| Objective1.This study was conducted to investigate the independent indicators of failure tocontrol bleeding, early rebleeding, early mortality and overall mortality in cirrhoticpatients after cessation of initial esophageal variceal bleeding(EVB) and evaluate theeffect of infection on the outcome of cirrhotic patients with EVB.2. To explore thevalue of leucocyte esterase reagent strips in the rapid and bedside diagnosis ofinfectious ascites.Methods1. We prospective collected67EVB inpatients from November2010to2011July,rejecting10cases does not meet the requirements, the rest57patients were includedin the study, clinical information and examine results were protocoled indetail andarranged, prognosis of the patients were followed and recorded.2. From November2010to August2011,115ascitic fluid samples were collected from102consecutivepatients, including80cases from cirrhotic patients and35cases for other reasons.Abdominal paracentesis was performed and the Multistix10SG was tested, then wecompare the results with PMN count. A ROC curve was elaborated to define whichcut-off point was more reliable to be used as a positive result of the reagent strip forthe diagnosis of infectious ascites.Results 1. Presence of bacterial infection is associated with prognosis of cirrhotics withesophageal varices bleeding In our research, bacterial infection was documented in17(29.82%) patients, early rebleeding was documented in14(24.56%) patients,17(29.82%) patients died during a follow up of6months(8cases died in6weeks). In themultivariate analysis,the presence of prophylactic antibiotics(OR=0.197,P=0.026)was protective for failure to control bleeding, age(P=0.012) was shown to beindependent predictors of failure to control bleeding; Bacterial infection(P=0.002)was independent predictors of early rebleeding; The number of patients withrecurrent hemorrhage in our study decreased over time: rebleeding within6weeksoccurred in58.33%of all recurrent hemorrhage cases,which account for24.56%ofall cases with variceal bleeding, and40.5%of early rebleeding episodes occurred inthe first5days. The frequency of bacterial infection (P<0.001), rate of failure tocontrol bleeding(P=0.013), rate of early rebleeding(P=0.017), Child-Pugh score(P=0.020)were higher in the group of death, and these differences were statisticallysignificant; Bacterial infection(OR=23.802,P=0.008)was shown to be independentpredictors of six-week mortality; Prophylactic antibiotics(OR=0.237,P=0.043)wasprotective for bacterial infection.2. Leucocyte esterase reagent strips in diagnosinginfectious ascites PMN was≥250in30ascitic fluid samples,26samples of them werecorrectly diagnosed by leucocyte esterase reagent strips. We define degree1+in the colormetricscale as the cutoff point for the diagnosis of infectious ascites, Multistix10SG showed asensitivity of96.67%, specificity of98.5%, NPV of87.88%and PPV of98.65%. But leucocyteesterase reagent strips can not be applied to deep dyeing ascities like bloody ascites or chylousascites.Conclusion1.Bacterial infection was predictors of early rebleeding and mortality, andprophylactic antibiotics was protective for bacterial infection. The condition ofcirrhosis EVB patients should be closely observed in six weeks, and the first fivedays is ciritical, aged patients should receive particular attention.2. According to our basical research, the application of leucocyte esterase reagent strips is a rapid,accurate method for the diagnosis of ascitic fluid infection, it is qualified to be ascreening tool of infectious ascites. |