Patient readmission is one of the most pressing problems in healthcare today. This common performance metric is so important to effective healthcare that the federal government now imposes penalties for providers with excessive readmission rates. While it’s becoming more common for healthcare systems to use data to improve outcomes, efforts to understand readmission are still in their infancy. Today’s models for readmission give us warning, but they don’t give us meaning.
Patients that return for treatment within thirty days (a common benchmark) represent a disproportionate amount of our total healthcare spending, even though they are only a fraction of the total population of healthcare consumers. And with private insurance, public programs, and the basic social safety net covering these costs there’s common cause in reducing readmission.
The interest in addressing patient readmission is simple enough. A healthcare system that sees patients over and over for the same problems is not addressing patient needs. Figuring out why these patients’ needs are not being addressed is a more difficult problem.
The true challenge of patient readmission is the ability to address issues at the individual level, rather than the population. Today these targeted programs are conducted using traditional risk models. Historical patient data is analyzed to create a scoring system, identifying which patients are likely to return within the thirty day window. Providers can then direct more resources to high-risk patients, hopefully leading to better outcomes.
Risk scoring helps hospitals to target specific patients, but what it doesn’t do is indicate why a patient is high-risk. For that, health systems will have to employ not just predictive, but prescriptive solutions. Methods that not only identify which patients are likely to be readmitted, but also which factors are affecting that level of risk.
Making the Change
Implementing prescriptive models for readmissions will change healthcare as we know it. Beyond just scheduling more resources for high-risk patients, prescriptive predictions identify the most impactful factors making a risky patient, or even the combinations of factors that put a patient at risk.
Armed with that knowledge, providers will be able to tailor each treatment guided to prevent readmission and improve patient outcomes. Targeted treatments will then lead to more effective patient care, better outcomes, and dramatically reduced readmission rates.