Aiming for the Stars: Leveraging AI to Improve Low CMS Star Ratings

August 13, 2024

min read

Understanding the 2024 CMS Star Ratings and Their Impact

‍

Since the release of the 2024 Star ratings from CMS about one month ago, every conversation we’ve had across the healthcare ecosystem has included at least some mention of them. Everyone is talking about the ‘beating’ that MA Plan Sponsors have taken through a tricky combination of the Tukey outlier deletion to the removal of COVID era protection. With 58% of plans rating below 4 Stars (up from 38% in ‘22 and 49% in ‘23) and new competition from disruptive health plans (such as Devoted Health), there has never been more urgency to improve Star ratings, thus "Aiming for the Stars".

It’s become obvious that what worked in the past to drive or maintain Star ratings isn’t going to necessarily work going forward. As an example, one can look to long-time 5 Star legend Kaiser Permanente, who didn’t land the designation for ‘24. Times are tough.

‍

The Role of Modern Technology in Improving CMS Star Ratings

‍

The reality is that all Star measures are not created equal and when we look at the non-CAHPS scores that measure the quality of care, we realize that moving these metrics takes a fresh approach. The good news is that there is help on the horizon with new and innovative solutions that can change the game. Cloud-based solutions, robust data strategies, and AI/ML have made it even easier than ever to innovate and drive significant progress without major investment or lift from our friends in IT. This is also the bad news, because these technologies will rapidly create new benchmark cut-points, and plan sponsors who allow rigid, old-school processes and technology to stand in the way of change will find their way to the bottom of the pack.

What is your strategy to leverage these modern technologies to improve outcomes, experience and Star ratings?

‍

The Importance of Addressing Non-CAHPS Scores in Star Ratings

‍

So what do we do? Many plan sponsors have bundled their non-CAHPS strategies into basic categories such as preventative care, wellness, medication, chronic conditions and transitions/readmission. While no metric is easy to manage, the “transitions and readmission” category is particularly tough and drove many plans to a lower Star rating in ‘24. In fact, the ‘Transitions of care’ and ‘Plan all-Cause Readmission’ metrics had the lowest average of the entire Star ratings for ‘24 (averaging 2.8 and 2.5 respectively).

The problem is that these are hard metrics to impact as the landscape is complicated. Plan sponsors deploy care management strategies in the marketplace where clinicians are in high demand and their workload is demoralizing. The beneficiary is equally frustrated as often these care management strategies might check the box but lack empathy and understanding. These metrics are Star destroyers and there has to be a better way.

This is where modern technology can play a major role and plan sponsors are scrambling to understand what tools are available to aim at these Star metrics. Natural Language Processing opens up the door for AI-enabled, cloud-based technologies to incorporate context into a plan of care. This new ‘context’ can do so much. When done properly, AI-generated ‘context’ can identify and alert a care team to new barriers or conflicts within a plan of care. For example, if an individual is depressed due to the loss of a loved one, their chances of missing critical medications almost doubles. If an individual lacks transportation, their chances of making that appointment are slashed by a third. Imagine the impact of living in a food desert.

AI can provide an extra set of ears that allow care managers to scale like never before. The end result is a win-win-win scenario as the beneficiary, care manager, and the plan sponsor all benefit from better outcomes. A recent study found that AI-enabled contextual care reduced readmissions by 35%. The same study found that care managers were able to manage 6x their standard workload! This is what happens when we leverage modern tools to ‘humanify’ the patient experience.

This technology has never been easier to implement. Gone are the premise based solutions that require an army of on-site engineers. Browser based cloud solutions are the way of the future and accelerate innovation in an exponential manner. Just like the stories of Blockbuster, Redbox, and Netflix, we will look back at this time in history and realize that THIS was the moment where plan sponsors either adopted modern technology to improve outcomes or they stuck with the same old video cassette. It’s time to leverage AI contextual care and to aim for the Stars.