Home health providers typically have to learn about their patients’ social determinants of health on the fly, quickly picking up on risk factors like access to transportation and proximity to healthy food following a hospital discharge.
One in-home care giant is turning to artificial intelligence for a head start.
Earlier this month, AccentCare Inc. revealed that it’s using a tool from clinical-AI company Jvion to better understand patients’ risk profiles and social determinants of health post-discharge. By leveraging the tool, the Dallas-based provider hopes it can further reduce its readmission rates, which is a positive for AccentCare, its patients and its referral partners alike.
“Obviously, something that’s important to us, our patients and our referral partners is readmission risk,” AccentCare CMO Dr. Anna Loengard told Home Health Care News. “And so what’s interesting to me about Jvion is that they will give us a readmission risk score for each patient, with recommended interventions.”
As a company, Advent International-backed AccentCare delivers home health, hospice, care management and personal care services across more than 240 locations in dozens of states. The provider simultaneously ranks as one of the country’s largest home health and hospice providers, with its end-of-life care capabilities bolstered following a merger with Seasons Hospice & Palliative Care.
In general, social determinants of health are all the non-medical variables that ultimately influence a person’s well-being. Payers and providers have been paying more attention to social impacts on health for years, but the focus has only sharpened of late with the U.S. health care system’s steady shift toward value-based care.
One in six patients has an unplanned acute care readmission within 30 days of discharge, according to statistics from the U.S. Centers for Medicare & Medicaid Services (CMS). Roughly half of America’s value-based care hospitals were penalized by CMS under its “Hospital Readmissions Reduction Program” for those high rates in 2020.
But while readmission rates are higher than what health care stakeholders would like, a patient’s return to the hospital is often avoidable.
In fact, as many as 75% of all readmissions could be prevented or reduced in severity with the right interventions, care coordination and communications, past research suggests.
AccentCare could have taken several different approaches to better analyzing social determinants of health. It decided to test out Jvion and AI to avoid placing any additional burden on front-line staff, Loengard said.
“Do we need clinical staff to actually code social issues as ICD-10 codes?” she said. “I just think that I have zero interest in adding to the burden of my front-line staff in that way. If we can actually learn this in ways that are behind the scenes, I think that we’ll be much more successful.”
Loengard had previously worked with Jvion back when she was the chief medical officer for Queen’s Healthcare System in Hawaii.
In addition to identifying patients with possible social determinants issues, AccentCare plans on using Jvion to flag which of the provider’s locations regularly care for more complex individuals, she noted.
“It’ll provide a valid way of discerning who is particularly complicated,” Loengard said. “And which of our branches have more complicated patient loads.”
Dr. John Showalter, chief product officer at Jvion, described a couple of different ways the AI tool can be used.
In one example, Jvion can help flag which patients don’t have access to public transportation or don’t live near a pharmacy, something that could hinder their ability to pick up life-saving medications. In that instance, Jvion would recommend enrolling such patients into a prescription-delivery service.
“To give another example, data might show that a diabetic patient lives in a food desert, in which case they wouldn’t have access to the nutritional food they need to stay healthy,” Showalter told HHCN. “In this case, the patient could be enrolled in a food delivery service.”
Currently, AccentCare is in the early stages of integrating Jvion. The provider hopes to get its program up and running within the next three months or so.
“It’s going to take a little while to get this implemented because obviously we have to share data with them and then figure out how to bring that data back in, in a way that’s usable,” Loengard said.
CMS has been encouraging providers to use artificial intelligence to cut costs and improve efficiency, even kicking off an “AI challenge” in 2019.