Interoperability Isn't Failing — It's Underfunded (Herko Coomans)

[00:00:00] Dear listeners, welcome to Faces of Digital Health, a podcast about digital health and how healthcare systems around the world adopt technology with me, Tjasa Zajc. When it comes to interoperability, Europe has done a lot of progress in trying to create a unified European health data space.

[00:00:25] So I decided to dig deep into where interoperability stands at the moment, because if you ask most people in health tech why interoperability keeps failing, you can get a technical answer. Herko Coomans, International Digital Health Coordinator at the Dutch Ministry of Health and Interoperability Lead within the Global Digital Health Partnership, disagrees.

[00:00:52] As he says, interoperability is a wicked problem. It's defined by incomplete, contradictory and shifting requirements, rather than by unsolved engineering. At the same time, he says that interoperability is a basic part of the healthcare IT infrastructure, but is not treated or funded as such.

[00:01:20] So in this discussion, you are going to hear how is the role of governments and healthcare ministries changing with the adoption of the European health data space, what progress are we making with EHDS in Europe, and are the deadlines set by EHDS realistic?

[00:01:43] This is the first in a short series of discussions related to the interoperability challenges and progress in Europe. If you haven't yet, make sure to subscribe to the podcast. Also, check out our newsletter, which you can find at fodh.substack.com. That's fodh.substack.com.

[00:02:06] There's already a lot of content related to the current AI challenges, policy, interoperability, governance. And after this series, I will make sure to create a reflection and an overview of the issue with prominent expert leaders from Europe. Now let's dive in the discussion with Herco.

[00:02:43] Herco, hi, and thank you so much for joining me on Faces of Digital Health to discuss interoperability in Europe and more broadly. You are the International Digital Health Coordinator at the Dutch Ministry of Health. You are also a very important member of the Global Digital Health Partnership, where you focus on the interoperability side of things.

[00:03:10] So that's what we are going to focus on in this discussion. In 2023 at HIMSS Europe in Lisbon, I asked you for an interview for Faces of Digital Health. Three years later, my time has arrived. So welcome. Thank you. Thank you. And thank you for having me. And thank you for your patience. Yes, absolutely. Perseverance is a lesson here. So let's talk about interoperability.

[00:03:37] Anyone that works in health tech, at least for a year, will probably understand that interoperability is very important, that we run into problem where data is in silo. So, but regardless of the fact that's quite a straightforward finding, it seems that we still can't really crack interoperability.

[00:03:59] So what would you say is the interoperability topic that should be front and center in 2026? What exactly are we not getting and who's not getting it? Yeah, I think I first want to push back a little on the fact that you say that we're not cracking interoperability yet. Because we're using it already in so many places. Systems can actually speak to each other way more than they could 20 or 30 years ago.

[00:04:28] We're designing interoperable systems from the very get-go. There's a global exchange of health data happening. So I'm not sure if we are not cracking it. I do think, though, that it's a wicked problem. And the definition of a wicked problem is that it's difficult or impossible to solve because of incomplete and contradictory and changing requirements that are often difficult to recognize.

[00:04:56] And that's almost the definition of interoperability. And interoperability isn't an all-or-nothing problem. We have a lot of interoperable health data already that is actively used. And it's not that we're not interoperable because not everything is standardized or harmonized yet. Also, not everything needs to be standardized and harmonized because interoperability is not a goal in itself.

[00:05:24] It's a means for better health data, for better health outcomes. As for your real question is what should be front and center, I think it's been front and center for a while and that still needs to be is implementation. And implementation support. And it's a marathon. It's not a sprint. And we're not at the finish line yet, as you said.

[00:05:49] So I do think that implementation and implementation support and the resources for that should remain front and center. And one of the interesting things that's happening here in the Netherlands now is that healthcare executives are asking the ministry to be more directive because, frankly, they have to make a thousand decisions a day.

[00:06:13] And with the standards and interoperability issues, a lot of smart people have already looked at these same problems that they're facing and have developed a working solution, have field tested it.

[00:06:30] So basically, they're telling the ministry, please be more directive on what we should standardize because that takes away our need to make those decisions for each and every one of us again. Because it is facing the same problem everywhere.

[00:06:48] No matter what type of health system you have, if you have an NHS model or a devolved private model or anything in between, everyone is facing the same decisions and the types of decisions to make. And standards help you make the decisions that everyone else has already made for you so you can move faster and you don't have to make all those decisions over and over again.

[00:07:12] So I think implementation and implementation support is critical there and remains critical. And that includes the educational part. Start with building interoperability into the health system by design and into the health processes by design. And that will, like I said, that will free up headspace and time for the healthcare executives and healthcare professionals to do what they want to do,

[00:07:38] is build the health system to provide right outcomes and not be concerned with all these technical details that smart people have already solved for them. Let's unpack that a little bit. So we're going to talk about the role of ministries. We're going to talk about implementation and funding. A few months ago, you wrote that technically speaking, interoperability shouldn't really be a problem. It's an infrastructure funding crisis.

[00:08:07] So let's unpack that for a little bit. So what did you mean when you mentioned that? Yeah, that's a big story perhaps. But during the COVID pandemic, every health system around the globe was under extreme pressure.

[00:08:28] And we found that the problems that are part of each of our health systems were enlarged and exacerbated. And the lack of having the right information at the right place at the right time to be able to make split-second decisions on life and death was really visible. So everyone was looking for how can we solve this?

[00:08:56] So right during and after the pandemic, a lot of resources went into digital health and telehealth and the digital health infrastructure so that we're building up our national capabilities to be able to deliver better health data and better digital health services. That window of opportunity is rapidly closing.

[00:09:20] So the funding and resources are now being directed to other areas like sovereignty and defense and cost of living and energy crisis. And there's plenty of crisis to direct resources into. We also see that a lot of countries have and are redesigning their health systems like the legal frameworks to include digital and data into the pillars of health systems.

[00:09:48] Everyone is starting to understand or has already understood the value of this and that any future-proof health system needs to include the infrastructure and the requirements and the legal grounds and safeguards and guardrails to be able to deal with health data to deliver better health care and a more efficient health system by itself. So we have resources in this field.

[00:10:16] We have legal mandates. We have political will because of the visibility of the problems during the pandemic and right after. Those are stars that haven't been in this alignment in a very long time.

[00:10:33] Add to that that the industry has shown that the technology is able to deliver on the requirements and on the needs at scale, both during and after the pandemic. And now with AI, that's even been accelerated more and more. So all those elements are already in there. So technology isn't the biggest problem.

[00:10:59] Sustainability of this momentum and of the decision-making and of the leadership and of the resources definitely is. Like you said in the introduction, my lens is the policy and government governance lens. So I look at our role as sort of public caretaker or public responsible organization.

[00:11:25] And as governments in interoperability, we've been funder mostly from an implementation perspective. So when we were implementing these standards, we also funded the development of specific capabilities in those standards that are useful for our own health organization or health system. And through that... Can you explain that a little bit? So what exactly did you fund in that sense?

[00:11:56] Because if we just slowed down a little bit in the past, when standards weren't as mandated as they're also now starting to be with the European health data space, the challenge with interoperability was that one vendor used potentially one standard, another vendor used a different standard, and then nobody wanted to invest in the bridge building between the systems.

[00:12:21] And that wasn't defined anywhere in the procurement process. So how is that changing? And just in relation to the fact that you said that the Dutch ministry is also heavily investing in the implementation part and support of standards. I think this is an ownership problem. So who owns the problem of the integration?

[00:12:48] And before in implementation, we were focusing mostly on product implementation. And we needed that product implementation phase. So we needed to have interoperable patient workflows. So these were very much... We need an EMR, and we need an image system, and we need a patient portal or a patient app, and it needs to be able to handle this data flow. So we looked at DICOM, and we looked at SNOMED, and we looked at FHIR,

[00:13:18] and we looked at IHE profiles, and P for medication. So we looked at these different elements so that we could include them in our requirements. In many different ways, some do this through the legal requirements. Some do this through the infrastructure requirements that they have for their national health IT infrastructure. Some have it through procurement requirements. When they are doing a public procurement for health services,

[00:13:45] they can include you have to be able to communicate in this way. And the Netherlands doesn't have public health care procurement or a national health information infrastructure. So we do this through incentives and mostly financial incentives and quality of care requirements. The shift that we're now making is that as we have implemented these at the point of care right now,

[00:14:14] that we do have interoperable capabilities in our health systems like modern EMRs and national platforms, and the EHDS builds on top of that and expands that to cross-border services as well. Well, because of that, we are now facing more of an integration issue, is that when all these new developments and all these new innovations come into play,

[00:14:41] we want to be able to integrate them into our national platform. We didn't have that national platform before. We haven't procured the national platform because we have no idea what that looks like now, or in a year or in 10 years. But it is something that's emerging as more of a concept than a technical thing. But integration into that is becoming more and more a requirement.

[00:15:08] And integration means that these products and services that are interoperable need to be able to work together more. So as governments, one of the major things that's happening is, like you said, with EHDS, but also with national legislation that Estonia, I think, pioneered years ago, but that other countries have built as well, like Finland and Denmark and Germany and France

[00:15:36] and the Netherlands and Spain and Belgium and all these other European countries have national legislation that build interoperability into the requirements for in different ways. But also outside of Europe with the 21st Century Cures Act in the US, with the My Health Record legislation in Australia, with the incredibly huge Ayushman Bharat digital mission in India,

[00:16:02] which is scale-wise, it's for us in the Netherlands, it's unfathomable. It's hundreds of millions of people and hundreds of thousands of healthcare providers and organizations. But they're all doing the same thing is in mandating the use of these standards. And my issue with that is that because we're mandating them,

[00:16:28] we also have a public accountability requirement and responsibility. So, for instance, our parliament is asking us, the Minister of Health, what she is doing to ensure that our nursing workforce can work with these standards and deliver the quality of care that they are required to do

[00:16:54] if SNOMED and NANDA are not really linked together very well. So, the parliament is asking our minister, what are you doing to ensure that NANDA and SNOMED work together to ensure that our nurses can do their job properly? That is a question that would have never been asked five years ago if we didn't have that legal mandate. So, that also means, as a consequence,

[00:17:22] that our role in the governance of these standards needs to extend into that domain as well. So, we're no longer just consumers and clients of these products. We need to be actively involved in their roadmap and in their development. And the other side of that coin is also the sustainability part. We need to ensure, as a responsibility, a public responsibility now,

[00:17:52] that these standards are still there and openly and freely available for all to be used in the next 20 to 30 to 50 years because this is what we are building the health system of the future on. Now, it is funded through member countries, which are governments. HL7 International has a few country affiliates that are funding parts of them.

[00:18:22] But what they are doing is they're funding a part of HL7 International's capabilities to ensure that we have Dutch fire profiles aren't necessarily interoperable and reusable in other countries. But we call them international standards because they are built on fire. But a Dutch Met My Fire profile doesn't necessarily work in Germany, which is only a few miles away from here.

[00:18:52] So, basically, that's another interoperability issue then. It's not a solution. It's another issue. So, part of the legal mandate that we as governments now have means, and the EHDS is organizing this at the European level as well, is that we now have these processes to consult the ecosystem,

[00:19:19] the national ecosystem, and that includes vendors and patients and funders and healthcare providers and healthcare professionals, and academic experts and all those stakeholders in this ecosystem to define the national data and interoperability needs. Right? So, this is what we do. It sounds like there's a lot... It almost sounds as if working for the Ministry of Health today

[00:19:49] on the interoperability side is much more stressful than in the past. If the parliament's looking at you, the ecosystem is asking you for guidance. Yeah. I'm not saying it's stressful. I'm saying it's a lot more fun. How is that? Yeah, no, of course. Because... So, we have new responsibilities, but this is part of the maturity process that we're all in. The standards organizations have been...

[00:20:18] I'm not saying begging, but pushing governments to put this in their mandates for a lot of years. Because this is the maturity curve that we're in. And we can do this as governments because of all the amazing work that the SDOs have done in the last 20 to 30 years in getting us to this level of maturity

[00:20:47] that we're now able to mandate this. Interoperability never starts with a law. It's always the formalization of a set of capabilities that have been built by the industry, by the healthcare professionals, by the standards development communities, and that we are now formalizing. But part of that formalization is that government takes a bigger role

[00:21:15] and has a bigger public responsibility and accountability for that. Now, so we do this at the national level with these mandates that we now have. We do this at the supranational level with the EHDS. And the EHDS board is the new problem owner at the EU level to be able to address these issues. So they were building a governance framework at the European level

[00:21:43] to coordinate our interoperability need and data need requirements. The question, though, is who has this at the global level? So in the Netherlands, we have what we call public holdership. And we call it holdership. You can call it public stewardship. We call it a holder because it's not an owner in the legal sense,

[00:22:11] but it is that stewardship responsibility. And it starts at the health system level and then goes down to who's the funder, who is the maintainer, who is the developer, who is the authorizer. So it's got all these roles. And we're defining for the standards in our national health system, we're defining each of these roles and responsibilities

[00:22:38] and assign them to the appropriate organization and people. And then ensure that they are adequately resourced to be able to provide those capabilities to our health system. In Europe, we're doing that with the EHDS. But at the global level, we're not doing that. And we used to look at the WHO as the most likely candidate to be able to build those capabilities

[00:23:07] and have that governance framework. But with the departure of the US and Argentina has also left, and the US being the most dominant market in this space, we're not looking at WHO anymore as the only likely candidate to do this. So this is where the GDHP starts to come into play because then the G20 countries are saying,

[00:23:34] we need to fix this problem at the global level. And how do we look for? So they looked at the GDHP, this group of now 44 countries and WHO that are working together. And they said to the GDHP, we need you to be able to coordinate a global effort to be able to adopt a standard as widely as possible.

[00:24:04] And they said, use the International Patient Summary, the IPS as your, not your pilot, but as the first standard that you're doing this with. So back in 2021, the G7 statement that was during the UK G7 presidency, they actually said, we want to ask you to advance the adoption of the International Patient Summary.

[00:24:35] And the G20 then picked that up and advanced that and pushed us further. And the interesting thing about the International Patient Summary is that it is by definition a global standard. And it's not just one, it's a collection of at least five standard bodies that are providing artifacts for that. It's an HL7 FHIR implementation profile. It's got a CDA profile.

[00:25:03] It's an ISO information model. It's got the SNOMED global patient set that provides the terminology. It's got the IHE implementation profile. Behind the ISO model is a European SENELEC model. And the IPS FHIR implementation guide is the one that's being used in the European Patient Summary as well. So there is a feedback loop

[00:25:32] between the International and the European Patient Summary. There's 40 countries involved in the Global Digital Health Partnership. I showed the website earlier. And the countries that are not part of WHO anymore are there. But still, this is a body or an organization that can recommend things. You can't really mandate people to do anything. So how do you define success?

[00:26:02] And how do you make sure that you get the buy-in of people to actually work in the direction of interoperability and other issues related to digital health that should be more coordinated globally? Yeah. Technically or legally speaking, the GDHP doesn't exist. It's just a meeting of people. Technically, that's it. And that's by design.

[00:26:31] It's not meant to be another secretariat that has a different mandate. We're not a UN organization or like an OECD or WHO type of organization. The decisions that we can make is what the collective can make. So we have the mandate of every member by itself. The thing is, if we all make similar decisions,

[00:26:58] then we have a real impact. There's over 40 countries in the United. I think it's 44 right now. And together, we have more than 40% of the world population. That is both, oh, look at us, we're big enough to matter. But also, it's the market size. So you can look at the GDHP

[00:27:27] as a market shaping mechanism. So if all those countries, including India and Nigeria, which is a big country, and Brazil, which is a big country, and the US and Canada and the UK, and Japan and Australia and South Africa and Zambia, and if all those countries make similar decisions for their own jurisdictions, then we are pushing the market

[00:27:56] globally in a certain direction. So especially with interoperability, where we're looking at harmonization of products and services so that they are actually very similar, not just aligned, but very similar in design and in how they work. Those are decisions that should come from a motivation

[00:28:24] that is very close to the local needs. So we cannot let some other organization than each of our own organizations make those decisions. They are better informed decisions if we share our rationale and our successful implementation needs and requirements and the resources and the investment that we need to make into those decisions with each other. And that helps us

[00:28:54] make very similar decisions for each of our jurisdictions. That is by design that we don't mandate that type of decision to something else like a GDHP secretariat or a governance body. And basically, Europe is the same in this sense. So it's very complementary to what we're doing. We're going to definitely focus on Europe and EHDS

[00:29:24] and maybe even the EU AI Act a little bit more. But before that, if we stick with the GDHP for just a little bit, it's also an organization where chairmanship moves from country to country. You, so the Netherlands, was chairing since 2004 to 2006. What would you say was the biggest success of the chairmanship and where do you hope that Portugal

[00:29:53] will continue to work with the chairmanship that they took over in 2026? So, we are, we've, we're very proud of the fact that we're in a row or a line of GDHP chairs that started with Australia and then became India then became the US and then the Netherlands. So we feel like we're like a mouse dancing with the elephants

[00:30:23] and saying, oh, look at how much noise we're making. And so, I'm very proud also that Portugal is continuing that tradition that the size of your country doesn't matter in your ability to provide leadership in this space. So, I think our approach to the GDHP chair and with the chair comes also the secretariat so that moves every two years

[00:30:52] to the member country and this is again by design so that we don't have a separate secretariat that then becomes bigger and starts to do the things that the country members should be doing themselves but that they delegated it out to something else. I think one of the main things that we tried to do was foster the safe space and the community part. The GDHP is a very interesting group

[00:31:22] for outside organizations because it has decision makers so we have top government digital health executive leadership and their top experts and specialists all working together. So, we have chief digital officer of the Australian Digital Health Agency. We have the deputy national coordinator for health IT from the US. We have

[00:31:52] the CEO of Canada Health Infoway. We have the secretary for information for health information and digitalization of Brazil. of Brazil. We have the vice minister for digital health from 44 countries, 44 important stakeholders. We have director levels, we have director general level, we have vice minister, we have politicals,

[00:32:23] and their top experts. But the thing is with the GDHP and what makes it really fun for us is that it's a very informal group. We don't sit behind our flags and share our national positions, but we share our failures, we share our challenges, we share, of course, we share our successes as well, but then we discuss how can we leverage those successes into what everyone else is doing. So, for instance, when Canada

[00:32:52] was drafting their new interoperability legislation that is currently also being processed in their parliament, government, they interviewed a lot of GDHP member countries and what type of legislation do you have and what's working for you and what is your rationale behind that and they fed that directly into, okay, then if that works there, then how can we leverage that in our Canadian federated provincial setting

[00:33:22] to work because there's other countries that have that type of setup as well. In essence, a really nice global repository of international challenges that people can look into. I think what really makes it interesting is it's the decision makers that are there and that their main goal is to make better informed and internationally validated decisions.

[00:33:52] So, we're not there to have the GDHP make a decision for all of us but we're helping each other because there is no GDHP that can make that decision for all of us. our flow that we enhanced or evolved from what the US has been building on top of what India had been building on top of what Australia had been building was very much that flow of first we make a state of play, so what are all the countries doing,

[00:34:22] then we get to the best practices, which is always nice but this is where a lot of it stops, then from those best practices we distill what's the decision making guidance that you can make, so if you're dealing with this topic like cyber security for medical devices, what are the best practices, but then we turn that into a practical guide for policy making and decision making and investment making at the public level.

[00:34:52] what type of decisions do you need to make and what has already worked in this peer group so that you can make these decisions faster and then if that's necessary but especially for interoperability, then we look at how can we harmonize the decisions that we're making so that each of us are making the same decisions that we're implementing the same standards for the same use cases with the same structures behind that and one of the things that you see

[00:35:21] there is with the IPS, it's a global standard that's being maintained and managed by the standards organizations themselves that are joined in the Joint Initiative Council, the JIC, and they're doing an amazing job there and Canada is using the IPS to implement the Canadian patient summary and Brazil is using the IPS to implement a huge system that will allow all its citizens

[00:35:51] to have a patient summary at their beck and call that they can use everywhere but also internationally. Saudi Arabia used the IPS to ensure that the annual Hajj pilgrimage meetings where millions of people come to Saudi Arabia to go to Mecca and these are elderly and frail and sick people mostly so there's a high chance of medical encounters that they have an

[00:36:21] IPS available through a QR code so that the data that's in Indonesia or in Malaysia or in Oman gets transferred to that hospital live by scanning that QR code and that at the point of care they know what the allergies and the previous problems are and what type of medication they're already taking in the language of that healthcare provider

[00:36:50] so this is a good example of having one standard or one set of artifacts and if we all make the same decision then you get real interoperability. I feel like you're still answering the first question which was a little bit provocative in terms of is progress happening and you now demonstrate it very nicely how yes progress is happening we just have to be patient because it does take time which kind of brings me to

[00:37:19] EHDS so the European health data space is now going from the excitement phase where it felt like this is an ambitious idea hard to implement but the right direction to the phase when things need to be defined on a technical level and the implementation phase which you also mentioned is the hardest part. in interoperability so give me your insight your insider look into

[00:37:49] where EHDS is at the moment and how you see the progress being done. My insider view isn't that special every country is working really hard in ensuring that we are complying with the requirements and for a lot of things this happens from different starting points so some countries already have national authorities the Netherlands

[00:38:19] doesn't have that national authority so we have to take that long path in getting the national legislation through setting up the organization defining all the roles and tasks and responsibilities and then actually recruiting people and setting up processes and building a real organization from scratch so everyone is scrambling and doing their best to make this happen so that by 2029 we can actually exchange

[00:38:48] patient data and medication information and images across the borders but also have these European capabilities available to our national health system so someone in Groningen which is in the north of the Netherlands can exchange data with someone in Maastricht which is in the south of the Netherlands using these same capabilities that we have for when we exchange the same information with Berlin or with Madrid or wherever

[00:39:18] we want to exchange it with so as a consequence everyone is now focusing on what does this mean for us and for some that's a discovery journey because we don't know yet a lot of the things we need to do we don't know yet because we haven't done it so it's not like what so for the Netherlands it's having

[00:39:47] that national authority right the health data access body and the health data authority we don't have that so we're still discovering everything how do you build something like that so other countries have this but it's they have to change it into EHDS compliant forms of what they have Belgium already has their health data authority Denmark has a

[00:40:17] long history of having a health data authority but they are pivoting and merging a lot of other organizations into their new digital health Denmark organization to be able to comply with the regulation and the requirements there and this is the process in each country so we have to make it fit in our health system so it's not a European thing but it's a very national implementation thing and a lot of resources

[00:40:47] go into that because not a single country has a simple national health system and since healthcare has always been a national competence interoperability isn't built into the European health systems because that's not how Europe worked and this EHDS isn't a health

[00:41:16] regulation it's a single market and a data and privacy regulation so a lot of the effort and the challenge is how do we build something that's uniquely national by definition and attach to that this European layer that is common to all of us but has a different implementation in each country this takes away a lot of the

[00:41:46] focus that we used to putting resources into this big European common project that is now being redirected into how does this work in our country if we have to implement it at a pace by the way that has that's unprecedented we chose 2027 2029 2031 deliberately but there's not a country

[00:42:16] that doesn't feel that this is an ambitious and challenging timeline but it is also because we feel that the result of this the capabilities that this delivers is something that everyone needs as soon as possible so this is what we call the stick after we've done the carrot thing but yeah I think at the European level we see a lot of resources being redirected to the national implementation so this I think is what

[00:42:46] is that sort of the European excitement is dropping but if I look at what this is generating as national excitement in all these different countries similar energies and similar dynamics starting and already happening there yeah when you were talking about the ambition and the ambition goals and deadlines and how people feel they're potentially a bit unrealistic it basically reminded me of planning

[00:43:15] through objectives and key results where the objectives always need to be super inspirational but the key results need to be defined in a way that you don't if you achieve 100% that means that the goal wasn't set ambitiously enough so if you achieve 60 to 80% that's already a good result and this is how I'm reading that EHDS is moving when it comes to the deadlines and everything we've committed to this as governments we're on

[00:43:45] the path to ensuring that we fulfill all the requirements that's our commitment it is in a way it is part of the European data moonshot and again you also mentioned the AI Act there's also the Data Act and the data governance and the digital services and the digital markets act and the interoperability act for the public interoperability there's the NIS and the CER the cyber security

[00:44:15] regulation so at the ambition level Europe has really exceeded itself in creating this sort of the European data union idea and yeah so health is an important part of that and you have to and I think this is important that's why I'm stopping you from saying something but intervening yeah so why start with health

[00:44:44] in this whole data European data union moonshot and that is because of Brexit because the answer to Brexit has been more integration between the EU member states right so making it more difficult for countries to get out of the European Union not because we add extra barriers but because we add extra incentives to be

[00:45:14] integrated like the euro was a big one of them the data integration is another big one of them start with health because that touches every single one of us so we start with you being able to get your medication with the European prescription that you have everywhere you are in Europe right that's the so you actually experience the benefits of this integration

[00:45:44] the next one is if you need care in another country then your data is already available the next step is that when someone has a rare disease then we can use the European data infrastructure and capabilities to do better and faster research for cures and treatments so it's compounding all those benefits that build on top of the integration layer that

[00:46:14] we have to build both economic and social and personal benefits from that European project so that is why we start with health which is arguably probably the most complex and most difficult one to start with this is also why we start with a regulation which is the biggest and strongest and most impactful instrument that we have in the European Union it's common law

[00:46:44] in

[00:47:26] in in a year how do you feel that pressure of speed for change because of AI it's the pressure is really really high there's two elements to this the positive element is that we see evidence of this speed everywhere

[00:47:56] in the development of new medications and treatments in understanding the interactions in our human bodies with the protein folding that alpha fold is doing with genetic research with all these areas that used to have a longer 10 20 year runway that is now incredibly shortened and therefore bringing a lot of

[00:48:25] hope and expectation that this will then follow through in the next phases of development but I think that there's areas where this really works well and there's areas where the capabilities that AI brings may augment but not replace a lot of the things that we're doing so in practical care delivery decision support is still

[00:48:56] in its early phase let's just say it like that there's pockets of evidence where we see that it really works especially in with images and things that have a lot of historically already structured and huge amount of training data available but not everywhere the thing I don't think it needs to do everything for everyone all at once so the expectation is in my opinion somewhat

[00:49:25] inflated not just somewhat quite a lot inflated yeah I explain AI to other people now as it's not magic but it's the magician it goes through all these motions and then it drapes the very shiny very thin blanket over the person in the chair there's a flash and there's a bang and there's smoke and then the person in the chair is gone right that is AI everyone says

[00:49:55] oh that's amazing that's a great trick that's magic the thing is AI is the magician and not the magic because the AI thinks or makes it very plausible that the person in the chair is actually gone and it cannot reason at least generative AI models it cannot reason where the person is because the result is the person is gone so what AI is doing is

[00:50:25] it brings things very plausibly plausibility isn't the same as truth it's close to truth but it's not the same so we can verify things to make it more trustworthy or we can see where but what we need to do is we need to know where is the person that was in the chair we know that part of this is a trick but for healthcare

[00:50:55] we need to know where is that person so that is a different type evidence based work that AI mimics a lot of these steps in but cannot do everything just because of the nature of these models basically what I was asking is more earlier we talked about the pressure on and the spotlight on the healthcare

[00:51:24] ministry for so many things so from that perspective I'm wondering how does AI impact your work at the end of the day there's still someone at the ministry that needs to make a decision that needs to issue a guidance own that guidance and stand by it which is basically the biggest challenge who's gonna take the responsibility and the Netherlands has its own historic story about why interoperability is a sensitive topic I can link some of

[00:51:54] the episodes on this in the show notes so interoperability is there an impact is there no impact there's definitely impact like you said it makes becoming interoperable easier especially because AI is very good at that interface between humans and machines and computers so it can make it way easier for humans

[00:52:24] to interact with machines instead of all those drop down boxes and the you need to select I need to have this term or this code in this field so a lot of what you see now is focusing on that interfacing like ambient scribes and those types of technologies which I think is really great that's one end the other end is that we can leverage interoperable data in

[00:52:54] AI way better now again because of how especially the generative AI models are built they don't work really well with ontologies inside the models so they cannot reason through SNOMED for instance so you can ask it please give me a SNOMED code for this patient and it'll give you a SNOMED code that looks like a SNOMED code that smells like a SNOMED code but that you have to verify

[00:53:23] if it's actually true so that's one of the interoperability problems that we're facing that again it's very plausible but it's not the same as truth and for interoperability the whole reason for interoperability is that we encode the truth as the healthcare professional and experts as they see it so that's the gap I think that we cannot just expect AI to solve

[00:53:53] but that we have to find that interface for yeah absolutely primary sources are key one last question for the Netherlands the priority for AI therefore is two things one is AI readiness for organizations and people so AI skills but the other big one is validation so how do we know that the AI actually delivers on what it promises that it's delivering

[00:54:22] so the validation part of AI is something you have to measure it against something else but that's what we're investing big in in the Netherlands okay that's good to know I had two episodes exactly on those primary sources when it comes to data standards and the challenge of validating AI if every output is different so yeah again just a call to action to other discussions as well before we wrap up

[00:54:52] OECD recently published a report where they actually quantified what benefit interoperability could bring so their assessment is that connecting fragmented healthcare data could equal to 2.7 to 6.6 percent of annual health expenditure which is huge it's also a number and now people love numbers we love to just grasp onto something

[00:55:21] and now just use that as an argument for new initiatives so I'm wondering how did you see that report how do you see that this quantification is going to impact anything yeah we saw the report we were part of the GDHP was part of the group that co-developed the content of that report but OECD did a really amazing job the numbers that you mentioned are directly from three

[00:55:51] other studies or four other studies that they reference two in Finland and one in UK and one in Canada if I'm correct so it's based on those four countries but if you look at that and extrapolate that to maybe a global number it's in the somewhere between the two and the ten billion US dollars which is like you said it's an unfathomable number so it does a couple of things one is it

[00:56:21] validates the many assumptions that every country has made to be able to make the investment use case to our ministries of finance to be able to invest the billions into what we're doing and have done for the past 10 years similar to why are we doing the EHDS what will it actually deliver it's and the thing I think that's mostly interesting

[00:56:50] and recommend everyone to read is that this means it's a shift from funding projects in different areas for implementing these standards and capabilities like governance and standards maintenance and translation and all those other capabilities that you need for interoperability to this needs to be part of a

[00:57:20] core infrastructure which part of it needs to be a public infrastructure and a public set of capabilities and I a healthier workforce it is

[00:57:49] having the data available for secondary use and the economic effects of all those things so it's extrapolating the wider effect of interoperability which is really interesting to follow up on because maybe that could be another podcast here with OECD because it shows that interoperability isn't just a technical health issue it is a

[00:58:19] public responsibility issue it is a market access and industry shaping issue it is a patient safety issue which brings it into clinical governance and medical education so it touches so many things and this report clearly defines where these effects are visible and where they ripple into and it's not unique interoperability is

[00:58:49] just one of the facets and aspects that have a similar effect but like you said it's a really interesting and important one and this report really shows that one of the things that are really about this report as well is it

[00:59:21] to shift to to to sustainable infrastructure funding at the global level as well it brings a lot of new opportunities for that as well kirk kirk i think we covered a lot so thank you so much for joining me for this discussion and i'm sure we are

[01:00:04] enjoyed the show do leave a rating or a review wherever you get your podcast subscribe to the show or follow us on linkedin additionally check out our newsletter you can find it at fodh.substack.com that's fodh.substack.com stay tuned