Using predictive modeling to keep patients at home
Healthcare providers are financially responsible for an increasing number of at-risk patients. Once these patients are discharged, providers lose contact with them until they return to the ER. Home monitoring coupled with a predictive analytics model housed in the cloud helps providers intervene before patients return to the hospital.
According to research done in 2014 by Avalere, of the 49 million Medicare beneficiaries, 63 percent of them have three or more chronic conditions, said Greg Gordon, Philips’ vice president of product management and marketing. These patients tend to be elderly and frail and will have at least one hospital admission during the course of a year.
“When they’re hospitalized and discharged … 70 percent of them go home with no assisted technology to help them stay connected to the healthcare organization which is increasingly at financial risk for their well-being,” he said.
Post-discharge interventions such as calling patients or sending nurses to visit patients in their homes may be effective but not scalable over time. Telehealth can be effective for the highest risk patients when applied for short periods of time but cost prohibitive for every at-risk patient that is discharged, said Gordon. A more cost-effective alternative that may also result in better health outcomes and avoided financial cost is a new cloud-based solution built on top of the #1 medical alert service in the U.S. – Philips’ Lifeline. This use of a private pay service accepted by seniors may assist at-risk organizations to leverage what seniors already know.
That new solution – CareSage – uses an algorithm based on the combined service usage data from Lifeline patients and from electronic health records to predict with a high degree of certainty which patients are at risk for emergency transport in any upcoming 30-day period.
“The important value of this technology,” said Gordon, “is it allows providers to only focus those expensive and valuable interventions on the individuals that the algorithm has identified as potentially at increased risk for transport, which may be an indicator of a change in status. It allows them to manage a large population over a long period of time, but only focus their interventions on those who need them.”