinic randomized controlled trials and case-control studies were accepted in searches of several drug therapy for GVB involving hemostatic control rate,variceal remission rate,rebleeding rate,incidence rate of hepatic encephalopathy and death,and the literatures were searched and screened,and then the final included literatures were evaluated for quality and data extraction.Clinic randomized controlled trial were evaluated by Cochrane collaboration’s tool for assessing risk of bias version 5.1.0.And the Newcastle-Ottawa Scale which was for assessing the quality of nonrandomized studies was carried out.Reviews Manger 5.4.1software,R language software 4.1.3 and R Studio 1.1.463 were used in statistical complement each other,and the pairwise,internal,inter-group and overall statistical heterogeneity were analyzed,while network meta-analysis were operated by ADDIS 1.16.8software.Results:A total of 38 literatures were included,including 10 randomized controlled studies and 28 case-control studies.The comparative analysis of active bleeding control rate of mixed therapy showed that in the comparative analysis of mixed therapy,the better rankingVprobability was improved Sandwich-therapy,and the worst was endoscopic ligation.The hemostatic control rate of improved sandwiches and traditional Sandwiches-therapy was better than that of endoscopic ligation,and the OR value was statistically significant.There was no statistical significance among improved Sandwich-therapy,TIPS and traditional Sandwiches-therapy.Indirect comparative analysis showed that endoscopic ultrasound treatment was better than endoscopic ligation in the inconsistent model,and the OR value was statistically significant,which had no statistical significance compared with improved Sandwich-therapy,BRTO and TIPS.The comparative analysis of response rates of mixed treatments showed that improved sandwich,BRTO and TIPS had higher ranking probability,with similar ranking probability and no difference in 95%CI among them.Traditional sandwiches had the lowest ranking.The remission rate of improved Sandwich-therapy was significantly higher than that of traditional Sandwich-therapy and endoscopic ligation,and the OR value was statistically significant,.In indirect comparative analysis,the response rates of endoscopic ultrasound therapy were not statistically significant compared with traditional Sandwich-therapy,improved Sandwich-therapy,endoscopic ligation,BRTO,and TIPS from the inconsistent model.The OR value and 95%CI of response rate between endoscopic ligation and BRTO and TIPS were not statistically significant.In the comparative analysis of the combined treatment of rebleeding rate,BRTO had the lowest rebleeding rate with a high probability,followed by TIPS,and endoscopic ligation had the highest rebleeding rate with a high probability.BRTO rebleeding rate was significantly lower than traditional Sandwich-therapy,endoscopic ligation and TIPS,and the OR value was statistically significant.TIPS rebleeding rate was significantly lower than traditional Sandwich-therapy,improved Sandwich-therapy and endoscopic ligation,and the OR value was statistically significant,respectively.The rebleeding rate of traditional sandwich and improved sandwich was lower than that of endoscopic ligation,and the OR value was statistically significant.There was no significant difference in the rebleeding rate between traditional Sandwich-therapy and improved Sandwich-therapy.In indirect comparative analysis,the rebleeding rate of BRTO was lower than that of endoscopic ultrasound treatment and improved Sandwich-therapy,the OR value was statistically significant.The rebleeding rate of endoscopic ultrasonography was lower than that of endoscopic ligation,and the OR value was statistically significant.In the comparative analysis of mixed treatments for hepatic encephalopathy,the incidence of encephalopathy in TIPS ranked first with a high probability,while the incidence of improved sandwich was the lowest.The incidence of hepatic encephalopathy in TIPS was higher than that in the modified sandwich comparison,and the OR value was statistically significant,but not statistically significant compared with the endoscopic ligation.In indirect comparative analysis,inconsistent model analysis was adopted,and the results showed that there was no statistical significance between traditional Sandwich-therapy,improved Sandwich-therapy,endoscopic ligation and BRTO.In the comparative analysis of combined treatment mortality,BRTO had the lowest mortality,followed by TIPS and traditional sandwich,and endoscopic ligation had the highest mortality(0.59).The mortality rate of endoscopic ligation was higher than that of BRTO,the OR value was statistically significant.Compared with traditional Sandwich-therapy,TIPS and BRTO,mortality was similar and not statistically significant.In indirect comparative analysis,endoscopic ultrasound treatment was not statistically significant compared with improved Sandwich-therapy,endoscopic ligation and BRTO.There was no statistically significant difference in mortality between BRTO and improved Sandwich-therapy.Conclusion:1.Improved Sandwich-therapy and traditional Sandwich-therapy were significantly better than endoscopic ligation in terms of active bleeding control rate and varicose relief rate.2.In terms of delayed rebleeding rates,BRTO was significantly lower than TIPS,traditional Sandwich-therapy,improved Sandwich-therapy,and endoscopic ligation.The rebleeding rate of TIPS was significantly lower than that of conventional Sandwich-therapy,improved Sandwich-therapy,and endoscopic ligation.3.The incidence of hepatic encephalopathy in improved Sandwich-therapy was significantly lower than that in TIPS.4.The mortality rate of BRTO was significantly lower than that of endoscopic ligation.