Artificial intelligence will not, by itself, solve the problems of Slovak healthcare; without quality data and infrastructure, however, we cannot use it meaningfully. A panel of experts agreed that an AI‑ready hospital is not a label, but the result of thoughtful digitization, standards, and work with people. The discussion brought examples, dead ends, and a realistic vision of what to do right away and what definitely not to do.
Data, standards, and interoperability
The panelists agreed that the core issue is data quality. Today, healthcare often records 'essays' that are hard to process and compare across departments and hospitals. This is exactly where interoperability and international standards such as HL7 or clinical vocabularies like SNOMED and LOINC help, so that information has an understandable structure. It is also important that data be collected at the source — many already exist in the devices and do not need to be transcribed.
In parallel, practical solutions are emerging: automated generation of discharge summaries from hospitalization data with physician oversight, intelligent DRG support, reading images from CT/MR and pathology, and digitized medication with reduced dosing errors. The question of interpreting results remains open — data analysts need to understand clinical context, otherwise incorrect conclusions are a risk. There was also a call for open data: until now, political will has been lacking, but NCZI is to start regularly publishing datasets so that the professional community can use them as well.
People, processes, and cultural transformation
The biggest obstacle is not the technology, but the change in mindset. Resistance to change is natural, especially in teams with a higher average age; therefore it is necessary to work gradually with concerns, provide tutorials, internal communication, and emphasize the benefits in day-to-day work. Forms should not be endless — the system should require only the necessary inputs and fill in the rest from source data or automation. The goal is for digitization to make work easier and not increase the administrative burden.
When processes and systems are well set up, time can be saved at scale. According to the HIMSS EMRAM reference model, digitization at a level close to Stage 6 can reduce administrative work by roughly 30 percent — but this is not about layoffs; it is about more time for the patient and higher safety. The state is simultaneously investing in equipment, hospital information systems, cybersecurity, and central data processing; what the patient does not see forms the basis for everything 'visible.' The result of an AI‑ready approach is a hospital where technology and people pull together in the same direction, and data turns into better decisions, not additional obligations.
