Hospital margins 2024: 54% lose money on patient care
We summed the patient-service margin of all 5,857 usable FY2024 hospital cost reports. 54.1% came out negative. Only 30.2% of the same hospitals lost money overall — the 24.7% in between were rescued by a single line item.
5,857 US hospitals filed a usable FY2024 cost report. On patient care alone they lost $41.3 billion — a −2.7% margin on $1.505 trillion of net patient revenue — and 54.1% of them were in the red. The same hospitals reported +$137.9 billion of net income, a median total margin of +5.6%. The difference is the cost report's "other income" line: $188.6 billion, or 4.6× the size of the patient-care loss. The finding is not "US hospitals are losing money". It is "US hospitals cannot make money treating patients". Those two sentences lead to opposite conclusions.
One line separates two pictures
A cost report carries three rows that matter: net patient revenue, operating expense, net income. The first minus the second is "net income from service to patients". Add other income and subtract other expenses and you get net income.
We computed two margins:
- Patient-service margin = net income from patient services ÷ net patient revenue
- Total margin = net income ÷ (net patient revenue + other income)
Same hospitals, same year:
| Measure | Aggregate | Median | Negative |
|---|---|---|---|
| Patient services | −2.7% (−$41.3B) | −1.5% | 3,168 (54.1%) |
| Total | +9.2% (+$137.9B) | +5.6% | 1,768 (30.2%) |
1,445 hospitals (24.7%) fall between the two lines: they lose money on patient care and make money overall. Only 45 go the other way.
Those 1,445 are not outliers. They are the normal case. Reading them as mismanagement misjudges a large set of financially sound institutions.
Negative five years running
One year could be a post-pandemic artefact. Five years say otherwise:
| Fiscal year | Filings | Net patient revenue | Patient-service margin |
|---|---|---|---|
| 2020 | 5,791 | $1,089B | −4.9% |
| 2021 | 5,828 | $1,226B | −2.9% |
| 2022 | 5,824 | $1,292B | −4.6% |
| 2023 | 5,863 | $1,397B | −3.0% |
| 2024 | 5,866 | $1,505B | −2.8% |
Revenue grew 38% in five years. The margin never turned positive. This is the cost structure of the industry, not a bad year.
Rural hospitals lose more — and still survive
Split by location and the table earns its keep:
| Group | Hospitals | Patient-service (agg / median) | Negative | Total margin (median) |
|---|---|---|---|---|
| Urban | 3,186 | −1.2% / +1.0% | 48% | +6.1% |
| Rural | 2,671 | −4.4% / −4.1% | 62% | +5.2% |
Rural hospitals lose more on patient care — 14 points more of them are in the red. And yet their median total margin is +5.2%, 0.9 points below urban.
Put plainly: rural hospitals are not failing. They cannot earn a living from patient care. Those two readings behave very differently in lending, M&A and supplier credit decisions.
By facility type the spread widens:
| Type | Hospitals | Patient-service (median) | Negative |
|---|---|---|---|
| Rehabilitation (RH) | 391 | +10.3% | 25% |
| Short-term acute (STH) | 3,154 | −1.1% | 54% |
| Children's (CH) | 70 | −0.1% | 51% |
| Long-term acute (LTCH) | 322 | −2.2% | 54% |
| Critical access (CAH) | 1,368 | −5.2% | 65% |
| Psychiatric (PH) | 537 | −0.2% | 50% |
Critical access hospitals are the only group where more than 60% lose money on patient care — yet their median total margin is +5.5%, identical to short-term acute hospitals. That number alone says CAH survival does not depend on patient-care profitability.
How deep the losses run
Bucketing the margin shows losses are not "near zero":
| Patient-service margin | Hospitals | Share |
|---|---|---|
| Below −10% | 1,808 | 30.8% |
| −10% to −5% | 612 | 10.4% |
| −5% to 0 | 756 | 12.9% |
| 0 to 5% | 733 | 12.5% |
| 5% to 10% | 591 | 10.1% |
| Above 10% | 1,366 | 23.3% |
Thick at both ends, thin in the middle. Nearly a third of hospitals lose more than 10% of patient revenue, while nearly a quarter earn more than +10%. This is not a distribution wobbling around zero; it is a split one.
We cut a finding we could not stand behind
Our first pass grouped hospitals by the cost report's "Type of Control" field and produced: for-profit +13.6%, non-profit −3.0%, government −13.5%. A dramatic, publishable result.
Then we sampled names. Of 134 hospitals with "HCA" in the name, 36 carry control code 2 (non-profit), 66 code 4 (for-profit corporation), 7 code 11 (hospital district). Of 53 "Ascension" hospitals, 49 are code 1. Of 110 "Mercy" hospitals, 81 are code 1.
So the field is unreliable at the individual-hospital level. Group averages off it would be precise and meaningless. We deleted those three rows.
What this file cannot answer
- What "other income" is made of. Investment returns, government payments, 340B, philanthropy and non-patient business all land in one line. Which of them carries the $188.6 billion is not broken out here.
- Who is about to fail. Cost reports are self-reported, fiscal-year-specific and published with a lag. They describe the year that ended; they do not predict closure.
- Whole-industry totals. Our denominator is net patient revenue plus other income, not audited total revenue.
FAQ
How can patient care lose money while the hospital profits?
Is this the same as the rural hospital closure wave?
Why 5,857 reports and not every hospital?
- Source: CMS Hospital Cost Report, FY2024 final file, 5,857 usable filings; the five-year series uses the matching FY2020–FY2024 files.
- Method: patient-service margin = net income from service to patients ÷ net patient revenue; total margin = net income ÷ (net patient revenue + other income). Identity check: patient-service income + other income − other expenses = net income held within $1,000 for 5,799 of 5,857.
- Excluded: filings with net patient revenue ≤ 0 or missing net income.
- Limits: self-reported, fiscal-year-specific, published with a lag; other income is not itemised; federal hospitals are largely absent.
Signup comes with a key and an MCP config: query cost-report fields by state, bed count, facility type or fiscal year — net patient revenue, operating expense, other income, net income — returned per hospital, with the same definitions used here.
2,000 free reads included with signup; searches and misses are not counted.