The poem: Well, it's not it's not the usual doggerel you see with this sort of thing. It's a quatrain…
Curbside Consult with Dr. Jayne 1/17/22
I’ve done several projects in the last couple of years that involve health IT interoperability. Each has been challenging in its own way. There are varying state requirements for data exchange and those have been a factor in some of the projects, with lots of extra time and effort spent trying to obtain patient consent when an opt-in strategy is in play. There are also plenty of requirements about protecting information that has been identified as sensitive, including that related to mental health, reproductive health, and the care of minors. Given all those considerations, for most of the projects my consulting efforts have probably been 80% focused on the operational and governance aspects as opposed to the technical ones.
Not that I’m a stranger to the tech piece. I started working with my first health information exchange in the early 2000s and those days certainly were an adventure. We were using it to share records within our own organization due lack of institutional support for an enterprise EHR. Since a given patient might have three or four charts on the system depending on where they sought care, we were using the HIE to try to create some semblance of a comprehensive patient record. It wasn’t elegant, but it got the job done, and we managed to reduce some duplications and identify some controlled substance reconciliation issues along the way.
Fast forward. Although the information superhighway may have been smoothed with the technology equivalent of a new coat of asphalt, there are still some steep grades and dangerous curves. There is a tremendous amount of trust that when vendors say their solutions are interoperable that they truly are. However, as with nearly everything in the healthcare information technology world, the devil is in the details. A lot of organizations have consolidated their enterprise purchases around a handful of vendors under the assumption that such decisions would bring greater interoperability and easier data sharing. There is quite a bit of variation though as organizations might not be on the same versions of a given platform.
There is also a lack of attention to the operational differences between organizations that might choose not to share certain types of data for a variety of reasons. Business goals are high on this list – reducing patient leakage, trying to consolidate all of a patient’s care at a single health system, preserving high-margin service lines, and more. Often these issues don’t become hot topics until interoperability projects are well underway and they can essentially derail even the most well-planned technical project.
A recent study published in the Journal of the American Medical Informatics Association looked at the interoperability limitations that are found even when organizations have the same EHR vendor. Although overall data exchange is somewhat easier, there are still struggles with data normalization and reconciliation. The authors looked at nearly 70 oncology sites that were using one of five EHRs and calculated interoperability scores for sharing with the same EHR as well as scores for sharing with a different EHR. They included 12 specific data elements, equally split between medications and laboratory tests, which are standardized within oncology practice.
Not surprisingly, same vendor sharing had stronger interoperability scores than sharing between different vendors. However, the results should be enlightening for anyone who hopes to do these types of projects. The authors noted that, “Reliable interoperability requires institutions to map their data to the same standards and ensure that mapping practices are consistent across institutions.” They also noted the importance of looking at potential interoperability of specific data elements. For example, there may be different levels of interoperability when looking at medications as compared to lab results or imaging studies. Even within those categories, organizations need to look at how interoperability looks for common medications and laboratories versus less common or rare examples.
Although some might think that greater standards for interoperability measures might be the answer, the researchers are concerned that this might lead to vendors and their clients focusing their attention on those elements that are being monitored rather than the overall picture. They noted, “it will be important to ensure that certification does not replace poor interoperability with poor interoperability except for a few chosen data elements.” We certainly saw this type of behavior in the Meaningful Use era when there was a tremendous degree of focus on checking the boxes even if it was at the expense of quality patient care and user satisfaction.
I’d like to see a similar study performed looking at primary care interoperability as opposed to a subspecialty such as oncology. Primary care is the core of healthcare and where the greatest exists for interoperability so that we can use existing data, avoid duplicate and wasteful studies, manage overlapping medications, and provide a comprehensive plan of care for individual patients. I’d like to see how easy it really is for my Epic-using internal medicine physician to get information from the radiologist across town who is on a different instance of Epic, as well as from the surgeon who is using Cerner in the hospital and NextGen in their office, and the telehealth vendor I used when I was out of state and the county health clinic where I might have had a COVID test. Despite what integrated delivery networks are hoping for, many of us choose the best physician for our situation regardless of whose logo is on the door.
I’d also like to see how it plays out for emergency department encounters since patients don’t always get to choose which facility they’re taken to in an acute situation. In fact, COVID is running so rampant in my city right now that one municipal ambulance district is refusing to take patients anywhere but the two closest hospitals unless the patient requires specialized pediatric care. Given the time it takes to clean the vehicles after transporting COVID-positive patients and get them back into service, they’re trying to avoid long runs and decrease their turnaround time. There’s a tremendous number of patients that seek care at our city’s other major health system, which makes the need for solid interoperability even more important.
What has your experience been with interoperability, either with the same vendor or different ones? Leave a comment or email me.
Email Dr. Jayne.
I’ve personally been fairly impressed with Epic-based tools such as CareEverywhere and Healthy Planet for information sharing. However, these tools are only as strong as the network effect they have: if the other organizations are not bought in to use this technology because they are on, say, Cerner products, there are huge limitations. I have had experience with hospital partners leveraging Epic-based interoperability technology in addition to their own NextGen technology, and while we appreciated the collaboration, I recognize that it is neither productive nor sustainable to document in two different systems, even if different information is documented in each.