AI in healthcare in northern Europe suggests future pathways

Efficiency

Currently, the Netherlands has 1.4 million people working in healthcare, roughly 14% of the total working population. But with the prediction that one in four people in Europe will be over 65 by 2050 and thus require more complex care, those numbers are far from sufficient.

Managing patients requiring complex care is expensive and requires the healthcare sector to shift from cure to prevention. AI can be used across so many different elements of healthcare, from a chatbot that helps people to live more healthily to an AI system that can diagnose diseases quickly to AI systems that predict the best treatment for a specific patient.

“The Netherlands is a forerunner in AI imaging for radiology and pathology diagnostics. This helps speed up the diagnostic process of diseases and therefore can help eliminate waiting lists,” says Jeekel.

Workforce gap

On the one hand, a lot of developments in AI are still viewed with some suspicion by healthcare professionals but might turn out to be a solution to the glaring staff shortage in healthcare. The workforce gap is only expected to widen in the Netherlands, Europe, and even worldwide.

Estimates are that the overall demand for healthcare workers will rise to 18.2 million across Europe by 2030. This means that the current availability of healthcare professionals will not meet the current or projected future need.  

“This number actually translates to having about a quarter of the Dutch labour force working in healthcare in 2030, just to keep the sector running,” Jeekel says. “That’s obviously impossible, so it’s vital we start working smarter. That is where AI can play an important role in the area of prevention, reducing care burden, and optimising processes. It is a labour-saving technology that ensures work can be done more efficiently.”

Support healthcare professionals

AI could be the silver bullet the healthcare sector needs because it can also be used to improve the quality of care. For example, the PersOn programme in the Netherlands uses AI to analyse all available data about the type of cancer and the individual patient to predict the expected outcome of treatment. With this information, doctors and patients are better equipped to decide on a treatment for that particular person to have the highest quality of life.

Another example of the smart use of data within healthcare is the use of AI to diagnose allergies in children, using DNA from nasal cells to determine whether a child has an allergy. This was developed during a study by the University Medical Center Groningen, the Medical University of Hanover, and the AI company MIcompany. It contributes to a better understanding of complex diseases and offers opportunities for innovative diagnostics in the future.

“In addition, a smart system can also support healthcare professionals with the right information about and for a patient,” says Jeekel. “Such a system can search a database of articles on a condition much faster than the professional could.” And then could there be AI used to schedule emergency ambulances? Possibilities seem endless and promising.

Examples of AI in healthcare

Other northern European healthcare sectors are also developing and applying AI.

For instance, Danish startup Robo-Doc is developing AI-powered robots to assist surgeons in performing complex procedures, such as spinal surgeries. Norwegian company Algeta, acquired by The Bayer Group, has developed an AI-powered system that can identify the best treatment options for prostate cancer patients, which is already used in many hospitals across the country. Sweden and Finland are already using AI solutions to help healthcare providers detect cancer at an early stage. These applications can also be found in Germany.

Meanwhile, in the UK, the Imperial College London is using AI to develop a diagnostic tool for tuberculosis. The London researchers are training an AI system to analyse chest X-rays and detect signs of the disease.

Sharing knowledge

One of AI’s major challenges, especially in healthcare, is the transparency of algorithms, says Jeekel.

“AI has not yet been widely adopted in healthcare, so we see a certain fear of it amongst professionals and board members. There needs to be more clarity on the technology and its possibilities and challenges for healthcare organisations. That is why addressing legal, ethical, and social issues is crucial before building AI systems,” he says.

The US and China are well advanced in applying AI in healthcare, he adds: “These are large countries with more clout as a result. Europe consists of many different countries, each with its own research and applications of AI. To stay competitive in this area, it is necessary to join forces. That is already happening with certain European laws and regulations, like the AI Act, but in terms of knowledge sharing, we need to cooperate much more in Europe.

“It is not just about sharing knowledge and use cases, but we need to start to jointly determine which legal, ethical, and societal issues need to be addressed and how we, as European healthcare, are going about that. This is crucial for adoption and trust in AI to ensure transparency and understandability.”

Accessible, affordable and high-quality healthcare

A broader understanding of AI within healthcare can accelerate its deployment. At the moment, innovation projects often take too long in healthcare, losing momentum.

“There are too many different parties involved,” says Jeekel. “And while it is crucial that, for example, the board of directors is knowledgeable, we must also demystify AI a bit. We are not talking about a terminator that comes running along, but about a system that can significantly improve the quality of care and reduce workloads. We need a little more speed in adoption to benefit from it.”

In 2022, an Integral Care Agreement was signed in the Netherlands, comprising agreements between the Ministry of Health, Welfare, and Sport, and a large number of parties in the care sector. The agreement aims to keep healthcare accessible, affordable and high quality for the future.

Jeekel is convinced AI can contribute significantly to this: “When an AI imaging solution can diagnose diseases faster and better, and another AI system can determine the best treatment, there is a higher chance of recovery for the patient. Moreover, waiting lists can be eliminated by saving time in this process, and healthcare providers can ensure and improve the quality of care provided.”

Prevention

But Jeekel also has high hopes for AI in the field of prevention in the coming years: “With the right triage, you can reduce the influx to hospitals, for example. GPs can also be supported. But what about the possibility of personally helping citizens with AI live healthier lives?

“This is already possible in terms of technology – think of all the wearables that collect data from us. This data can be combined with other data sources, such as healthcare data. Then an AI system can provide personalised advice that helps someone stay healthy and thus keep people out of healthcare longer.”

The potential of this kind of application is enormous, Jeekel believes, but the challenge lies in making it available and accessible to everyone, including those less digitally literate.

Important in this development, as in fact with all AI applications in the healthcare sector in northern Europe, is to involve healthcare providers at an early stage when developing algorithms, says Jeekel.

“Too little time and budget are currently allocated to free up healthcare providers for this,” he adds. “But when you want to design an AI system that supports the caregiver in their work process, that caregiver must be involved to make it successful. So there has to be time and opportunity for that. Currently, the Dutch healthcare sector is not considering this enough.”

Adoption and scale

In terms of technology, great strides are currently being made in the field of AI in healthcare in northern Europe. Where we, as Europe, can and will differentiate ourselves from North America and China is by being more human-centered AI, Jeekel says. A lot is being invested in this area in northern Europe.

Take the AI Hub in France, where some €1.5bn has already been invested. There are also many initiatives in Germany. Also, they have a government strategy for startups to ensure they can make more significant strides in digital health applications (DiGa), get investments, and grow faster if their innovations work well.

“We still lack that in the Netherlands; the speed of innovation and implementation that allows companies to take the next step and reinvest and innovate further. We do need that, otherwise we will miss the boat,” says Jeekel. “For investors, the Dutch AI climate is, for this reason, much less interesting than, say, the UK or Poland, where innovation is applied and developed much faster.”

Northern European countries currently do share some knowledge and expertise on AI in healthcare, but this needs to be expanded, according to Jeekel: “There are countless wonderful projects, for example, in Poland and the Nordic countries. When we see something working well, we should adopt it faster, provided it fits into our healthcare system. It is crucial for the future of the healthcare sector to start using Europe’s scale in this area.”