As I was getting ready to see my next patient, I began what has now become a familiar ritual for doctors and nurses everywhere, washing my hands, doing my part to not become a vector for disease. As I toweled off my chapped hands for the umpteenth time today, a thought occurred to me: how come we’re not better at preventing hospital acquired infections (HAIs) despite all this hand washing? Across the country doctors and nurses are scrubbing our hands until they are impossibly chapped, yet HAIs abound. We’re doing our part but not really making much of a dent. I wondered if there might be a way that our IT systems and the immense quantity of data we collect on each and every patient might lend a helping hand…

Current state of the art prevention for HAIs comes down to making sure that we’re completely fastidious in keeping clean (washing hands), identifying high-risk conditions (patients with indwelling catheters, recent surgery) and starting the right antibiotic at the right time. Despite this, we still have unacceptably high rates of HAIs.  The Centers for Disease Control estimates that each year over 1.7 million cases contribute to approximately 100,000 deaths costing our healthcare system an estimated $5 to $11 billion dollars.

So how might we use the mountains of data that we collect continually on all of our patients? As I thought about it, I realized that just like preventable hospital readmissions this is another example of a great use case for Predictive Analytics (PA) –  a mature BI technology that is just now finding its way into medicine. We currently approach the HAI problem by looking at it backwards. That is to say, we use the “retro-spectroscope” to see what classes of patients developed an unacceptable rate of HAIs. We then categorize these types of patients as “high risk” for acquiring HAI and set up labor and cost intensive monitoring systems to screen everyone in this category.

Yet, I and every other clinician know that not everyone in this high-risk group will develop a HAI.  This is where predictive analytics comes into play. The ability to sift through large pools of data and recognize previously unseen connections in that data allows for these previously largely undifferentiated patients to be risk-stratified. Once we know who is truly at high risk we can intelligently apply interventions to prevent disease. Will I stop washing my hands before every patient, no, but in a not too distant future, I hope that I can then use those hands to check my computer to see which of those patients I should scrub extra hard for!