Interoperability Woes a Symptom of a Bigger EHR Ailment

November 10, 2014

Understanding the current lack of EHR interoperability is difficult. But perhaps Genesis 11 can help clarify the problem.

Making a diagnosis is a high-brow way of saying that we think we know what is to blame for a person's medical affliction. Ideally, when a physician assigns the blame for a problem, it is on the basis of basic medical knowledge, evidence-based guidelines, and an orderly ("scientific") thought process. In reality, the blame is often assigned incorrectly. It's not possible to know everything. Old habits may suggest a bacterial infection when it's actually viral. Anecdotal experience, while satisfying, is simply not reliable.

Physicians do not have a monopoly on laying blame. Everyone does it. The patient or a family member may blame a runny nose and cough on not wearing a jacket and getting a chill. Some blame illnesses on bad luck. Some believe that they are being punished for having gone against God.

But just because one can lay blame doesn't make their conclusions valid and it doesn't necessarily provide a framework for doing something predictably constructive. Some kind of belief underlies any attempt to lay blame, or make any other decision. The scientist and physician believe in the scientific method and also believe that adhering to the method produces better, more consistent results than alternate belief systems. Belief in folk remedies or religion is certainly valid from the perspective of the believer, but tends to lead to conclusions and actions that differ from those reached by physicians.

It's no secret that EHRs are ailing. It's the constant topic of discussion in the media and within professional groups like the AMA. One quarter of EHR users are so unhappy they are looking for a new system to replace what they have.

What's to blame for the problems and dissatisfaction? Lately, "lack of interoperability" has bubbled to the surface as the thing to blame. If it truly warrants blame, there should be an answer to the question: Why blame this above all or at all? What's the evidence that lack of interoperability is really to blame?

This question rarely gets asked, which is probably good because there really isn't a good answer. Digging deeper will reveal that lack of interoperability is a symptom of a more serious, more pervasive ailment. It is not, in and of itself, a root cause of the ailment.

Considering the problem from the perspective of Genesis 11 may aid in understanding the current lack of interoperability.

11:1- In the days of the paper chart the whole world had one language and it was the narrative progress note.

2 - As time passed, the people found the land of computers and settled there.

3 - They said to each other, "Come, let us adopt one of this land's new computer systems."

4 - The people chose systems from different vendors. Each vendor said, "We have our own ideas. We will build you a health record system that reaches to the heavens that will do everything we think you need."

5 - In doing so, they abandoned the people's common language of the narrative note and instead each began speaking their own dialect of data.

6 - The deity worshiped by computer vendors observed that, "They have confused their language and will not understand each other," and was pleased. "It is good to protect each vendor's proprietary interests."

7 - The deity sent word through his prophets that the systems should be called BHRs (Babel Health Records) - because their builders have confused the language of the whole world and precluded straightforward communication between systems.

If interoperability is really as important as some believe it is, there are only two ways to reliably achieve it:

1. A national health scheme with a single management and one system to rule them all. This would eliminate the need for interoperability.

2. A new breed of EHRs that use the narrative note (with embedded contextual metadata) as the common language. Communication between systems would take place by exchanging narrative notes that could then be used for any purpose that the recipient cared to devise. Done in the way that I have suggested, interoperability would be a direct consequence of the architecture, not something that had to be added on.