Takafumi Naiki1, Nobuaki Nakayama, Satoshi Mochida, Makoto Oketani, Yasuhiro Takikawa, Kazuyuki Suzuki, Shin-Ichiro Tada, Takafumi Ichida, Hisataka Moriwaki, Hirohito Tsubouchi. 1. Department of Gastroenterology, Gifu University Graduate School of Medicine, Gifu Department of Gastroenterology and Hepatology, Faculty of Medicine, Saitama Medical University, Saitama Department of Digestive and Life-Style Related Disease, Health Research Course, Human and Environmental Sciences, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima Department of Gastroenterology and Hepatology, School of Medicine, Iwate Medical University Department of Internal Medicine, School of Medicine, Keio University Department of Gastroenterology, Shizuoka Hospital, Juntendo University, Tokyo, Japan.
Abstract
AIM: In Japan, the indication for liver transplantation in patients with acute liver failure (ALF) is currently determined according to the guideline published in 1996. However, its predictive accuracy has fallen in recent patients. Thus, we attempted to establish a new guideline. METHODS: The subjects were 1096 ALF patients enrolled in a nationwide survey. All patients showed a prothrombin time <40% of the standardized value and grade II or more severe hepatic encephalopathy. A multiple logistic regression analysis and receiver operating characteristic analysis were performed in 698 patients seen between 1998 and 2003 to identify significant parameters determining the outcome of patients. The extracted parameters were graded as numerical scores. An established scoring system was validated in patients seen between 2004 and 2008. RESULTS: Six parameters were identified and graded as 0, 1 and/or 2; the interval between disease onset and development of hepatic encephalopathy, prothrombin time, serum total bilirubin concentration, the ratio of direct to total bilirubin concentration, peripheral platelet count and the presence of liver atrophy. When the prognosis of the patients with total score of 5 or more was judged as "death", the predictive accuracy was 0.80 with sensitivity, specificity, positive predictive value and negative predictive value greater than 0.70. The values were similarly high in patients for validation. CONCLUSION: Novel scoring system for predicting the outcome of ALF patients may be useful to determine the indication of liver transplantation, since the system showed high predictive accuracy even after validation.
AIM: In Japan, the indication for liver transplantation in patients with acute liver failure (ALF) is currently determined according to the guideline published in 1996. However, its predictive accuracy has fallen in recent patients. Thus, we attempted to establish a new guideline. METHODS: The subjects were 1096 ALFpatients enrolled in a nationwide survey. All patients showed a prothrombin time <40% of the standardized value and grade II or more severe hepatic encephalopathy. A multiple logistic regression analysis and receiver operating characteristic analysis were performed in 698 patients seen between 1998 and 2003 to identify significant parameters determining the outcome of patients. The extracted parameters were graded as numerical scores. An established scoring system was validated in patients seen between 2004 and 2008. RESULTS: Six parameters were identified and graded as 0, 1 and/or 2; the interval between disease onset and development of hepatic encephalopathy, prothrombin time, serum total bilirubin concentration, the ratio of direct to total bilirubin concentration, peripheral platelet count and the presence of liver atrophy. When the prognosis of the patients with total score of 5 or more was judged as "death", the predictive accuracy was 0.80 with sensitivity, specificity, positive predictive value and negative predictive value greater than 0.70. The values were similarly high in patients for validation. CONCLUSION: Novel scoring system for predicting the outcome of ALFpatients may be useful to determine the indication of liver transplantation, since the system showed high predictive accuracy even after validation.