Free Our Minds
As a young Army Lieutenant, I remember standing in the cupola of many armored personnel carriers clutching a laminated map and compass, trying to find the right spot to breach the opposing force's minefield during training. There were many risks - "enemy" artillery, bad intel, mechanical failure - but the most pressing risk was often just navigation... in the dark, with a column of vehicles behind me, bouncing over the terrain in a tracked vehicle. I wonder how different it must be for today's Lieutenant? My experience must be primitive by comparison. Heck, the APCs may drive themselves to the staging area. Without the challenge of navigation, today's leaders must have more time for important things: what is the significance of the lack of artillery fire? Are the soldiers properly prepared? Are the explosives being kept safe until they are needed? What have we missed? What a difference!
It is the same in cars. While a pleasant voice provides turn-by-turn direction, we use the extra mental bandwidth for a host of things... to talk with our children, pay more attention to traffic safety, listen to the radio... (and hopefully not to text).
A similar transformation is underway for healthcare operators. Predictive analytics are starting to solve the small problems so humans can focus on the tough ones. The computer can prompt the operator with an as-good-or-better decision. And, it can do it faster than the human. This doesn't replace the operator. Not even close. Instead, it creates much-needed oxygen for bed managers, staff coordinators, transporters, unit admins... to focus on the 20% of decisions which cannot be automated, the complex comorbid patients, the late night change of plans, the referrals from afar, etc.
That's the potential of predictive analytics in healthcare: free our most important assets to focus on patients and the most difficult decisions. That's where their brainpower counts most.
For example, what if a bed placement analytic could prompt bed managers with the optimal placement for all pending ED and OR patients? What if it understood not just the physician preference hierarchy and bed attributes, but also considered the likely state of the hospital over the next 48 hours? What if the analytic suggested 80% of the placements in seconds rather than hours? What if staff could quickly accept or deny each suggestion and focus on the remaining 20%?
This is starting to happen. It's in the "flip phone" phase not yet the "smart phone" phase but it's happening. And, there are dozens - maybe hundreds - of such predictive analytics that may be built and deployed.
Free our minds.