Contents
The HHS Office of Inspector General recommended that Pinnacle return $31,227,884 to Medicare after its PDPM audit found unsupported payment codes, unnecessary skilled services and documentation failures. Reviewers found that 99 of 100 sampled claims failed Medicare requirements. Pinnacle disputed the findings and every recommendation. For healthcare staff in assessment, therapy or billing, the dispute reaches familiar ground: what the resident’s chart says, what the assessment records and how those entries become money.
The Office of Inspector General reviewed Medicare Part A claims for services Pinnacle provided during 2020 and 2021. Pinnacle Multicare Nursing and Rehabilitation Center is a 480-bed skilled nursing facility in the Bronx, New York. The November 2025 audit says the agency selected it after data analysis showed a significant increase in reimbursement following the change to the Patient Driven Payment Model.
The audit covered 3,371 claims totaling $50,828,595 in Medicare payments. The Office of Inspector General selected a simple random sample of 100 claims, representing $1,605,609, and sent the claims and supporting records to an independent medical review contractor.
The Office of Inspector General identified $1,059,443 in overpayments within that sample. Medicare paid for claims the reviewers said the charts could not support. The report contains Pinnacle’s written response alongside the agency’s answers.
The Patient Driven Payment Model bases payment classification on resident characteristics and clinical needs. It replaced the RUG-IV system’s therapy-minute classifications on October 1, 2019. The payment question moved from the volume of therapy toward the conditions and needs recorded for the resident.
PDPM combines six payment components. Physical therapy, occupational therapy, speech-language pathology, nursing and nontherapy ancillary services each have a classification based on the resident. A separate component covers resources that do not vary with patient characteristics.
The Minimum Data Set assessment supplies the clinical inputs for those classifications. Section GG records usual self-care and mobility performance. Those assessment inputs feed the Health Insurance Prospective Payment System code reported on the Medicare claim. PT and OT have separate payment components but share the first character of the HIPPS code. The bill carries the result of several clinical decisions inside a short code.
Those decisions draw on different people. Nurses and aides observe the resident across shifts. Therapists assess function and treatment needs. Physicians supply diagnoses and orders. A registered nurse coordinates or conducts the assessment, and contributors certify their portions. The coding decision can therefore be compared with observations made elsewhere in the care record.
The Office of Inspector General found unsupported HIPPS codes in 95 sampled claims. Nursing classification errors appeared in 88 claims and speech-language pathology errors in 63. PT/OT errors appeared in 38 claims and nontherapy ancillary errors in 31. A single claim could appear in several categories.
The Office of Inspector General’s nursing example concerned a resident whose classification reflected septic shock and moderate depression. Reviewers said the chart did not support the depression level. They found no documented investigation of the mood distress, safety interventions or communication of the distress to the physician. The physician’s progress notes recorded no depressed mood.
Pinnacle answered that Section D of the MDS should record the resident’s own mood-interview responses, without physician notes overriding them. It argued that follow-up care planning was separate from the coding that determined payment:
OIG’s interpretation improperly conflates care planning with reimbursement criteria.
The Office of Inspector General replied that Pinnacle had omitted the assessment manual’s instruction for a clinician to evaluate the relevance of reported symptoms. It maintained the nursing finding.
In the Pinnacle OIG audit, the aphasia disagreement concerned whether the diagnosis was active during the assessment period. The screening note listed aphasia, but OIG said it did not document an effect on the resident’s functioning or care. The note also said speech therapy was not indicated at that time.
Pinnacle argued that an active aphasia diagnosis could support the speech-language classification without active speech therapy. It pointed to care plans, staff communication assessments, interviews and other clinical notes as possible support. The facility maintained that the resident’s diagnosis met the requirements.
The Office of Inspector General’s answer accepted that the resident had aphasia. Its reviewers disputed the evidence that the diagnosis was active: the note did not show an impact on function, a need for medical treatment or a need for nursing monitoring. That was the stated basis for retaining the finding.
The active-diagnosis disagreement matters because the corrected claim amount then feeds the extrapolated overpayment.
The Office of Inspector General’s PT/OT example concerned a claim classified around osteoarthritis as the primary reason for skilled services. Reviewers concluded that the medical documentation instead supported major depressive disorder as the primary reason. Their correction changed the therapy classification. The diagnosis chosen to explain the stay carried a payment consequence.
The nontherapy ancillary component pays according to qualifying conditions and services recorded in the assessment. In Pinnacle’s example, the facility coded a need for intravenous medication. The Office of Inspector General said the resident had multidrug-resistant organisms but that the corresponding treatment did not require IV medication.
For someone familiar with the medication record, the useful comparison is the treatment need documented for the resident against the treatment need used to classify the claim.
The Office of Inspector General found 54 claims involving residents it concluded did not require skilled nursing services. These findings could overlap with the coding findings. A rate-code correction asks how much a covered claim should pay. The skilled-need review asks whether the resident required the services being billed.
One example put the contradiction in plain view. The assessment described daily skilled needs for bed mobility and wheelchair training. The medical record described the resident moving independently. OIG concluded that the documentation did not justify the claimed need.
Pinnacle disputed the skilled-need findings too. It argued that residents’ age, acuity, frailty and other conditions required skilled professionals, and that the documentation adequately supported the care even when it did not use the reviewers’ preferred format.
For the claims examined here, the medical record needed to support the skilled services and the assessment behind the HIPPS code. OIG identified two claims with documentation deficiencies: one lacked sufficient medical records supporting skilled need, and another lacked the MDS assessment supporting the billed code.
Pinnacle already had written procedures for clinical review, coding and record retention. OIG’s finding was that staff did not consistently follow them.
The Pinnacle PDPM refund recommendation used the lower end of the Office of Inspector General’s estimated overpayment range. Its point estimate was $35,713,836. The agency chose the lower bound as its conservative recommendation for repayment.
The calculation began with the sampled claims. Reviewers compared what Medicare paid with what they concluded the documentation supported, producing full or partial overpayments for individual claims. The Office of Inspector General then used those results to estimate overpayments across the claim population.
An actual recovery is money returned to the program. As of October 18, 2026, OIG’s published recommendation status listed the refund recommendation as open and unimplemented. It marked the recommendations for internal review and additional staff training as implemented and closed on March 11, 2026.
The audit assigns the Centers for Medicare & Medicaid Services, acting through a Medicare contractor, the decision on overpayments and collection.
Pinnacle challenged the pandemic rules applied to its claims, the sampling and extrapolation, and the qualifications of the medical reviewers. It argued that reviewers overlooked emergency flexibilities, questioned whether the sample fairly represented its claims and said the reviewers lacked relevant skilled nursing experience. The Office of Inspector General maintained its findings after considering those objections.
Pinnacle was the first completed audit in OIG’s PDPM series. The same work-plan entry listed seven further audits in progress.
For a PDPM qui tam case, useful evidence connects the resident’s condition, the assessment entry, the paid claim and the people who knew or controlled the entry. The health care records supply the comparison: nursing observations, therapy notes, diagnoses, orders and the original or changed MDS assessment. HIPPS validation and billing records show what reached the claim.
Pinnacle’s audit makes those comparisons concrete. The physician’s note, the aphasia screen and the medication record each tested a different input to payment. We work with insiders to connect such discrepancies to who chose the classification, what they knew and what Medicare paid.
A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers.
If you work in assessment, therapy or billing and know why a payment code tells a different story from the resident’s chart, talk with our whistleblower experts about that difference. We build your strongest skilled nursing whistleblower case around the clinical records, payment decisions and your inside knowledge. We bring in Medicare fraud whistleblower lawyers who understand PDPM claims and stay with you from the first assessment of your concerns through the case.