Q&A: CEO Charles Jaffe on broadening HL7's horizons
HL7 – not just for IT anymore. That thinking is the catalyst behind a triptych of recent moves designed to open the standards process to more health professionals, notably caregivers.
Ideally, pulling in a new group of professionals will open the feedback loop, particularly to those concerned with usability and workflow, but by no means limited to that. According to Charles Jaffe, HL7 CEO, tapping into their minds and, indeed, day-to-day work experiences will also yield specialist knowledge that bolsters decision support.
Government Health IT Editor Tom Sullivan spoke with Jaffe about those initiatives, very positive initial reactions, and what the future holds for HL7. Hint: Genomics, and mobile health.
Q: I’ll start by asking about HL7 recently opening up some of your intellectual property to the caregiver community…
A: We are going to make, at an absolutely nominal charge, a caregiver membership available immediately for health professionals – physicians, nurses, therapists, pharmacists and so forth, to broaden the base of HL7, increase domain expertise and, hopefully, a small percentage of them will actually become involved in the development of standards. And not simply usability, workflow, so on and so forth.
Q: So what’s the end-goal of this broadening?
A: It really falls into two categories. One is that it’s not an economic driver. In other words: I’m kind of hoping that we recoup enough to pay for the administrative cost, but if not, it’s an expense for outreach.
It certainly is in our interest to bring in a new group of people whose reason for interest in this is because of their concern about usability, workflow, domain content, and so forth. A really simple example is that a pediatrician needed an electronic health record that they thought more accurately reflected their specific needs for caring for children. But in another sense and across all domains so would the allergist and immunologist. The most incorrect diagnosis in a medical chart is allergy to some medication. The immunology community is very concerned about that because in the CCD there are lists of things that glibly are referred to as allergy which really fall into the category of ‘I think I don't like this medicine.’ If I had a dollar for every one of those, I probably could pay for you for lunch. The allergy community is up in arms over this nonsense.
Q: Any other communities with a similar sentiment?
A: We as the United States spend billions of dollars avoiding whatever it is that someone alleges they have an allergy to, when they absolutely do not. An example is the X-Ray Contrast injection which is called x-ray die, it’s not die at all. It’s clear as water. In a vile, you couldn’t separate out the two. They are opaque to x-ray and when injected into an artery they often cause some symptoms – not an allergy. It’s not related to iodine, shellfish, or shrimp. It’s not even the kind of immune reactions we characterize as allergy. And yet all of this information was known in 1971 and 40 years later we’re still asking the question about your food intolerance, which is totally unrelated.
This is the reason that you can get articles that are reproduced in all the big city newspapers that say ‘well, EMRs are not effective.’ So if electronic medical records are simply a substitute for paper, then there’s probably little benefit, other than you can read it. But if the definition of electronic medical record embraces decision-support, the tens of thousands of people who go into their family doctor today in the U.S. who have a diagnosis of strep throat when they really have Viral Pharyngitis to get a broad spectrum antibiotic even in lieu of the very narrow antibiotic that could be prescribed for a strep throat, we have bacterial resistance, we have antibody development, we have reactions to the medications, real or imagined and we have millions of dollars in cost. Decision support would really enable the caregiver to say – or perhaps to learn rather – I guess 40 year olds don’t get strep throat very regularly. Maybe I need to reinvestigate this. And if I have a strep test and it is positive, well then I’ll give them the appropriate antibiotic and not a broad one.
Q: You’re talking about decision support based on a variety of specialties?
A: So if we had allergists, immunologists, at the table we would help improve the content, the workflow, the business rules, for an electronic medical record. And the same is true for other diseases, other specialists, other verticals, where having caregivers present really improves our disease management. At some point, we will have medical records with well-embedded decision-support as part of their functionality. HL7 has been focused on this for well over a decade. It’s difficult, it’s expensive, and you need to know reliable source for content. There are content providers, and we need standards for integrating that content into workflow. We need better algorithms, we need better decision support and we want the caregivers at the table – and so our decision to make the IP free, our decision to embrace a larger stakeholder community beyond the IT and technical folks or the few people who have a clinical background and a technical one, that’s not enough. So we bring in the people who care about the workflow and the content, let them look at our profiles, our domain models. Those two announcements really went hand in glove.
Q: The first one was opening up the IP to caregivers, the second being late last month you made EHR domain models available for no-cost licensing. Is the recent announcement that you are connecting HL7 CDA to clinical trials data also a piece of that puzzle, or something else entirely?
A: At HL7 we have a domain model called BRIDG (Biomedical Research Integrated Domain Group), in which we have modeled clinical research into the larger domain of medicine and so one terminology, one exchange language, and so forth. This is a long-standing collaborative effort between CDISC [Clinical Data Interchange Standards Consortium] and the National Institutes of Health and HL7, and a very successful one. So that’s not a new interest. We have other vehicles to bring the research community into the organizations because many countries around the world use HL7 for electronic submission of clinical research data. Needless to say, we embrace those. What we think is these purely clinical people who understand the context of healthcare, will not participate, so for that that reason we’ve made our profiles and domain models available, for them to criticize, for them to improve, for them to show us concepts and important elements that we hadn’t really addressed.
Q: And has anybody started participating in that?
A: Yes. As a matter-of-fact we’ve had over 200 people already – and it’s only eight or nin days – download the profile. I spoke at a meeting this weekend and we had two-dozen clinicians, and I’d guess the majority were MDs, ask where they could log on and join as HL7 caregiver members. So we’re doing our level best to get the Web site up and functioning. There’s a skeleton Web site up, we’re loading the content onto it, and expect to have that available by the end of the month. That’ll be less intriguing for IT people well aware of HL7 and the value it brings to their work. But it’s a new idea so the American College of Physicians,
we’re also reaching out to the big international nursing associations this summer, to try and integrate them into the process. While they want to learn for us, we also want them to influence the things they do because their input is really critical.
Q: Building on the caregiver initiative, the EHR domain models, what’s on the horizon for HL7?
A: This project to make the intellectual property cost-free to the user is just a pilot. We’ll evaluate the success of the program, the implications to our business model, and we may expand it to more HL7 intellectual property. At the same time we’re expanding some of other key programs. In the not-too-distant future, for example, your genetic profile will help guide your caregiver in both diagnosis and treatment. And so we have a large genetics and genomics workgroup that is developing the specifications so that genomic data becomes part of your record in a seamless, uniform fashion.
We also have a new workgroup interest in mobile health. It’s on everyone’s mind, what sorts of standards are needed for delivery of content securely if healthcare is transacted in a mobile environment. These discussions of putting your chest x-ray on a phone don’t help too much. I can tell you, I could barely recognize it as a chest x-ray but perhaps on a larger device there could be some value to that. More importantly, it’s a way to deliver decision support. Physicians are reluctant to say ‘Sorry, Tom, I don’t know the answer to that and I’m going to go look in a book.’ I bet that doesn’t happen too often. But if you have a mobile device at the point of care, they might be willing to look something up, and say ‘I didn’t know that drug interacted with the ones you’re already taking.’ So we think mobility is a key issue and we’re jumping in with both feet.