Retrospective risk adjustment is dead? Hardly

A few weeks ago, CMS released its 2027 Medicare Advantage Advance Notice for payments for Medicare Advantage (MA) plans. The headlines covering it were grim: According to industry analysts, CMS is holding rates flat in 2027 and is going to eliminate chart reviews. Based on this news, the share price of United Healthcare, the largest health insurer in America, plummeted and remains down by almost 20%.

There are two problems with the reporting that rates are flat and risk adjustment is dead:

1. Neither is true
2. Insurers have the tools necessary to adapt to what CMS is proposing

Are rates really flat?

Hardly. Rock Health Capital portfolio company HealthMC did a fantastic job of explaining how the “rates are flat!” headlines overlooked CMS’s proposal of 2.5% increase. Yes, you had to read some footnotes to figure this out, and yes, a 2.5% increase is lower than prior years. But it’s certainly more than 0%.

The trickier bit to untangle is the reporting on changes to chart review—or retrospective risk adjustment.

But first: What is “Risk Adjustment”?

Medicare Advantage was designed to overcome the limits of traditional fee-for-service Medicare and to incentivize private Medicare Advantage health plans to deliver quality health care outcomes at more predictable costs. Health plans get paid a (relatively) fixed amount per member, with additional incentives for quality and health outcome measures.

“Risk adjustment” in Medicare Advantage is a mechanism for CMS to adjust the payments it makes to plans up or down based on each enrollee’s documented health status and expected medical costs: Plans are paid somewhat more for sicker plan members. Retrospective chart review—the process of looking at each member’s medical records over the prior quarter or year—is the process by which plans may identify potentially overlooked health problems.

Closing a risk adjustment loophole

In 2024, the Wall Street Journal reported that under existing risk adjustment rules, plans had identified diagnosis in charts—and received payment from CMS—in patients who never received treatment for those conditions. This was and is permitted under the existing rules for retrospective chart reviews. This raises the potential for overpayment from CMS. But in its 2027 Advance Notice, however, CMS didn’t propose eliminating risk adjustment altogether.

Instead, CMS is proposing to close what appears to be a loophole: Instead of identifying and submitting potential diagnoses based solely on a patient’s chart, a health plan will be required to (if the Advance Notice is adopted as new regulation) link a diagnosis in a chart to an encounter (i.e., treatment by a doctor).

Under existing rules, bigger plans with more resources have had the ability to throw bodies at the retrospective chart review problem and to dig deeper into charts, review those charts, and submit more potential diagnoses to CMS. This meant that bigger plans had a competitive advantage over smaller plans—even though those smaller plans might be more nimble and effective at delivering high quality patient care.

So, CMS isn’t penalizing MA plans as a whole so much as it’s trying to level the playing field in a manner that improves competition and benefits consumers. In case this was in doubt, Chris Klomp nailed it:

“We do not want risk adjustment to be a source of competitive advantage for health plans.”
— Chris Klomp, CMS deputy administrator and director of Medicare

So what? Doctors aren’t medical coders (and never will be)

Yes, the proposed change will, based on CMS’s modeling, reduce aggregate payments to MA plans (because that’s the goal). CMS knows that under their proposal there will be fewer charts linked to encounters than there are simply “unlinked” charts. But just because plans have to link a chart to an encounter doesn’t mean chart review / retrospective risk adjustment will suddenly disappear.

On the contrary, plans can’t afford to give up on retrospective chart reviews. In the day-to-day of patient care, doctors don’t exactly prepare their notes in a chart and generate an ideally-coded encounter all in the same moment in every single encounter. The ideal world of “prospective review”—when the chart/physician notes and encounter/claim line up perfectly in real-time—is and always will be more the exception than rule.

In the real world, chart notes and claims/encounters happen in all sorts of sequences. Correctly identifying and linking these will remain a complex, non-discretionary task for as long as MA exists. CMS isn’t proposing that payers suddenly bear all of the costs for care of their sicker patients: They want MA plans to identify the cost of caring for less healthy patients in a manner that is more strict, uniform, and transparent across the industry.

To us at Rock Health Capital, this points to stronger long-term footing for MA as a whole—and reinforces our belief that payers will need robust tools like HealthMC to meet CMS’s higher bar.

What’s next?

The final rule is forthcoming in a few weeks. Stay tuned, healthcare nerds. There’s sure to be a lot to wade through.

Rock Health Capital is an early-stage venture fund focused on digital health. We would love to hear from you. Get in touch!