# Prediction of major liver-related events in the population using prognostic models

## TL;DR

This paper reviews models that predict liver-related events, offering better risk assessment than traditional methods focused on fibrosis detection.

## Contribution

The paper highlights models specifically designed to predict liver-related events rather than just detecting fibrosis.

## Key findings

- Models like CLivD, dAAR, and CORE directly estimate the risk of future liver-related events.
- Traditional models like Fibrosis-4 were developed for fibrosis detection, not event prediction.
- Combining fibrosis screening with LRE-focused models can improve preventive care strategies.

## Abstract

Liver disease poses a significant global health burden, with steatotic liver disease related to metabolic dysfunction and/or alcohol use being the most prevalent type. Current risk stratification strategies emphasize detecting advanced fibrosis as a surrogate marker for liver-related events (LREs), such as hospitalization, liver cancer, or death. However, fibrosis alone does not adequately predict imminent outcomes, particularly in fast-progressing individuals without advanced fibrosis at evaluation. This underscores the need for models designed specifically to predict LREs, enabling timely interventions. The Chronic Liver Disease (CLivD) risk score, the dynamic aspartate aminotransferase-to-alanine aminotransferase ratio (dAAR), and the Cirrhosis Outcome Risk Estimator (CORE) were explicitly developed to predict LRE risk rather than detect fibrosis. Derived from general population cohorts, these models incorporate either standard liver enzymes (dAAR and CORE) or risk factors (CLivD), enabling broad application in primary care and population-based settings. They directly estimate the risk of future LREs, improving on traditional fibrosis-focused approaches. Conversely, widely used models like the Fibrosis-4 index and newer ones, such as the LiverRisk and LiverPRO scores, were initially developed to detect significant/advanced fibrosis or liver stiffness. While not designed for LRE prediction, they have later been analyzed for this purpose. Integrating fibrosis screening with LRE-focused models like CLivD, dAAR, and CORE can help healthcare systems adopt proactive, preventive care. This approach emphasizes identifying individuals at imminent risk of severe outcomes, potentially ensuring better resource allocation and personalized interventions.