Just-in-time insights at the point-of-care

I’ve been a doctor long enough that I gone from clunky looseleaf notebook charts and metal flip charts hanging on a spinning rack at the nursing station to the paperless word of the Electronic Medical Record (EMR). I can remember the time when I daydreamed about having an EMR. I thought, “…if only I had this information electronically, I could take better care of my patients.”

But the daydream became a nightmare when the EMR actually arrived. The hope and the dream of being able to make better decisions was derailed by endless screens to view, mind-numbering data entry, care-givers focused on workstations rather than on patients, and a thousand other disappointments. What happened? Why was I and so many other physicians, nurses and other providers so disappointed?

I think we thought the EMR would help drive us to better decisions. We confused the EMR with the power and insight that Artificial Intelligence, Maching Learning, Predictive Analytics, Personalized Analytics and Prescriptive Analytics could bring.

So now, nearly a decade into EMR experience, some of our systems are bringing these tools into the world of healthcare delivery. Some of the tools are coming through the EMR, but most are coming around, over, above, or ignoring the EMR. And that’s not necessarily a bad thing. The good news is that healthsystems are embracing these tools, driving them into their big data, and gleaning for new insights.

Carolinas HealthCare System is in the 3rd year of deploying predictive analytics at the point-of-care. Working with their software partner, Predixion Software, they focused on readmissions. Hospital Readmissions (Unplanned within 30 days of discharge), have been a focus for all hospitals in America, due to a penalty program launched by Medicare. Using real-time clinical and utilization data, over 200 case managers use predictive analytics at the point-of-care. From the time of admission, and every hour thereafter, the risk of readmission, personalized for that patient, their location, their care team, and hour by hour as the care and clinical condition progresses.

For the first time, analytics moved from the back office, business analysts, to the front line clinicians; and better yet, in a real-time format. Now case managers and discharge planners know who is at high risk, and the variables that drives that risk rate for that particular patient. On a daily basis, over 200 case managers utilize this information as they focus on giving the patient the best transition of care plan for their particular need.

We shared this information recently at a national healthcare executive meeting, and after the meeting a leader approached us. His question showed us that he had quickly processed the impact of what he had just seen to what might be. “Do you foresee the day when physicians will get feedback in the middle of a procedure to change course?”

Without hesitating for a moment, I answered: “Absolutely”.

Going on to explain, the ability to collect and store information is here: EMR, Enterprise Data Warehouses, Analytic Platforms.


The ability to analyze volumes of data quickly is here. Big data analytics is now a present tense phenomena.


The ability to integrate into work-flow.


Ok, here for a few, but potential for many.

Think GPS and driving through a busy city with road blocks and detours. Real-time, location based, personalized, changing directions for the driver. Your GPS tells you to avoid the traffic ahead, to turn left due to the detour, to change your speed due to the road conditions. Its just a matter of time, until Garmin or TOM-TOM are names I recognize at the bedside.

Hey, that’s a good idea. I should probably give them a call.