Looking for your highest risk patients? Predictive analytics will help you find them

In my last blog, I talked about a branch of analytics called predictive analytics, which help identify patients who are at highest risk for being hospitalized.

Focusing on such patients can help dramatically improve their health outcomes—and yield significant cost savings. But first you have to identify and prioritize the individuals in need of help.

Population health management programs that are powered by predictive analytics help health care providers identify and stratify patients, often with high-, medium- and low-risk designations. Once sorted by risk, patients can be targeted for intervention with evidence-based treatment programs.

One example is Steward Health Care System, which received its designation as a Pioneer Accountable Care Organization (ACO) in December 2011. The leaders of the Boston-area integrated system knew care management would be key to improving the cost and quality of care for the more than 35,000 lives in the ACO population.

One of the first-and most important steps-toward formulating a strategy was analyzing data.Jeremy Orr, MD, MPH

Analytics were crucial in determining the highest-impact opportunities for reducing the cost and increasing the continuity of care. One of Steward’s most important findings was that a high number of readmissions were going to skilled nursing facilities (SNFs) rather than inpatient facilities.

To reduce readmissions, Optum’s population health management programs helped Steward target the critical transition point from acute care to home. They introduced care managers to address patients’ chronic or complex medical conditions and ensure patients followed recommended plans of care, and provided a nurse line for telephonic access and support to patients. A new readmission prevention program connected with patients while still in SNFs, then provided post-discharge follow-up.

In the game of baseball, there’s a corollary to what Steward did. In the eyes of most modern baseball statisticians, a “walk” is as good as a base hit. It may take less skill, and it may be less exciting, but the result is the same. A player is on base, and once that happens, that player can score. Think of finding and intervening with your sickest patients as the equivalent of earning a walk in baseball. It’s a relatively easy way to quick results.

In my next post, we’ll continue to look at how to use data proactively to manage population health.

To read more about providers who are using analytics to increase the coordination of care and reduce costs download this Optum eBook, “Moneyball Analytics: Connecting and leveraging the best data across the health care continuum”

–Jeremy Orr, MD, MPH, Chief Medical Officer, Optum Analytics

Leave a Reply