What are codes for anyway?

Feb. 22, 2017 / By Rhonda Butler

I don’t know about you guys, but around this time of year, when it seems like spring will never come again, ever, the restless imp in my head starts questioning things I normally take for granted—a “what are you doing taking up space on the planet” kind of thing. It can be both irritating and interesting. No subject is spared, including the work I do to maintain ICD-10. While it is impossible to capture a conversation with one’s own restless imp, what follows is sort of a reconstructed dialogue on codes and coding.

Restless imp: While I am on the subject [of your job]… what are ICD-10 codes doing taking up space on the planet? They seem kind of silly and pointless.

Me: Silly—I get that. They read like a strange translation from Chinese to English by way of Esperanto, and the annual code update costs the industry time and money. But pointless—no. Codes are like a combination of a SKU and Cliffs Notes, used for healthcare. Tons of healthcare transactions for payment, quality and research would be impossible without them.

RI: What do you mean “impossible”? Seems like codes are a waste of time now that we have the EHR. Why bother with the Cliffs Notes version when we have the whole medical record in electronic form?

Me: We use the Cliffs Notes version because we need the information in a form that is compatible with all the current systems and processes that use healthcare information—payment, quality, research. Coding is a necessary compromise between the differing data needs of payment, quality and research, based on the historical limitations of computers and the healthcare transactions designed around those limitations.

RI: Translate, please!

Me: We code medical records because we need coded data, and our current systems are not smart enough to work off the EHR directly. We use ICD-10 codes because our current legal and regulatory infrastructure requires them. Coding is a simplification of the medical record because the downstream systems and processes were built around that level of simplification.

RI: “Coded data” is just a fancy term for information. Anything can be data—the English language is data, pictures are data, an EKG reading is data. So why not just throw out the current system and replace it with something better?

Me: Just throw out the current system? Just!? Really, imp, don’t be a….well, imp. The world isn’t that simple. It’s like saying you can just throw out U.S. currency and make everyone switch to Bitcoin. A transactional system like healthcare is no less embedded than the U.S. currency system, and no less resistant to ideas of revolution. Our healthcare data collecting and processing system is an ecosystem, so it makes more sense to talk about evolution, not revolution. And a system that evolves must be allowed to evolve over time. This is what programs like HHS alternative payment model programs like MACRA (The Medicare Access and CHIP Reauthorization Act) are doing on the planet. They are all about testing possible alternatives to the current model, so the industry can evolve in a generally beneficial direction—more like a breeding program than natural selection, but still an evolving system.

RI: All right already, so we can’t throw out codes and we have to let the system evolve. I can see that. So coding is just a means to an end, not an end in itself?

Me: There you go again with just. Yes, coding is a means to an end—we code medical records so we have coded data for the downstream transactions of payment, quality and research. In the bad old days of paper-only records, codes felt more than that, like they were preserving something that would otherwise be lost. Now we are looking at mountains of digital information, but technologists and the industry are only beginning to explore whether unstructured data mining can take over some of the role that codes now play in so many critical healthcare data transactions. Coders are like geologists, taking core samples of the most interesting bits of the mountain. Codes allow us to do the best we can with what we have until the industry finds something that it can implement without fatal disruption of the current infrastructure. The whole deal is slow and it’s messy—but, hey, that’s life outside this bony head of ours. Change rarely conforms to our wishes for speed and efficiency.

You know, imp, you should give the industry a little credit. Healthcare is trying to do all the right things—become more data-driven, more value-driven, and more patient-centered—all at the same time. It’s a lot to take on. So patience, my dear imp. We’re working on it. Now leave me alone, for a while, I have some documents to review.

RI: You are such a drag to hang out with sometimes. Maybe I’ll look into hibernation. If bears can do it…

Me: Sweet dreams.

Rhonda Butler is a clinical research manager with 3M Health Information Systems.