Cross-sector care you can touch

How will CGM present itself at DMEA 2026?

E. Oesterhoff: At DMEA 2025, all CGM companies were united on one stand for the first time. This year, we will show how these systems support the healthcare landscape in an interoperable system. In doing so, we are emphasising that outpatient practices, hospitals and rehabilitation clinics are not separate worlds for us, but part of a digitally networked ecosystem. Health policy is increasingly focusing on ‘digital before outpatient before inpatient’. This can only be implemented if the different IT systems function smoothly and interoperably. This is what CGM stands for. CGM also stands for the greatest possible involvement of patients. Electronic patient records and patient portals hold the digital ecosystem together like glue and enable data and documents to accompany patients seamlessly through the facilities. Our stand in Berlin also follows this philosophy: we present all systems on an equal footing. And at the centre are our two experts, Nina Bougatf and Jana Schmidt, who use a patient journey to show what interoperability and artificial intelligence (AI) can achieve in healthcare when they are thought of in a comprehensive way. DMEA stands usually only show excerpts of healthcare. At CGM, you get the complete picture.

J. Schmidt: This comprehensive approach is also reflected in how CGM is organised with regard to AI and interoperability. As far as the AI part is concerned, for which I am responsible, there is no isolated AI team for each product. Instead, we have an international and cross-sector AI team that works closely with users to develop AI use cases, which are then incorporated into various products. We want to showcase this centralised AI development at DMEA – and then, of course, provide concrete examples of the benefits/added value this brings to individual products.

Which AI-supported care processes will be on display at the CGM stand?

J. Schmidt: One example of a product that is already market-ready is the documentation assistant that we offer for physician information systems (PIS) and hospital information systems (HIS). This assistant listens in on doctor-patient consultations and makes specific suggestions for documentation – not a copy-paste version, but one that is deeply integrated into the HIS or PIS. We also showcase AI assistants that provide an overview of patients and those that support doctors with discharge letters. These assistants are currently in the pilot phase. We would be delighted to find interested institutions at DMEA that would like to work with us on the pilot in the coming quarters.

But that sounds like individual AI assistants. What is the central, connecting element of this approach?

J. Schmidt: We are not developing these assistants as standalone products. If we stick with hospitals, which are the focus of DMEA, what we will be building in the medium term is a new application called HealthAssistant. In future, users will be able to contact this HealthAssistant and hand over certain tasks to virtual agents. These agents will then carry out the delegated tasks automatically. This could be the creation of a medical report, a discharge letter, coding or simply searching for individual file entries, findings or diagnoses. We will have the HealthAssistant with us at DMEA as a prototype. And we look forward to receiving as much feedback as possible so that we can finalise this solution quickly.

E. Oesterhoff: A platform like this, implemented properly, is very convenient for users. But it also has an operational dimension: hospital IT managers and CIOs are currently being inundated with AI solutions. The end result is dozens of individual solutions, each of which works great on its own, but which are extremely difficult to administer when taken together. A platform approach is extremely helpful here. As a hospital IT department, I basically only have to roll out the platform, and the AI services made available via the platform can then be selected as needed. Incidentally, these do not always have to be LLM-based tools. Good algorithms retain their importance. LLM AI is not suitable for everything, and it also has a significant ecological footprint. We also do not believe that LLMs will solve all interoperability problems. Semantically clean data storage is essential – for AI tools as well as for cross-sector care scenarios.

DMEA is always a forum for digital policy. CGM is one of the major German healthcare IT players. What discussions do you expect this year?

E. Oesterhoff: There is a lot to discuss. We now have an updated digital strategy, which is very good. In the outpatient sector, appointment portals and patient management will be a major topic in the coming months and years. Germany is establishing a primary care system, and this will not be possible without digitally supported patient management. We do not believe that the entire issue of patient management and scheduling can be handled via the telematics infrastructure (TI). We believe that patient portals such as our m.Doc portal can meaningfully expand the telematics infrastructure and the electronic patient record (ePA). Nevertheless, we should not downplay the TI, but rather handle as much as possible and as sensibly as possible via the TI. You can also see this for yourself at our DMEA stand. CGM is one of the most important providers of TI infrastructure components.