Precision medicine is more hype than reality right now — but, at the same time, the incredible potential it holds for the future is even greater than all the buzz teases today.
That’s what I came away with from the Precision Medicine Summit in Boston this week.
Let’s look into the distant future: A patient walks into a hospital to meet with clinicians who run tests and pinpoint a biomarker for, say, Alzheimer’s. Then a gene surgeon does some on-the-spot genome editing. The patient walks out with that Alzheimer’s-free-for-life feeling.
“Primary care and genome sequencing will come to the forefront to identify which patients can benefit in a future where genome editing is widespread,” said Ross Wilson, principal investigator at the University of California Berkeley’s Institute for Quantitative Biosciences.
Just how widespread can precision medicine get? Well, Eric Dishman, who spearheads the NIH’s All of Us program said the program is starting off with the goal of attracting 1 million American participants but is already thinking about how to scale that into the billions globally.
Getting genomic data into an EHR
The grand vision is to democratize research and apply more brainpower per problem to the most vexing medical issues.
Before we can get there, though, a lot has to happen to hammer out data gathering and sharing capabilities, retool the healthcare system so it’s much more adaptable to change and ultimately modernize IT infrastructure to support precision medicine and all the data that entails.
Robert Green, MD, a medical geneticist and physician-scientist at Brigham and Women’s Hospital and Harvard Medical School predicted skirmishes, if not all-out war, over genetic and genomic screening practices: with clinicians and patients on one side, calling for as much information as they can possibly get, versus public health officials and others, warning about the unforeseeable consequences of over-screening.
Among the reasons that people are refusing to participate in genetic testing is fear of discrimination by life, disability or long-term care insurance companies, according to Mayo Clinic Department of Laboratory Medicine and Pathology attorney Sharon Zehe. She added that the whole scenario puts providers in an awkward position because even among patients who are willing to undergo screening, many don’t want that data to live in their medical records.
Not that getting genetic data into a medical record is exactly easy. One of the fascinating accounts at the conference was Washington University genetics fellow and bioinformaticist Nephi Walton explaining how it took nine months working with Epic to include genetic results into the EHR. “You can make a human in that time,” Walton said to laughter from the audience as he turned to a slide with a baby picture.
Precision medicine architecture emerging
While it’s true that today’s EHRs and IT infrastructure are not ready for the big data needs of precision medicine — and I saw that the same thing is true about population health last month — at least one architecture is emerging.
Indeed, the strategy of harnessing FHIR standards, with mobile phones as middleware and a common data repository outside the EHR, is an apt way to manage the demands of precision medicine, said John Halamka, MD, CIO of Beth Israel Deaconess Medical Center. The idea is to maximize what patients already have in their homes.
That approach also gives patients more control over who can and cannot share their data, including researchers, which India Hook-Barnard, director of strategy and associate director of precision medicine at University of California, San Francisco, said it is both the right thing to do and sound science.
But even the architecture Halamka described and giving patients more control over data sharing will not conquer all precision medicine challenges, of course. Michael Dulin, MD, director of the academy for population health innovation at the University of North Carolina Charlotte said simply dumping a whole heap of genomic data on top of the already broken healthcare system, replete with huge variances and medical errors, may actually yield worse outcomes than we have today.
“We have to use technology, we need AI,” Dulin said. “We cannot do this without it.”
Walton noted that first we need simple artificial intelligence and machine learning algorithms just to clean up healthcare’s messy data so it’s suitable for more sophisticated AI tools.
Becoming ‘precision health’
What was perhaps the boldest prediction to emerge from the conference came from Bryce Olsen, global strategist for Intel’s Health and Life Sciences unit: Patients will start asking for precision medicine in the second half of 2017 – though many of them will not even realize what they’re requesting.
“Patients are going to demand that doctors get a better understanding of underlying drivers of disease and defects in their tumor. We’re going to see this for cancer first,” Olsen said. “Doctors that don’t have good answers will see patients bounce.”
I’ll add one more to the mix: Precision medicine, in both term and concept, will be supplanted by the phrase precision health – and, yes, this is distinct from how I’m seeing digital health become digital medicine.
“Precision health,” said Megan Mahoney, chief of primary care in Stanford’s population health division, “is a fundamental shift to a more proactive and personalized approach that empowers people to live healthy lives.”