Population health as the indicator, gaps in care as the call to action

Most of you who are in the upper half of the age demographic brackets will surely remember a time when all cars came equipped with a dashboard full of gauges. I am not referring to the electronic gauges displayed graphically on an electronic console; rather I allude to true vacuum-sealed gauges, with painted numbers around the periphery, and a phosphorescent needle as the measure. For those of you under the age of thirty please feel free to Google this; honest we did not make it all up!

My favorite indicator was always the engine temperature gauge. It seemed to be the king of all the dashboard monitors; since it was relaying engine status to me continuously. The problem was that in those rare cases where the needle worked its way up to the “H,” the gauge had no way of telling me why the engine was overheating. All this mechanical reporter could do is simply state the fact that I was overheating.

It is one thing to know that your engine is overheating; it is quite a different thing to know exactly why the rise in temperature is occurring. Yes, there is not enough coolant reaching the engine; but why? Is the water pump failing? Does the radiator have a hole in it? For that matter, has one of the hoses sprung a leak or has a clamp come loose at the end of the hose?

A population health gauge tells you that something is wrong with a selected population grouping’s engine, but it fails to tell you why! Assume your diabetic population gauge indicates that A1c control is getting worse over time. You now have a population-level indicator, but no sense as to why the change! Is it that the patient churn has replaced well-controlled patients with more poorly-controlled diabetics? Are your diabetics simply failing to get an annual A1c test, and compliance is really the issue? (Note: NQF 0059 defines “poor control” as A1c over 9% or no A1c in the past year). Have your care coordinators simply dropped the ball on their outreach efforts? The population gauge was never designed to give you a reason!

The one thing we can ascertain with confidence is that somewhere we have failed our diabetic population through a gap in care. To reach the answers as to why; we have to be able to drill down through the data that fed that gauge reading. It is only when we get to the next level of granularity that we begin to glean the information necessary to determine where the gaps exist. We can look at subsets of the diabetic population to see that we have allowed these patients to become non-compliant and they are simply foregoing their annual A1c.

In this scenario, we would launch a massive outreach to patients and providers to create awareness as to the importance of the annual test. But is awareness enough? Through the next level of drill-down, we could extract the knowledge to know exactly which patients were not compliant and which providers (by peer-to-peer benchmarking) had the worse rates on non-compliance. Now we can take our knowledge and act upon it.

The care managers are given lists of individual non-compliant patients to call in for appointments and A1c tests. The medical director is given a list of providers who need additional counseling. What is the end result? We have now taken action upon our knowledge and started to move the diabetic population health needle in the right direction on our gauge!

By acting on our population health indicator through individual drill-down into patients and providers, our organization has evolved. We have gone from data to information to knowledge to action. As we evaluate the result of this transformation; we can evolve the decision support and care models to prevent a recurrence, and in doing so attain wisdom!

Old radiator hoses could be fixed with duct tape; the same cannot be said for your healthcare analytics!

Learn how to predict hospital readmission using artificial intelligence.