Risk scoring systems to predict in-hospital mortality in patients with acute variceal bleeding due to HCV-induced liver cirrhosis

Abstract

Background: this study was designed to validate and to compare accuracy of the prognostic scores; mainly Child Turcotte Pugh (CTP), creatinine-modified Child Turcotte Pugh (CTP-Cr), MELD, albumin bilirubin score (ALBI), and AIMS65, for the predicting clinical outcomes in cirrhotic Egyptian patients presenting with acute variceal bleeding (AVB). Methods: Retrospective single center study involving 725 patients presenting with AVB due to liver cirrhosis and HCV infection either alone or mixed with HBV infection. In hospital mortality prognostic scores were calculated; mainly CTP, modified CTP-Cr, MELD, ALBI, AIMS65. The endpoint is either patient improvement or death. Results: 725 patients were included over 1-year period. 547 (75%) survived and 178 (25%) died. Patients presented with hematemesis (515/71%), melena (120/16.5%) or hematemesis and melena (90/12.5%). Those with hematemesis for the first time were 241 (33%) and recurrent attacks were 484 (66.8%). The non-survivors had significantly more incidence of shock on presentation, more blood transfused units, history of NSAIDS intake, more ICU admission days and were more likely to be Childs C. Child, modified CTP-Cr, MELD, ALBI and ALMS65 scoring systems showed significant difference between survivors and non-survivors. Conclusion: Liver specific scores (Child, MELD) and gastrointestinal bleeding scoring systems (ALBI, AIMS65) are useful in predicting clinical outcomes of AVB in cirrhotic patients. CTP-Cr score had the highest prognostic capability of in hospital mortality. Presence of active bleeding at time of endoscopy, more complications, old age, shock and higher CPT-Cr score are additional independent predictors of in hospital mortality.