How close are we to precision medicine being commonplace, widespread – perhaps even the standard of care? How near are we to an environment where genomic advances are routinely part of the equation when it comes to treating patients?
"The changes are happening so quickly," said India Hook-Barnard, associate director of precision medicine at University of California San Francisco. "I think the world is going to be a very different place in five years."
Suddenly we seem to be nearing a critical mass where precision medicine is no longer just a buzzword but is actually widely integrated into the way physicians deliver care.
"In my opinion, we've well gone over critical mass," said Joel Diamond, MD, co-founder and chief medical officer of Allscripts subsidiary 2bPrecise.
"I'll give you an example. I was seeing patients in my own practice a few days ago, and a young woman came in for a regular physical exam. She was a nurse. I asked her about pap smears, mammograms – normal screenings.
Since her mother and two aunts all died of breast cancer and another relative passed away from ovarian cancer, Diamond’s patient was naturally concerned about getting BRCA tested and a bit nervous. Knowing what we know today, it would be unconscionable for a physician to not counsel such a patient to undergo genetic testing, said Diamond.
"The chance of that woman not having a BRCA mutation is almost zero," he said. "And if she has that mutation – and she has a daughter and a sister – she has an upwards of 80 percent chance of having breast cancer and a 40 percent chance of having ovarian cancer."
The challenge, however, is that such a test would not necessarily be prescribed by a physician who's not up-to-speed with the new wave of genomic insights. And even if it was, there's no guarantee that the test is easy to order, and there's little chance the insights it revealed – most likely tucked away in the patient's electronic health record as a PDF – would be available to other clinicians along the continuum of care. And there really is no methodology for screening genetic tests.
Long story short: Despite the massive leaps forward in knowledge since the completion of the Human Genome Project in 2003, the healthcare industry has yet to grapple with the big changes to people, process and technology needed to truly capitalize on it.
From EHRs to interoperability to clinical decision support, healthcare as an industry does not have the technological infrastructure for precision medicine in place yet.
But that's all about to change — and quickly.
Just what is precision medicine, anyway?
Before exploring this topic further, it's worthwhile to settle on a definition of just what precision medicine is. As defined by the National Institutes of Health, it's "an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person."
"It's about genomics, but it's certainly not only genomics," said India Hook-Barnard, director of research strategy and associate director of precision medicine at University of California San Francisco.
"To my mind, it's very much about the data," added Hook-Barnard. "A variety of different kinds of data that impacts health, and then being able to analyze and use that. It's genomics, but it's also your microbiome, your metabolome, the environment and socioeconomic factors. Being able to measure how often you exercise, how do you sleep, the quality of your water and your air. Those things can have huge impacts on your health and we all know that.
The interplay of all that information with individual’s genomics can, in turn, enables researchers and clinicians to gain a better understanding of what therapies or interventions might be most appropriate for a particular individual.
“That's where we want to try to get," she said.
EHRs are not enough
Healthcare is moving toward a precision medicine future state faster than many might think. According to HIMSS Analytics, "precision medicine usage has tripled in the last five years." More than 25 percent of U.S. Hospitals have recently purchased precision medicine technology and 22 percent have gone live with an installation.
The biggest growth area for precision medicine has been academic and specialty hospitals, which together represent 18 percent of inpatient sites, HIMSS Analytics said more than half of such facilities have precision medicine technologies in place.
That's unsurprising, given that academic medical centers have both the mission and vision to innovate on leading-edge treatments, and the resources to build out the IT infrastructure necessary to handle the demands of precision medicine. Because EHRs are not enough.
"Not only are they not enough but they shouldn't be enough," said Diamond. "EHR companies are very big on saying they're compliant with Smart on FHIR technologies. And that's all great. I applaud that, it's wonderful, at least there's a standard for getting the information in there.”
Or put another way: The good thing about EHRs is you can put just about anything in there but the bad thing about EHRs is you can put just about anything in there.
"The fact that you could insert genetic info in an EHR doesn't mean it's in a form that's acceptable and usable for doctors at the point of care,” Diamond said. “I don't think it should be in the EHR.”
When a genetic panel, exome or a full genome is ordered, for instance, the return will likely include hundreds or maybe even thousands of variants of unknown significance.
“Science has not yet determined whether they're pathological or not. To store those in the EHR, you'd have to be crazy,” Diamond said. “I would not want to store something that today might be pathological, and tomorrow a paper comes out and says that it is, and then I'm responsible for digging it up and notifying the patient. There's some danger in the EHRs being the repositories for that kind of information."
From her perspective, Hook-Barnard sees EHRs paying lip service to the new avenues of precision medicine and working on making it happen, by and large, as glorified billing systems.
On the other hand, "there's a lot of innovation in the consumer space and in a lot of cases moving more rapidly," she said. "If EHR vendors don't make some changes, they may get bypassed."
Precision medicine architecture emerging
While many others in the health IT industry agree that electronic health records and today’s existing technologies are not robust enough for the large data sets and analytics that precision medicine will require, Beth Israel Deaconess Medical Center CIO John Halmka, MD, said new innovations will happen outside the EHR.
“We will get beyond MU, the last 8 yrs of clinician burden, and connect an ecosystem around our EHRs,” Halamka said. “We can turn EHRs into your friend rather than your foe.”
Halamka said an architecture is already emerging to advance precision medicine: Apple’s HealthKit running on a smartphone that uses FHIR APIs to link to EHRs.
Beth Israel’s BIDMC@Home is one such example. Describing the app as a Facebook-like next-generation portal, Halamka said it includes a patient care plan with information about diet, medications and exercise.
“The app takes data showing in the EHR and offers reminders on variance, such as ‘your weight went up 7 pounds and you’re a congestive heart failure patient, you’re accumulating fluid and could end up in the ER soon,’” Halamka said.
Data, data everywhere
Expect to see more apps and tools like that in the years ahead at providers large and small. But in the meantime, some big hurdles remain, beyond technology challenges.
Most especially, there's a data problem. To wit: How much is there, where is it, who has access to it, how does it move from one clinical setting to another?
"All of these tests are being ordered Wild West-style. It's unusual to have policies around what they're ordering, how the information is recorded, where it's going,” Diamond said. “It's multiple labs, multiple different forms – PDFs, Word docs, etc. – sitting in proprietary document folders in the EMR. Because people haven't put any methodology around it.”
Such a methodology that includes understanding who’s ordering which tests, and where, and knowing where it all exists, would be a logical first step to putting some efficiency in the process, Diamond noted. Add to that the profusion of patient-generated information that factors in when tailoring treatments and the problem becomes that much more acute.
"People are increasingly going to have their own data that they're going to want to use and bring to their physician," said Hook-Barnard.
There's IoT data, Fitbit insights, information from your iPhone or other mobile apps, sequencing and genomic results from consumer-focused tests such as 23andMe.
The challenges involve "being able to connect the electronic health records, and thinking about how to get the data in and out of the EHR,” Hook-Barnard said. “How does that data get into the patient's record? And then how is that going to be used so the physician can access it?"
Make precision med a focus of CME
Doctors have some work to do on this front too.
With the field of genomics progressing exponentially, many clinicians – a good number of whom probably learned little if anything about precision medicine techniques – will need to get up to speed quickly (and stay up to speed) as the scientific breakthroughs keep coming.
Diamond said its time for physicians to start educating themselves. That means knowing what genetic variants to test for and when, whether the results of those tests are actionable, sometimes even what a test looks like and how it's performed.
"I'll bet if I asked 80 percent of doctors if they know about BRCA testing they would – but then if I asked, 'How do you order the test?' they wouldn't," said Diamond. "'Do I order it from Quest? Do I do a mouth swab? Do I use blood? I think I'll just refer them to a genetic counselor' – which is probably the wrong answer."
He suggested, instead, that doctors acquaint themselves with the website of the American College of Medical Genetics and Genomics, which publishes a list of the genetic variants that are clinically actionable.
"At the very least they can start with that list of 80 or 85 or whatever they're up to now," he said. But beyond that learning more about pharmacogenomics, for instance, "is an easy one. It doesn't require people to go back and revisit medical school criteria and start learning all these complicated nomenclature.
"We all know how to look at drugs and find out if people are allergic to those drugs or there's a drug-drug interaction, or drug-condition interaction," he said. "It's not a giant reach for us to understand that some of those things also have a genetic determinant to them and we should start looking at that."
Patients need education too
Patients and consumers also have a steep learning curve ahead, said Hook-Barnard, who sees a need in the near term, for professionals help them separate the proverbial wheat from the chaff.
"I think there will be a role in helping patients understand the data that they have, what's meaningful and what's not," she said. "In the consumer space, which of those things are real, and useful, and which of them are junk science?
"In some of those cases, it's going to be a real challenge," she added. "When is a patient bringing in information that really does need to be considered and you can help them do that, versus when are you very legitimately able to say, 'We're not ready to use this effectively yet,' or 'This company is not doing anybody any favors' and it's more of a snake oil situation. I think there's going to be some challenges for physicians during that transition."
Beyond the risks of fly-by-night companies peddling tests that might only muddy the waters, it's worth remembering that not every potential genetic mutation should necessarily be tested for – and that, when those tests are performed, the resulting information can sometimes be complicated to process.
"People are feeling empowered: Knowledge is power," said Jessica Langbaum, principal scientist at Phoenix-based Banner Alzheimer's Institute. "Individuals are not always aware of the considerations they should keep in mind before deciding to learn their genetic risks for certain diseases, such as Alzheimer's.”
Langbaum pointed to emotional considerations, including how someone might react to receiving this kind of information.
"People should take a step back and ask, by having this information, what would it do for them – what would it change in their life, would it help them in any way? Are they actionable results?” she added.
Physicians will also have to be ready to help patients interpret the result of these tests as more and more people make use of direct-to-consumer testing, she said. "Even though the companies are trying to make the information clear to consumers, people are still confused about their risks for actually developing the disease."
And in some cases, it might be better not to know. For instance, "we don't advocate for people learning their genetic risk of Alzheimer's disease, because outside of perhaps participating in a clinical trial, there's nothing we can tell them to do differently," said Langbaum. "There's no medication they can take. We'd only suggest they learn that information if it's something that would help them make a choice about participating in a research study or something along those lines."
That said, on the larger question about precision medicine, it's here and gaining steam every day.
"It's happening," said Hook-Barnard. "Now it's a question of how do we do it well, and have patients engaged in it as partners. Those technologies – a lot of them are already at work in other industries and now it's just a matter of having them a part of healthcare and how we can use them to improve health."
Email the writer: email@example.com