I’ve written previously about the idea of skinny data in healthcare instead of big data. It’s an important concept that I think many are putting into practice. Today in a LinkedIn thread discussing our post on “What If EMR Interoperability Was Mandatory?” Rolando Merino, MD suggested what I think could be called skinny healthcare interoperability.
If you are convinced that your idea is good, and everyone you pitch it to – resists; then re-evaluate your idea. I think that there must be bit of a danger to the idea. We need the market to provide incentives to developers and users. Without it, there will be no innovation. What would be the advantage for a big insurance company to have all their beneficiaries records to the next biggest competitor? To require things that are not meaningful or useful from people that are used to being very meaningful and useful (healthcare providers and software developers) causes stifling of the creative process. For example: software that saves lives (like APACHE in the ICU?) every time, all the time, that does not even need any interconnectivity at all, would be out. I think that careful evidence based recommendations of what information actually saves lives, need to be made.
Trust me, there is no way anyone in health care can ignore such evidence if you can get it. Exchange for the good of exchange, and taking a Fabian approach to wait and see, is a waste of resources and intellectual power. Study it carefully, find the evidence, make your recommendations to save lives, and you will be a medical scientist / informaticist. Everyone is resisting for a good reason. You are talking of Terabytes of very complicated data that needs superbly trained individuals to generate, store, distribute, and interpret it – sometimes emergently at the point of care. There is just not enough brain power for all that, there are barely enough trained professionals to do what we do now, and we are doing pretty well. Find out who needs what, where, when, and why (exactly).
In medicine we have learned our lessons well. We follow the scientific process. And, in America, we like a little competition to determine what is best too. Determining good is like saying that a runner that runs this fast, this distance, is good (because we said so), to determine the best, however we have a race, and we determine the winner that we measured and witnessed. That is the evidence: recorded time and distance. Just requiring everything, all the time, everywhere, because it is a good idea, just does not cut it.
I love the idea that evidence that something saves lives, people can’t ignore it. This is a powerful concept that should be considered more fully in any healthcare application. It’s a hard thing to prove, but he’s right that if you do so it’s impossible to ignore.
One thing I don’t think Dr. Merino addresses is that healthcare interoperability isn’t just about saving lives. In many cases interoperability is about lowering costs. In fact, in many cases it’s getting information somewhere faster. In these cases, there’s no risk to care, but there can be a decrease in cost. Not every change in healthcare has to save more lives. If it maintains the quality of the healthcare at a lower cost, that should be embraced as well.
The problem is that the lower costs are for the healthcare system as a whole and for the patients in that system. That doesn’t provide much motivation for the institutions that will lose money.
I do think it’s worth considering whether we’re being too ambitious in our attempts at healthcare data interoperability. Could we focus on a small subset of healthcare data sharing that would lower costs and/or save lives? I like the idea of skinny interoperability. I’ve long said if we could just connect the end points, we’d have made a lot of progress towards intereroperability of healthcare data. We could work on the standards for all the data and the quality of the data later. Right now, I’d just love to see all the healthcare organization connections made.