Companies such as Netflix and Amazon have been using predictive data for years to recommend movies and products to their customers based on their past purchases and behavior.
And now, health care organizations are following suit by using data from within and outside health care settings to predict and reduce readmissions, says data expert Jay Bhatt, DO, MPH, the chief health officer of the Illinois Hospital Association (IHA).
In one example, the IHA is using housing and transportation data alongside de-identified data from member hospitals to find patterns among patients who use health care more frequently.
“If you know which patients are likely to utilize health care more, then big data insights can help you develop unique care models to optimize their health,” Dr. Bhatt says. “You can allocate the right resources to that population and use data to achieve the triple aim of lower costs, better care and better population health.”
Hospitals and health systems have the opportunity to combine their own data with public data, such as social media, to obtain a more complete picture of patient health outside the care setting, Dr. Bhatt notes. For instance, Researchers in Arizona and Texas recently found that air quality data, hospital claim data and local Twitter chatter about asthma could predict the volume of asthma-related emergency room visits.
Improving public health
Data can also be a powerful tool for improving public health. When Dr. Bhatt worked as Chief Innovation Officer for the Chicago Department of Public Health (CDPH), he oversaw a project in which researchers used data to help uninsured women receive breast cancer screenings.
Using information from a behavioral factors surveillance survey, a predictive model by Civis Analytics, and U.S. Census data, the CDPH identified 5,000 likely uninsured city residents most at risk for premature mortality from breast cancer. The department then mailed the women information on mammography screening and follow-up care available at a nearby hospital. In the first week after the mailings went out, nearly 80 women contacted the hospital to inquire about screening.
CDPH also partnered with the University of Chicago and the City of Chicago’s Department of Innovation and Technology last year to develop a statistical model to identify homes at high risk for lead contamination and help predict pregnant women and children’s risk for lead exposure.
In the future
Dr. Bhatt sees data playing an ever-greater role in health care in coming years, with physicians using data to guide the questions they ask patients and their treatment plans.
Health systems will use data not only to guide patient care decisions, but also staffing and practice protocols and resource allocation, Dr. Bhatt says.
“In the future, I think we’ll see an interprofessional care team that includes a data scientist,” he says. “Someone who’s helping the care team identify points of opportunity and insight based on the data.”