When it comes to business intelligence and analytics in healthcare, it seems to me that there is a significant divide between our strategy and what actually concerns those at the front lines of care. I recently attended a “Big Data” conference and heard a lot about the promise of advanced analytics as it relates to things like predicting adverse events and the potential for things like natural language processing to tap into the massive stores of clinical notes and other unstructured data. These are exciting developments, and analytics practitioners are understandably eager to delve into these areas. “Advanced analytics” and its associated technologies are “where we are going” and we need to spend time developing capabilities in those areas.
The problem is that in every organization we have been in, whether they are behind the analytics maturity curve or relatively sophisticated in the tools and data they have available, without fail we hear the same thing from at least some of the physicians, nursing unit managers, and/or administrators: “I can’t even get basic volumes”. How can that be? Most organizations we work with have had self-service business intelligence tools in place for years. Even those that haven’t have at least had reporting teams who should have been able to supply volumes reports with some level of basic filtering to everyone by now right? It seems that everyone has moved past self-service BI and on to dashboard/visualization tools that offer an engaging user experience in the form of visually appealing and easy-to-use BI applications. Certainly volumes are a part of those? And some organizations have even moved past the visualization applications and are now focusing on near-real-time clinical decision support that leverages medical device integration, predictive analytics, and machine learning. Those organizations must have checked the “everyone can get basic volumes” box right?
Unfortunately that is not often the case for two main reasons: first is that volumes (and more) are available to everyone but not everyone knows about it. We’ll address that in the next blog post (hint: it’s all about marketing). The second reason basic volume data is scarce is because the EDW/BI team moved on from the “boring stuff” too quickly.
Let’s face it, producing a grid of volumes by department or nursing unit or diagnosis code is not what little kids dream about doing when they grow up. I guess developing interactive dashboards isn’t exactly rescuing someone from a burning building either, but if you put a boring grid next to a beautiful dashboard they would at least choose the latter if forced to pick one or the other. My point is that grids of volumes are boring, and EDW/BI developers are drawn to things that are visually appealing and/or cutting-edge. But basic business intelligence, which can sometimes be as unexciting as simple grids of volumes, is just what is needed for many in your organization and it allows them to better understand their patients and run their business.
We need to make sure we raise the level of capability for everyone before we move on to “bigger and better” things. That doesn’t mean we can’t begin developing more sophisticated visualization tools, or start doing some R&D related to advanced analytics in order to be ready when the time comes, it just means we shouldn’t ignore those who still need the basics while we invest all of our energy in what’s new and exciting. So make sure you are always making the rounds with your customers to ensure nobody is being left in the dust, which ultimately hurts the reputation of the EDW/BI team as those under-served physicians and administrators can be some of the most vocal critics.
In the next post we’ll address the other reason people in your organization can’t get basic volumes data: they don’t know it’s there!