Physicians will have learn what the automated systems can and can't do, and recognize their own roles as the developers of the algorithms that smart systems will use, says a proponent of the Internet of Things.
Enough sensor-equipped, collision-detecting, self-braking cars will be on U.S. roads by 2022 to eliminate one third of all traffic highway deaths, the Insurance Institute of Highway Safety predicted last year.
Talk about technology bending the cost curve of healthcare. "That's with today's technology," says James Mault, MD, chief medical officer of Qualcomm Life, who highlighted this prediction at Qualcomm's annual Connect conference in San Diego, CA last month.
"That's 11,000 human beings that within the next five years will not be dying each year. What's going to happen to the trauma centers of the hospitals that are making a ton of money on motor vehicle trauma?
"Your budget is going to change because… people aren't crashing into each other."
'Unnecessary' Trauma Centers Could Harm Care Networks
Mault is an unabashed champion of the disruptive impact of technology, and his employer sponsors the Tricorder XPrize, a $10 million contest to develop a handheld device that can diagnose 13 health conditions and capture five vital signs in real time, anywhere.
Prize winners will be announced early next year.
Mault also isn't afraid to fire back at critics of some digital technology, such as American Medical Association President James Madara MD, who in June accused the healthcare technology industry of peddling too much digital snake oil.
"From ineffective electronic health records to an explosion of direct-to-consumer digital health products, to apps of mixed quality—it's the digital snake oil of the early 21st century," Madara told the AMA's House of Delegates at its annual meeting.
Ubiquitous sensors and the Internet of things may be poised to reduce the highway body count. They already enable passenger jets to fly themselves with minimal pilot intervention.
The same underlying technology will soon disrupt many other aspects of healthcare, Mault says.
"Let's talk about the intensive care unit in the operating room, where we are still practicing no differently than we did 50 years ago," he says. "We're looking at a patient and, without any real objective information, [thinking,] 'maybe he's ready to wean off the ventilator. I'll dial it down and we'll watch and see.' "
In such a scenario, it's easy to imagine a patient then breathing with difficulty, followed by physicians dialing the ventilator back up.
Hacking Healthcare Device Interoperability
"Would you consider that to be a smart ventilator, or a not-so-smart ventilator, because you've got a doctor or a nurse who is not able to see what a continuous sensor is going to be able to see and watch, and go, 'well, based on 100,000 sets of data, I know this patient is ready to wean, and here's exactly how we're going to wean this patient based on how he's doing.' All of these systems will become automated."
Failsafe: Human Beings
Of course, as the recent movie Sully demonstrated, pilots still have to take charge of a plane in an improbable crisis. Likewise, in the case of Mault's ventilator example, physicians would still intervene, but only as necessary.
"The failsafe is still going to be the human being," Mault says. "But potentially I could be in a central hall monitoring four different people getting their automated anesthesia, and then it's going to show me this one's gone a little south. I better go in there and make an assessment."
Physicians will have learn what the automated systems can and can't do, and recognize their own role as the developers of the algorithms that smart systems will use, says Mault.
"Machine learning is going to start making these algorithms smarter. When I have ventilator data from a hundred thousand patients and then a million and then ten million patients, that ventilator weaning protocol, that automated system, will get smarter and safer," he says.
Both Madara and Mault criticize the current way medicine is practiced, in that it does not rely enough on the base of available evidence.
Evidence-based Practice Centers Reduce Costs, Improve Care
Madara points to "digital so-called advancements that don't have an appropriate evidence base, or just don't work that well—or actually impede care, confuse patients and waste our time."
Mault, who will probably end up debating Madara about the pros and cons of digital health somewhere down the road, points out that current care relies too much on "episodic and sporadic" ambulatory patient data, acute-care EMRs which only look at snapshots of data, and a general continued acceptance of trial-and-error care.
Mault serves on the University of Michigan Medical School admissions committee and sees the move to sensors, asynchronous care, remote care, and exception management-based care as essential medical school curriculum going forward.
It's difficult to see how that will also be open to debate.
Scott Mace is the former senior technology editor for HealthLeaders Media. He is now the senior editor, custom content at H3.Group.