Healthcare organizations are increasingly trying to crunch data to find patients who are at risk of being hospitalized and returning to the hospital soon after going home. But finding them doesn't mean they'll all respond to the same kinds of interventions.
A not-for-profit group that works closely with the North Carolina Medicaid agency has developed a data-driven transitional-care program intended to connect the right patients with the resources that are most likely to work.
The physician-led Community Care of North Carolina, which coordinates care for 1.4 million Medicaid beneficiaries, created an algorithm two years ago based on transitional care delivered to more than 100,000 patients. The process produces an “impactability score.”
“Traditional models look at risk,” said Dr. Tom Wroth, CEO and chief medical officer of North Carolina Community Care Networks, the Medicaid arm of CCNC. The impactability score, Wroth said, “predicts prospectively which patients are going to benefit from which transitional-care interventions.”
With this approach, CCNC has significantly reduced hospitalizations for some of the state's sickest patients. In March it was selected as the winner of the inaugural Hearst Health Prize, a $100,000 award recognizing outstanding achievement in managing or improving health.
Since the program began in 2008, hospital admission rates for Medicaid recipients with multiple chronic medical conditions have declined by 10%, according to the organization. Readmissions for the same group have dropped by 16%.
For example, a 6-year-old boy had more than 30 emergency department and hospital visits in a year because of seizures. One episode resulted in a 16-day hospital stay. Under the transitional-care program, a care manager helped the boy's mother find a new pediatrician and get a referral to a neurologist who initiated a new treatment plan. A few weeks later, the boy was stabilized with the help of medication and went seizure-free for almost a year.