Clean claim rate is the first number a practice can move and the last one to reach the bank account. It counts the claims that leave the billing system with no human touch, so it falls days or weeks before cash slows down. What follows is the definition HFMA uses, the arithmetic behind the number, the reject codes that drive it, and the reason days in A/R follows it by one to two months.
What this covers
- HFMA defines the clean claim rate as claims that pass edits with no manual intervention divided by claims accepted into the billing system, which makes it a pre-submission measure.
- A claim that fails the payer’s front-end edits is a rejection, not a denial, and it never receives a control number, per the CMS Medicare Learning Network.
- Medicare may take 30 days on a clean claim and 45 days on one that needs development, and the clock pauses while the contractor waits on your practice.
- Days in A/R divides a balance by a three-month average daily revenue under HFMA FM-1, so a fix made this month shows up in the metric slowly.
- Track rejection reasons by code, because the durable fix sits at the source: eligibility, prior authorization and provider enrollment.
A rejected claim is not a denied claim. The payer never put it in front of an adjudicator, and the filing clock kept running while it sat in the work queue.Medicare’s front-end edits run before a claim gets a control number.
The number gets argued about because two teams can report the same figure and mean different things. One counts claims that cleared the scrubber, the other counts claims the payer paid on the first submission. The definitions and timelines below come from HFMA’s MAP Keys, the CMS Medicare Learning Network fact sheet on claim status, the Medicare instructions for other-than-clean claims, and the current X12 code lists.
What the Clean Claim Rate Counts
HFMA’s MAP Keys define the clean claim rate (CL-1) as the number of claims that pass edits requiring no manual intervention, divided by the number of claims accepted into the claims processing tool for billing. The numerator counts primary, secondary and tertiary 837 claims that cleared the scrubber untouched. The denominator counts what entered the system, so claims submitted straight to a payer without running through the edits are excluded.
The measurement point sits inside the practice, before submission. Nothing in the definition involves the payer’s decision, the payment, or the appeal.
| Measurement | Numerator | Source of the data | What a weak number tells you |
|---|---|---|---|
| Clean claim rate (HFMA CL-1) | Claims passing edits with no manual intervention | Claims processing tool or scrubber | Data quality at intake, coding and charge entry |
| Front-end rejection rate | Claims the payer rejects before adjudication | 999 and 277CA acknowledgement reports | Format, enrollment and authorization problems |
| First-pass acceptance rate | Claims paid or appropriately adjusted on first submission | 835 remittance advice and payer portal | Coverage, medical necessity and clinical documentation |
The three measures do not move in step. A practice can post a 98 percent scrubber rate and still lose one claim in ten after adjudication.
Set the pair side by side and the diagnosis gets easier. A weak scrubber rate points at intake and charge capture. A weak first-pass acceptance rate next to a strong scrubber rate points at payer policy, authorization and documentation. Teams that run medical billing services watch both, because the second number is where the money actually stops.
A single claim can be clean at the scrubber, rejected by the payer’s front-end edits, and denied after adjudication. Each event lands in a different report on a different day, so a dashboard showing one number is showing one third of the story.
The Arithmetic Behind One Point
The formula is a division, and the useful part is what the remainder represents.
| Clean claim rate | Claims needing manual work per 1,000 | Same figure per 100 |
|---|---|---|
| 98% | 20 | 2 |
| 95% | 50 | 5 |
| 92% | 80 | 8 |
| 88% | 120 | 12 |
| 85% | 150 | 15 |
Manual intervention is a biller opening the claim to add an authorization number, correct a rendering provider identifier, recheck eligibility for the date of service, or attach documentation a payer asked for. Each event consumes staff time that would otherwise go to aged A/R, and each one delays the claim.
Work the direction of the movement. Moving from 88 percent to 95 percent on 1,000 monthly claims removes 70 rework events a month, and those 70 claims reach the payer on the first pass instead of the second.
A clean claim rate improves if you stop counting claims. HFMA excludes claims never accepted into the billing system from the denominator, so a practice that parks problem claims outside the queue can report a better number while the backlog grows. Report the number of claims sitting in the queue next to the rate.
Why Days in A/R Follows the Clean Claim Rate
Days in A/R is a balance divided by an average, and that shape explains the delay. HFMA’s FM-1 key defines net days in A/R as the net A/R balance divided by average daily net patient service revenue, where the revenue figure is a three-month daily average. One month of rejects changes the balance gradually, so the metric moves in the direction the clean claim rate pointed a month or two earlier.
The claim-level timeline makes the same point. A clean electronic claim reaches payment status in as little as 14 days with one Medicare contractor, and Medicare owes interest when a clean claim is not paid within 30 calendar days of receipt. A claim that fails the front-end edits has no clock at all until it is corrected and resubmitted.
The file goes to the clearinghouse or the Medicare contractor, which returns a 999 acknowledgement on the batch.
The payer checks each claim against the HIPAA implementation guide requirements. CMS allows up to three days for both edit levels together.
Claims with errors come back on a 277CA acknowledgement and carry no control number until they are corrected.
Someone opens the claim, fixes the field and resubmits. The filing deadline does not move while that happens.
A clean electronic claim has a 14-day payment floor with Medicare. A claim that needs development can run 45 days under the Social Security Act standard.
Add the days a rejection sits in a work queue and the arithmetic is plain. A rejected claim commonly reaches payment four to eight weeks after the clean claim submitted beside it. Days in A/R cannot react to that until the balance ages, so the clean claim rate leads and days in A/R confirms.
Rejections, Denials, and the Codes That Separate Them
A rejection means the claim never reached adjudication. A denial means an adjudicator saw it and paid nothing or less. CMS describes two levels of edits: the file-level 999 acknowledgement, which can reject an entire batch, and the claim-level check against HIPAA implementation guide requirements, which returns a 277CA and rejects only the claims with errors. A rejected claim has no document control number or internal control number.
The reason codes tell you which of the two events you are reading. Verify each code against the current X12 lists before you build a report on it, because X12 retires codes and a retired code quietly corrupts a category count.
| Code | Wording | What it means for the work queue |
|---|---|---|
| CARC 16 with RARC MA130 | Claim/service lacks information or has submission/billing error(s) | The claim is unprocessable. MA130 states no appeal rights are afforded, so correct it and submit a new claim. |
| CARC 197 | Precertification/authorization/notification/pre-treatment absent | The authorization was missing, invalid or not on file for those dates. Fix the authorization before resubmitting. |
| CARC 288 | Referral absent | A referral the payer requires is not on the claim. |
| CARC 18 | Exact duplicate claim/service | Often self-inflicted. Resubmitting while a claim is still in the edit process creates one, per CMS. |
| CARC 27 | Expenses incurred after coverage terminated | Coverage ended before the date of service. Recheck eligibility before the resubmission. |
| CARC 29 with RARC N211 | The time limit for filing has expired | Provider liability under most contracts. N211 states you may not appeal the decision. |
Separate the queues by code and the follow-up changes. A CARC 16 claim needs a correction and a new claim. A CARC 197 claim needs a call to the payer or a retro authorization request. A CARC 29 claim usually needs proof that an earlier submission was received, and the denial management workflow for it looks nothing like the workflow for a coding denial.
CARC 15, the older wording for a missing or invalid authorization number, is deactivated. X12 stopped it on May 1, 2018, and the active code for a missing authorization is 197. When a report still groups authorization problems under 15, that report is reading a retired code and its counts cannot be trusted.
Where Behavioral Health and ABA Claims Break
Authorization and units cause most of the front-end failures in this niche. One state Medicaid billing guide for applied behavior analysis shows the pattern: adaptive behavior treatment requires prior authorization, the initial assessment is limited to one in every 180 days, a continued service request has to reach the program by the last authorized date with a recommendation to file 5 to 15 days earlier, and an incomplete request is pended with five business days to supply the missing information before a technical denial.
Unit arithmetic matters just as much. Weekly hour limits apply to the combined units of the treatment codes 97153, 97155 and 0373T, so a schedule that runs past the authorized units produces claims the payer edits out, and an authorization that expires mid-month splits one month of service into two billing problems.
Provider identifiers are the second source. State programs require each service to be billed with the NPI of the provider who performed it, not the supervising clinician, and a rendering provider who is not enrolled with that payer fails the front-end check no matter how clean the coding is. One missing enrollment produces a rejection on every claim for that clinician.
- Authorization number on the claim matches the dates of service billed.
- Units billed fall inside the authorized units for that authorization period.
- Rendering provider NPI is the person who performed the service, and it is enrolled with the payer.
- Diagnosis code on the claim matches the diagnosis the authorization was issued for.
- Continued service request filed before the last authorized date.
- Eligibility verified for every date of service on the claim, including the last one.
The Routine That Moves the Number
Two habits do most of the work: reading the acknowledgement report every day, and grouping rejections by reason code instead of by biller.
The 277CA arrives within about 25 minutes of a file submission with one Medicare contractor and stays retrievable for 60 days. Working it the day it lands turns a two-week delay into a same-day correction.
Log the reason code rather than a free-text note. Codes group into causes, and causes can be fixed once.
Count the top three reasons each month. A single intake field that is wrong generates the same rejection on hundreds of claims.
A wrong tax identification number or a missing enrollment is corrected once in the system. A wrong modifier on one claim is corrected once on one claim.
The clean claim rate tells you this week whether last month’s fix worked. Days in A/R confirms it a month or two later.
Sample the queue and check whether the correction was made in the system or only on the claim in front of the biller.
A practice that reports the rate without the reject volume behind it keeps meeting the same problems on the remittance advice. The number is a management tool only when someone owns the cause behind it.
Who does the work matters less than whether the fix reaches the source. Our team reports the scrubber rate and the first-pass acceptance rate together, because a practice can only manage the one it can see.
Clean claim rate questions
HFMA defines the metric without publishing a numeric target in the MAP Keys, and the widely quoted 95 percent is an industry convention rather than a benchmark. A defensible target depends on where you measure. Set the floor from your own data: take the best quarter your practice has posted at the scrubber, treat that as the baseline, then set a separate and lower floor for first-pass acceptance after adjudication.
No. The clean claim rate counts claims that clear your edits without manual intervention, measured before submission. First-pass acceptance counts claims the payer pays or appropriately adjusts on the first submission. A claim can be clean and still deny for authorization or medical necessity. Report both, and expect the second number to be lower. The gap between them measures payer policy risk.
Weekly at the scrubber, monthly for the payer side. Rejections come back within days, so a weekly number tells you whether last week’s change worked. Days in A/R and the denial rate belong on a monthly report, because both move slowly. Report by payer and by provider when you can, since a single practice-wide number hides the payer that causes most of the rework.
Yes. Medicare requires a claim to reach the correct contractor within 12 months of the date of service under 42 CFR 424.44, and the contractor’s receipt date is what counts. A claim the payer rejected as unprocessable is not a filed claim. One Medicare contractor states plainly that resubmitting corrected claims that were unprocessable falls inside that same 12-month requirement. The deadline does not pause for a work queue.
It measures something else. CMS reported a 93.45 percent claim accuracy rate and a 6.55 percent improper payment rate for Medicare fee-for-service in its 2025 supplemental data, and improper payment includes underpayments and missing documentation rather than only denials. Use that as context for how often payment rules are missed across the program, not as a target for your practice.
Rank them by volume, dollars and whether the cause repeats. A missing rendering provider enrollment produces a rejection on every claim for that clinician, so it outranks a one-off address error worth a few dollars. The reason codes do this work for you: CARC 197 and 288 point at authorization, 27 and 177 at eligibility, and 16 at data quality. Work the largest repeating cause first.
The bottom line
A clean claim rate is not a score for the billing team. It measures how much of the practice’s revenue arrives on the first attempt, and it responds to changes in intake, authorization and enrollment faster than any other number on the report. Move it, and days in A/R follows.
Where is your clean claim rate leaking?
Send us a month of claims with the acknowledgement reports and the remittance advice behind them. We will separate the clean claim rate from the first-pass acceptance rate, group every rejection by reason code, and show you which cause to fix first. Our team runs medical billing services and denial management for behavioral health and specialty practices.
Request a free clean claim rate auditPayment rules, code lists and contractor instructions change, so confirm current requirements with each payer before you rebill a claim; this is billing guidance, not legal advice.
