Researchers employ a machine learning technique known as random forest analysis and found that it significantly outweighs traditional methods in predicting whether patients with cirrhosis are at risk of death. Gastroenterology.
This gives you a crystal ball. Hospital teams, transplant centers, GI and ICU services will help triage more effectively. ”
Dr. Jasmohan S. Bajaj, corresponding author of the research
Important findings:
- Data analyzed from 121 hospitals around the world were part of a cleared consortium.
- This model was consistently implemented in high- and low-income countries.
- It was verified using data from US veterans and remained accurate.
- The tool maintained strong performance even when limited to just 15 important variables.
- Patients were accurately grouped into high-risk, low-risk categories, making the model scalable and clinically practical.
Explore the working model here: https://silveys.shinyApps.io/App_Cleared/.
This paper is one of three recently published studies on the topic in the journal of the American Gastrointestinal Association. One is a global consensus statement on organ damage including liver in patients with cirrhosis, and the second study identified specific blood markers and complications that affect the risk of in-hospital death, focusing on biomarkers of liver failure.
“Live disease is one of the most underrepresented causes of death in the world. Alcohol, viral hepatitis and slow diagnosis are the main drivers,” says Bajaj. “When someone is hospitalized, it's often because everything is upstream — prevention, screening, primary care — has already failed.”
sauce:
American Gastrointestinal Association
