Written by: John Lynn
The following is a guest blog post by Dana Sellers, Chief Executive Officer of Encore Health Resources. Dana’s comments are in response to my post titled, “Skinny Data Solves Specific Problems While BIG DATA Looks for Unseen Problems.” For more context, also check out my post on Skinny Data in Healthcare, and my video interview with Dana Sellers.
You did a great job of nailing down the kinds of problems our industry can tackle with BIG DATA on the one hand and smart, skinny data on the other in your blog last Thursday, “Skinny Data Solves Specific Problems While BIG DATA Looks for Unseen Problems.” We here at Encore Health Resources were particularly intrigued when you asked whether skinny data would be enough for ACOs, or whether hospitals will need full enterprise data warehouses – EDWs – to meet the demands of ACOs.
I’d love to take a shot at that. As I’m sure lots of your readers know, an EDW is a collection of enterprise data based on the best guess of what an organization thinks it will need over the long run. So it’s bigger than skinny data (only what we know we need now) but smaller than Big Data (every bit of data available). So now we get to your question…do hospitals need an EDW to meet the demands of participating in an ACO?
If you’ve got one, great! In large part, we know what measures ACOs want a hospital to report. If you already have a mature, well-populated EDW — fantastic! Pull the needed data, calculate the required measures, and go for it.
If not, start with skinny data. Many organizations find that they are jumping into ACOs before they have a mature EDW. So this is a great example of where skinny data is a great idea. The concept of skinny data lets you focus on the specific data required by the ACO. Instead of spending a long time trying to gather everything you might need eventually, focus on the immediate needs: quality, readmissions, unnecessary ED visits, controlling diabetes, controlling CHF, etc. Gather that quickly, and then build to a full EDW later.
Think about a skinny data appliance. One of the problems I’m seeing across the country is that organizations are rarely talking about just one ACO. These days, it’s multiple ACOs, and each one requires a different set of metrics. I talked with an organization last week that is abandoning its current business intelligence strategy and seeking a new one because they didn’t feel the old strategy was going to be able to accommodate the explosion of measures that are required by all the ACOs and commercial contracts and Federal initiatives coming down the road. The problem is that you don’t have to just report all these measures- you actually need to perform against these measures, or you won’t be reimbursed in this new world.
One way to deal with this is to establish a sound EDW strategy but supplement it with a skinny data appliance. I doubt that’s an official term, but my mother never told me I couldn’t make up words. To me, a skinny data appliance is something that sits on top of your EDW and gives you the ability to easily extract, manipulate, report, and monitor smaller subsets of data for a special purpose. As the demands of ACOs, commercial contracts, and Federal regulations proliferate, the ability to be quick and nimble will be critical — and being nimble without an army of programmers will be important. One large organization I know estimates that the use of a smart skinny data appliance may save them several FTEs (full time equivalents) per year, just in the programming of measures.
Bottom line – I believe skinny data will support current ACO requirements. Eventually, an EDW will be useful, and skinny data is a good way to get started. Many large organizations will go the EDW route, and they will benefit from a skinny data appliance.
John, as always, I love talking with you!