• Jun Lu, Zhihui Tu, Zhen Zhang, Shumei Wang, Zhuoqi Liu, Xiaohui Lu, Jun Zhang, and Daya Luo.
• Department of Intensive Care Unit, Infectious Disease Hospital Affiliated to Nanchang University, Nanchang, Jiangxi, China.
• Medicine (Baltimore). 2024 Oct 25; 103 (43): e40246e40246.

AbstractThe objective of this study was to investigate the predictors and predictive model construction of the progression of HBV-Pre.Acute-on-chronic liver failure (ACLF), a total of 133 patients with HBV-Pre.ACLF was divided into the progressive group (52 patients) and the recovery group (81 patients) according to whether they progressed to ACLF or not. The clinical parameters N%, L%, PLT, ALT, TBiL, ALB, Cre, Na, NH3, CRP, AFP, prothrombin time (PT), international normalized ratio (INR), FIB, and their rate of change at baseline were analyzed in the 2 groups. The independent risk factors for HBV-Pre.ACLF progression was found by univariate and multivariate analyses, and a predictive model was constructed. The clinical parameters ALB, FIB, Na, combined alprostadil treatment and MELD, and MELD-Na scores at baseline were significantly different between the 2 groups (P <.05), while ALT, TBiL, Cre, CHE, NH3, N%, L%, PLT, INR, and PT were not significantly different (P >.05). The change rates of Na, CHE, PT, FIB, CRP, Cre, PLT, and the ratio after to before of N% were significantly different between the 2 groups (P <.05), while the change rates of ALT, TBIL, NH3, AFP, L%, and the ratio after to before of INR were not significantly different between the 2 groups (P >.05). Univariate and multivariate analyses showed that baseline ALB, Na, FIB, combined alprostadil therapy and the rate of change of Na and PLT were protective factors for disease progression, and the rate of change of PT, CRP, and the ratio after to before of N% were independent risk factors for disease progression. The novel model was LogitP = -6.051 + 4.049×ΔPT + 0.626×ΔCRP + 4.527×the ratio after to before N% and its area under the curve was 0.944 (95% confidence interval: 0.900-0.988) predicting progression of HBV-Pre.ACLF, and the best cutoff value was -0.22. The patients with a higher logitP score (> -0.22) had an increased risk for progression to ACLF (P <.05). The novel model logitP shows good predictive value for the disease progression of HBV-Pre.ACLF.Copyright © 2024 the Author(s). Published by Wolters Kluwer Health, Inc.

19 keywords...

hide…