POSTED : March 2, 2015

There is an old adage that simply, yet eloquently, states, “An ounce of prevention is worth an ounce of cure.”  What does that have to do with my healthcare analytics?  It all depends on your perspective and your desired outcome.

“Predictive analytics” has become a new buzzword in healthcare.  Yet today’s fad quickly becomes yesterday’s memory; remember PHO’s and EPO’s?  Let’s face it; there is something sexy to the word “predictive.”  It implies a level of mystique, a certain magic, and it is tempting to want to know what the future holds in store!

To see how this piece of motherly applies to our current discussion; let us look at the varying types of analytics available to us in the new golden age of healthcare IT.  First off, we have retrospective analysis.  You simply look at the past and hope to learn from it.  If the results are bad, we wring our hands quietly; if the results are good, we high-five in the conference room.  Second, we have prospective analytics.  Now we are looking at the present and attempting to identify those interventions that will positively affect the future.  Finally, we have predictive analytics, we forecast the future.  We then try to affect change to influence it.

Predictive analytics is only as trustworthy as the current data you possess to form a foundation for analysis.  If your data quality is suspect or not all-inclusive, your results will be inaccurate.  This model of analysis is also highly dynamic.  Each day brings a new slice of baseline data, so your forecasts have to be run repetitively.  Also, predictive analytics is subject to an infinite number of variables.  Introduce one change, like 100% diabetic foot exam compliance; and the results vary widely.  Thus, each time you implement a new intervention; the prediction effort must occur anew.

Prospective analysis peers into the current state of a population’s health.  Outliers are identified and drilled down to the individual provider and patient.  Gaps in care are identified, or undesired provider behaviors are tagged, and interventional strategies are formulated and executed.  Based upon standards of care, we all know what that compliant diabetic looks like.  We just need to assure that they consistently receive the right care, from the right provider, in the right setting, at the right time.  Sure, we cannot predict the exact outcome for that patient; but we know that the intervention will yield positive outcomes in the future.  Going back to our diabetics; if every single one of them is receiving an annual foot exam, surely our level of lower limb amputations will decrease.

So, let us return to the wisdom of our forefathers.  Predictive analytics is identifying the pound of cure, hoping to derive the ounces of prevention that will bend the cost/quality curve.  Prospective analytics is identifying the ounces of prevention, knowing that the pound of cure will come.  So the question falls upon each of us.  Will we embrace the ounce of prevention or the pound of cure?  Your grandmother would be proud of you!

Tags: , ,

Blog Archive