Skip to content

Predictive Analytics on In-Hospital Mortality of ICU Patients with Heart Failure (Y2 Individual University Project)

Notifications You must be signed in to change notification settings

fabecode/Analytics-on-Heart-Failure-Mortality

Repository files navigation

Analytics-on-Heart-Failure-Mortality

Predictive Analytics on In-Hospital Mortality of ICU Patients with Heart Failure

Hospital admits critically ill patients into Intensive Care Unit (ICU) for close monitoring and medical intervention. In general, ICU patients have higher risk of death compared to non-ICU patients. A real dataset from a hospital is publicly available (see data01.csv). The data contained ICU patients with heart failure and 51 variables. The dataset and summary statistics were explained in Li F., et al (2021) research paper1. Patients who died are coded outcome = 1 in the dataset. They proposed a new risk scoring model Nomogram to assess mortality risk.

The American Heart Association Get With the Guidelines – Heart Failure (GWTG) is an established risk model popularly used to assess the risk of mortality among heart failure patients at hospital. This simple model was explained in Peterson P.N. (2009).

This project evaluates the predictive performance of Logistic Regression, Random Forest, GWTG and Nomogram.

About

Predictive Analytics on In-Hospital Mortality of ICU Patients with Heart Failure (Y2 Individual University Project)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages