Embedded Analysts: More Than Data

I have worked the Alberta Children's Hospital (ACH) for over two years, and it has been a highly rewarding experience in many aspects. That said, despite measurable progress effective use of data in healthcare remains slow on the uptake, and for reasons that may seem trivial to my fellow data-ites.

I believe Alberta Health Services (AHS) is lucky to have hundreds of motivated analysts/data scientists who love what they do, are willing to tackle complex data sets, and in some cases leverage advanced statistics/modeling to answer questions never before attempted. It's exciting stuff, but frankly it is only useful if healthcare professionals (i.e. the clients) are able to access, understand and apply this knowledge.

Embedded analysts - as the name suggests - are embedded in healthcare settings and as such, often find themselves in face-to-face meetings with doctors/nurses, demoing new software products out on the units, and more generally, providing a tangible, personable link to the expansive, yet mysterious AHS data repository. 'Mysterious' is an exaggeration of course, but only just in some cases.

Embedded analysts have been in and out of vogue since AHS had a data shop, but for much of the past 20+ years the typical service model has been a central hub of data analysts accessible by phone, or an online data request tool. Data was delivered from a far away place, from a person you did not know, and it seemed there were few opportunities to ask "dumb" questions. This is not to pick at AHS of course because in fact many organizations still function this way, but I think it is important to emphasize the expanded role embedded analysts can fill in our healthcare system.

This brings me to one of the most common questions I am asked: "how can I logon to Tableau"? To paraphrase: "how can I see the graphs you made"?

For those not up on their big data buzzwords, Tableau is a powerful visualization software which is especially valuable for operational and quality reporting sans statistical sophistication. Since this tool is required on an enterprise scale at AHS there is a server running Tableau that connects analyst generated "Tableaus" with the doctors and nurses who wish to view them.

I think many analysts would be shocked to learn how few of their clients actually view their dashboards, and not because the desire to do so is missing, but simply there is a "dumb" question, or two, in the way; simple obstacles that could be overcome with a little trust. After all, if I am able to perform a number of complex surgeries on a child, or insert a central line into the heart, I should already know how to logon to Tableau, right?

To conclude, I believe that embedded analysts play a key role in health care analytics because of their unique ability to build trust, consideration, and mutual understanding. With these elements in place we all can be confident that the hard work invested in our data products is having the intended impact.