John Halamka: 'Create the Fear of Missing Out'

29 March 2024
John Halamka_Mayo_Clinic

Can you briefly explain your role at Mayo Clinic?

 "Many academic centers that collaborate with industry and/or government are not as flexible. Perhaps it's different in Europe, but in the U.S., collaboration between these groups is hindered by various rules related to safety, privacy, and compliance agreements. This often results in delays in forming collaborative partnerships. Mayo Clinic's CEO and President, Gianrico Farrugia, came up with the idea in 2019 to start a platform that bundles a set of technologies, policies, and business processes to enable frictionless collaboration between organizations on a global scale. My task, when I started in 2020, was to develop that idea and realize the platform. I then sought a way to make all valuable healthcare information from 100 years of Mayo Clinic, free of rights and restrictions, available in 'the cloud' for various innovators. All of this is aimed at achieving groundbreaking innovations."

Where does your passion for Mayo Clinic come from?

"I have worked in academic healthcare for over forty years, and typically, the professor drives these types of organizations. But at Mayo, it's entirely different. It's all about the patient from the very beginning. Mayo Clinic is an organization with 79,000 employees worldwide, all working collaboratively every day, asking, 'How can we provide the best care for our patients?' This culture enables us to build the kind of innovations needed. Mayo Clinic is a fairly flat organization, where each of those 79,000 employees plays an equal role when it comes to seeking improvements for patients worldwide."

How do you connect with people worldwide, and how do you reach them with your innovations?

 "Initially, Mayo Clinic started with its innovations for its ten million patients and demonstrated that its model worked, and that data could be ethically and responsibly used to drive innovations. Therefore, we reached out to our colleagues in Brazil (Hospital Albert Israelita Einstein), Israel (Sheba Medical Centre), Canada (University Health Network), and just last week, our colleagues in South Korea (Seoul). We will be traveling through Asia and Europe in the coming months, asking if countries can safely and reliably make their data available in the cloud, of course, under their own control, for innovative web services. This approach has given us a reach of more than 30,000,000 patients in a short time, and hopefully, soon more than 100,000,000 patients. Once the innovations are realized, the challenge is to integrate them into the workflow of healthcare professionals. In the U.S., our models are used in radiology Picture Archiving and Communication Systems (PACS). With the help of specific algorithms and AI (always under human supervision), radiology images can be rapidly interpreted and translated into diagnostics. The beauty of this is that it's not limited to the U.S. and can be used worldwide. This is our practical approach. We collect data globally, create models from it that are universally applicable and accessible. It's important to mention that our applications do not replace medical specialists, but they enhance their capabilities by, for example, taking on the heavy mathematical computation, allowing medical specialists to spend more time with patients."

How is it possible that examples of the above applications were developed in Dutch hospitals over fifteen years ago but have not yet led to medically validated products in clinical practice? Do you have any idea why it's taking so long to implement them?

"For every innovative application you want to implement, you're dealing with (combinations of) different technologies, and different regulations can be a determining or constraining factor. This can also vary from country to country and culture. In the U.S., for example, it's determined that every algorithm (AI) is considered a 'device' because our FDA states that it's not a 'drug.' This means that every AI application must go through an approval process, which translates to a cost of one to two million dollars per algorithm. Sometimes, randomized clinical trials and rigorous market research further slow things down. In summary, the application of innovations is dependent on a lot of (delaying) regulations. Mayo Clinic has already delivered one hundred and eighty validated models, of which only eighteen can actually be used due to the regulatory delays. But, as I mentioned earlier, the cultural component also plays a role. In some countries, there's even debate or doubt about the application of AI, even after a positive approval process, because patients wonder if AI is safe. This was the case in the beginning, but with a sufficiently high standard of algorithms, this fear is being eradicated. On top of that, there's the current state of compensation in medicine. If we find (in practice) that we are not curing the patient but just reducing their annual check-up costs, then what would be the financial reward for the hospital? It's an important aspect that has to be considered if a model is developed but no innovation process is initiated or continued. So, I think the most important thing is to make it an international effort, which can only be realized by collaborating and adhering to a universal set of rules."

Do you think that people still doubt the safety of AI applications?

"Initially, when AI models were new, patients had some doubt about the safety. But now, if you compare it to human interpretation, AI's capabilities in, for example, detecting lung cancer nodules or retinal findings, are far superior. Of course, there are always concerns about privacy. For instance, even with an EEG, the AI model can detect not only if you're asleep or awake, but also what you're thinking. That's a big leap, but we want to do it in a responsible and legal way."

You say you want to make healthcare available to all people across all demographics using AI. How do you intend to do that?

"My background is the digital transformation of many sectors (from banking to education, and from supply chains to governments). We saw a common trend, which is that all sectors are moving towards value-based systems. Also, hospitals should adapt in that direction. This means that services should be available to everyone and should be designed with equality in mind, giving the same opportunities to everyone. So our approach is to make medical information available on a global level for free in 'the cloud,' but under the control of the owner, of course. We then make use of various platforms to enable young and creative minds to develop the innovations we need. Mayo Clinic's goal is to make healthcare data available for them to use and build, so we no longer have to be confined to the space and resources of a large organization. It enables us to collect data from many people who have already been patients. We use that data, on the one hand, to innovate and create global applications, and on the other hand, to better understand the diseases we're dealing with, such as long COVID or glioblastoma. All our results are available to everyone at all times, and we offer full transparency. This makes it a true public-private partnership. It's an approach I've been using for many years now and it has led to over a billion users of the database. With all that data, we can offer health to the world."

You say that it should be available to everyone. Does that mean that the whole world is ready for your model?

"It doesn't mean that it's all up and running now. It's a big cultural shift. Right now, it's only the first six months in some countries, but Mayo Clinic will be traveling around the world to ask each country whether they are ready to do the same. In two years, more than one hundred million patients should be able to benefit from the innovations we've made. But, on the other hand, it will also make Mayo Clinic even better because we'll be on the receiving end of all those data in different languages and from different cultures. When this happens, we will ask the countries to contribute data. We will integrate that data, translate it to a universal standard, and give back what we found from it. It will be a network of nodes, all securely connected. This is what we are currently building in collaboration with Google."

You are therefore putting Mayo Clinic at the center of the world, but isn't that also a danger?

"In the sense that Mayo Clinic is a 'command and control center'? That is not the case. Our goal is not to be the best. Instead, our goal is to offer health to the world. Mayo Clinic's position is to build a global public-private partnership. The more different partners we can involve, the better it is. That's how you create a community. That's the way to succeed and deliver the best possible health to the world."

How do you deal with, for example, cybersecurity, given that data will be collected globally? What are you doing to prevent data leakage and cyber threats?

"Cybersecurity is an important aspect of everything we do. It starts with securing the individual cloud providers. Of course, it is also about creating a secure connection between cloud providers. On top of that, every data store is monitored and secured with the highest possible standards, far beyond current standards in many countries. Next, we make use of blockchain technology. All access and transactions are logged on blockchain. This allows us to see where, for example, health records were consulted and by whom, in real time. Finally, we anonymize data. We strip the data of all patient-specific identifiers. It's a layered defense system, a multilayer cake of security. It requires both ethical and technical defense to be effective. We are in the process of creating a layered defense mechanism."

The European GDPR was set up in 2016 to set standards for data protection and privacy in the European Union. What is your opinion on this?

 "It's important to respect a patient's right to determine who is allowed to access their data. GDPR and other local laws must be maintained. It is important to emphasize that we are not seeking to move data from a local environment to the U.S. or another country. Our objective is to make our approach global, and we will put everything together on a European level, following the European directives for the patient. That doesn't mean that all data have to be stored in Europe, but you can follow the example set by the EU, which is putting data in the hands of patients. It's exactly the way we want to develop our platform and ensure all data and applications have high integrity."

Innovation that knows no borders. What are you most proud of so far?

"I am most proud of Mayo Clinic's staff, who have dedicated themselves to improving the care for patients. We are fortunate to have the best possible caregivers. It's also impressive to be the President of a team that is able to make this much progress. It's about more than just high-quality care. We are very well aware of all the data and their impact on healthcare. This way, we not only offer data and applications for the Netherlands but also around the world. We are contributing to global healthcare."

What would be your key takeaway message for your colleagues in the Netherlands, both in healthcare and the political landscape?

"I can't speak for every organization. It's a global challenge. Our work on digital health is a global project. Not only does Mayo Clinic see it this way, but we are working with 1,800 other institutions on a worldwide level to build what I described. In addition, we're working with Google, which provides the technology to make all the required information available. I would like the Netherlands to join the group and use Mayo Clinic's platform, offering access to data, research, and global applications, but also contributing to the network of institutions, which will make it even better. We are all strong together. We are here to create an organization for the people. We build value-based services, and we'd like the Netherlands to participate."

John Halamka is slated as one of the keynote speakers at the prestigious ICT&health World Conference 2024 in the Netherlands. Together with the Mayo Clinic Platform, he extends a warm invitation to healthcare leaders and professionals worldwide, urging them to unite in shaping the future of healthcare at this groundbreaking conference. Their shared vision aims to harness the power of technology to transform patient care and propel global progress in the field.