HCI: So, in essence, you had experience leading a complex volunteer-driven international community….You've been in this role leading HL7 for a little over six months. Any surprises so far?

Dunscombe: I don't think so, because in my academic work I've done global research. The U.S. government is absolutely fantastic in supporting us, but I think what does continue to surprise me is the lack of interest and investment from certain governments around the globe. It surprises me that people have not realized that for the new healthcare system that is safe and economic to operate, you need your data interoperating properly.

HCI: Is that tied to whether those governments have enough resources to do this? Or are there countries that are well off, but somehow they're not catching on to the importance of it?

Dunscombe: You need money, and you also need the intelligent client in the government, right? You find situations where you've got one of those two things but not both, and that's where it falls down. There are some very developed, high-GDP countries that are not getting this still because they don't have the expertise within the government to recognize that it's important and actually leverage it.

HCI: HL7 does an annual State of FHIR survey with Firely. Are there some positive signs of momentum in the most recent survey results? 

Dunscombe: Absolutely. I’m actually really interested as we get to r6 — we’re becoming normative. That's where it will take off further. So we are seeing momentum. The U.S. is doing fantastic work in terms of getting the right skills within the ecosystem. But globally we probably need about half a million people with really heavy FHIR expertise. As this takes off across the world, I think the U.S. is in the best position because it's really invested in this, along with Spain. Funnily enough, that's the other place that's really invested in skills. I think we're going to see a sort of hockey stick upwards as this becomes normative and people start to really leverage it. We're looking at CMS and others who are really starting to deploy this at scale.

HCI: During your keynote talk at the WEDI conference earlier this year, you spoke about the importance of not only creating standards, but doing the work to scale interoperability, and you said the FHIR accelerators are a good example of how to do that. The most recently announced was the Caliper accelerator for medical device interoperability. So I wanted to ask: has the work of the accelerators evolved over time? Have there been some learnings about what works best in terms of how they bring their work forward?

Dunscombe: Yes, there have. It was almost Darwinian in the beginning. We had the accelerators. We worked out what's worked, what hasn't worked, and what we should standardize. We have found over time that there are certain things that they need to do consistently to feed the learnings back into the standard, but also there are certain things that need to vary by accelerator. What we've done is create a playbook that's being used to create the new accelerators. You've just heard about Caliper. You'll shortly be hearing about our quality accelerator, which we announced at the CMS Connectathon. 

We're codifying what works and things that need to be consistent, including governance. But we need to keep the flexibility. For instance, Caliper is really different. We've got so many stakeholders. We need to be able to deal with the flexibility of managing everybody from the domestic wearable manufacturers through to medical device manufacturers. I would say it’s the broadest ecosystem of stakeholders that we've got. And it’s the one I'm most excited about. 

HCI: Is there a clear threshold for when one of the groups has reached the point where it's ready to launch as an accelerator? Or is it sometimes determined that something's not quite ready for prime time yet?

Dunscombe: We ask for letters of commitment from stakeholders that want to be involved either financially or to be at the party. Once we know that we've got a fully rounded ecosystem of stakeholders that have committed, we can form it. But unless you have a fully rounded ecosystem of those stakeholders to actually take an accelerator forward, you're not going to come out with the right answer. 

HCI: Could you talk a little bit about that quality accelerator? Who are the stakeholders involved and what are some of the issues they will focus on?

Dunscombe: I will say that we've got great support from government entities —  CMS, CDC, and others. And some U.S. national quality bodies are interested. We've got industry really interested. We've got a really nice ecosystem forming. I think the challenge with the quality accelerator is putting the parameters around it. We've had to do a lot of workshopping because quality runs from the quality of care through to the data quality. Quality can mean so many things.

HCI: Is part of it about using FHIR to report quality measures?

Dunscombe: I agree. There are aspects of it, as you say, with a much fuller recorded interaction with the patient, of which some will be structured into FHIR and some will be unstructured. I was just talking with NCQA on this around things like social determinants of health. Then if you look at Caliper — how does patient and citizen data from the wearables that they may choose to contribute play into quality, too? 

The patient-reported outcome measures are something that will really be coming because there is a sense of the quality of care in a very clinical sense, but also the overall perception of care from the patient really needs to be considered, too.

HCI: Are there other areas where accelerators might make sense, or where we could see more progress on interoperability? I’m thinking of things like like laboratory data…

Dunscombe: Yes, I think there are. In fact, I was just talking about dental as one. Some of that may be done in work groups.

I think AI is going to end up giving us a number of opportunities around the accelerators, and obviously there will be an AI theme to most of them. We have done some really good work to make our standards more machine-readable and ingestible by AI. But I think the opportunity that AI offers us is multi-fold — everything from supporting us in developing the standards internally and we’re starting to use it for that — to the way it is being used within health systems and by individual patients. I’ve had some really good discussions the last few weeks around how far HL7 goes into the standards for AI.

HCI: And you just named Dan Vreeman your chief AI officer. So does he have an assignment to try to figure all that out?

Dunscombe: He does. He's been doing a lot of work initially on enabling us internally and also enabling FHIR to be ingestible by AI, and that's great. That work has gone to ballot. A lot of these discussions are going to be emergent. We don't have all of the answers, and you can move too soon and become hyped by some of these things. He is really balanced and measured, and making sure that he intersects at the right point with AI. 

I think we also need to be very careful that we  indicate that our primary objective is to get the data right for AI. But some of the the questions that I've been asked around lifecycle management for AI and oversight for AI, data quality for AI, I think over the next 5 to 10 years we're going to have to see what emerges in terms of who leads on certain aspects of that. It may lead to partnerships or other work. I was just talking with some academics this morning about lifecycle management for AI. I think standards bodies are going to have to create some standards around that.

HCI: Here in the United States, we tend to be very U.S.-centric in our focus and most of us know little of what's going on in Europe and elsewhere in the world in terms of health data standards, interoperability and regulation. You're in the UK and lead an international organization. For our U.S.-based readership, is there a different emphasis on health data interoperability in Europe than there is in the U.S.?

Dunscombe: I love working with the U.S. It's got the most absolutely fantastic policy. The issue is that data is locked into systems, and also it made the digitization journey early. It's got different EMRs in different places with different data models. It’s a patchwork that's hard to mesh together. 

A lot of Europe is still a quite green field, so they can learn from what has happened in the U.S. They have a big aspirational policy to make people's record available across Europe. I think that they have things going for that. The social healthcare systems mean that providers aren't going to block data in, but the downside is GDPR [General Data Protection Regulation], which can be interpreted very differently across Europe. Every region has its challenges and every region is having to deal with either being not very digitized and lots of legislation, or in the U.S., a lot of history and a free market, and they all have their pros and cons. I would say Australia at the moment is doing some pretty good work. I think they've lined their ducks up. And I think the Nordics have done some pretty good work, but they're only the size of some of your cities, those countries. So you can't compare, right? 

HCI: Well, what about Great Britain? As far as these policies and data sharing, did Brexit break Britain off from Europe?

Dunscombe: Yes, it did. It’s quite funny because the incoming prime minister is somebody I worked with really closely and I know really well because he used to be the mayor of Manchester and used to be in charge of the data. So he will take it in a different direction, I believe, because he was always incredibly close to the work I did with HL7 and openEHR.

Brexit has pulled us away from the European Health Data Space. We do have in the UK a single patient record tender that's just gone out. I have a lot of friends who are working on that from the NHS side, but the aspiration is to have one single patient record across the whole of at least England, if not the UK. Again, though, like the U.S., we were very early to digitize. We're a patchwork of systems, so that will be a job to do.

HCI: Any other initiatives coming up or observations about things I haven't asked about?

Dunscombe: One thing I've been reflecting on recently is that we're going to see a lot more people needing skills in interoperability, not just FHIR, but all of the interoperability standards. And one question I've been asking is, how do we create the education ecosystem with universities and institutions to bring on board that talent, but also get the next generation from universities on board? We're doing some work called “Break the Code” and creating university chapters so that we we get the next generation in. That’s going to be kicking off in September. I'm really asking those questions about what the next phase requires, and I think skills is probably one of my big agenda items.