Part I — Situation overview

Two mutually reinforcing announcements arrived on 2 September 2026. In an interview given to Telex, Zoltán Kaló, appointed director general of the National Health Insurance Fund Manager in June, would transform the institution into a value-based, patient-centred purchaser of services; it currently manages a health insurance budget of some 5,000 billion forints — on ten- to fifteen-year-old servers. The practical content has three elements. One: the medicine manufacturer receives the full subsidy amount if its product delivers the undertaken outcome in real care as well, and otherwise has to pay back part of the subsidy. Two: the same logic for hospitals — higher financing where patient satisfaction indicators are better, lower financing where complications are frequent. Three: tougher bargaining over medicine prices, because according to the director general there is a product representing an item of billions which, on present knowledge, Hungary purchases at two and a half times the price paid by western European countries; for the expensive medicines used in rare diseases he plans joint negotiation with the Visegrád countries. The institution is also the consortium leader of an EU-financed project bringing together 13 public payers. The director general called the prime minister’s earlier promise of 500 billion forints a year of additional support for the sector “very probable”, but indicated that this relates to the sector as a whole, not to the institution’s development budget, and that the state of public finances is a constraint.

On the same day it was announced that on the night of 3 September 2026 the database of healthcare-associated hospital infections and epidemics by institution would become public on the website of the National Centre for Public Health and Pharmacy. Beatrix Oroszi, the national chief medical officer, emphasised twice at the background briefing: the aim is not to establish a ranking of hospitals, and the data are not suitable for that either. The data disclosure builds on the database of the National Nosocomial Surveillance System, with uniform definitions and validated methodology, on an interactive interface, with five-year time series, updated annually and also in the form of a downloadable database. The chief medical officer highlighted that an institution’s infection indicators are strongly influenced by the institution’s profile, the condition of the patients treated, the complexity of the care and also by the intensity with which infections are sought and confirmed microbiologically — which is why indicators aiding interpretation will be published alongside the infection indicators. A third piece of news runs in the background: the health minister has ordered an extraordinary inquiry at the National Institute of Oncology, where patients may, on suspicion, have been directed to private care.

MIAK’s reading: the two announcements are two halves of the same principle — public money has to be tied to what measurably improves the patient’s condition. This is the central proposition of MIAK’s health programme, so the direction is to be supported. The risk, however, lies precisely in the quality of the measurement. International experience is unambiguous that raw institutional data without risk adjustment distort — the centres treating serious cases appear bad — and that outcome-linked payment can also create bad incentives, from changing coding practice to patient selection. The right direction is not enough in itself; the details of implementation decide whether patient safety improves, or only the statistics.

Part II — Foundations in the literature

The interpretive frame is given by three sources. The first is Improving Healthcare Quality in Europe by Reinhard Busse and co-authors (a summary work of the European Observatory on Health Systems and Policies), which calls public reporting and quality-linked financing the two most contested of the quality-improving instruments: the literature examining their effect is extensive, but the evidence is not unambiguous, and unintended consequences are documented for both instruments. The second is the OECD’s Health at a Glance: Europe 2024 volume, which publishes the European hospital infection data and shows that the difference between countries narrows significantly between the raw and the risk-factor-adjusted values — that is, raw comparison in itself misleads. The third is the pharmaceutical market chapter of the volume Health Systems Governance in Europe, according to which member state pricing systems increasingly aim not merely at lower prices but at value for money, and it is this approach that frees up public money for other care. The three sources together say the same thing: data transparency and outcome-based payment are effective instruments, but only together with their methodological conditions. The detailed treatment of the literature — author by author, with quotations — can be found in section 6.4 Literature in detail.

📖 Source: Busse et al.: Improving Healthcare Quality in Europe; OECD: Health at a Glance — Europe 2024; European Observatory on Health Systems and Policies: Health Systems Governance in Europe

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures. The starting point of the position is that both announcements point in the right direction, and that publishing infection data is a defensible step that is rare even in international comparison. The proposals do not call the direction into question, but fix three points of implementation at which, according to international experience, most systems fail.

3.1 Compulsory, primary display of risk-adjusted values (within 90 days of the launch of the data interface)

On the public infection interface the risk-factor-adjusted indicator should be the value displayed primarily, with the raw figure beside it in a secondary place — not the other way round. The adjustment should extend at least to the patients’ age, the length of hospital stay, the use of invasive devices and comorbidities; in addition, the indicator of microbiological testing intensity should appear for every institution, because without it the hospital that looks less appears better. The interface should indicate the uncertainty range for every institutional value. This proposal does not slow publication: the methodology of the National Nosocomial Surveillance System already collects the variables needed for the adjustment, and according to the announcement indicators aiding interpretation will also be included — MIAK asks that these be not supplements but the main view. MIAK’s programme point E3 records this same requirement for waiting list data, precisely so that publicity does not push institutions towards the simple cases.

3.2 Open, machine-readable, time-series data disclosure and compulsory side-effect monitoring (simultaneously with the introduction of the outcome-based elements)

Alongside the display interface, the database should also be available in a freely downloadable, machine-readable format, with a documented data dictionary and unchanged identifiers, so that the data of earlier years can be linked. This is the condition for the profession, research institutes and civil analysts to be able to work with it too — not instead of the authority, but alongside it. In parallel with this, the introduction of the outcome-based financing elements should be accompanied by compulsory side-effect monitoring watching at least three phenomena: change in coding practice (the same care with a different severity classification), the shift of patient composition between institutions (the diversion of complex cases), and distortion of patient satisfaction measurement (selective sampling). The result of the monitoring should be public annually. MIAK’s programme point E2 prescribes uniform, standardised digital health data management, and A1 the general principle of machine-readable, queryable publication — the two meet here.

3.3 A public international reference price basket and publication of the aggregate price difference (before the next medicine price negotiation cycle)

For the medicine price negotiations the health insurance fund should publish the basket of reference prices: which countries’ prices it takes as a basis, with what weighting, and when it updates them. The difference between the gross and net price arising from reimbursement agreements — which is a business secret in individual contracts — should be published in aggregated form, broken down by active substance group, so that at least the order of magnitude is visible. The announced two-and-a-half-fold price difference is an item of billions even for a single product; if the phenomenon is systemic, it is the cheapest source available in the care system, because it requires neither a new tax nor a new institution. The direction of joint Visegrád negotiation is right, and the international literature supports it too: a larger number of patients gives a stronger bargaining position. The publicity of the reference basket strengthens rather than weakens this position, because it makes it predictable for the manufacturer too what we measure its offer against.

The three measures are linked by a single principle: measurement improves care if the measured figure cannot be manipulated more cheaply than the care itself. If a hospital can improve its infection indicator by testing less, then that is exactly what it will do. If it is easier to raise the patient satisfaction score by selective questioning than by real change, then that is what will happen. Risk adjustment, open data and side-effect monitoring together ensure that the cheapest path to improvement is actual improvement. The warning of Busse and co-authors (see 6.4.1) is about precisely this.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Patient safety The publicity of institutional infection data strengthens the internal argument for improvements; according to international estimates a significant part of hospital infections is preventable A raw, unadjusted list punishes the centres treating serious cases, and shakes trust precisely in the best institutions
Financing Outcome-linked medicine subsidy shifts the risk onto real effectiveness; the reference price basket may bring immediate savings The complexity of contractual reimbursements may create an administrative burden and legal disputes if the rule of measurement is not fixed in advance
Organisation of care Building patient satisfaction and the complication rate into financing gives a measurable quality incentive Patient selection and shifts in coding: the diversion of complex cases to institutions not covered by the measurement
Data policy Machine-readable, time-series publication also brings professional and civil analysis into play If the data are available only on a display interface, publicity remains formal, and the original aim — encouraging improvements — is not achieved

The main question to be weighed is the relationship between measurement and trust. The chief medical officer’s concern — that publication may reduce trust in hospitals — is not unfounded, and the international literature knows it too: public reporting usually worsens the perception of the institutions concerned in the short run, while the improvement in quality appears more slowly. It does not follow from this, however, that the data have to be withheld, but that everything depends on the quality of the presentation. If the reader sees on the first screen the adjusted value, the uncertainty range and the institutional profile together, then the data explain. If they see the raw percentage in a form that can be ranked, then it misleads. The proposal tips over to the risk side if the side-effect monitoring is left out: the measurement distortions of outcome-based financing build into the system not immediately but over two or three years, and by then they are hard to dismantle.

Part V — Measurability and summary

5.1 What is worth following? (proposed KPIs)

MIAK proposes the following performance indicators (KPIs — Key Performance Indicators). These are proposed indicators, not governmental undertakings.

  • Development of the risk-adjusted infection rate: at institutional level, in a five-year time series. The substantive aim is not the improvement of the national average, but the reduction of the spread between the worst and the best institution.
  • Microbiological testing intensity: how many tests there are per thousand care days by institution. If the infection rate falls but testing intensity does too, that is not improvement but a deficit of measurement.
  • Shift in patient composition: the annual change in the distribution of complex, high-risk cases between institutions. This is the earliest signal of patient selection.
  • Divergence from the reference price: for what percentage of subsidised products the Hungarian net price exceeds the average of the reference basket, and by how much. The two-and-a-half-fold divergence is today a single known case — the question is how many such there are.

5.2 Summary

MIAK’s request in a single sentence: the infection data by institution should appear in risk-adjusted form, as a downloadable database and together with an uncertainty range, and side-effect monitoring should be built in alongside outcome-based financing before its introduction — because bad measurement is not merely useless but actively harmful, and data disclosure that has lost its credibility cannot be restarted for years. Publicity of the medicine reference price is, by comparison, the easiest step: it requires no new institution, and may bring measurable savings even in the first year.

Of MIAK’s foundational values two are in play here. Data-drivenness is what warrants the whole package of proposals: public financing is just if the money goes where the patient’s condition measurably improves — but for this the measurement itself also has to be correct, otherwise data-drivenness merely accelerates the bad decisions. And transparency is the yardstick of implementation: an interactive interface from which data cannot be downloaded is information, not publicity. The two values point in the same direction here — publishing good data is at once a professional and a democratic requirement.

Part VI — Justifications and further sources

6.1 The framing of the press, spectrum by spectrum

The economic band gave the fullest picture. Portfolio processed the two announcements in two separate articles: in one it set out in order the institution’s need for resources and the elements of the change in the financing model, highlighting that the 500 billion of additional funds relates to the sector as a whole, not to the insurance fund’s development budget — this is the clarification missing from the other papers; in the other it described the technical construction of the infection database, from the five-year time series to downloadability. Economic framing typically examines feasibility, and here it is also the most accurate reading in substance.

The liberal-left and public affairs band chose the personal register. Telex published a full-length interview with the director general, and this is the most valuable source in the field: it is here that the passage appears on lobbying by MPs, on the role of consultancy firms and on the consequences of taking over the László Batthyány-Strattmann Foundation, which the other papers did not carry. HVG took over the same strand in shortened form, putting the medicine price difference in the headline. On the infection data disclosure, 24.hu placed the chief medical officer’s disclaimer at the head of its framing: a ranking is not the aim, and the data are not suitable for a patient to select the safer hospital before their operation. This sentence is the most important limiting statement in the field, and MIAK’s proposal 3.1 follows precisely from it.

The conservative band did not bring this topic into top focus in today’s issues; the framing comparison is therefore confined on this day to the economic and the liberal-left/public affairs bands. ATV carried a related strand: it reported on the extraordinary inquiry ordered at the National Institute of Oncology, concerning the suspicion of directing patients to private care — this case belongs here because it is the other side of the same question: what the public sees of the care system when there is no regular data disclosure.

6.2 Facts and data

Indicator Value Source
Health insurance budget some 5,000 billion forints interview with the director general of the health insurance fund, 2 September 2026
Announced additional sectoral resources 500 billion forints a year (for the sector as a whole) ibid.
Known medicine price divergence for a product representing an item of billions, 2.5 times the western European price ibid.
Annual number of hospital infections in the EU some 4.3 million cases in acute care hospitals OECD: Health at a Glance — Europe 2024
EU average point prevalence 7.1 per cent of patients acquired a hospital infection (2022–23) ibid.
Preventable share 35–70 per cent of hospital infections are avoidable with adequate infection control ibid. (on the basis of a WHO estimate)
Cost burden the cost attributable to hospital infections is as much as 6 per cent of the public hospital budget ibid.

Two statements are to be highlighted from the OECD data, because they bear directly on the interpretation of the present Hungarian publication. The first: measured prevalence is strongly influenced by the frequency of testing — countries that test more find more infections, so raw comparison partly measures the quality of detection, not patient safety. The second: when the data are adjusted for patients’ individual risk factors, the differences between countries narrow significantly. The same logic applies at institutional level, and it is precisely this adjustment that MIAK’s proposal 3.1 asks for.

One Hungarian datum deserves particular attention: according to the OECD volume, Hungary is in the leading group for the availability of alcohol-based hand sanitiser dispensers, but shows the lowest value for the actual use of hand sanitiser. This is the case where the infrastructure is there but the practice is not — and it is precisely the institution-level data disclosure now starting that can make this visible.

6.3 Policy dimensions

  • Healthcare (programme points) — the principle of real-time, risk-adjusted institutional data disclosure (programme point ID: E3), the uniform, standardised digital data management without which time-series linkage is not possible (programme point ID: E2), and the use of public health data for prevention (programme point ID: E4);
  • Transparency and anti-corruption policy (programme points) — the general requirement of machine-readable, queryable publication, which both the infection database and the medicine reference price can satisfy (programme point ID: A1);
  • Public administration and e-government (background material) — the operation of the data disclosure interface, the maintenance of the data dictionary and the institutional responsibility for the annual update;
  • Economy (background material) — medicine price negotiation as a direct budgetary item: reducing the divergence from the reference price frees up funds without drawing in new resources.

6.4 Literature in detail

6.4.1 Busse et al.: Improving Healthcare Quality in Europe

The volume systematises the quality-improving instruments, and treats public reporting and quality-linked financing in a separate chapter. The summary assessment is markedly cautious:

“The two strategies are probably the most controversial ones discussed in the book as there has been considerable debate about the potential unintended consequences of both strategies.”

In the chapter on the use of quality indicators the authors put it even more sharply: quality information is an instrument, and if used inappropriately it can cause serious harm — especially when strong incentives are attached to it, because in that case the manipulation of indicators and patient selection are also among the possible responses. On risk adjustment the volume describes a three-level classification proposal: the highest level handles every factor outside the clinician’s control that influences the outcome, whereas the lowest level disregards several important factors.

Translated to the Hungarian situation: the methodology of the data disclosure now starting provides an adequate basis, because the surveillance system works with uniform definitions and, according to the announcement, interpretive indicators will also appear. The risk lies not in the measurement but in the presentation and in the incentives. As soon as the complication rate has a financing consequence — and according to the announced model it will — every one of the unintended effects described in the volume is activated. This is why MIAK proposes that side-effect monitoring should be not part of the evaluation but a condition of introduction.

📖 Source: Busse et al.: Improving Healthcare Quality in Europe

6.4.2 OECD: Health at a Glance — Europe 2024

The OECD volume publishes the results of the European point prevalence survey in its chapter on hospital infections, and the methodological note is the most important part from the point of view of the Hungarian publication:

“The predicted HAI prevalence adjusts for individual patient risk factors (patient age, length of hospital stay, use of invasive medical devices and patient comorbidities) which increase the risk of HAIs.”

The data themselves are telling too: the EU average was 7.1 per cent of patients in 2022–23, with the lowest values in Latvia, Romania and Bulgaria (below 4 per cent) and the highest in Cyprus and Greece (above 12 per cent). The volume records at the same time that the differences partly reflect the frequency of testing, and that the spread of the adjusted values is substantially smaller. Two practical conclusions follow from this for the Hungarian interface. One: the list of variables to be adjusted for is not a theoretical question but internationally fixed — age, length of stay, use of invasive devices, comorbidities — and can therefore be demanded of the Hungarian data interface as well. Two: the testing intensity indicator has to be published alongside the infection rate, because without it the hospital that tests most thoroughly appears the worst.

📖 Source: OECD: Health at a Glance — Europe 2024

6.4.3 European Observatory on Health Systems and Policies: Health Systems Governance in Europe

The volume’s pharmaceutical market chapter examines how member states’ room for manoeuvre in price and reimbursement policy has been transformed within the EU regulatory framework. The essence of the finding is that the aim of price regulation has gradually shifted:

“increasingly, national price and profit control regimes aim not only to deliver lower prices for patients, but also value for money.”

The difference is not semantic. Purely price-cutting logic looks for the cheapest product; value-proportionate logic asks how much health gain we get for a forint. The present Hungarian announcement — outcome-linked medicine subsidy, where the manufacturer pays back if the product does not deliver in real care — belongs precisely to this second family, and is a defensible direction within the volume’s framework. At the same time the chapter records that the demand side, that is, price and profit regulation, has remained a member state competence, and EU harmonisation in this field is not to be expected within the foreseeable future. From this follows point 3.3 of MIAK’s proposal: if pricing is a national competence, then using the room for manoeuvre is also a national responsibility — the publicity of the reference basket and joint Visegrád negotiation are both instruments of this.

📖 Source: European Observatory on Health Systems and Policies: Health Systems Governance in Europe

6.5 International comparison

There are two instructively differing models for institution-level quality data disclosure. In the United Kingdom the publication of cardiac surgery results began after the Bristol paediatric cardiac surgery affair of the late 1990s, and the effect was twofold: mortality indicators improved, but risk avoidance was also documented — some surgeons avoided the gravest cases in order to protect their own statistics. The answer was given not by withdrawing publication but by refining risk adjustment and by publishing interventions with very small case numbers in aggregated form. In the United States the hospital comparison interface of the federal health insurance programme has from the outset worked with risk-adjusted indicators and an uncertainty range, and does not place institutions in a ranking but in three categories (better than average, in the average range, worse than average) — this solution technically preserves the information while making tabloid ranking harder.

The Scandinavian quality registries show a third path: the data are public at institutional level, but the system’s most important function is professional feedback, so the hospital sees its own data continuously, in comparable form, before they reach anyone else. The present form of the Hungarian publication — five-year time series, downloadable database, annual update — stands closest to this family, and MIAK’s proposal would strengthen this model too: publicity should be an instrument not of shaming but of professional comparison.

Healthcare

  • E2 — Digital healthcare system
  • E3 — Transparency of waiting lists
  • E4 — Prevention data programme

Transparency and anti-corruption policy

  • A1 — Public money dashboard

Proposed new programme point: An international medicine reference price basket — for the Healthcare area: public fixing of the reference countries, weighting and updating regime forming the basis of subsidy price negotiations, and publication of the aggregate gross–net price difference at the level of active substance groups.

6.7 List of sources

Press sources (MIAK press monitor, 3 September 2026 — topic 3):

Knowledge base references (specialist books):

  • 📖 Busse et al.: Improving Healthcare Quality in Europe
  • 📖 OECD: Health at a Glance — Europe 2024
  • 📖 European Observatory on Health Systems and Policies: Health Systems Governance in Europe

Note: the local file path of the books does not appear in the visible text of the blog — only the author and the title. The file path is an internal matter of the generation process, not the reader’s.

MIAK internal materials:

  • MIAK policy area: Healthcare (programme points; programme point ID: E2, E3, E4)
  • MIAK policy area: Transparency and anti-corruption policy (programme points; programme point ID: A1)
  • MIAK policy area: Public administration and e-government (background material)
  • MIAK policy area: Economy (background material)
  • MIAK press monitor, 3 September 2026 — topic 3, score: 88/100

Supplementary public data sources:

  • ECDC — European point prevalence survey of healthcare-associated infections and antimicrobial use (2022–2023)
  • OECD — Health Care Quality Indicators (HCQO) dataset

Generation metadata