What is the Anahp Observatory?
The Anahp Observatory is the annual report of the Associação Nacional de Hospitais Privados (Anahp, Brazil's National Association of Private Hospitals) with market, care and institutional indicators, including financial ones, for its member hospitals [4]. The data comes from the Anahp Hospital Indicators System, fed by the hospitals themselves [3]. Its indicators include denial and the collection period, measured with definitions published by Anahp itself [1].
Anahp also publishes the Balanço Observatório Anahp, a shorter quarterly report. It combines a selection of hospital indicators with financial data on health plans (operadoras), taken from the open data of the Agência Nacional de Saúde Suplementar (ANS, Brazil's National Supplementary Health Agency, the federal regulator of private health plans) and processed by the consulting firm Arquitetos da Saúde [3].
Both reports are available in the publications area of Anahp's website, with download after registration [4]. What they measure fits into the broader set of revenue cycle KPIs, and many Brazilian hospitals use them to compare their denial and collection period with the sector.
Observatory and Balanço: frequency, sample and sources
The two publications serve different purposes and should not be cited as if they were the same.
| Observatório Anahp | Balanço Observatório Anahp | |
|---|---|---|
| Frequency | Annual | Quarterly |
| Data period | Full previous year (the 2025 edition covers 2024) [1] | Latest quarter or accumulated year (the 9th edition, March 2026, covers full-year 2025) [2] |
| Content | Market, care and institutional indicators, including financial ones, with definitions [1][4] | Selected hospital indicators + health plan data + price adjustments and costs |
| Hospitals | Members reporting to the system (2024: 169 members, 119 reported [1]; 2025: 194 members, 138 reported, 71.13% [6]) | Average of members: 194 in December 2025, 199 in Q1 2026 and 203 in Q2 2026 [2][3][7] |
| Data sources | Anahp Hospital Indicators System | Same system (hospitals) + ANS open data processed by Arquitetos da Saúde (health plans) [3] |
| Recent editions | 2025 and 2026 [4] | 8th (January 2026, data through Q3 2025), 9th (March 2026, full-year 2025), June 2026 edition (Q1 2026) and 11th (September 2026, Q2 2026) [2][3][4][5][7] |
Each edition may revise prior-year figures, and the sample changes. The rule is to always cite the edition, the period and the denominator. In the Observatório Anahp 2026, the most recent annual report, the 2025 figures match those in the March Balanço: initial denial of 15.93%, accepted denial of 1.72% and an average collection period of 77.35 days, with data from 138 of 194 members (71.13%) [6]. The 11th Balanço, from September 2026, covers Q2 2026 and changes the initial denial method [7].
What do the hospital numbers say for 2025 and 2026?
Anahp hospitals became more efficient in patient care, but cash flow got worse: the collection period rose, initial denial stayed at the highest level in the series and margins fell [2][3].
| Indicator (average of Anahp hospitals) | 2022 | 2023 | 2024 | 2025 | Q1 2026 | Q2 2026 |
|---|---|---|---|---|---|---|
| Initial denial (% of receivables from health plans) | — | 11.89% | 15.89% | 15.93% | — | 11.69%*** |
| Accepted denial (% of gross health plan revenue) | 0.78%* | 1.17% | 1.96% | 1.72% | 1.61% | 1.35% |
| Average collection period (days) | 73.51 | 76.38 | 69.91 | 77.35 | 73.84 | 69.54 |
| Average payment period to suppliers (days) | 45.25 | 49.40 | 50.70 | 48.30 | 48.38 | 52.78 |
| EBITDA margin | 11.26% | 11.89% | 11.68% | 11.05% | 9.45% | 10.60% |
| Occupancy rate | 76.64% | 76.85% | 78.97% | 77.79% | 76.85% | — |
| Average length of stay (days) | 4.29 | 4.10 | 3.99 | 3.76 | 3.72** | 3.78** |
| Health plans' share of gross revenue | 83.42% | 80.56% | 78.80% | 76.47% | 80.21% | — |
* In 2022, measured on total net revenue (old basis, not comparable). ** New patient-day concept in 2026. *** From 2026, measured on amounts received from health plans in the month (new method, not comparable with 2023–2025). Sources: [1][2][3][6][7].
See the numbers as a table
| Period | Average collection period (days) | Accepted denial (% of gross health plan revenue) |
|---|---|---|
| 2021 | 68.56 | — |
| 2022 | 73.51 | 0.78% (old basis: total net revenue) |
| 2023 | 76.38 | 1.17% |
| 2024 | 69.91 | 1.96% |
| 2025 | 77.35 | 1.72% |
| Q1 2026 | 73.84 | 1.61% |
Three readings follow from the table. First, in 2025 hospitals took 77.35 days to collect from health plans and 48.30 days to pay suppliers, nearly 30 days financed with their own cash or credit [2]. Second, total expenses absorbed about 93% of net revenue in 2024 and 2025 and 95.09% in Q1 2026 [2][3]: with thin margins, every point of denial comes straight out of the bottom line. Third, the Observatório Anahp 2025 shows the problem goes beyond denial: in 2024, hospitals collected 88.61% of what they billed health plans over the previous five years, and invoices more than 90 days overdue equaled 61.53% of average monthly billing [1].
In Q2 2026, the collection period was 69.54 days, against 78.49 in the same quarter of 2025, and accepted denial was 1.35%, against 1.70% [7]. The supplier payment period was 52.78 days, already on the new denominator [7].
Anahp's denial has two denominators
Anahp publishes two denial indicators, each with its own denominator. Initial denial (a management figure) is the amount denied by health plans while the claim is still under negotiation, measured on receivables from health plans [1][2]. Anahp notes that this indicator does not include amounts not yet billed, which could make receivables even larger [1][6]. From 2026, Anahp divides initial denial by the amounts received from health plans in the month: the 2026 series (11.69% in Q2) is not comparable with 2023 to 2025 [7]. Accepted denial (an accounting figure) is what the hospital does not receive after audit and negotiation, measured on gross health plan revenue [2]. In US terms, they are close to the initial denial rate and the final denial write-off rate.
Three common mistakes follow. The first is dividing 1.72% by 15.93% to conclude that “the hospital recovers 89% of denials”: the bases are different and the calculation is meaningless. The second is comparing the 2022 accepted denial (0.78%) with 2023 onward: until 2022 the denominator was total net revenue [2]. The third is treating accepted denial as the total loss. Anahp itself notes that hospitals also accept discounts in settlements to get invoices paid, which the indicator does not capture [2]; in a survey of 85 hospitals, the allowance for doubtful accounts totaled R$ 1.8 billion in 2024 (R$ 1.4 billion in 2023), and withheld denials, R$ 5.8 billion in 2024[8].
The difference between denial types is in the guide Initial, final and accepted denial, and the formula for a hospital to measure its own is in the guide Denial rate.
What are the reading traps in each edition?
Errors when citing Anahp figures usually come from mixing series, periods or concepts. The table summarizes the most common traps.
| Trap | What happens | How to read it correctly |
|---|---|---|
| 2021 collection period cited as 2024 | 68.56 days is the 2021 value; 2024 was 69.91 days [1] | Check the year in the original table |
| Mixing annual and quarterly series | Each quarter has its own series: collection period of 75.80 (2023), 72.32 (2024), 74.31 (2025) and 73.84 days (2026) in Q1 [3], and 76.64, 70.22, 78.49 and 69.54 days in Q2 [7] | Compare a quarter with the same quarter of prior years |
| Using a partial accumulated figure as the full year | Through Q3 2025, Anahp reported initial denial of about 18% and a collection period of 78.5 days [5]; the year closed at 15.93% and 77.35 days [2] | Always cite the edition and period |
| 2026 length of stay | In 2026, a patient-day began to count any patient cared for during the hospital day, including same-day admission and discharge; before, it was the midnight census [3] | Do not compare 3.72 days with the earlier series without a caveat |
| 2026 supplier payment period | The denominator changed from total gross revenue to total expenses (excluding depreciation, provisions and financial expenses) [3] | Treat 2026 as the start of a new series |
| 2026 initial denial | The denominator changed from receivables to amounts received from health plans in the month; Q2 2026 was 11.69%, against 15.63% in Q2 2025 [7] | Treat 2026 as the start of a new series; do not read the gap as a drop in denials |
| Growing sample | Members went from 120 (end of 2023) to 203 (Q2 2026); Anahp warns that heterogeneity affects the figures [2][3][7] | Read the trend, not tenth-of-a-point changes |
| Average as a target | These are averages of Anahp member hospitals, not of the hospital sector as a whole | Compare the hospital with itself, using Anahp as context |
Anahp's average collection period (the Brazilian equivalent of days in A/R) is counted from the invoice date [1] and therefore does not capture the time before the invoice is issued, such as billing submission windows imposed by health plans. To measure the full cycle, a hospital must add the time between discharge and claim submission. The calculation is in the guide Average collection period.
What the health plan numbers tell the hospital
Medical-hospital health plans have recovered: in 2025 their net result was the highest in the series, and in Q1 2026 more than 80% of them were profitable [2][3].
| Medical-hospital health plans | 2023 | 2024 | 2025 | Q1 2026 |
|---|---|---|---|---|
| Members (millions) | 50.88 | 51.60 | 53.09 | 52.93 |
| Premium revenue (BRL billion) | 272.81 | 304.61 | 336.66 | 86.07 |
| Loss ratio | 86.8% | 83.4% | 81.5% | 80.8% |
| Net result (BRL billion) | 1.93 | 10.29 | 23.43 | 6.02 |
| Net result ÷ revenue | 0.71% | 3.38% | 6.96% | 7.00% |
| Health plans with positive result | 76.3% | 73.3% | 79.2% | 80.4% |
Source: ANS open data processed by Arquitetos da Saúde, published in the Balanço reports [2][3]. The 2023 share (76.3%, 426 health plans) is from the March 2026 Balanço [2].
Two details change the reading. The 2025 operating result was BRL 8.78 billion (2.61% of revenue); the rest of the net result came mainly from financial income [2]. And the gain is concentrated: the ten largest health plans have about 43% of members and about 71% of the result [2]. For Q1 2026, Anahp estimates that, without an extraordinary provision by one large health plan, the sector result would have been about BRL 7.6 billion [3].
These figures are not the same as those released by ANS: the series excludes dental plans and benefit administrators, and members are adjusted for organic change [2][3]. When citing, write “ANS, via Balanço Observatório Anahp”. The full reading is in the guides Health plan results and Loss ratio. The full series is in Industry data.
How to use the Anahp Observatory in hospital management?
Anahp figures are context, not targets, and have three practical uses:
- Calibrate your own indicator. Calculate the hospital's initial denial, accepted denial and collection period with the same denominators Anahp uses. Only then does the comparison make sense.
- Prepare negotiations with health plans. Health plan results and loss ratios, together with the hospital's collection period and denial, support arguments on price adjustments and payment terms. Break the data down by health plan, because the sector average hides who pays worst.
- Set the revenue committee agenda. With each Balanço, compare the sector trend with the hospital's.
What we see in operations
At Accrua, we use the Anahp Observatory as a reading reference, not as a target. When we compare a hospital with the average, the first check is the denominator: initial denials are measured against amounts receivable from health plans through 2025 and against amounts received in the month from 2026, and accepted denials against gross health plan revenue; the average collection period starts at the invoice. We rebuild the hospital's metrics on the same base, period and counting rule before putting them side by side.
A hospital's own average also hides the spread. In the diagnostic, we break denials and collection time down by health plan, plan type and claim type, from the history of claims sent. That breakdown is where the differences an average folds into one line become visible.
When reading each edition, we check the year, whether the figure is annual or quarterly and whether the sample changed. A quarterly figure compared with an annual one, or a swapped denominator, changes the conclusion about the same hospital.
Where to start
- Download the latest edition of the Balanço and the Observatory from Anahp's website and note the period, sample and denominators.
- Recalculate the hospital's indicators with the same definitions: initial denial on receivables (or on amounts received in the month, to compare with 2026), accepted denial on gross health plan revenue, collection period from the invoice date.
- Keep the annual and quarterly series separate in the hospital's spreadsheet.
- Break down denial and collection period by health plan before comparing with the average.
- Review the comparison with each Balanço, recording any definition changes announced by Anahp.