To what extent has technical progress already been made, enabling the digitalization of healthcare services and treatment in every hospital?
Technically speaking, digitalization in the healthcare sector could be much more advanced. Many medical care processes are either still handled on paper or are carried out in data silos within institutions in areas that do not communicate with one another. From a technical point of view, they could all be easily interlinked. There are no real issues associated with digitalization not being technically feasible or solvable. However, the description of a particular disease entity, for example, at a district hospital in northern Germany must be analogous to that of a university hospital in another part of Germany. This ensures that patient data is comparable and exchangeable between institutions. And there are indeed huge challenges associated with the interoperability of data. These are particularly demanding in the healthcare system.
Are there also barriers in people’s minds?
Yes, these do exist. Naturally, we also have to use persuasive efforts to make people understand that the extra effort is worthwhile when it comes to digitally recording data and also entering descriptions of diagnostic findings, medical reports, etc. Standardizing these processes and mapping them in IT systems initially leads to extra work for the doctor in the daily routine. We can only be successful here if we can prove that this additional effort leads to therapeutic success for patients and also improvements in the doctor’s daily clinical practice.
How do you envision tackling this problem?
I believe we need to use smaller, specific projects to showcase how we can improve patient care while at the same time assisting doctors in making their work even more streamlined, efficient, and better. If we can demonstrate this in selected projects or for disease entities in selected clinical processes, then I believe we will be able, bit by bit, to successfully implement and enforce these digitalization processes in other areas as well.
Is it true that you already have a concrete project in the cancer field with some university hospitals that plan to work with you?
Eight university hospitals have joined what is known as the HiGHmed consortium, committing themselves to making their data interoperable from a technical standpoint, so that it can be exchanged across institutional boundaries. This consortium includes the Charité and university hospitals in Heidelberg, Göttingen, and Hanover, just to name a few. Based on three clinical case studies, we would like to systematically demonstrate that this data exchange leads to added value in medical care and research. One of our case studies is in oncology, where we are planning to build virtual molecular tumor boards across institutions in order to standardize and improve the treatment of cancer patients. As for other case studies, such as an early warning system for infections, we hope to make more accurate, earlier prognoses. This will enable us to monitor outbreaks of multi-resistant pathogens, for example, either at one location or across locations.
In Germany, the federal government has initiated a “decade against cancer.” To what extent can digitalizationhelp conquer cancer?
I think cancer is a perfect example of an area in which digitalization is indispensable. Why? Huge amounts of data are being generated, particularly in cancer research. One example here is genome sequencing, which generates enormous amounts of data on individual patients. This data needs to be linked to clinical data from a treatment context. And naturally, we do not merely have one patient whose data we review in isolation. We have hundreds, thousands, and even tens of thousands of patients with similar tumors at single institutions, nationwide, across Europe, and even worldwide. This data is only rudimentarily interlinked, let alone analyzed in an integrated manner. There are incredible challenges in this area, but also opportunities to use machine learning methods and technologies, i.e., artificial intelligence. Here, we are trying to learn how to better characterize tumors in a fully automated way, in order to better predict how we can efficiently treat these tumors.