Saskia Peerdeman: 'AI has become our co-pilot'

Healthcare will improve through the use of AI. Of that, Saskia Peerdeman is convinced. "AI has become our co-pilot," the director of the Amsterdam UMC noted optimistically during Zorg & ict. In her keynote, she did issue a slight profit warning: "Artificial intelligence is still developing, we are far from there yet."

Saskia Peerdeman, who in addition to being a member of the board of directors of Amsterdam UMC is also a professor of neurosurgery and dean at the Faculty of Medicine VU, became fascinated by AI when she learned about its applications in her own field, neurosurgery. The board of directors of Amsterdam UMC has now named data-driven work and the application of AI in the hospital as one of its spearheads for the coming years. The ultimate goal? "We want to move toward a system in which we can provide better care through AI."

Data maturity

The road there still consists of cobblestones, but the beginning is there, Peerdeman observes. "When we switched from handwritten and therefore often unreadable files to Epic in 2010, it yielded an enormous increase in data, with unimaginable possibilities. The consistency was still unclear, though. Since then, we have made great strides within Amsterdam UMC, on our way to data maturity. But, we are far from there yet."

Personalized care

Care, research, education and valorization are the four core tasks of Amsterdam UMC. With all the data collected during the care process of patients, predictions about their health are becoming increasingly accurate. As a result, decisions about treatment and care can be better substantiated. Peerdeman: "This improves the quality of care and allows us to offer more personalized care."

As examples of AI in healthcare, the director mentions the "consulting room dashboard," in which all patient data is immediately accessible. In addition, there are the dashboards for better care pathways, such as the SlicerDicer. With this tool, patient data can be selected and viewed, not only at the population level, but also, for example, for complications.

Hybrid care

Amsterdam UMC is also committed to implementing hybrid care in the near future, the director said. Upcoming projects include a digital (outpatient) clinic, which includes digitizing the discharge process, and Amsterdam UMC Thuis, where patients monitor themselves and take simple measurements, such as blood pressure. In addition, virtual nursing is on the rise. This involves the use of smart mattresses, smart plasters and continuous monitoring in normal care.

Simulation education

In addition, research and education have taken off in terms of data-driven work, far follows Peerdeman. "Amsterdam UMC wants to move toward 'learning health care,' in which we use data to improve care, and share data for care and research." Within education, data-driven learning is already being used adrift, she says from her position as dean. "Students are learning about, from and with data. For example, we have the Edu-Cow for online bedside teaching, an E-health learning and living lab, and we are betting on DUTCH, the healthcare education of the future, where students receive education in a simulated environment.

Legal bottlenecks

The academic hospital is sharing data with the regional healthcare organizations OLVG and Antoni van Leeuwenhoek, among others, in the so-called Health Data Space Amsterdam, a regional healthcare data infrastructure in which hospitals and other healthcare providers in the Noord-Holland/Flevoland region will reuse each other's medical data pseudonymized. "Such a regional plan may sound simple, but there are many bottlenecks and questions, such as legal ones," says Peerdeman. "Because who does all that data actually belong to? It takes time to figure all that out properly."

In addition, Amsterdam UMC is one of the regional carters of Health-RI, a national project funded by the National Growth Fund. The teaching hospital is also represented in consultations engaged in the development of CumuluZ, a digital platform that aims to make all health data system-independent available to healthcare providers and patients.

Ambitious

Peerdeman is not blind to the challenges ahead, she says, such as about data quality. OLVG, for example, uses different codes for certain conditions than Amsterdam UMC. She also sees the sensitivities in areas such as privacy. "Do we also share the data with the social domain and who gives permission for that? " Still, she is hopeful. "It sounds ambitious, but by 2028 we want to have already implemented a lot."

Change challenges

Despite the great benefits of all the AI-driven capabilities, Peerdeman warns of the impact these changes can have on nurses. "We come up with all kinds of nice things, but on the shop floor you're not doing that, there it's just 'hands on the bed.' So we definitely have to pay attention to that change side."

Nor should the fact that there are sometimes four generations working in an organization who are far from all having the same digital skills be overlooked, the executive says. "Let's embrace those differences," she advocates, "from policy makers to GenZ."

Bumps and opportunities

When asked if AI will ultimately make healthcare more affordable, Peerdeman does not dare say a resounding yes. "Macroeconomically it will make a difference, but overall, I'm not sure. People will always be needed to check things, although administratively it does save time. But I think it's incredibly cool to go forward with this. Yes, I see bumps, but also, on the contrary, lots of opportunities!"