OIG flagged a hospice group as high-risk. 45% failed review.
Summary
Medicare's hospice benefit is meant for enrollees with a life expectancy of 6 months or less. HHS's inspector general built a targeted sample: new FY2021 hospice enrollees who had no hospital or emergency room visit in the 18 months before starting hospice care, and who were still alive 6 months after starting it. Of 100 such enrollees OIG reviewed, the records for 45 did not meet Medicare's hospice eligibility documentation requirements. Extrapolated across the 46,767-enrollee population that fit this profile, OIG estimates Medicare could have saved $255.1 million by screening for it -- and CMS has agreed to consider doing so.
Two ways the eligibility case fell apart
The failures split into two categories. For 21 of the 45⧉ failed periods, the enrollee's medical records simply didn't clinically support a terminal-illness diagnosis -- $251,067 in unallowable payments. For the other 24, the records were missing required eligibility documentation entirely -- $294,432 more. Together, the sample's 45 failures totaled $545,499 in payments classified as unallowable, out of $1.2 million reviewed.
View data as table
| Terminal illness not clinically supported (21 periods) | 251,067 |
|---|---|
| Missing required eligibility documentation (24 periods) | 294,432 |
A $545,499 finding, extrapolated to $255 million
's sampling frame -- every FY2021 enrollee nationwide who fit the no-recent-hospital-visit, survived-past-6-months profile -- came to 46,767 initial certification periods⧉, tied to $580.5 million in Medicare hospice payments. Extrapolating the sample's 45% failure rate across that population, 's statisticians produced a point estimate of $255.1 million in potential savings, with a 90% confidence interval running from $204.9 million to $308.8 million -- meaning the estimate implies roughly 44% of all the money paid to this specific population went to claims whose records didn't document eligibility, not that the enrollees were confirmed ineligible or that anything improper occurred.
View data as table
| Lower limit | 204,918,573 |
|---|---|
| Point estimate | 255,113,489 |
| Upper limit | 308,786,496 |
CMS agreed -- with a caveat
's recommendation was narrow: that work with hospice Medicare Administrative Contractors to treat this specific enrollee profile -- no recent hospital or ER visit before hospice -- as a high-risk category⧉ worth targeted eligibility review, potentially before or shortly after payment. concurred, but its written response committed only to sharing the report with hospice MACs 'to use in their risk analysis and work planning to determine whether the risk area identified in this audit report should be prioritized' -- agreement in principle, without a firm commitment to build the review procedure recommended.
The takeaway
- A hypothesis-driven sample found a near-coin-flip failure rate. didn't sample hospice enrollees at random -- it targeted a specific profile it suspected was risky, and 45% of that targeted sample failed eligibility review.
- The extrapolated number is a meaningful share of the money at stake. $255.1 million against $580.5 million in total payments to this population -- 's own math implies close to half of that spending may not have been properly earned.
- agreed with the finding but not yet to a fix. Its response commits to sharing the audit with contractors for prioritization discussions, not to building the prepayment review procedure recommended.
's targeting criteria -- no inpatient or ER claim in the 18 months before hospice, plus survival past 6 months -- is a risk-based proxy, not a diagnosis of fraud. In its written comments, explicitly stated that ' did not identify any instances of fraud' in this particular audit, distinguishing it from separate fraud enforcement -- the same comment letter notes recently suspended payments to nearly 450 Los Angeles-area hospices, which received about $600 million in Medicare payments in 2025, over a credible allegation of fraud unrelated to this audit. This report's findings are documentation and clinical-support gaps in the medical record, not a determination that any enrollee was actually healthy or improperly admitted to hospice; 21 of the 45 failures involved a medical review contractor finding the record insufficient to support a terminal-illness prognosis, while the other 24 were missing required paperwork such as election statements or signatures, unrelated to clinical eligibility. This audit covers FY2021 claims only and does not estimate savings for hospice enrollees outside this specific profile.
Sources(1) ▾
- HHS Office of Inspector General, Medicare Could Have Saved $255.1 Million Related to Hospice Services for Certain New Hospice Enrollees (Report No. A-06-22-09003) (2026-06-18) — statistical audit of Medicare hospice payments to a specifically-targeted high-risk population: new FY2021 hospice enrollees with no inpatient or emergency room claims in the 18 months before starting hospice care, who then survived at least 6 months into hospice care. Issued June 18, 2026. Read in full via curl with a browser user agent and pdftotext -layout, including the statistical sampling appendices. oig.hhs.gov · original document
Comments
Always open. Logged-in readers can annotate paragraphs in place.
's Office of Inspector General built a targeted sample⧉ to test a specific hypothesis about Medicare's hospice benefit, which is reserved for enrollees with a life expectancy of 6 months or less. It looked at new FY2021 hospice enrollees who had no inpatient hospital or emergency room claim in the 18 months before starting hospice care -- since a patient's health typically declines with hospital visits before a terminal diagnosis -- and who were still alive 6 months after starting hospice. Of 100 such enrollees reviewed, the records for 45 did not meet Medicare's hospice eligibility documentation requirements.