Sneak Peak: Improving Patient Engagement and Outcomes with Predictive Analytics

My Folder

Gregory A Spencer, MD (Chief Medical & Chief Medical Information Officer, Crystal Run Healthcare), Louis G. Cervone, Jr. (Director of Business Intelligence, Crystal Run Healthcare)

Session Overview

Crystal Run has been developing an analytics-supported approach to patient care using disease-specific metrics, costs, trends, simulation and predicted outcomes to motivate patients, improve outcomes, and decrease costs. Using Diabetes patients as an example, Crystal Run has focused on costs and complications for high-risk and low-risk cohorts by provider and location. From the patient perspective, they are then able to look at patient-specific lab results and costs, care recommendations, and a comparison of current status to predicted complications. During office visits, they will have the ability to share risk levels and care simulations with patients based on their current status and show the predicted impact for of making lifestyle changes such as smoking cessation, improving HgA1C levels, losing weight, decreasing blood pressure, etc. The predictive model is based on multiple regression analysis and Charlson risk scores and compares the individual patient based on their current status to the experience and clinical progression of similar patients. The session will explore questions such as, “How can we measure health?” and “What factors can predict health?”. The answers to these questions are at the core of building an effective predictive analytics model.

Whether you are on the go or have time for a deep dive, we offer multiple ways to view the content.