Hospitals are constantly looking for ways to work more efficiently. Relieving staff workloads, streamlining processes and making the most of the capacity that is available. Zuyderland Medical Centre is tackling these ambitions with a data platform and a predictive platform. Guido Frijns, Manager of Data and Artificial Intelligence, explains: “By using data and technology smartly, we can deploy staff more effectively and avoid unnecessary costs, with plenty of positive side effects.”
One of the biggest challenges lies in the Emergency Department (ED). By nature, it’s unpredictable and serves as a key junction within the hospital. Patients often move on to other departments, which can lead to overcrowding, unnecessary transfers and additional staff deployment. Guido: “That’s why it’s so relevant to improve forecasting in the ED: how many patients are expected, what symptoms will they present with, and what level of care will they likely require.”
Dataplatform
To address the challenges in the ED, Zuyderland developed a data platform. Accurate forecasting is only possible with consistent and reliable data. Guido: “We moved away from SAP Business Warehouse and, with Conclusion as our development partner, transitioned to Microsoft Fabric in the Cloud. This gives us a flexible and scalable platform that brings together all cure, care and mental health data. It improves data quality, simplifies analysis and is widely used across our organisation.”
Data from various systems is uniformly defined, consolidated and made available centrally. Creating a single, trusted source of information for the entire hospital, with strong attention to compliance.
“A single flexible and scalable cloud platform”
Technically sound and practically relevant
The dashboards and AI models within the platform were developed in collaboration with healthcare professionals and managers, ensuring optimal use of consistent and reliable data. Guido: “By gathering use cases from stakeholders, we identified which models deliver the most value in terms of efficiency and quality.”
The dashboards are logical, fast and intuitive. They’re suitable for nurses, specialists, support staff and administrators, and tailored to current information needs at various levels. This means hospital data is fully accessible, secure and immediately applicable in practice. The platform enables faster reporting, more consistent decisionmaking and is the technological foundation for predictive models and future healthcare innovations.
“Forecasts are broken down by location, specialty and triage code”
Predictive platform
Built on the data platform, Conclusion and Zuyderland developed a key predictive model: the ED Forecasting Platform. It allows the hospital to predict patient inflow to the ED with a high degree of accuracy - around ninety percent.
Guido: “This platform forecasts which patients will arrive and what triage level they’ll require. The model factors in a wide range of variables, such as season, day of the week, weather conditions, public holidays and events like marathons or concerts.” Forecasts are broken down by location, specialty and triage code, enabling targeted planning both in the short and medium term.
First time right
This allows Zuyderland to align staffing and resource planning with expected demand weeks in advance. Even bed capacity can be forecasted. Guido: “We can plan more precisely. Better assess what’s needed in terms of staffing. Even in cases of unexpected absence, like a sick colleague. Everything runs more smoothly.”
Strategic decisions are also better supported, such as hiring temporary staff. It’s the ‘First time right’ principle at its best: Zuyderland gets processes right from the start, avoiding the need for corrections or adjustments. Guido: “The result is less ad hoc decision-making and greater calm throughout the organisation.”
Patient Flow
The predictive platform is also valuable for clinical departments. By forecasting patient flow from the ED, departments can prepare accordingly. Guido: “This kind of predictive model doesn’t stand alone. It’s part of a chain. Departments can prepare for peak demand by scheduling extra nurses or planning earlier discharge rounds.”
This application provides insights up to three months ahead with eighty percent accuracy. It also factors in care intensity, allowing staffing to be even more precisely matched to patient needs.
The combination of data and predictive platforms has led to clear improvements at Zuyderland, from saving time and costs to greater staff satisfaction and better patient care. Guido: “Care is now more predictable, more efficient for staff and more cost-effective. The foundation for this success is ever-improving data quality. That’s what we stand for, that’s what makes us different. And we keep improving every day.”
About Zuyderland
Zuyderland Medical Centre is a top clinical hospital with five locations in Limburg, in the south of the Netherlands. It has around 980 beds and handles approximately 35,000 inpatient admissions, 38,000 daytreatments and over 200,000 first outpatient visits annually. With over 4,500 employees and nearly 300 independent specialists, it is one of the largest hospitals in the Netherlands.
