Topol, the founder and director of the Scripps Research Translational Institute, is often considered one of the leading thinkers on how technology will impact the practice and business of medicine.
Harnessed correctly with deep learning, Topol believes the volume of healthcare data being created on a daily basis through technologies like genetic sequencing and biosensors could usher in an era of “Deep Medicine,” delivering healthcare in a way that’s more personalized, more safe and more empathetic.
During a presentation at Nvidia’s GTC 2019 Conference in San Jose, Topol sketched out a vision of a world where this multi-modal data is collected and assessed from “the pre-womb to the tomb” and run through clinically validated algorithms that can reveal personalized insights and aid in more accurate diagnoses.
He ran through some of the exciting research utilizing AI algorithms to diagnose high potassium using a wearable, screen for diabetic retinopathy, as well as his own effort – detailed in a New York Times essay – to use an algorithm to guide his daily diet.
Ultimately though, these new technologies must be to the same kinds of standards and prove our their clinical effectiveness in real-world clinical care settings.
“One of the important points I want to underscore is how there is no exceptionalism for AI in medicine,” he said in his speech.
“When we put them into the implementation phase we need to have very careful surveillance because there’s different venues, different people, different everything and we need to keep careful eye on that.”
In the near term, Topol said deep learning will enable “keyboard liberation” for doctors, freeing them up to do more of the human work of healthcare. A research effort led by Topol for NHS England to help plan for the future healthcare system, found that one minute of keyboard liberation per patient could lead to massive increases in consultation and outpatient clinic time.
“Right now we have human scribes, we’re going pre-Gutenberg for our solution,” Topol said in an interview at the conference. “We have machines that can do that better than human scribes and they should.”
There’s plenty of room for the transition to deep medicine to go the wrong way, Topol warned. Exponential improvements in productivity – combined with existing misaligned financial incentives – could lead to healthcare delivery more comparable to a factory system than a deep human-centered experience.
For Topol, a prior cautionary tale on the current pathway toward healthcare’s de-personalization has been the implementation of EHR systems, and its corresponding effects on patients and physician burnout
“Doctors didn’t stand up then, they just rolled over and played dead while this abject failure was taking over and encroaching on the doctor-patient relationship,” Topol said.
“The problem is too many people think that’s what digital medicine is. But that’s actually what digital medicine isn’t. It has to be patient centered, which means if you have electronic records, it’s owned by the patient and it’s for the patient – not for billing purposes.”
Topol positioned the idea of clinician activism as key to keep things moving in the right direction and act as a bulwark against the kind of business incentives that have led to some of the current issues in American healthcare.
He pointed to the influx of a younger and more diverse workforce entering healthcare as encouraging progress, evidenced by the recent mobilization of doctors against the National Rifle Association’s message for clinicians to “stay in their lane.”
Privacy is another concern and many experts agree that HIPAA is underequipped to handle the new types of data being created and utilized in new ways.
Serious investment and thought by policymakers (along with severe legal consequences) is a piece of the answer, Topol said. But key to the solution is moving toward a truly patient-centered system where an individual owns and controls their own personal health information, instead of it sitting on disparate servers and vulnerable to cyber-attack.
That idea intersects with one the central healthcare debates happening in the public forum: the move towards a socialized, single-payer system.
“I don’t care if you want to call it Medicare-for-all, but we need to say every deserves healthcare as a civil right, just as every citizen deserves to own their medical data,” Topol said.
Picture: Kevin Truong, Getty Images