Two women sitting in front of a computer, doing research in a medical laboratory

Operationalizing responsible AI in health care

SAS helps Erasmus MC ICU improve patient care.

Real-time health and insights for clinicians

Erasmus MC achieved this using • SAS^® Viya^® on Microsoft Azure • SAS^® Visual Text Analytics • SAS^® Visual Machine Learning • SAS^® Intelligent Decisioning • in partnership with Notilyze

As the largest university medical center in the Netherlands, Erasmus University Medical Center (Erasmus MC) is a cornerstone of Dutch health care. Located in the heart of Rotterdam, the hospital is internationally recognized for its cutting-edge research, top-tier education, and specialized care for the most critically ill patients. With more than 14,000 employees and a mission to improve individual patient care and public health, Erasmus MC is constantly seeking ways to innovate responsibly.

But like many health care institutions around the world, Erasmus MC faces mounting pressure. “There’s nothing more rewarding than saving a critically ill patient’s life and witnessing their return to society,” says Dr. Michel van Genderen, Intensive Care Unit (ICU) and Internal Medicine Physician and associate professor. “But rising demand and a shortage of health care professionals are making that mission harder than ever.”

To fully implement a model at the bedside, you need trust. And that is what we’re now currently building. Dr. Michel van Genderen Internist-Intensivist Erasmus MC

Trust is the foundation for AI use and adoption

For Dr. van Genderen, trust isn’t just a prerequisite for AI adoption in health care – it’s the cornerstone. As an ICU physician, he knows that every decision can mean the difference between life and death. “We will only use AI-based solutions if I know that they are safe, explainable and trustworthy,” he says. “That’s nonnegotiable.”

This mindset is what led Erasmus MC to partner with SAS and the Delft University of Technology to create the Responsible Ethical AI in Health Care Lab (REAiHL). The lab’s mission is to translate high-level ethical principles – like those outlined by the World Health Organization – into practical, bedside-ready technologies. “We are reshaping these high-level ethical values into more tangible ones so that they can be implemented in practical technological solutions,” Dr. van Genderen explains.

Founded in 2023, the REAiHL has been recognized as an official lab by the Netherlands’ National Innovation Centre for Artificial Intelligence. REAiHL is working to create a framework for ethical AI in health care that will serve as a model for medical centers and regions around the world.

“Together,” says Dr. van Genderen, “we are not only building AI technologies, but also shaping the future of health care with a steadfast commitment to upholding ethical principles and prioritizing the well-being of patients in every innovation.”

I’m proud to work with SAS because we share the same agenda: making the world a better place. This isn’t about innovation for innovation’s sake. It’s about doing what’s right – so when I look a patient in the eye, I can say with confidence that what we are doing is safe and beneficial. Dr. Michel van Genderen Internist-Intensivist Erasmus MC

Explainability and transparency are essential

The AI model governance capabilities of SAS Viya play a critical role in this effort. It provides the transparency, monitoring and governance needed to build clinician confidence in AI tools. “To fully implement a model at the bedside, you need trust,” Dr. van Genderen emphasizes. “And that is what we’re now currently building.”

For health care providers to feel confident integrating AI into clinical workflows, explainability and transparency are essential. “If I don’t understand it, I won’t use it,” Dr. van Genderen says. “Clinicians need to know how a model works, why it makes certain recommendations and whether it’s performing reliably.” With tools like SAS Viya, Erasmus MC can provide that clarity, ensuring that AI operates as a trusted partner in patient care, rather than a black box.

“With SAS, we can monitor and explain models. That gives clinicians the confidence to adopt AI at the bedside,” Dr. van Genderen says.

Deploying SAS Viya in conjunction with the Microsoft Azure tenant of Erasmus MC also supports data security and compliance with European regulations. “We feel more in control of our data,” he adds. “That’s critical, especially with the European AI Act.”

Erasmus MC – Facts & Figures

14,000+

employees

2x

increase in optimal ventilation treatment

1,233

hospital beds

Driving results and more human care

The partnership has already delivered real-world benefits:

Reduced administrative burden, freeing up time for patient care.
Increased adherence to treatment protocols through real-time insights.
Improved outcomes and treatment plans for patients.
Greater clinician trust in AI, leading to broader adoption.

One example is the dashboard that provides real-time data insights across the ICU. Developed in response to the growing complexity of patient care and the need for data-driven decision-making, the dashboard aggregates information from 40 ICU beds into a single, accessible interface. This allows clinicians to instantly assess which patients are being treated according to protocol and which require intervention.

The dashboard not only improves adherence to clinical guidelines but also helps prioritize care more effectively. “Now, when you’re treating one patient in one ICU box, you also know what’s happening with patients A, B and C,” he adds. “That’s how we can better titrate our attention and improve outcomes.”

In addition to the ICU dashboard, Erasmus MC developed a dashboard designed to monitor patients who needed mechanical ventilation. “We noticed twice as many patients had positive outcomes when they were treated with optimal mechanical ventilation settings. Both nurses and clinicians took these real data-driven insights into account when they were making treatment plans,” Dr. van Genderen explains.

Most importantly, these tools are helping restore the human connection in health care. “When I ask ICU patients what they remember most, it’s not the machines, it’s the time doctors and nurses spent with them,” Dr. van Genderen shares. “By reducing administrative tasks, we’re giving that time back to providers.”

“By using data-driven technology,” continues Dr. van Genderen, “we save time with the decision-making, and we can spend more time caring for patients, and that is the true value of connected health care.”

AI and the future of health care

Dr. van Genderen envisions a future where AI helps solve some of health care’s most pressing challenges. “We’re building transatlantic networks to validate models and share best practices. Within five years, I believe we’ll see AI solutions that meaningfully improve care quality and reduce burnout.”

“I’m proud to work with SAS because we share the same agenda: making the world a better place,” Dr. van Genderen says. “This isn’t about innovation for innovation’s sake. It’s about doing what’s right – so when I look a patient in the eye, I can say with confidence that what we are doing is safe and beneficial.”

Next Steps

The results illustrated in this article are specific to the particular situations, business models, data input, and computing environments described herein. Each SAS customer’s experience is unique based on business and technical variables and all statements must be considered non-typical. Actual savings, results, and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software. Brand and product names are trademarks of their respective companies.