Let’s Talk Data
Data is poised to save us from our fallible, proned-to-bias selves, enabling increasingly valid and substantiated decisions. Healthcare is looking to data-driven decisions will cure what ails it – reducing costs, make us safer and improve the experience. The driving forces are in place. Payment reform is on the way. The data is there. Healthcare providers can look to more and more data from years of activity recorded in EMR systems. It may eventually be the ultimate competitive differentiator.
It’s a complex and arduous journey, with some very really and challenging barriers to overcome– but we’ll get there. One leg of this journey are the conversations we have on our data. Before a data-driven decision, there is, in all likelihood, a data-driven conversation. I’ve heard it said that data allows sane conversations. Listening to FiveThirtyEight’s podcast on this year’s U.S. election, I’ve heard some very sane, data-driven conversations (on a topic where sanity seems to be in short supply…)
For those not familiar, FiveThirtyEight uses statistical analysis to tell stories about politics, sports, science and health, economics and culture. It was founded by Nate Silver, author of The Signal and the Noise: Why So Many Predictions Fail But Some Don’t. He was also Fast Company’s Most Creative Person in Business 2013 (…a statistician, the most creative in business—how cool is that?) Nate gained fame by doing incredibly accurate analysis of elections. As a young statistician he saw politics as area of huge opportunity for someone like him—an area rife with opinions (based on very little data).
Listening to their conversations, here are some of the questions the FiveThirtyEight staff of ‘statistical journalists’ ask of their polling data:
- How is this metric performing against expectation?
- Does this show us anything about how we set expectations?
- Does the current value represent a shift from the last period? If so, to what can we attribute that change?
- How can we break this performance down? (Demographics, workflow etc.)
- What can this tell us about what to expect next period?
- Where in the lifecycle is the effort being measured? How does the current value reflect that?
- What does it show us that we don’t understand? If there’s something else we want to know, do we have the data?
These are just a few examples. Perhaps this can help you reflect on how you and your colleagues ask questions of your data, the quality of those conversations, and get the most out of them as you progress on your journey.
If you’d like to bring Draper & Dash into the conversation on your healthcare data, contact us at firstname.lastname@example.org