Improving our ability to predict health outcomes will substantially decrease the costs of health care while simultaneously improving health outcomes. Individuals who go to the hospital for major in-patient procedures can have many different health outcomes. Hopefully the recovery fully without incident, but there is often also substantial risk of post-operation kidney failure, admittance to an ICU and death.
In this project we are working to be able to predict each of these outcomes before they occur. Through collaborations with the UCLA hospital system, we are are analyzing electronic medical records along with within surgery data (e.g. heart rate, blood pressure, drugs administered, labs order) in order to improve predictions algorithms. Our goal is to be able to identify individuals at the highest risk of these negative outcomes so that proper preventative actions can be taken to reduce the outcome risk.