Using predictive data tools may provide new insights into how to treat so-called complex patients, those with multiple medical conditions who drive a disproportionate level of healthcare costs, speakers said this week at an ECRI Institute-sponsored conference.
Intervening early with complex patients could reduce overall treatment costs and predictive data tools may enable such early intervention, speakers noted.
One example of such use is advising cardiologists to avoid using certain contrasts in heart imaging with patients who are at risk of renal failure.
Dr. Stephan Finn, director of analytics and business intelligence at VA Puget Sound Health Care System, said the VA is focused on using predictive analytics to identify patients who are at risk of renal failure, or those who are at risk of committing suicide.
Seemingly surprising data points can provide insight into certain conditions, speakers said. Ralph Muller, CEO of University of Pennsylvania Health System, said his hospital's analysts use social media data to help determine risk of readmission. The process, he said, it's just as accurate as genomic data analysis.
But there are some issues to think through. One problem revolves around data access, or the lack of enough data to find the necessary correlations to aid care, providers agreed.
Follow Darius Tahir on Twitter: @dariustahir
