Predictive Algorithms, Artificial Intelligence: Tools for Early Intervention


According to a January 2017 report from management consulting firm NewVantage Partners, approximately 95 percent of Fortune 1000 companies have made significant investments in big data initiatives within the past five years.

The senior living profession is no exception. United Church Homes collects demographic and health data of its residents that can be used to predict the health status of a particular resident and improve care, said Chuck Mooney, senior vice president of senior living services at United Church Homes.

“To me, data analytics is like the unicorn — often talked about, rarely seen,” Chuck said. “Our cloud-based software platforms allow for a number of things. We’ve invested in hiring Clinical Informatics Manager Kathy Ely. She’s really our resident in-house PointClickCare (PCC) expert. The idea is to figure out how to use the data that we collect in more intelligent ways to, perhaps, predict falls, monitor diabetic status and be able to predict future illness for residents with diabetes.”

According to the PointClickCare blog, much data analytics comes from electronic health or medical records. This can include nursing assessment information, medication usage, hours and types of therapies and more. Documenting relevant data is one side of the coin. The other is the ability to analyze, understand and apply the data that has been collected. EMRs contain a treasure trove of data that can be leveraged to drive positive health outcomes among residents, increase staff engagement and drive operational efficiencies.

Every EMR contains a wealth of resident information, all of which can be analyzed to better understand individual resident care needs, to more efficiently schedule staff on different shifts and even to document whether certain meals have fallen out of favor among residents, according to PointClickCare.

The ultimate vision for UCH is to partner with a college gerontology program to provide an avenue for research-based data collection. Predictive algorithms and artificial intelligence technology can, for example, monitor a person’s movement patterns during a defined period of time. A change in these patterns may indicate a potential problem and, once identified, can be addressed quickly and efficiently.

“All providers in long-term care are required to collect the minimum data set (MDS),” Chuck said. “The MDS record is a very extensive clinical record that draws from the medical record and also from other sources, such as social work notes and chaplain/spiritual services. It captures all that information into a standard data set that is then reported to both the state and federal levels.”

Casper reports are quality metric reports generated from these data sets that have over 2,000 metrics with standard definitions. The reports lend themselves to benchmarking relative to UCH’s peers in senior living.