International Partnership Works on Predictive Analytics

A coalition of businesses, governments and universities has come together to work on predictive analytics modules to address issues such as the opioid crisis and hospital readmissions.

Bill Lucia, CEO of Texas-based HMS, which provides technology, benefits and care management services to insurers, recently spoke to Healthcare Innovation about how his company is partnering with the Digital Health Cooperative Research Centre (CRC) in Australia, Stanford University and Southern Methodist University.

 Here is how the unusual partnership came together: Lucia said he and colleagues met executives from the Australian CRC at an international insurance fraud conference. The CRC has grant money from the Australian federal government to find ways to use technology to reduce costs and improve the quality of healthcare for Australians and others. “We joined their effort as the largest commercial entity supporting the program,” he said. There are 40 partners working on it, including many of the large Australian universities. The idea is to take the research done and have it commercialized so it helps Australians and others. HMS helped recruit some U.S. universities, including Stanford University Medical Center, SMU, and University of Texas-Dallas and University of Texas-Arlington. “We wanted to fund Ph.D. or master’s level students to perform research on a rich set of data,” Lucia said, “particularly projects that would be impactful if we could build predictive models coming out of the research data.”

HMS works with 42 Medicaid programs, state governments, and many other payers, including Medicare Advantage and employer-sponsored insurers. “We went to a number of state and managed care clients and said if we could de-identify a data set for research, would you be interested in being part of it, and also become early adopters of the output and drive the research questions we can work to solve,” Lucia said.

Four states signed up as did one regional Medicaid managed care plan. Combined they offer de-identified claims data on about 4 million lives over five years of history and across several regions of the country.

The Stanford research is led by Dr. Tina Hernandez-Boussard, associate professor in Medicine, Biomedical Data Science and Surgery at the Stanford University School of Medicine. Her team  working on opioid issues plans to use the historical data to develop a model that will determine at the point of prescription who is most likely to develop opioid abuse syndrome, Lucia said. “And once someone has developed opioid abuse syndrome, what is the most efficacious treatment to get them off opioids, and out of that spiral, and who are the people most likely to relapse. We also want to study the impact of opioids on the third trimester of pregnancy. This data will be harder to come by, but a pressing issue is what are the things in the community that predict who is most likely to overdose and most likely to die from an overdose.”

Ultimately, the researchers would publish their results in journals but the predictive models would be marketed through HMS’ predictive analytics offerings that it sells to payers and other healthcare entities including state governments. “We want to integrate it into the existing opioid analytics in our toolset today,” Lucia said.

Researchers at SMU are going to take a similar approach to hospital readmissions.

Commenced in 2018, Australia’s Digital Health CRC calls itself the largest digital health research cooperative in the world with funding from government, universities and businesses totaling more than $160 million. When the partnership was announced, Dr. Victor Pantano, Digital Health Cooperative Research Centre CEO, said, “Our country recognizes that by working with private industry and academia, we can advance the health of all people worldwide. We look forward to the results of this ground-breaking data analytics’ research and its ability to solve some of the most critical health issues of our time.”