Data-driven tool enables patients to balance recovery expectations

Before deciding whether to go through with a medical procedure, it's important for patients to balance their expectations on recovery. A new data-driven tool being developed allows them to do just that.

A team led by Cynthia Rudin (pictured), a professor at the Massachusetts Institute of Technology Sloan School of Management, has created a tool that enables patients to obtain predictions for surgical recoveries, the school announced this week. Rudin and the other researchers--MIT Ph.D. student Fulton Wang, and Tyler McCormick and John Gore, M.D., both from the University of Washington--took data from patient surveys on sexual function in order to create a personal recovery curve from prostatectomies.

They wanted to model what a patient's recovery would be like, while making it personalized, Rudin said in an interview with FierceHealthIT.

It was important to provide patients "with something to use to manage expectations. Something visual," she said. They wanted to include, not only the information curve, Rudin added, but also the uncertainty around the curve to provide patients' with as much information as possible.

"It gives them a choice in what treatment to get," she said. "If you don't have correct information about what your recovery is going to look like, it's harder to know which direction to take."

The patients are able to get the information through an interactive tool on an iPad, Rudin said. Through the app, they provide their age and sexual function level, and it gives them a set of personalized curves. The researchers are in the process of getting permission to do a clinical trial, Rudin said.

While their focus was on prostatectomy, Rudin said the model can be used on other surgeries and recoveries from medical emergencies, such as strokes.

"The general goal of big data to help with predictive modeling in healthcare has major implications for health," Rudin said. "There's no paper that can have this full trajectory. Being able to better forecast recovery could have a major influence on patients' decisions going forward."

These types of data-based models are already being implemented at Rush University Medical Center--which is using predictive analytics for quality and efficiency improvement efforts. In an interview with FierceHealthcare, Lac Tran, Rush's CIO, spoke about using predictive analytics to improve care for patients at risk for stroke and cardiac arrest.

To that end, the University of Pittsburgh Medical Center has taken things one step further by developing models that take analytics beyond patient claims to household data, like shopping preferences.

To learn more:
- read the announcement