Surgical Bypass To The Gallbladder Or Bile Duct For The Palliation Of Jaundice Due To Unresectable Carcinoma Of The Head Of The Pancreas: A System Review And Meta-analysisl

Objective To compare the efficacy of cholecystoenterostomy（CCE）versuscholedochoenterostomy（CDE）treatment of jaundice due to unresectable carcinoma ofthe head of the pancreas.Method Standard literature search strategy for the databases was performed.Literatures which were up to the follow standards would be included.1.The type of thetrial are randomized controlled trials, quasi-randomized control trials, prospective trials,cohort studies, retrospective studies. Published form is unlimited and languages arelimited in Chinese or English.2.The diagnosis of the object of the studies must beunresectable Carcinoma of the head of the pancreas and obstructive jaundice.3.Intervention measure is cholecystoenterostomy and choledochoenterostomy as theexperimental group and the control group. Literatures which were up to the followstandards would be excluded.1. laparoscope was used.2. gastrojejunostomy wasperformed as an addition surgery.3.Literatures were not included the main outcome orincluded the incomplete main outcome. The author search the Cochrane, PubMed,EMBASE,Ovid,Springer,CNKI databases with the terms of pancreatic carcinoma,cholecystoenterostomy, choledochoenterostomy and gain literatures. Generalinformations were collected from the literatures which were searched.and up to thestandards. All literatures were evaluated. The funnel plot was drawed by the mainoutcome to assess the publication bias of literatures. The main outcome is recurrentjaundice and the secondary outcomes are median survival time, post-operativemorbidity, mortality. Review Manager5.0was used for statistic analysis. Chi-squaretest was used for the statistical heterogeneity. Fixed effect model would used when P≥0.10which means no statistical significance. Random effects model would used when P＜0.10which means statistical significance. Statistic analysis was based on all data of literatures and the forest map were drawed. Qualitative data use the odds ratio(OR)and95%confidence interval(CI) and quantitative data use the mean difference(MD)and95%confidence interval(CI) for statistical description.Result376literatures were searched basing on standards. A total of twelveliteratures were choesn which included two random controlled trial and ten retrospectivestudies. The results of quality assessment showed the12studies were quality of medium.There is may potential bias in studies. The meta-analysis result show that random effectmodel was used on jaundice rate for P＜0.10. The CCE are significantly higher thanCDE in recurrent jaundice rate. OR=5.00（95%CI：3.12，8.03）. Fixed effect model wasused on post-operative mortality for P＞0.10. The CCE are significantly higher thanCDE in post-operative mortality, OR=1.90（95%CI：1.56，2.23）, P＜0.00001. Randomeffect model was used on median survival time for P＜0.10. The CCE are significantlylower than CDE in median survival time, MD=-2.15（95%CI：-2.51，-1.80），P＜0.00001.Fixed effect model was used on post-operative morbidity for P＞0.10. Thereare no significantly difference between two groups in post-operative morbidity,OR=0.92（95%CI：0.68，1.23）,P=0.56.Conclusion CDE is much better surgical billiary bypass than CCE in reducing thejaundice due to unresectable carcinoma of the head of the pancreas. Although CDE aremore complexity, it is various of advantage on recurrent jaundice rate, median survivaltime and mortality.