I read with interest the news item last week about the Vermont Health Information Exchange. While that state’s exchange is under the microscope for issues with spending and data quality, there are plenty of other HIEs out there that are struggling with being useful at the point of care. Since my work spans multiple states, I’m able to see what is available to clinicians and how well it integrates with clinical workflows. There are variables whether the exchange is truly an exchange – namely whether data can be pushed or pulled or whether it’s little more than a view-only repository.
In the Vermont situation, 91 percent of interviewed stakeholders think that the state needs an HIE, but only 19 percent of interviewees feel that it is meeting their needs. One of the major barriers cited is Vermont’s opt-in policy, which limits the number of patients whose data is present for sharing. At this point, only about 20 percent of patients have opted in.
I had my own adventure with opt-in in my early days as a CMIO, when we created a private HIE to share data among physicians affiliated with our hospital. Although there wasn’t a specific state law that forced us to be opt-in, there were case law citations that prevented us from assuming all patients gave us permission to share data. We were able to maneuver through it over time by having all participating practices add language to their new patient consent forms that permitted sharing through the HIE. The practices also had to go back and have new consents executed for existing patients, and that took time.
Our vendor was subsidizing the interfaces because we were a beta client for their new HIE platform. Our hospital was picking up the rest of the tab, so there was no cost to the community physicians. My staff and I did countless road shows trying to convince physicians that this was a good thing to be part of, but at the same time, our CIO spent a lot of time trying to kill it simply because it wasn’t his idea and it was being executed by clinical leadership rather than IT leadership.
We ended up being live for quite a few years until our state HIE began to take shape. In all, it was an exciting time, but very different from the environment we’re in now, where interoperability is at least a little bit easier.
Despite having been live for several years, my own state HIE still struggles. It doesn’t communicate with our state immunization registry, which reduces its utility for primary care and urgent care physicians. All the immunizations sent to the HIE are strictly added as read-only data element, and there is no mechanism for resolution of duplicates or for reconciling with the immunization registry. A physician looking to validate immunizations on the HIE also has to go to the registry, and since the registry actually functions as a source of truth, why not just go there in the first place?
Our HIE doesn’t store any diagnostic imaging, only PDF report documents. Sometimes these are useful when an existing finding is well described and can help serve as a comparison, but there are entirely too many radiology reports out there with “clinical correlation recommended,” which means the reading radiologist isn’t going to stick his or her neck out by providing a specific diagnosis. When we find unusual things on an x-ray or CT, we’re hard-pressed to understand whether they’ve changed from previous. Instead of being able to provide the patient with immediate reassurance, we’re left giving him or her a copy of our films on a CD for them to take to their primary physician or the pertinent specialist to get it sorted.
The consultants evaluating the Vermont HIE recommended that it provide quality reports to support data-driven care. Our HIE doesn’t do any kind of reporting either, which to me seems a waste of a good population management tool. We’re in the midst of the worse influenza season we’ve seen in the last decade and yet can’t leverage that data for real-time reporting or surveillance. We have to wait for data to be reported to the state health department, then for it to be parsed and sent back to us in static form.
The Vermont HIE review also revealed concerns about patient matching and the function of its master patient index. We struggle with that in my state as well. Our state HIE’s program for identifying potential duplicate patients and merging them feels like it’s virtually non-existent. Since the matching algorithm appears to use address as one of its criteria, when I search for patients I find records that are clearly the same patient but are treated as unique individuals because they have different addresses, even if the rest of the demographics are the same.
We don’t tolerate that level of records duplication in my current practice, and in my former life at Big Health System, we had aggressive policies in place to identify, validate, and merge duplicate patients in our system on a regular basis. There’s no reason the HIE can’t do the same, especially with subpopulations that are known to be transient, such as college students, migrants, and homeless persons.
Another general concern around HIEs that plays out across the country is the sustainability of their funding models. Many are heavily subsidized with state funds and others are cobbled together with a variety of funding sources.
I worked with a practice recently whose HIE is struggling with funding. Practices are either required to do a full integration with the HIE at a cost of more than $40K and then pay a couple hundred dollars per provider per year to stay connected, or if they don’t want to do a full integration, they can pay a steep annual fee for providers to have web portal access. My client’s practice has a residency program with many rotating providers along with a number of locum tenens providers who fill in at their rural clinics. The fee for portal access is strictly per provider, with no regard to resident, full-time, or part-time status. For residents who are only in clinic for a couple of half days a week during a four- to six-month rotation, it’s too costly. For part-time physicians and those who are functioning in a job share situation, it’s not cost effective. We attempted to negotiate a break with the HIE, but were unsuccessful.
In my own practice, where I’m surrounded by Epic hospitals, I’m waiting for the advertised Share Everywhere functionality to start making an appearance. Although it was to be included in their November release for MyChart, I haven’t been inundated with patients whipping out their phones to give me access codes so I can see their records and send a note pack to their Epic-based care team. I’d be interested to hear from anyone who has seen it in the wild or used it to access patient information.
How satisfied are you with your HIE options? Email me.
Email Dr. Jayne.