On Thursday evening the database of hospital-acquired infections became public and searchable on the website of the National Centre for Public Health and Pharmacy (NCPHP). A government decree issued in May obliged the NCPHP to publish detailed data on healthcare-associated infections. In the three months since then, it has developed a methodology for analysing the infection data, created a database broken down by hospital, and built an interactive platform on which anyone can look up the data that interests them.

In the first round, several large datasets have become searchable: the database of outbreaks that have occurred in hospitals; the detailed data of a comprehensive study carried out in 2022; and data on Clostridium difficile, an infection that spreads mainly in hospitals and causes severe diarrhoea. Databases on the so-called multidrug-resistant pathogens, which are resistant to strong antibiotics, and on bloodstream infections will be published later. The state has been collecting these data for more than twenty years, but until now only a very small part of them was accessible, and even that only in a form that was very hard to understand.

We will report in a separate article shortly on the more important details that can be read out of the newly published data.

It has been a long road to the point where the state gave up its secrecy and publicly acknowledged the scale of the problem at all: under the Orbán governments a great deal of energy was invested in sweeping the issue under the carpet and keeping the data secret. Several civil society organisations, experts in favour of transparency and the independent press played a major role in the change, including Direkt36’s 2023 series of articles and documentary film. At the time we were able to obtain detailed data only through lengthy litigation, and we analysed and published it in the face of resistance from the NCPHP. The articles in our series called Project Semmelweis can be viewed here.

According to estimates, tens of thousands of infections occur in Hungarian hospitals every year, and they can cause suffering and serious complications for patients. It is dramatic that they catch these infections in the very place where they went to be cured: in hospital.

Join the Direkt36 supporters’ group and get insight into investigative journalism! Details →

The number of cases will never be zero, but some of them are certainly preventable. A whole range of factors influences the spread of hospital-acquired infections: for example the cleanliness of hospital rooms and surgical instruments, the use of hand sanitiser, and also how quickly healthcare workers detect an infection in the hospital and whether they are then able to isolate the patient properly. Multi-bed wards, run-down bathroom facilities, the shortage of healthcare workers and even poor air conditioning all help the spread. Invasive procedures and long hospital stays increase the chance of infection, as does the condition of the people admitted: many of them are elderly, weakened and struggling with a range of underlying conditions, and are therefore also more susceptible to infections.

In other words, most Hungarian hospitals are fighting the spread of infections in conditions that are far from ideal. Earlier national statistics also show this: in the years before the coronavirus pandemic the figures rose relentlessly, during the pandemic years the situation became even worse, and afterwards the infection numbers stabilised at a high level. No significant improvement has been achieved in recent decades.

As chief medical officer Beatrix Oroszi said at a press conference on Wednesday, they are preparing a paradigm shift in this area. The NCPHP, which has come under the control of the new Ministry of Health led by Zsolt Hegedűs, has worked hard in recent months to make the data publishable with the appropriate explanations attached. According to Oroszi, in addition to the minister the National Hospital Directorate General (OKFŐ) is also committed to reducing hospital-acquired infections. “This is a good cause, everyone needs to get behind it,” she declared.

On Tuesday the NCPHP also organised a briefing for hospital directors, who, according to Oroszi, received the change optimistically. One aim of the new system is precisely that the institutions themselves should get to know each other’s data, be able to assess their own situation and, with the help of this, develop their infection control work further. The health government would like hospitals to be able to achieve visible improvement within a few years by learning from each other’s good practices.

The NCPHP’s experts are noticeably afraid that people without the relevant expertise will misinterpret the data. At the press conferences held in recent weeks they said dozens of times that the performance of individual hospitals cannot be compared or ranked purely on the basis of the data. They tried to explain how the statistical model used for the database works and the limits of interpreting the data. They also encouraged hospital managers to communicate honestly in the press about their own situation. They do not want patients’ trust in healthcare, or even in particular hospitals, to decline as a result.

What is spreading in hospitals?

Pathogens spreading in hospitals are typically ones that are not killed even by concentrated disinfectants and that easily resist strong antibiotics. They are known as multidrug-resistant bacteria, and by infecting elderly or immunocompromised people lying in hospital beds who are already struggling with illness, they can cause serious symptoms.

There is a wide range of them, from multidrug-resistant Acinetobacter baumannii (MACI), which can cause pneumonia, wound infections and urinary tract infections, to the bacterium MRSA, which can cause abscesses and wounds to fester and heal with difficulty. The latter is one of those pathogens which, multiplying in the body, can enter the bloodstream and cause even fatal blood poisoning (sepsis). Common pathogens also include Clostridium difficile, a spore-forming bacterium that causes foul-smelling, mucousy diarrhoea and even severe inflammation of the intestines, and against which alcohol-based hand sanitiser is not effective either.

Pathogens can be spread by droplet transmission, faeces, wound drainage, through medical equipment and, to a not insignificant extent, by the insufficiently disinfected hands of a doctor, a nurse or a visitor. Without proper hygiene, pathogens can linger on medical equipment, hospital toilets, door handles, shower curtains, bed linen, nurse call buttons – almost anywhere. Any invasive procedure, such as surgery, as well as a tube, catheter or cannula sticking out of a patient, poses a risk of infection. The longer a person spends in hospital, the more likely they are to become infected. And even to die as a result.

This is a problem not only in Hungary but in hospitals all over the world. More and more superbugs are appearing and spreading. The difference lies in what countries’ healthcare systems do about the problem: whether they tackle it and keep it under control, or resign themselves to the situation and sweep it under the carpet.

Not only the billions spent on treatment are at stake, but lives too: according to a globally accepted estimate, if hospitals do it right, 20-30 percent of infections – and in certain cases as much as 50 percent – can be prevented.

It was taboo, they were afraid, they kept it secret

Many years of a strongly anti-transparency attitude have now been replaced by more open communication.

The NCPHP has been collecting data on hospital-acquired infections since 2004, as hospitals are required to report every single case they detect. For decades, however, the NCPHP was content to produce for the public only an annual national summary of the situation, which was not only difficult to interpret but even hard to find on the authority’s website.

In the mid-2010s a civil society organisation called the Hungarian Civil Liberties Union (HCLU) began fighting for detailed data broken down by hospital to be published. The system resisted; health leaders considered it an outright bad idea for everyone to be able to see the detailed data, fearing that people would be frightened if they found out how many infections occurred where.

They were reluctant even to admit that the problem existed. “The confidence in Hungarian healthcare workers is undermined by rumours that our doctors and nurses do not wash their hands before procedures. Patients are safe in Hungarian hospitals, and we will defend the honour of Hungarian professionals against such unfounded accusations at every occasion,” the Ministry of Human Capacities wrote in a 2016 statement.

The HCLU launched a lawsuit for the data and won it, but it could do nothing with the database it received, because the spreadsheet with its tens of thousands of rows was useless on its own. This is where Direkt36’s journalists entered the story. With the help of the data analysis and research team of the international journalism organisation OCCRP (Organized Crime and Corruption Reporting Project), we managed to turn the data into a manageable database.

In parallel with this, at the end of 2022 we also tried to request more recent data from the NCPHP, expecting it to be a mere formality since a court had already ruled that these data had to be released. The NCPHP refused, however, and we had to go to court again, with the help of the HCLU’s lawyers. We won the case, but the NCPHP did not let the matter rest there either and appealed against the ruling. “Disclosing detailed, institution-level data can be dangerous and can mislead the public and patients,” they argued. In the autumn of 2023 the NCPHP finally lost the case once again, and so it was obliged to make its databases public.

These still had to be analysed, however, which ran into a series of obstacles. Direkt36 therefore asked for the help of Tamás Ferenci, one of the country’s most renowned biostatisticians, who, after lengthy work, developed a unique statistical model that sought to filter out the distortions caused by differences between patients and between individual hospitals, and thus made it possible to assess and compare the performance of individual hospitals objectively.

Ferenci helped because, in his view, “the detailed analysis of data on hospital-acquired infections helps to identify opportunities for improvement and motivates better performance”. “It is true that such rankings can also cause trouble if the general public does not interpret the figures correctly, but the answer to that is not to refrain from producing them, but to produce them well: for example to publish results with appropriate explanations, put in context and with the necessary statistical corrections,” he said.

For his calculations Ferenci took into account a range of important factors in addition to the infection numbers, and used them to adjust the data. For example the differing patient volumes of hospitals, and even the differences arising from hospitals’ various departments, as well as – by incorporating a public indicator called the case-mix index – the differences between the patients admitted to hospitals. Ultimately, therefore, anyone browsing the hospital rankings published in Direkt36’s articles was getting a picture of hospitals’ performance. The results can be seen in this article, where a detailed video and written explanation of Ferenci’s method can also be found.

We also explained the situation behind the ranking: we conducted more than thirty interviews and background interviews and reviewed hundreds of pages of technical material. We also presented the situation of individual hospitals close up and looked into the practices of other countries. The investigation and the data analysis took more than a year, and in addition to a 2023 series of articles we also made a documentary film about it, titled Washing Our Hands.

Zsolt Hegedűs, an orthopaedic specialist who is currently the health minister of the Tisza government, was quoted in the articles and appeared in the film (at the time the Tisza Party did not yet exist in its present form, and Péter Magyar had not yet appeared on the scene). Among other things he stressed that “in the homeland of Ignác Semmelweis we are still not doing enough to protect our patients from infections with a simple hand wash”.

He emphasised that human lives depend on how seriously healthcare takes infection prevention. And treating infected patients costs hundreds of millions of forints, which could be saved. “And this is swept under the carpet in Hungary. It is a scandal and a disgrace,” he said at the time.

Join the Direkt36 supporters’ group and get insight into investigative journalism! Details →

After the articles and the film were published, a lively public and professional debate began about hospital-acquired infections and about whether the data can be analysed. Our series of articles, which was given the name Semmelweis Project, won the Transparency-Soma Award in 2024, the most prestigious Hungarian prize for investigative journalism, and was also shortlisted for the European Press Prize.

We tried to get every hospital concerned, as well as the NCPHP, to respond to the results of our work and the questions that arose. Most of the hospitals did not respond at all at the time, and the NCPHP – sticking to its earlier position – wrote that since hospitals and their patients are different, in its view “reports on healthcare-associated infections are unsuitable for comparison, and arranging institutions in any order is meaningless”.

A complete turnaround

The NCPHP’s attitude changed after the change of government and Hegedűs’s appointment as minister. The government decree of May 2026 already mentioned obliged the institution to arrange for the data to be made public in a way that makes them easy to understand. This was one of the first measures of the newly formed ministry.

After that, work began at breakneck speed at the NCPHP to be able to publish the data of recent years (up to 2025). The analysis of the data was carried out by György Surján, head of the NCPHP’s health monitoring department, who also involved Tamás Ferenci in the work; at one of the press conferences Ferenci remarked ironically that he had moved over from the “rebel side” to the other one. It was the two of them who developed the so-called risk-adjustment model, which is designed for Hungarian conditions and seeks to correct a range of factors that distort data analysis.

For example the fact that hospitals treat very different kinds of patients, who – because of their differing conditions – are more or less susceptible to infections. In other words, the differences between patients have to be filtered out for a valid data analysis to be produced. For this, Surján and Ferenci used a non-public database, the data collected on the so-called homogeneous disease groups (HBCs), Hungary’s diagnosis-related group system, which sorts hospital patients into some 750 different groups. With their help they created the indicator that is able to take into account the patient mix of hospitals. The institutions were then sorted into three groups according to – to put it simply – how serious the cases they treat are and how complex the care they provide is. For there is a big difference between a small-town hospital operating with a few smaller departments and, say, a large university hospital performing a series of complicated procedures. (Ferenci has also produced a detailed explanation of how their model is constructed and what pitfalls need to be avoided when interpreting the figures. This has also been published on the page presenting the data, and can be read here.)

Although this statistical model is based on much finer and more detailed calculations, it closely resembles the one that Ferenci developed years ago while helping Direkt36 – and which the NCPHP called misleading at the time. The institution’s main objection – as we quoted above – was that in its view hospitals cannot and must not be ranked against one another, because that is counterproductive and causes harm. This was also stressed several times at the press conferences held over the past month, so on this point their position has not changed.

The results of Ferenci and Surján’s new model can also be seen on the interactive platform, hospital by hospital. It is easy to read from the indicator how individual institutions stand compared with the average – and with their own earlier situation. The aim is essentially to filter out anomalies: where the data deviate greatly from the average, where they are outliers, it is worth hospitals looking at what lies behind them, and if they identify some kind of error, they can then correct it. At the same time, striking deviations can also mean that the institution concerned is doing something very well – for example that it is able to detect far more cases than the average, or that it prevents infections more effectively.

According to the chief medical officer, with the published data “we do not want to create fear, there is nothing forward-looking in that”. Instead she would like 20-30 percent of infections to be prevented even in the short term, as she considers this a realistic goal. In the first year or two, however, the number of infections in hospitals can be expected to rise further – paradoxically, the NCPHP’s experts would even welcome this, because it would mean that institutions have begun to detect and report infections more effectively.

Join the Direkt36 supporters’ group and get insight into investigative journalism! Details →

An important question, after all, is how reliable the data have been so far. It is possible that some hospitals did not report to the NCPHP all the infections they found, while others may not even have found some of them, because they carried out too few laboratory tests. The new database therefore also contains figures on this: from the rate of stool samples taken and blood tests performed it is possible to infer how much a given hospital tests for infections. It may be that some hospitals perform worse on paper than others, when in fact these institutions may simply be detecting infections more effectively – in other words, ultimately doing their job better.

Beatrix Oroszi also said on Wednesday that what is now being published is unique even by global standards – whether one looks at the abundance of data, the user-friendly interactive platform or the calculation methods.

The database is available on the NCPHP’s website. We will report on the details in a separate article later.

Illustration by Péter Somogyi (szarvas) / Telex