Initially of 2026, HL7 appointed Rachel Dunscombe as its new CEO, succeeding Charles Jaffe, M.D., Ph.D., who led the interoperability group for almost twenty years. She not too long ago sat down with Healthcare Innovation for a wide-ranging interview about HL7’s FHIR evolution and its function within the age of AI.
Earlier than becoming a member of HL7, Dunscombe served as CEO of openEHR Worldwide, the place she led international adoption of open well being knowledge requirements and strengthened partnerships with governments, requirements our bodies, and healthcare organizations worldwide. OpenEHR is a nonprofit group that publishes technical requirements for an EHR platform, together with area‑developed medical fashions to outline content material. She’s additionally held positions throughout the UK well being and know-how panorama, together with main the NHS Digital Academy, serving as an advisor to the UK Secretary of Well being and contributing to nationwide coverage by means of the UK Authorities AI Council. As well as, she holds a visiting professorship at Imperial School of London and serves as on the board of administrators for the Digital Well being Society.
HCI: For people who find themselves much less aware of openEHR Worldwide, may describe a number of the a number of the work you led there earlier than coming to HL7?
Dunscombe: My openEHR involvement truly began with HL7 as a result of I used to be utilizing HL7 v3 RIM. The storage side of that was sunsetting and I used to be searching for one thing that I may use for my knowledge, so I adopted openEHR. It’s a knowledge mannequin for storing your knowledge, so it retains the whole lot constant, and an apple means an apple. As a CIO, I had used that, and as an instructional, I began doing a little analysis into it. Then I received a cellphone name sooner or later from Professor David Ingram, who mentioned, “, we’d like any individual to take over.” So I went to the board assembly, and it was one thing that I agreed to do. The reason being that it had supplied me, together with HL7 requirements, such worth. OpenEHR is massive in Australia and Europe and different components of the world.
Another excuse that it was so essential to me was that I used to be one of many early adopters of AI. Over a decade in the past, I had an information science staff because the CIO, and that made me actually recognize the consistency in knowledge and why we’d like it. In order that’s why openEHR turned essential to me, as a result of we have got to get the information proper earlier than we are able to do AI.
HCI: So, in essence, you had expertise main a posh volunteer-driven worldwide neighborhood….You have been on this function main HL7 for slightly over six months. Any surprises to date?
Dunscombe: I do not assume so, as a result of in my tutorial work I’ve completed international analysis. The U.S. authorities is completely implausible in supporting us, however I feel what does proceed to shock me is the shortage of curiosity and funding from sure governments across the globe. It surprises me that individuals haven’t realized that for the brand new healthcare system that’s protected and financial to function, you want your knowledge interoperating correctly.
HCI: Is that tied as to whether these governments have sufficient assets to do that? Or are there nations which are properly off, however in some way they are not catching on to the significance of it?
Dunscombe: You want cash, and also you additionally want the clever consumer within the authorities, proper? You discover conditions the place you’ve got received a type of two issues however not each, and that is the place it falls down. There are some very developed, high-GDP nations that aren’t getting this nonetheless as a result of they do not have the experience inside the authorities to acknowledge that it is essential and really leverage it.
HCI: HL7 does an annual State of FHIR survey with Firely. Are there some constructive indicators of momentum in the latest survey outcomes?
Dunscombe: Completely. I’m truly actually as we get to r6 — we’re changing into normative. That is the place it would take off additional. So we’re seeing momentum. The U.S. is doing implausible work when it comes to getting the precise abilities inside the ecosystem. However globally we in all probability want about half one million folks with actually heavy FHIR experience. As this takes off the world over, I feel the U.S. is in one of the best place as a result of it is actually invested on this, together with Spain. Funnily sufficient, that is the opposite place that is actually invested in abilities. I feel we’ll see a kind of hockey stick upwards as this turns into normative and folks begin to actually leverage it. We’re taking a look at CMS and others who’re actually beginning to deploy this at scale.
HCI: Throughout your keynote discuss on the WEDI convention earlier this 12 months, you spoke concerning the significance of not solely creating requirements, however doing the work to scale interoperability, and also you mentioned the FHIR accelerators are an excellent instance of how to try this. Probably the most not too long ago introduced was the Caliper accelerator for medical system interoperability. So I wished to ask: has the work of the accelerators advanced over time? Have there been some learnings about what works finest when it comes to how they bring about their work ahead?
Dunscombe: Sure, there have. It was nearly Darwinian to start with. We had the accelerators. We labored out what’s labored, what hasn’t labored, and what we should always standardize. Now we have discovered over time that there are particular issues that they should do constantly to feed the learnings again into the usual, but additionally there are particular issues that must range by accelerator. What we have completed is create a playbook that is getting used to create the brand new accelerators. You have simply heard about Caliper. You will shortly be listening to about our high quality accelerator, which we introduced on the CMS Connectathon.
We’re codifying what works and issues that should be constant, together with governance. However we have to hold the flexibleness. As an illustration, Caliper is de facto totally different. We have so many stakeholders. We’d like to have the ability to take care of the flexibleness of managing everyone from the home wearable producers by means of to medical system producers. I’d say it’s the broadest ecosystem of stakeholders that we have. And it’s the one I am most enthusiastic about.
HCI: Is there a transparent threshold for when one of many teams has reached the purpose the place it is able to launch as an accelerator? Or is it generally decided that one thing’s not fairly prepared for prime time but?
Dunscombe: We ask for letters of dedication from stakeholders that need to be concerned both financially or to be on the celebration. As soon as we all know that we have a completely rounded ecosystem of stakeholders which have dedicated, we are able to type it. However until you have got a completely rounded ecosystem of these stakeholders to really take an accelerator ahead, you are not going to come back out with the precise reply.
HCI: Might you discuss slightly bit about that high quality accelerator? Who’re the stakeholders concerned and what are a number of the points they are going to deal with?
Dunscombe: I’ll say that we have nice help from authorities entities — CMS, CDC, and others. And a few U.S. nationwide high quality our bodies have an interest. We have business actually . We have a very nice ecosystem forming. I feel the problem with the standard accelerator is placing the parameters round it. We have needed to do quite a lot of workshopping as a result of high quality runs from the standard of care by means of to the information high quality. High quality can imply so many issues.
HCI: Is a part of it about utilizing FHIR to report high quality measures?
Dunscombe: Completely, but additionally bearing in mind what’s coming with quite a lot of the pondering round cost by high quality outcomes. There’s a actual feeling that sooner or later high quality can be one of many measures that cost relies on.
HCI: I not too long ago interviewed Dr. Ari Robicsek, the primary chief medical officer at NCQA. He mentioned that he thinks AI is going to assist them change how they measure medical high quality — that the ambient instruments will enable for extra granular measurement of how a affected person is handled and also will assist get into patient-reported end result measures…
Dunscombe: I agree. There are facets of it, as you say, with a a lot fuller recorded interplay with the affected person, of which some can be structured into FHIR and a few can be unstructured. I used to be simply speaking with NCQA on this round issues like social determinants of well being. Then if you happen to have a look at Caliper — how does affected person and citizen knowledge from the wearables that they might select to contribute play into high quality, too?
The patient-reported end result measures are one thing that can actually be coming as a result of there’s a sense of the standard of care in a really medical sense, but additionally the general notion of care from the affected person actually must be thought-about, too.
HCI: Are there different areas the place accelerators would possibly make sense, or the place we may see extra progress on interoperability? I’m pondering of issues like like laboratory knowledge…
Dunscombe: Sure, I feel there are. In truth, I used to be simply speaking about dental as one. A few of that could be completed in work teams.
I feel AI goes to finish up giving us quite a lot of alternatives across the accelerators, and clearly there can be an AI theme to most of them. Now we have completed some actually good work to make our requirements extra machine-readable and ingestible by AI. However I feel the chance that AI gives us is multi-fold — the whole lot from supporting us in growing the requirements internally and we’re beginning to use it for that — to the best way it’s getting used inside well being methods and by particular person sufferers. I’ve had some actually good discussions the previous few weeks round how far HL7 goes into the requirements for AI.
HCI: And also you simply named Dan Vreeman your chief AI officer. So does he have an task to attempt to determine all that out?
Dunscombe: He does. He is been doing quite a lot of work initially on enabling us internally and in addition enabling FHIR to be ingestible by AI, and that is nice. That work has gone to poll. Numerous these discussions are going to be emergent. We do not have the entire solutions, and you’ll transfer too quickly and grow to be hyped by a few of these issues. He’s actually balanced and measured, and ensuring that he intersects on the proper level with AI.
I feel we additionally should be very cautious that we point out that our main goal is to get the information proper for AI. However a number of the the questions that I have been requested round lifecycle administration for AI and oversight for AI, knowledge high quality for AI, I feel over the following 5 to 10 years we’ll must see what emerges when it comes to who leads on sure facets of that. It might result in partnerships or different work. I used to be simply speaking with some teachers this morning about lifecycle administration for AI. I feel requirements our bodies are going to must create some requirements round that.
HCI: Right here in america, we are typically very U.S.-centric in our focus and most of us know little of what is going on on in Europe and elsewhere on this planet when it comes to well being knowledge requirements, interoperability and regulation. You are within the UK and lead a global group. For our U.S.-based readership, is there a special emphasis on well being knowledge interoperability in Europe than there’s within the U.S.?
Dunscombe: I like working with the U.S. It is received probably the most completely implausible coverage. The problem is that knowledge is locked into methods, and in addition it made the digitization journey early. It is received totally different EMRs elsewhere with totally different knowledge fashions. It’s a patchwork that is laborious to mesh collectively.
Numerous Europe remains to be a fairly inexperienced discipline, to allow them to study from what has occurred within the U.S. They’ve an enormous aspirational coverage to make folks’s report accessible throughout Europe. I feel that they’ve issues going for that. The social healthcare methods imply that suppliers aren’t going to dam knowledge in, however the draw back is GDPR [General Data Protection Regulation], which might be interpreted very in a different way throughout Europe. Each area has its challenges and each area is having to take care of both being not very digitized and many laws, or within the U.S., quite a lot of historical past and a free market, they usually all have their professionals and cons. I’d say Australia in the mean time is doing a little fairly good work. I feel they’ve lined their geese up. And I feel the Nordics have completed some fairly good work, however they’re solely the scale of a few of your cities, these nations. So you possibly can’t evaluate, proper?
HCI: Effectively, what about Nice Britain? So far as these insurance policies and knowledge sharing, did Brexit break Britain off from Europe?
Dunscombe: Sure, it did. It’s fairly humorous as a result of the incoming prime minister is any individual I labored with actually intently and I do know very well as a result of he was the mayor of Manchester and was accountable for the information. So he’ll take it in a special path, I consider, as a result of he was at all times extremely near the work I did with HL7 and openEHR.
Brexit has pulled us away from the European Well being Information Area. We do have within the UK a single affected person report tender that is simply gone out. I’ve quite a lot of associates who’re engaged on that from the NHS aspect, however the aspiration is to have one single affected person report throughout the entire of a minimum of England, if not the UK. Once more, although, just like the U.S., we had been very early to digitize. We’re a patchwork of methods, in order that can be a job to do.
HCI: Another initiatives arising or observations about issues I have not requested about?
Dunscombe: One factor I have been reflecting on not too long ago is that we’ll see much more folks needing abilities in interoperability, not simply FHIR, however the entire interoperability requirements. And one query I have been asking is, how can we create the training ecosystem with universities and establishments to convey on board that expertise, but additionally get the following era from universities on board? We’re doing a little work referred to as “Break the Code” and creating college chapters in order that we we get the following era in. That’s going to be kicking off in September. I am actually asking these questions on what the following part requires, and I feel abilities might be one in all my massive agenda objects.
