Multivariate analysis for prognostic factors of 282 patients with hepatic encephalopathy

Objective: To explore the prognostic factors of hepatic encephalopathy (HE) for survival. Methods: We retrospectively reviewed the clinical data of 282 hepatic encephalopathy patients from January 2012 to May 2014 in our hospital. Data base was built using Excel and SPSS 18.0. Multivariate logistic regression statistics were used to analyze 21 possible risk factors, including gender, age, ALB, ALT, K+, Na+, Cl-, PH, TBIL, TBA, PT, APTT, UREA, Ca+, Mg+, P, NH3, PaCO2, PaO2, HCO3 and TCO2. Results: The mortality rate of the 282 HE patients was 73.8%. There was statistical difference in Na+, NH3, APTT, UREA and ALT. These five factors indicated poor prognosis of patients with hepatic encephalopathy, whose overall percentage correct was 83.0% and a predictive model: logit(P)=1.638×Na+ (mmol/L) 2.295×NH3 (umol/L) 1.812×APTT (sec) 1.738×UREA (mmol/L) 1.315×ALT (IU/L) + 3.340 was developed. Conclusion: The prognosis of patients with hepatic encephalopathy have a variety of factors, hyponatremia, hyperammonemia, prolonged APTT, high urea and high ALT were found to be potential risk factors for prognosis of hepatic encephalopathy. Monitoring these factors may be helpful to detect the development of hepatic encephalopathy in time.