For healthcare organizations, challenges to interoperability include getting consistent and timely response from electronic medical record (EMR) vendor interface developers, financial costs of building interfaces, and technical difficulty of building interfaces. Undoubtedly, many have classified the state of interoperability in healthcare immature as vendors and providers continue to try to work together to make all systems interoperate, while the call for more direction from the federal government has been made.
What’s more, medical image exchange across the enterprise presents its own set of challenges, as historically, medical image storage has been under the control of individual picture archive and communications system (PACS) applications, requiring management of that data to be completely reliant upon system functionality. To this end, David Avrin, M.D., Ph.D., vice chairman, informatics, professor of clinical radiology, department of radiology and biomedical imaging at the University of California, San Francisco (UCSF), will be part of an interoperability panel discussion at the iHT2 Health IT Summit in San Francisco on March 3, 2015 (the Institute for Health Technology Transformation, iHT2, is a sister organization of Healthcare Informatics under our corporate parent organization, the Vendome Group LLC).
Dr. Avrin and others will discuss the best practices for data exchange at their respective organizations, addressing how organizations are working with regional health information exchanges (HIEs) and vendors, and what’s on the horizon for interoperability. Click here to register for the San Francisco event. Avrin recently spoke with HCI Associate Editor Rajiv Leventhal about what’s holding the industry back from being more interoperable. Below are excerpts from that interview.
Interoperability is clearly a huge healthcare buzz word these days. Where does your organization stand when it comes to data exchange?
We have been successful with integrating in-house vendor/provider applications. All of our principle apps—our PACS system, radiology system, Epic’s Radiant, and PowerScribe—are tightly integrated. Take abdominal imaging for example, a patient’s medical record at UCSF opens up in a second, and you are one click away from seeing the entire record and notes by date or by category. We have achieved that through a requirements and specifications process, and customer assistance with these vendors. Certain vendors know they need to interoperate. That works well.
Here’s what does not work well: as an academic medical center, we get patients from elsewhere, which might include Sutter Health, Oakland Children’s, Kaiser Permanente in northern California, or whoever. If you’re dealing Epic to Epic, for example, to Kaiser in northern California, sometimes with the right permissions you can get read-only access into Kaiser’s medical record, but you can’t bring over things like medical lab values automatically.
We also receive patients from community physicians, and in spite of meaningful use, we still have a terrible time with electronic records or any records across care boundaries from the outpatient world into the inpatient world and back again. It’s a mixed world, and it’s getting a little better because customers are insisting more, and vendors know they have to do a certain amount. But it’s far from perfect, far from what many hoped would happen. There’s a lot of finger pointing going around. It’s in the community hospitals’ interest to share data, but data really belongs to the patients, not the hospitals. Vendors say there aren’t well-accepted standards, and they also take advantage of the fact that the neither the customer base nor the government have done a good job with insisting to adherence to Health Level Seven (HL7) Version 3 or Logical Observation Identifiers Names and Codes (LOINC), or whatever. Outside of healthcare, they have had a simple time with electronic data exchange. We are still in the dark ages.
You mention finger pointing, so where does most of the onus lie?
The majority of the onus is still on the community hospitals now for two reasons: they haven’t done a good job of making demands, and also too many community hospitals view patient information as their proprietary property, their marketplace advantage, and are not motivated to share it. Vendors would go along if the customers were more progressive. Another thing is there is no national medical record number for patients. The Veterans Administration (VA) uses social security numbers for patients, yet every other industry does it differently.
Also, meaningful use was a great opportunity for the government to guide the adoption of certain standards and XML cloning of data. I think the federal government blew it with Stage 2, as interoperability was supposed to be a huge part of it. That was not well done. It was a real missed opportunity to force some standards development. HL7 could simply things, and it needs to recoup.
There’s a story I like to tell about a seminar that I teach at the medical school here at UCSF on problems with EMRs. So we go around the room to get background on where the students come from, their interests, if they work, etc. One person actually said he’s a CTO of an EMR company, so I said tell me more. He said his company provides 90 percent of the university clinic outpatient healthcare records in the U.S. I asked him how they got their solution certified, and he said for Stage 1 requirements, to transmit records electronically, all you had to do was put it in PDF and email it. That’s not interoperability. Interoperability is XML-tagged data.
Can Stage 3 fix some of this?
Well, there is the new 10-year interoperability vision now, but I don’t think any stage of meaningful use will do it. They simply blew it. In terms of making this an interoperable world, we haven’t gotten there and it’s our own fault. Providers need to be insistent, community hospitals can’t get in the way, and vendors, including the Epics of the world, need to be more responsive.
Regarding image exchange, what are the most advanced organizations doing?
One of the most advanced HIEs in the country is in Kansas City, so if you show up in an outpatient imaging center or hospital in the Kansas City metro area, they have an opt out strategy meaning you’re automatically opted in. Your images, your finger X-ray, and the report are available throughout the area. Is that okay with you? Ninety-five percent or more people say yes, and if you say no, here is what you have to read and sign. That’s what opt out is. They have one of the most successful HIEs in country that includes exam reports and image exchange, so we need to follow that model nationwide. You love the world where everything about patients pops up in a second, as you should.
How do you envision the state of interoperability five years from now?
We just want all vendors to be able to talk to each other with open standards, and I think that problem will be solved in a few years, less than five years. The biggest complaint you hear is that “I don’t want hospital X to get stuff from my archive without me knowing it.” If it really belongs to the patient though, that shouldn’t be a problem. But it is a problem. We haven’t achieved this permanent, built in transparency to get rid of requesting and moving the CD, but we’re getting there. Mark Kohli is a guy who has my job at Indiana University, and he refers to the International Standards Organization (ISO) seven layer model, where the top layer is programs that look like they’re talking to each other, with the bottom layer being an Ethernet cable. But there is an 8th and 9th layer, Kohli says, that is politics and money. And more than money, it’s about agreeing on technologies and a set of XML tags. There is a lot of politics involved, and we could have used help from the government there regarding meaningful use. Unfortunately, we didn’t get it.
Dr. Avrin will be talking more about data exchange at the iHT2 Health IT Summit in San Francisco next month. The entire conference agenda can be seen here.