A Predictive Analytics and Machine Learning Approach to Improving Hub Performance and Patient Outcomes

Patient Switching Behaviors Impact on Adherence and Engagement: A Predictive Analytics and Machine Learning Approach to Improving Hub Performance and Patient Outcomes

Today we have access to more data, from more sources than we could ever dream possible. Living in a digital world, we increasingly need the ability to efficiently and effectively process this data for insights and actions in order to be competitive. The life sciences industry can leverage this data using analytic tools and machine learning to rapidly identify patient behaviors and patterns – allowing us to predict “next best actions” in our quest to improve patient outcomes.

The key for pharma brands who increasingly play a role in supporting patients through their care journey is to think about how to implement predictions in the apparatus of patient and hub services. A prediction alone is not interesting. A prediction that enables an action and learns from the outcome of that action is what creates a high performance operation.

Prediction enables actions to be taken and existing resources to be better utilized. In the above image, we trace the path of one of the personas we created, Persona A, who originally discontinued Company A’s therapy. We developed a unified data set – consisting of demographics, income data, total Rx costs per year, estimated out-of-pockets, and total cost of care – to help train our model. Patients were identified from the database at the time of hub enrollment that matched Persona A and deployed/enrolled into the hub process. The results of our modeling showed a 98% accuracy rate in our ability to describe the type of patients, or personas, across the model.

In the above image, we demonstrate the economic impact of ACTICS BY EVERSANA™ vs. Company A’s hub performance.

In this white paper, EVERSANA’s Brigham Hyde discusses how predictive analytics and machine learning have the potential to transform healthcare by helping us identify diseases faster, decrease costs through precision therapies, improve clinical trial enrollment, and increase operational effectiveness. And, to learn more about the methodology behind this study, click here.