1.1 Enrollment Between October 2013 and June 2016, a total of 1143 chronic hepatitis B inpatients with liver failure (acute-on-chronic liver failure, ACLF) were enrolled from seven liver disease centers across China (infection department of the Third Affiliated Hospital of Sun Yat sen University, infection department of the First Affiliated Hospital of Zhejiang University, Shanghai Ruijin Hospital, Xiangya Hospital of Central South University, Qilu Hospital of Shandong University, Shandong Provincial Hospital, and Xuzhou Medical College). Among them, 936 were men, 207 were women, and the average age was 43.4 ± 12.6 (ranged from 18 to 65). The patients were followed up for >6 months, during which 464 died and 680 survived.
HBV-related ACLF diagnosis for all patients followed the criteria established by the 18th Asia-Pacific Association of Liver Research consensus on chronic-acute liver failure. The criteria include a history of chronic hepatitis B and acute flare-up of liver injury with clinical manifestations of jaundice (total bilirubin ≥ 85 umol / L) and coagulation disorder (prothrombin time international standardized ratio ≥ 1.5), ascites and / or hepatic encephalopathy within 4 weeks.
1.2 Inclusion criteria Male or female patients aged 18-60; HBsAg-positive history > 6 months, HBV DNA-positive, HBeAg-positive or negative; persistent hepatitis symptoms of fatigue, anorexia, abdominal distention, or yellow urine; gradual aggravation of jaundice over a short period of time; total serum bilirubin ≥ 85 umol / L or daily elevation ≥ 17.1 umol / L; abnormal coagulation function, and international standardized ratio of prothrombin time ≥ 1.5.
1.3 Exclusion criteria Patients were excluded if they: (1) had other hepatitis virus infection; (2) had human immunodeficiency virus infection, biliary, alcoholic liver, or autoimmune liver diseases; drug poisoning, liver tumors, or were undergoing liver transplantation, renal insufficiency, or long-term anticoagulant therapy related to renal diseases.
2. Observations and follow-up endpoints Clinical information and test findings for all patients at and after admission were collected weekly. The clinical information mainly consisted of the stages of hepatic encephalopathy. Laboratory findings included serum total bilirubin, albumin, creatinine, prothrombin time, prothrombin time international normalized ratio (INR), serum sodium ion concentration, liver size (B-mode ultrasound measurement), ascites, and pleural effusion (B-ultrasound measurement), as well as infection (peripheral white blood cell count, neutrophil ratio, and chest inflammation images). The end point of 180 days of follow-up was used to determine the survival rate. The death count included patients that rejected rescue treatment and were discharged from the hospital as well as patients that died during the hospital stay or within 180 days of follow-up.
3. Score calculation
3.1 MELD score Since all subjects had hepatitis B-related liver failure, the MELD score was calculated as 3.8×loge (serum bilirubin umol/L×0.058) + 1.2×loge (prothrombin time INR) + 9.6×loge (serum creatinine umol/L×0.011) + 6.4.
3.2 MELD-Na This score was calculated as MELD + 1.59×(135-Na+), wherein serum Na+ levels ≥ 135 mmol/L were treated as 135 mmol/L, ≤ 120 mmol/L as 120 mmol/L, and between 120~135 mmol/L as the specific value.
3.3 CTP score This score was calculated using the scoring standards for five indexes, i.e. grade of hepatic encephalopathy, ascites, total bilirubin, albumin, and prolonged prothrombin time. A score of 1, 2, or 3 was assigned to each index to reflect the severity of each condition, and the CTP score was the sum of the five indexes.
3.4 Chinese scoring system This system consists of seven clinical indicators: prothrombin activity, serum creatinine, hepatic encephalopathy, serum total bilirubin, liver size (B-ultrasonic measurement), ascites/pleural fluid (B-ultrasonic measurement) volume, and infection (peripheral blood leukocyte count, neutrophil ratio, and chest inflammation image). A score of 1, 2, 3, or 4 was assigned to each indicator to reflect the severity, and the sum of the seven indicator scores was used (see Table 1 for details).
4. Statistical analysis
(1) All numeral data were expressed as mean ± standard deviation (±s) and the difference was computed using the f test. (2) The count data were expressed as percentage (%), and the difference was assessed with the χ2 test. (3) Assessments of the short-term and long-term prognoses for patients with hepatitis B liver failure using the Chinese scoring system, MELD score, MELD-Na score, and CTP score were compared using receiver operating characteristic (ROC) curves and the ROC area under the curve (AUC). AUC > 0.7 was deemed to be of clinical utility and > 0.8 to be of good prediction accuracy. The AUC values were compared using normal Z tests and the ROC curve sensitivity and specificity were used to determine the best cut-off values for the score and the Youden index. SPSS l8.0 statistical software was used for all analyses. P < 0.05 was considered statistically significant.