Insufficient documentation caused 53.0% of Medicare fee-for-service improper payments in CMS's fiscal year 2025 report, more than any other error type. Among the service types a physician practice bills, non-fee-schedule lab tests lost the most to missing or incomplete records ($1,087,049,622), followed by minor procedures ($769,641,180).
CMS built the report from 49,061 sampled claims, of which 37,651 were reviewed. The claims cover the 12 months from July 1, 2023 through June 30, 2024. The overall improper payment rate was 6.55%, or $28.8 billion on $439.9 billion paid.
How CMS splits Medicare errors by cause
CMS sorts every improper payment into one of five error categories. Table A3 of the 2025 supplemental data gives each category's share of all improper payments, after the adjustment for rebilled inpatient claims.
| Error category | Share of 2025 improper payments |
|---|---|
| Insufficient documentation | 53.0% |
| Medical necessity | 15.3% |
| No documentation | 12.0% |
| Incorrect coding | 11.1% |
| Other | 8.5% |
Source: CMS, 2025 Medicare Fee-for-Service Supplemental Improper Payment Data, Table A3. Base: all improper payments, Parts A and B and DMEPOS.
Insufficient documentation means the provider sent records, but they did not prove the service met Medicare rules. No documentation means the provider sent nothing when CMS asked. Together the two categories make up 65.0% of improper payments. Incorrect coding, the category most billing teams watch, makes up 11.1%.
The overall rate from insufficient documentation fell from 4.6% in 2024 to 3.5% in 2025, according to Table A2. The overall improper payment rate fell from 7.7% to 6.6% over the same two reports. Part B alone had a rate of 8.4% on $114.1 billion paid, based on 11,920 sampled claims.
Practice service types with the most documentation errors
Table F2 lists the service types with the most improper dollars caused by insufficient documentation. It also includes hospital outpatient ($3,464,787,362) and skilled nursing facility inpatient ($2,313,469,722) claims, which the table below leaves out. The rate column shows the improper payment rate from that one cause, measured against all payments for the service type. The overall total in the table is $15,285,281,270, or 51.5% of improper payments in that table.
| Service type | Improper dollars from insufficient documentation | Rate from that cause |
|---|---|---|
| Lab tests, other (non-Medicare fee schedule) | $1,087,049,622 | 22.3% |
| Minor procedures, other (Medicare fee schedule) | $769,641,180 | 13.6% |
| Specialist, other | $567,474,991 | 21.0% |
| Ambulatory procedures, other | $252,054,200 | 19.5% |
| Oncology, radiation therapy | $148,078,138 | 11.0% |
| Chiropractic | $128,898,870 | 27.2% |
Source: CMS, 2025 Medicare Fee-for-Service Supplemental Improper Payment Data, Table F2. Base: Medicare fee-for-service payments for each service type, July 2023 to June 2024.
Chiropractic has the highest rate of the six, at 27.2%, but its dollar total is the smallest. Lab tests carry the largest dollar total because the service type is large. Table E1 puts the 95% confidence interval for the lab test total rate of 27.9% at 20.1% to 35.7%, so the true rate has wide room on either side of the estimate.
Which services fail mostly on documentation
Table E1 ranks Part B service types by total improper payment rate and splits each rate by error type. For the top-ranked services, documentation is the main cause.
- Chiropractic: a 30.4% improper payment rate, of which 89.5% was insufficient documentation.
- Specialist, other: 25.0%, of which 84.0% was insufficient documentation.
- Ambulatory procedures, other: 24.2%, of which 80.4% was insufficient documentation.
- Lab tests, other (non-Medicare fee schedule): 27.9%, of which 79.9% was insufficient documentation.
- Minor procedures, other (Medicare fee schedule): 17.6%, of which 77.6% was insufficient documentation, and 3.3% of overall improper payments in Table E1.
Other tests, other shows a different split. Its rate is 19.9%, and 51.0% of the errors were no documentation, against 46.3% insufficient documentation. For that service type, about half the errors came from records that CMS never received.
Lab tests, other (non-Medicare fee schedule), account for 4.6% of overall improper payments in Table E1, and specialist, other accounts for 2.3%. Chiropractic accounts for 0.5%, so its high rate sits on a small base.
Visits fail for a different reason
Evaluation and management visits show the opposite pattern. In Table F2, established office visits lost $140,235,348 to insufficient documentation, a rate of 0.9%. In Table F4, which covers incorrect coding, the same service type lost $556,500,710, a rate of 3.5%.
| Service type | Improper dollars from incorrect coding | Rate from that cause |
|---|---|---|
| Office visits, established | $556,500,710 | 3.5% |
| Hospital visit, subsequent | $361,156,694 | 7.1% |
| Office visits, new | $178,377,755 | 5.9% |
| Emergency room visit | $152,935,085 | 9.7% |
| Nursing home visit | $134,623,746 | 6.3% |
Source: CMS, 2025 Medicare Fee-for-Service Supplemental Improper Payment Data, Table F4. Base: Medicare fee-for-service payments for each service type.
Table E1 shows the same pattern by share. Of home visit errors, 56.2% were incorrect coding. For subsequent hospital visits the share was 51.4%, and for nursing home visits it was 49.6%. A practice that bills mostly visits needs a coding check. A practice that bills labs, procedures, or therapy needs a records check.
What a practice can check
The CMS rates come from Medicare fee-for-service claims, so they do not describe commercial payers. They do show where Medicare reviewers found gaps in the records. A practice can use them to pick which claims to sample first.
For lab and procedure billing, a practice can pull a sample of paid claims and match each one to a signed order and a note for the date of service. For visits, it can compare the level billed with the note. A mismatch in either direction counts as an error under the CMS method, so downcoded visits show up too.
Each sample also gives a number to track. If 3 of 50 sampled lab claims lack a signed order, the practice has a 6.0% record gap on that service, which it can compare with the CMS rates above.
Sources
Talk to us
Fifteen minutes about your practice and the systems you run.