APIs: A key weapon in winning VA-DoD interoperability
Chief Medical Officer
The U.S. Department of Veterans Affairs selection of the same EMR used by the Department of Defense has the potential to reduce fragmentation of patient records and enhance care. But, simply picking the same EMR will not be enough. In this multi-part series, we examine the role that robust interoperability must play in unleashing the value of EMRs and fulfilling the promise of better health for our military service members and their dependents. We also look at how these same lessons apply to civilian health systems.
“VA’s adoption of the same EHR system as DoD will ultimately result in all patient data residing in one common system and enable seamless care between the Departments without the manual and electronic exchange and reconciliation of data between two separate systems.”
— David J. Shulkin, MD U.S. Secretary of Veterans Affairs
When the U.S. Department of Veterans Affairs (VA) announced it was going to migrate to Cerner’s EMR, the same system being adopted by the Department of Defense (DoD), many assumed that this will result in seamless exchange of data that will improve care while reducing cost and hassle. Perhaps, but notice that Secretary Shulkin was careful to insert the word “ultimately” into the statement above. This is an acknowledgement that simply deploying the same EMR won’t fix the problems. In my opinion, if the VA and DoD don’t simultaneously pursue robust interoperability based on APIs, “ultimately” could easily become “never.” This is true for virtually all private health systems as well.
Active duty military, veterans and their families represent a large, diverse and sometimes highly complex patient population. The numbers are impressive. On the DoD side, there were 9.3 million active military patients in 2016, while the VA cared for 9 million veterans in 2014. This complex delivery system is comprised of the DoD, VA and importantly, private facilities and providers. With no common electronic medical record (EMR) and very limited interoperability, sharing information and coordinating care can be a nightmare for patients and providers alike. And, undoubtedly, care is less efficient and more expensive than necessary.
There’s no question an EMR strategy of “one platform to serve them all” is seductive, particularly when facing the kinds of challenges the DoD and VA deal with. Proponents of this view maintain that if all patients and providers are on the same EMR, interoperability will become a minor issue.
I used to believe this as well, but time and experience have taught me otherwise. In this multi-part series, we will take a close look at the shortcomings of the single-platform approach and why robust interoperability will be essential for success.
Using the same brand of EMR is not the same as being on the same EMR.
Let’s start with the reasonable, but incorrect, assumption that two installations of the same EMR can easily share data. Many people assume that EMRs of the same brand can share information as easily as we pass around documents or spreadsheets created by the same application. After all, it’s the same EMR program, right? Nope.
The hard truth is that every implementation of an EMR is different and even same-brand EMRs do not seamlessly connect. Visit a large health system and you will discover that, due to limits in scalability or as a result of mergers, they must deploy or maintain multiple instances of their EMR and that those different instances do not easily interoperate. This is not unique to Cerner and I want to be very clear I don’t mean to single them out. The same is true, to a substantial degree, for all EMR platforms. The current plan is for DoD and VA to be on the same instance which is good.
Of course, they will still have an enormous problem being interoperable with civilian health systems. You see, the problem is even worse when trying to share across different EMRs, which is precisely what the DoD and VA will need to do for their many patients who also receive care from non-military providers (and those patients who will still be on a DoD or VA legacy system during the transition).
In summary, choosing and deploying the same brand of EMR is not going to solve the DoD-VA interoperability problem any more than it has for other large health systems. EMRs, even those of the same brand, do not readily connect and exchange large amounts of information. Even more important, patients, whether in the DoD-VA system, the civilian health system or both, regularly cross these IT boundaries when seeking care. Their information should follow.
