Karl Lauterbach (SPD), Federal Minister of Health /photo alliance, Michael Kappeler
Berlin – Federal Health Minister Karl Lauterbach (SPD) hopes that Germany will become a pioneer in the development of artificial intelligence (AI) in medicine thanks to its emerging data infrastructure in the healthcare system. He explained this today in Berlin.
The Digital Law (DigiG) and the Law on the Use of Health Data (GDNG) would create enormous opportunities, especially for the development and further development of AI. “With what we are doing here, we probably have the largest data room in the world, in which this data is also interconnected,” said Lauterbach. “So I think this will take us far.”
In the future, health data from the electronic patient record (ePR) will be sent pseudonymously to the facility’s research data center (FDZ) at the Federal Institute of Medicines and Medical Devices (BfArM) to which any natural person in the EU can gain access for research purposes upon request.
This will open up enormous possibilities for training AI algorithms that rely on the largest possible quantities of high-quality data sets. “We want to establish Germany as a global leader in research,” explained Lauterbach.
As examples, he cited prediction models for certain diseases that are similar to weather forecasts: based on a large number of parameters, AI algorithms could calculate probabilities about whether and when a disease will emerge in a patient, without understanding the context. scientific in detail.
We know that atrial fibrillation increases the likelihood of stroke, especially with a certain constellation of frequency and amplitude of fibrillation, because this promotes the formation of blood clots.
In the case of Alzheimer’s disease, we know that the interaction of a whole range of factors – including certain genetic alleles, but also fat and insulin metabolism – can be an indicator of a patient’s likelihood of developing the disease.
Just as in meteorology weather forecasts at a given time and place can be made based on a large number of factors, from air pressure to wind speed and humidity, this will also be the case with diseases in the future. “All of this will be possible if we homogenize and merge the data.”
Furthermore, the new amounts of data will provide completely new opportunities for research. For example, there is still insufficient knowledge about the impact of long-term use of HIV medications on the development of neurodegenerative diseases. There is a significant benefit to research here.
AI will also bring big changes to care itself, he emphasized again. “This new medicine will not bypass any practicing physician, that is clear,” he said. In the future, there will be “digital doctor companions” who will be able to give patients assessments and explanations about their diagnoses, which in turn would ease the burden on human doctors.
The current version of ChatGPT can already explain medical connections. The AI could reach the level of experts in the near future. “This will lead to a completely new understanding of patients about their diseases,” he emphasized.
Furthermore, the emergence of AI in understanding complex relationships is already evident. Language models would soon be able to summarize content not based on probabilities, but rather to develop a deeper understanding of a patient’s health status based on data from an EHR, for example.
All of this will lead to more transparency in supply. The human factor must be maintained if, for example, patients value the advice of a human doctor in the case of serious diagnoses.
In the development of AI in outpatient care, he hopes for good cooperation with the National Association of Statutory Health Insurance Doctors (KBV), especially with the responsible counselor Sibylle Steiner.
The “ePA for all”, thus BMG The ePA with objection solution will be introduced from January 15, 2025. The ministry expects almost all those with legal health insurance to receive their files within the first four weeks, explained Sebastian Zilch, head of the sub-department “Gematics, Telematics Infrastructure, E-Saúde” from BMG.
Doctors are then required to upload documents and other data to ePA, but in most cases this must be done automatically. There will be no standardized check to see if they are compliant. “The requirement to comply with the law also naturally applies to doctors,” said Zilch.
However, there may be cases of sanctions, for example, if a doctor explicitly refuses to place content on the ePA. The patient can then contact the responsible Association of Statutory Health Insurance Doctors (KV) with a complaint.
Following the start of the ePA with objection resolution in January, the first expansion of the digital medication process is scheduled for mid-year. BMG itself can define additional structured use cases. Zilch mentioned laboratory results and an electronic patient record as possible next steps; © lau/aerzteblatt.de