The use of predictive modeling paired with cluster analytics on patient-reported data has the power to pinpoint populations that may need specialty services prior to care delivery, according to a new report published in eGEMs (Generating Evidence & Methods to improve patient outcomes).
The researchers, from Kaiser Permanente Colorado (KPCO), gave more than 6,000 patients questionnaires as they enrolled in the KPCO delivery system.
The questionnaires asked about general health status, conditions that interfered with daily activity, diseases like diabetes and asthma, medications and financial constraints, among other topics.
With that data in hand, the researchers used predictive models and cluster analytics and algorithms to reach conclusions about a population’s health needs.
They found, for example, that those who would have the highest costs of care include patients with fair to poor health, who required medications and who were enrolled in a low deductible plan.
The authors noted that previously using predictive models for this kind of researcher was the norm, but said that it has its limitations.
“Tailoring successful interventions to optimize efficient and effective care requires either focused predictive models designed around specific interventions ... or additional means of identifying meaningful subpopulations,” they wrote.
Those additional means, they added, are cluster algorithms, which can ID subpopulations and then parse out the highest-risk groups into subgroups with identifiable care needs.
“Anticipating care needs before they happen can improve system efficiency and minimize the adverse health outcomes associated with discontinuity,” the researchers concluded.
Health systems increasingly are using analytics as they see the benefits of data analysis for population health.
To learn more:
- check out the study