Every year, CMS runs a statistically valid sample of Medicare claims through its Comprehensive Error Rate Testing program to estimate how much of what it pays out shouldn't have been paid at all. The most recent numbers for hospice and home health are worth sitting with, not because the dollar figures are shocking on their own, but because of what's actually driving them.

Here's what the data shows, and what it means for how agencies should think about documentation.

The numbers themselves

The overall Medicare fee-for-service improper payment rate sits at 6.55 percent for the most recent reporting period, representing $28.83 billion. Within that, hospice services show an improper payment rate of 6.8 percent for non-hospital-based hospice, a projected $1.6 billion, and 10.8 percent for hospital-based hospice, roughly $190.6 million. Home health's improper billing came in somewhat lower than hospice's in the same reporting cycle, but it's far from a rounding error either.

What's actually driving these numbers

This is the part that matters most for agencies trying to learn something useful from the data. The root cause behind most of these improper payments isn't fraud in the sense most people picture. It's unsupported medical necessity, documentation that doesn't clearly establish why the level of care billed was actually warranted.

For hospice specifically, one of the most common, specific failure points is a missing or inadequate physician narrative explaining the clinical findings that support a life expectancy of six months or less. When that narrative is missing or too generic, a reviewer completes the claim as an insufficient documentation error, and the payment gets recouped, regardless of whether the clinical judgment behind it was sound. The documentation didn't fail because the care was wrong. It failed because the record didn't show its work.

Why this connects directly to the mismatch reviewers flag first

This is the same underlying pattern behind the diagnosis-versus-function mismatch that tends to be the first thing a reviewer's eye catches on any chart. The CERT data confirms it at scale: it's rarely that care wasn't warranted, it's that the documentation doesn't clearly show why it was. Individualized, specific, contemporaneous documentation is what closes that gap. A generalized certification or a narrative written to satisfy a checkbox doesn't.

Why this data matters more given the current enforcement climate

CMS is currently in the middle of a real increase in audit and enforcement activity, and this improper payment data is part of what justifies that heightened scrutiny in the first place. An agency with a real, but honestly-earned, documentation gap is sitting in the same statistical bucket as one with actual fraud until a reviewer looks closely enough to tell the difference. The CERT numbers are the evidence CMS points to when arguing for more scrutiny across the board, which means every agency inherits some of that scrutiny, deserved or not.

What actually reduces exposure here

Closing this gap isn't about writing more documentation, it's about writing documentation that specifically supports medical necessity in a way a reviewer unfamiliar with the patient could follow and agree with. That means individualized clinical findings tied to the specific patient in front of the clinician, not boilerplate language that could apply to any chart in the system.

See where your own documentation stands against this

We'd like to walk through how our QA process specifically checks for medical necessity documentation gaps before they turn into a CERT statistic.