BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
Influential CIOs, such as Marc Probst of Intermountain, whom we saw at National Health IT Week a couple of weeks ago, are sounding calls for interoperability standards, either to have ONC mandate them, or if that doesn't happen, to call on Congress to pass legislation requiring the standards. His metaphor was to force a standard, just as government forced the railroad track gauge. How do you think health IT can get to uniform standards most quickly?
Doug Fridsma: The wonderful thing about health IT standards is there are so many to choose from. We've got a lot of different standards that are out there that we could potentially choose. There is this tension between saying we just need to use this standard, versus creating a consensus-driven approach where the industry says, 'We believe this standard will solve our problem, we believe the government should require that because it will help everyone.'
There is a tension between top-down, government-imposed standards and bottom-up, industry-adopted standards. Standards are standards because people use them. There's a whole host of standards out there that are standards but no one uses them and they don't really advance where we want to go. There's a real need for us as a country, and that means public and private participation, to identify both the challenges we have with the current standards, and to constrain them in ways that are really going to drive toward interoperability.
Doug Fridsma, M.D.chief scientist, ONC
Is there one standard you think there is good consensus on right now at any level?
Fridsma: We've broken down the problem into five different building blocks, and for each of those building blocks, we've identified a collection of standards within them. This is a big country. There are small practitioners that are just struggling to get their health information technology out, and there are big academic medical centers that have lots of sophistication. So, it's challenging to come up with a single standard that serves the needs of everyone. We have to think of our job less as thinking there's only one way, and more as constraining the choices so the majority of people can succeed.
When it comes to standardizing meaning [of clinical data], one of the building blocks, we've identified four different vocabularies, each of which serves a single purpose. We've identified SNOMED, to help with clinical descriptions; RxNorm, to help us with drugs and medications; LOINC, to help us with laboratory tests; and ICD-10 to help us with administrative and billing transactions. Each one of those serves a specific purpose and we've tried to sort of focus those in. It's not 100 vocabularies. It's four, and each for a very specific purpose.
We've adopted essentially one standard to help us with the way we exchange clinical care summaries in document-based information. And that's called the consolidated CDA [clinical document architecture]. Consolidated CDA still requires more work. It has too many options in it and was developed in consensus with the industry, but now the industry is telling us … it doesn't interoperate. So, we need to go back to the industry and say, you know what, a successful solution allows you to both send and receive this information.
How important are informatics and data science to the work of ACOs, as well as to traditional large healthcare systems?
Fridsma: Again, if we go back to Karen DeSalvo's federal health IT plan, we need to expand the contributors to solve these really hard problems. It's not just about meaningful use. It's about the federal agencies and our private sector partners; those people who are skilled in informatics and those who understand data sciences are going to be critical to the next stage of how we manage this information.
Informatics professionals really have the ability to take the computer science and the data sciences, the technical infrastructure, and translate it into solutions that work in the healthcare environment. It's a very applied field in many ways. When it comes to things like the ACOs and business models that include shared risk, the only way you're going to be able to be profitable is to really understand your population. Who's at risk? You really need to put your dollar investments into keeping those people out of the hospital and healthy.
Congratulations on your AMIA [American Medical Informatics Association] position. What are you going to be doing there?
Fridsma: I've been a member of the AMIA community for over 20 years, and I went to some of the early conferences that they had and I did my training in the informatics community as well. In many senses this represents getting back to my roots. This is my tribe. These are the folks that I hang out with. And that's a good thing. And to have an opportunity to sort of lead, grow and create a vision for these people that I care about and this organization that I care about is really a privilege.
One of the things I see as being an untapped resource or opportunity is [that] AMIA is composed of a little over 4,000 really smart people that know how to use, manage and collect medical information in a way that makes it useful for patient care or clinical research.
We have not had as much engagement, I think, at the national level of conversation. Properly galvanized, the community that is represented in AMIA can be transformational. To me, AMIA isn't just about the academic researchers who are gaining new knowledge, but it's also about all the cool things that they do. We need to get better visibility, so people see the great things that the informatics community is doing.
Doug Fridsma trumpets value of patient-generated data
Missed interoperability chances harmful to consumers
2014 a year of preparation for health IT regulations