INTRODUCTION: Acute-on-chronic liver failure (ACLF) is defined by the European Association for the Study of the Liver-Chronic Liver Failure (EASL-CLIF) consortium and the North American Consortium for the Study of End-Stage Liver Disease (NACSELD) as an acute deterioration of cirrhosis with multiple organ failures and high short-term mortality. However, their diagnostic criteria differ. We aimed to compare these 2 criteria in the prediction of prognosis in hospitalized cirrhosis. METHODS: This was a prospective study of nonelectively hospitalized patients with cirrhosis (N = 468) from a single tertiary hospital between 2016 and 2018. Baseline characteristics, incidence, and types of organ failure and survival data at 7, 28, and 90 days were collected. Prognostic utilities of the 2 criteria were compared. RESULTS: One hundred thirty-seven of 468 patients (29.3%) had EASL-CLIF ACLF, and 35 of 468 (7.4%) had NACSELD ACLF. The 28-day transplant-free survival of ACLF was 58.4% using EASL-CLIF and 37.1% using the NACSELD criteria. In predicting 28-day mortality, the NACSELD criteria demonstrated significantly higher overall accuracy (92.0% vs 85.3%, P < 0.01), specificity (99.7% vs 84.0%, P < 0.001), and positive predictive value (97.1% vs 50.4%, P < 0.001) but lower sensitivity (49.3% vs 92.5%, P < 0.001) and negative predictive value (91.6% vs 98.5%, P < 0.001) than those of EASL-CLIF. The results were similar in predicting 7-day outcome. However, the overall accuracy became similar between NACSELD and EASL-CLIF ACLF criteria in predicting 90-day mortality (86.3% vs 88.7%, P = 0.27) because of the decrease of sensitivity and negative predictive value of NACSELD ACLF criteria. The prognostic performance of these 2 ACLF criteria was similar when applied to patients with or without hepatitis B virus infection as an etiology of cirrhosis. DISCUSSION: There are both caveats and utilities of NACSELD and EASL-CLIF ACLF criteria in prognosis prediction in patients with cirrhosis. NACSED criteria is highly accurate in predicting morality, whereas the EASL-CLIF criteria is more sensitive to identify patients who would benefit from liver transplantation.
INTRODUCTION: Acute-on-chronic liver failure (ACLF) is defined by the European Association for the Study of the Liver-Chronic Liver Failure (EASL-CLIF) consortium and the North American Consortium for the Study of End-Stage Liver Disease (NACSELD) as an acute deterioration of cirrhosis with multiple organ failures and high short-term mortality. However, their diagnostic criteria differ. We aimed to compare these 2 criteria in the prediction of prognosis in hospitalized cirrhosis. METHODS: This was a prospective study of nonelectively hospitalized patients with cirrhosis (N = 468) from a single tertiary hospital between 2016 and 2018. Baseline characteristics, incidence, and types of organ failure and survival data at 7, 28, and 90 days were collected. Prognostic utilities of the 2 criteria were compared. RESULTS: One hundred thirty-seven of 468 patients (29.3%) had EASL-CLIF ACLF, and 35 of 468 (7.4%) had NACSELD ACLF. The 28-day transplant-free survival of ACLF was 58.4% using EASL-CLIF and 37.1% using the NACSELD criteria. In predicting 28-day mortality, the NACSELD criteria demonstrated significantly higher overall accuracy (92.0% vs 85.3%, P < 0.01), specificity (99.7% vs 84.0%, P < 0.001), and positive predictive value (97.1% vs 50.4%, P < 0.001) but lower sensitivity (49.3% vs 92.5%, P < 0.001) and negative predictive value (91.6% vs 98.5%, P < 0.001) than those of EASL-CLIF. The results were similar in predicting 7-day outcome. However, the overall accuracy became similar between NACSELD and EASL-CLIF ACLF criteria in predicting 90-day mortality (86.3% vs 88.7%, P = 0.27) because of the decrease of sensitivity and negative predictive value of NACSELD ACLF criteria. The prognostic performance of these 2 ACLF criteria was similar when applied to patients with or without hepatitis B virus infection as an etiology of cirrhosis. DISCUSSION: There are both caveats and utilities of NACSELD and EASL-CLIF ACLF criteria in prognosis prediction in patients with cirrhosis. NACSED criteria is highly accurate in predicting morality, whereas the EASL-CLIF criteria is more sensitive to identify patients who would benefit from liver transplantation.