Data on Social Needs May Redefine Precision Healthcare

This article originally appeared on Health Leaders Media on June 15, 2017

By Tinker Ready

Data derived from the non-medical drivers of a patient’s health can improve quality of care and enrich the utility of so-called intelligent machines.

The small crowd the “On the Front Lines of Healthcare” event in Boston in early June, included a patient activist, a state health systems analyst, and even a doctor who was in town for a gastroenterologist meeting.

But the gathering, held in an airy space on 33rd floor of a downtown high rise, was not a professional or academic meeting. Organized by The Atlantic and the STAT, a national science and medicine publication, the public event offered an ambitious overview of a range of weighty issues.

Three big topics that resonated throughout the day: the social determinants of health, data, and machine learning—computers that can digest data and use it to answer questions about patient care.

For hospitals and health systems, the sessions presented different scenarios of how the three can meet.

The message: Data derived from the non-medical drivers of a patient’s health can improve quality of care and enrich the utility of so-called intelligent machines, such as IBM’s Watson.

The day started with what was described as a wide-angle lens on “The State of Health Care in America.” Audrey Shelto, president of the Blue Cross Blue Shield of Massachusetts Foundation, made a major pitch for improving care by addressing social as well of medical issues.

There needs to be greater recognition than “what’s involved in getting people healthy. And keeping people healthy only has a small amount to do with what happens in the clinical system,” she said.

“A lot of it has to do with the issues of poverty. It’s housing. It’s food. It’s all the things that make it hard to focus on your health.”

After the panel, Shelto told Health Leaders that, in Massachusetts, the rise of accountable care organizations and a Medicaid waiver are allowing state providers to better address social needs.

For example, the installation of an air conditioner might do more to help a person with asthma than anything that happens during an office visit.

Still, breaking down the silos that separate health, social services, and housing is going to be a challenge, Shelto said.

There might be a way to do that by figuratively linking that AC unit to another machine, the Watson Health supercomputer. Much like Watson beat two humans on the game show “Jeopardy,” Watson Health is an IBM supercomputer that can digest and analyze data, from individual patient records to the contents of millions of journal articles.

Watson Health offers products that promise to identify high-risk patients and suggest individualized care plans. Kyu Rhee, the chief health officer of Watson Health, said the predictive capacity of such programs would benefit from including information not found in clinical records.

“A lot of the data we have in health care is incredibly valuable, but it needs to be linked to the other data about the other determinants of health, whether social, environmental or behavioral,” he said. “Where you live (has) as much of an impact as what doctor you have. You could argue even more so. ”

And by adding information such as address or income, Watson may be able to do a better job of calculating what a patient needs.

On the population health front, Monica Bharel, MD, the Massachusetts Commissioner of Public Health, said she found that a lot of healthcare data sits in silos.

“We don’t always do as good of job in taking that data and turning it into digestible, actionable information,” she said.

Bharel called for “precision public health” that, much like precision medicine, uses data to target the most effective ways to address health issues. For example, Bharel’s staff collected data from various sources that might impact on the opioid epidemic: death records, prescriptions drug monitoring records, and data from corrections and mental health agencies.

They were able to learn that only a small percentage of non-fatal overdoses patients were treated with medications that reduce the subsequent risk of death and actual overdose deaths by 50%.

“I really think this is the future of public health,” she said.

Other sessions covered pending changes to the Affordable Care Act, opioid abuse, antibiotics, and medical education.