Your clean claim rate (the share of insurance bills that go out without an error) dropped, and someone went looking for who messed up. That was the wrong first move, and the number that triggered it was telling you less than you thought.
The clean claim rate looks like a report card on your billing team. The share of claims that go out without an error, clean on the first pass. So when it slips, the instinct is to look at the people preparing the claims.
But that number blends two completely different things, and only one of them is about your team. The other is about the payers, the insurance companies, and they get a vote that has nothing to do with how good your work is. Blended like that, your clean claim rate is lying to you.
What the number actually mixes together
A claim can come back not clean for two very different reasons. One is a real error on your side, a wrong code, a missing modifier, a bad ID. The other is a payer changing the rules, getting more aggressive, or applying a new edit that flags claims that were fine last month.
Both show up the same way in your clean claim rate. The number can’t tell them apart, so it presents payer behavior and team performance as one combined figure, and then everyone reads the combined figure as a judgment on the team.
That’s where it misleads. The number is real, but it points at your people while a large part of what moves it is the payers, who you don’t control and didn’t change.
What we found when we separated them
At one practice, the billing quality was 96.2 percent, genuinely strong work by the team. The clean claim rate told a gloomier story, and the gap between the two was the payers.
When we pulled the not-clean claims apart and sorted them by cause, a large block had nothing to do with the team’s accuracy at all. They traced to disputes the payers were raising, and those disputes were piling up in a category nobody had separated out.
In that category sat about $650K in disputed claims, blended into the same bucket as ordinary errors, invisible as its own problem. The team was being measured on a number that was dragging in $650K of payer behavior and presenting it as if the team had slipped. The team hadn’t slipped. The payers had moved, and the report had no way to say so.
Why blaming the team is worse than wrong
When the clean claim rate drops and leadership responds by pressing the billing team, two bad things happen. First, you spend effort fixing people who were already doing 96 percent good work, which produces almost nothing because the error was never the main driver.
Second, you miss the actual problem, the $650K of payer disputes that needed a completely different response, because you were looking at the wrong cause the whole time. The morale cost is real too. Nothing wears a good team down faster than being blamed for a number they don’t control.
Why the category stays hidden
The disputes hid in plain sight because the practice never split the category. Claims that came back got lumped together, errors and disputes in the same bucket, and the report totaled the bucket without ever asking what was inside it.
A category set up once and never questioned will quietly hold very different things under one label, and the total looks orderly the whole time. The $650K wasn’t buried by anyone on purpose. It was buried by a category that was never designed to tell errors and disputes apart.
What to measure instead
The fix is to stop watching the blended rate as if it were one thing and split it into its two real parts. Measure your billing quality, the share of claims that were actually correct when they left your hands, separately from payer behavior, the share that came back because a payer raised an issue.
Now you have two numbers that mean different things. One tells you about your team and is genuinely actionable by coaching and process. The other tells you about the payers and calls for a different response, working the disputes, escalating patterns, renegotiating where you can.
The number that already separates them
The split has a name in the standard, and knowing it changes the conversation with anybody who benchmarks.
HFMA defines clean claim rate narrowly: claims that pass all edits without manual intervention, divided by claims accepted into the claims processing system for billing. Read that carefully. It measures whether the claim survived the scrubber (the software that checks claims for errors before they’re sent). It says nothing about whether the payer paid it correctly.
First pass resolution rate is the stricter number. It counts claims paid in full on first submission, with no rework, no resubmission, no follow-up. The gap between the two metrics is precisely the territory this article is about.
And that gap is where underpayments hide. A claim can pass every edit, reach the payer, adjudicate (get the insurer’s decision), and come back short of the contracted rate. Clean claim rate counts it as clean. Your dashboard shows it as paid. Nothing flags a shortfall, because from the system’s point of view money arrived. Only reconciling payments against your contracted rates surfaces it.
So a practice can run a strong clean claim rate and a first pass resolution rate well below it, and never know the second number exists. If you track only the first, the entire category of correctly formatted claims paid at the wrong price is invisible to you.
There’s a measurement trap on top of that. Counting a claim as clean the moment your scrubber passes it isn’t the same as counting it clean when the payer accepts it. Measured at the clearinghouse, the figure flatters you. Measured at the payer, it comes in lower, by whole points.
How other fields handle a mixed signal
Any field that depends on measurement learns early to separate signal from noise, and to never act on a number that mixes the two. A scientist doesn’t read a result without accounting for the instrument. An engineer doesn’t treat a sensor reading as truth without knowing what else moves it.
What this means for you
You can check your own number. Take last quarter’s not-clean claims and sort them by cause. Put the genuine errors in one pile and the payer disputes in another. If the second pile is large, your clean claim rate has been telling you a story about your team that’s really a story about the payers, and you have been managing the wrong one.
The zero-dollar version of this problem, denials (claims the insurer refused to pay) posted as “paid” payments, has its own write-up: Your Claims Went Somewhere. Nobody Could Say Where.
Grab 30 minutes with us. Prep nothing. You will see what your clean claim rate has been hiding.
Questions people ask
What’s a good clean claim rate for a medical practice?
Aim for 95% or more of claims going out clean, meaning no errors on the first pass. But the rate mixes two things: real errors by your team, and insurance companies changing their rules or adding new checks. Sort the claims that came back by cause before you judge anyone. A strong team can sit behind a weak-looking number.
Why is my clean claim rate dropping when my billing team hasn’t changed?
The rate can drop without your team changing, because it mixes your team’s errors with changes on the insurance company’s side. An insurance company can change its rules, add a new check that catches claims that were fine last month, or raise disputes. All of it counts against the same number. So the rate can fall while your team’s work holds steady.
How do I separate my billing team’s errors from insurance company issues?
Take last quarter’s claims that didn’t go out clean and sort them by cause. Real errors on your side go in one pile: a wrong code, a missing modifier (the add-on that explains a code), or a bad ID. Insurance company rule changes, new checks, and disputes go in the other. A large second pile means the rate has blamed your team for something it doesn’t control.
Should I change billing companies if my clean claim rate is low?
Not until you’ve sorted the claims that came back by cause. If most of the gap traces to insurance companies changing rules or raising disputes, a new billing company inherits the same problem and the number won’t move. At one practice we measured, the sort found about $650K in disputed claims mixed in with ordinary errors. That money needed a different response entirely.