How can dialysis care become more predictive and personalized?

Dialysis, which removes waste and excess fluid from the blood of people with kidney failure, has been used for more than 80 years. While the underlying technology has reached a high level of maturity, dialysis centers continue to face significant challenges.

“Solutions are needed to structurally improve dialysis care, for patients, healthcare staff, and hospitals,” says John Gesquiere, general manager of Nipro Digital Technologies Europe. “However, the biggest evolutions in dialysis technology are behind us. The real innovation today comes from data. We want to evolve from a manufacturer to a digital player, enabling hospitals to organize care in smarter ways, through the use of data and AI.”

Nipro is a global supplier of dialysis equipment and related products, including machines and disposables. Through its day-to-day collaboration with hospitals and nephrologists, the company gains a direct understanding of the challenges healthcare professionals encounter. These challenges include:

• Too much administration: Nurses spend a considerable amount of time on administrative tasks.
• Disrupted planning: When a patient’s blood pressure falls during dialysis, hospitals often need to adjust their planning.
• Limited patient insight: Nephrologists need better visibility into patients’ conditions to monitor treatments more effectively.

These challenges are compounded by broader pressures on healthcare systems. Hospitals face increasing financial pressure, while nephrologists are caring for growing numbers of patients. Digital technology will be instrumental in helping to solve these issues.

John Gesquiere, general manager of Nipro Digital Technologies Europe.

Proximus NXT accelerates Nipro’s data strategy

For Nipro, data is the foundation for addressing these challenges. Their NephroFlow software platform already combines information from dialysis machines, laboratories, and patient records, giving care teams the insights they need to adapt treatment to individual patients.

The next step was to create a secure and robust data environment capable of bringing this information together and making it available for analysis. “Our teams have knowledge of AI and data science, but we needed a partner who could build the right architecture and help accelerate it,” says John Gesquiere.

Nipro turned to Proximus NXT for its expertise in cloud and data platforms through its subsidiary Codit. Together, the teams developed a data lake using Microsoft Azure Databricks, creating a central hub for Nipro’s growing digital ecosystem.

“Codit has helped us to lay the foundations for a secure cloud environment in which we can train and test our algorithms. They not only helped to build the technology but also acted as a sparring partner to put our ambitions into practice. Without their expertise in cloud and AI, it would never have gone so fast.”

The decision to build on Microsoft Azure was a natural one for Nipro, particularly given the sensitivity of the data involved. John Gesquiere explains: “We work with sensitive, personal data. This cloud offers the highest guarantees in terms of security and compliance.”

The architecture also recognizes the reality of healthcare IT environments, where some hospitals continue to operate partly on-premises despite the growing adoption of cloud technology.

John Gesquiere, general manager of Nipro Digital Technologies Europe.

How AI is used to anticipate dialysis complications

The first major application built on the new data foundation is an AI model designed to predict sudden drops in blood pressure during dialysis. This is a particularly relevant use case for care teams, according to Gesquiere.

“For nurses, that is one of the biggest challenges,” says Gesquiere. “AI will ensure that we can predict that risk, so that the nurse can intervene in time. That means less stress for the care team and more comfort for the patients.”

Tests will start in a number of Belgian hospitals. Initially, the algorithm will operate in the background alongside the care team, allowing Nipro to refine its model using real-world data.

“The tools are ready, the infrastructure is on point. We can now test our first algorithms live. This is a huge step forward for our digital transformation, and it brings us closer to our goal: more time for care, less time on administration.”

Data creates new opportunities for smarter care

The potential of Nipro’s data platform extends beyond predicting complications during dialysis. The same technology can eventually support operational improvements within dialysis centers. “With well-structured data and the right algorithms, numerous possibilities open up: think of planned maintenance at the right time or more efficient use of people and resources. In this way, we are building a smart, comprehensive solution that facilitates the work of healthcare professionals and enables patient-centric care models,” says Gesquiere.

Nipro also has a broader ambition: to expand the cloud platform globally together with Proximus NXT. Gesquiere explains: “We have a strategic foundation on which we want to build further. The cloud platform is scalable, allows innovations to be rolled out faster, and keeps costs under control. Our ambition is to bring together data from different countries – of course within a strict legal and ethical framework. Why? The broader the dataset, the more powerful our models become, and the more possibilities there will be.”

“I’m convinced that this approach will also provide solutions for countries currently facing a shortage of nephrologists. Ultimately, for us, digital transformation is about better quality of care and higher patient satisfaction,” concludes Gesquiere.

John Gesquiere, General Manager of Nipro Digital Technologies Europe