Advanced Search

Publications > Journals > Journal of Clinical and Translational Hepatology > Article Full Text


A Novel Backward Stepwise Logistic Regression and Classification and Regression Tree Model to Predict 180-day Clinical Outcomes in Hepatitis B Virus-acute-on-chronic Liver Failure Patients

  • Shima Ghavimi 
 Author information
Journal of Clinical and Translational Hepatology   2021;9(4):456-457

doi: 10.14218/JCTH.2021.00240

As the era of precision and personalized medicine is gaining exponential positive gain in the field of medicine, there is a positive shift towards a more evidence-based patient care approach for patients with hepatological diseases. One factor that is crucial in any physician’s decision-making efforts involves the application of novel innovative approaches that can enhance predicting survival outcome. Acute-on-chronic liver failure (ACLF) is a perfect example of how liver can rapidly deteriorate, and the hepatitis B virus (HBV) is one crucial culprit. Patients can experience organ failure that leads to their mortality, and in this article the authors clearly described the use of backward stepwise logistic regression (LR) and classification and regression tree (CART) analysis to derive two predictive models and then compared them with the model of end-stage liver disease (MELD) score for novel prognostic models of the 180-day outcome for patients with HBV-ACLF.

This innovative study showed the novel predictive models to be superior to MELD score, providing Hepatologists and Gastroenterologists with a new guiding technique in their evidence-based patient-centered care approach for their clinical treatment decision-making. Of note, one of the leading causes of chronic liver failure in Asia is HBV1 and the current leading therapeutical intervention for cure is liver transplantation (LT).2–4

However, due to decline of health of patients with ACLF, some of the patients become delisted from the organ transplantation list because they are not well enough to undergo LT. Currently, MELD score is the most used for patients on the LT list, although several other scoring modalities are available.5 Current and recent advancements in Hepatology generated the finding for HBV-ACLF having a window of 30 to 90 days, with respect to the scoring system used. In the article of interest, the authors introduced a predictive module of 180-day outcome for patients based upon multiple variables that are not currently used in the MELD score. Also, there are other scoring systems that have been introduced for HBV-ACLF recently; although, the authors created a CART system scoring that is easier for the clinician to interpret and, in turn, for the patient to receive the paramount care they deserve. Ultimately, the research demonstrated that both LR and the CART model appeared to perform better than the MELD score.

There are, however, some factors and limitations within the study (as properly mentioned in the article), ranging from not having a mid-data report and not having a larger cohort for studying their novel scoring models. I believe if a consensus can be developed by having multiple LT centers around the globe adapt this novel scoring module, we can achieve a better understanding and more precise scoring system developed for patients with HBV-ACLF requiring LT. These novel models which the authors have described and investigated may be helpful for Transplant Hepatologists, Gastroenterologists and Health Care teams who need to make essential clinical decisions for patients with HBV-ACLF.



acute-on-chronic liver failure


classification and regression tree


hepatitis B virus


liver transplantation


model of end-stage liver disease



None to declare.

Conflict of interest

The authors have no conflict of interests related to this publication.


  1. Davlidova S, Haley-Johnson Z, Nyhan K, Farooq A, Vermund SH, Ali S. Prevalence of HIV, HCV and HBV in Central Asia and the Caucasus: a systematic review. Int J Infect Dis 2021;104:510-525 View Article
  2. Cotter TG, Sundaram V. Living donor liver transplantation in acute-on-chronic liver failure: encouraging outcomes but with important caveats. Liver Transpl 2021 View Article
  3. Li X. Emergent liver transplantation for patients with acute-on-chronic liver failure. Dig Liver Dis 2021 View Article
  4. Vipani A, Lindenmeyer CC, Sundaram V. Treatment of severe acute on chronic liver failure: management of organ failures, investigational therapeutics, and the role of liver transplantation. J Clin Gastroenterol 2021 View Article
  5. Liu F, Zou Z, Shen L, Wu W, Luo J, Lankford S, et al. A prediction model for outcome in patients with HBV-ACLF based on predisposition, injury, response and organ failure. Sci Rep 2020;10(1):20176 View Article
  • Journal of Clinical and Translational Hepatology
  • pISSN 2225-0719
  • eISSN 2310-8819
Back to Top

A Novel Backward Stepwise Logistic Regression and Classification and Regression Tree Model to Predict 180-day Clinical Outcomes in Hepatitis B Virus-acute-on-chronic Liver Failure Patients

Shima Ghavimi
  • Reset Zoom
  • Download TIFF