Predictive Modeling for Bundled Payments
The following article was written by Danielle Eldredge, a freelance healthcare journalist
Nearly 500 care organizations are enrolled in the Bundled Payments for Care Improvement (BPCI) initiative. To help navigate uncharted bundled structures, progressive care providers are looking toward predictive modeling.
As defined by the Centers for Medicare and Medicaid Services, BPCI organizations will enter into payment arrangements that include financial and performance accountability for episodes of care. These models may lead to higher quality and more coordinated care at a lower cost to Medicare.
So why would a care provider voluntarily enroll?
The American Medical Association puts it like this: “Bundled payment arrangements are a type of risk-contracting. If the cost of services is less than the bundled payment, participating physicians and other health care providers retain the difference. But if the costs exceed the bundled payment, physicians and other providers are not compensated for the difference.”
In an effort to identify and minimize risk, innovative care providers are turning to predictive modeling technology.
“It’s the evolution of business intelligence,” said Cash Forshee, senior vice president of business development at Medalogix, a company that specializes in post-acute predictive analytics. “Predictive analytics can unveil scientifically driven insight unavailable to clinicians, which increases the opportunity to appropriately adjust care for the patient and reduce costs to the organization.”
Medalogix began as a way to reduce readmissions by helping post-acute providers give their patients better care and meet the requirements of the Affordable Care Act. Medalogix’s analytics combs through a health agency’s historical patient data to learn agency-specific readmission predictors and then identifies the top 20 percent of patients most at risk for readmission.
By understanding just who is at risk, caregivers can adjust care plans and reduce overall hospital readmissions.
This tactic has helped Alternate Solutions Home Care in Ohio reduce 30-day readmissions by 36 percent.
“Reducing readmissions has strengthened our relationships with referring hospitals but, more importantly, has improved our patients’ physical and emotional health,” said Chad Creech, Alternate Solutions’ chief development officer.
After realizing predictive modeling’s positive impact on readmission reduction, other post-acute organizations reached out to Medalogix about the possibility of a custom-bundled payment support model. The custom solution has since been created and deployed.
The bundled payment solution goes a step further than the readmission reduction solution. It considers diagnosis related group (DRG) data and DRG-payment allotments to ensure hospital clinical data and financial repercussions are incorporated in an organization’s risk analysis for patients.
“Using facility-specific data to understand who is most at risk, and then coupling that information with financial reimbursement information, we can help care providers understand where their resources will be best deployed for optimum outcomes,” Forshee said.
Some are concerned that this type of analysis, which is tied to financial reimbursements, could undermine care — influencing organizations to prioritize patients based on their financial risk as opposed to their clinical needs.
“There is an inextricable link between the financial and clinical risks in patients,” said Medalogix’s CEO, Dan Hogan. “Using scientifically driven insight to effectively prioritize care yields greater patient and organizational outcomes, which in turn reduces financial risk in value-based payment models.”
Medalogix is working to further enhance their risk identification service, adding coordinated care functionality to assist several care providers in understanding where a patient will receive optimum care through each phase of their care episode.