June 21, 2019 | Karl Poterack came to digital health through anesthesia. As a practicing anesthesiologist at Mayo Clinic in Arizona, Poterack began incorporating electronic medical records, digital health, and clinical informatics into his practice 15-20 years ago. "Eventually I qualified through just practice experience to be able to sit for the clinical informatics boards," he explains.
Now, while still a practicing anesthesiologist, Poterack spends the other half of his time in informatics administrative roles at Mayo Clinic. "I have a number of areas that I'm interested in, but one of the particular areas is the use of wearable devices," he says.
On behalf of Diagnostics World News, Mana Chandhok spoke with Poterack about the latest in digital biomarkers and digital health, and what needs to happen for wearable data to be useful in the clinic.
Editor's note: Poterack will be speaking at a shared session on Emerging Technologies at the Point-of-Care and Data-Driven Diagnostics at the upcoming Next Generation Dx Summit in Washington, DC, August 20-22. The conversation has been edited for length and clarity.
Diagnostics World News: Your career has evolved from medicine to a mix of medicine and informatics. From this vantage point, how do you see data changing medicine for physicians?
Karl Poterack: I really see this as part of a longer and broader trend of the democratization of knowledge and information. I graduated from medical school in 1985, and at that time it would have been unthinkable for a patient to come into a physician's office with an unusual disease and have a whole list of information and very pertinent questions that they had obtained from somewhere.
But now—and this isn't a new thing, this has been going on for 20 years or so—people come in with even an unusual disease and they've already researched it on the internet, and they've got a whole bunch of information, and they've got a whole list of questions. In some cases, people come to their physicians with things the physician doesn't know about the particular condition.
Now we have the ability for patients to go in and order their own blood tests. In Arizona it's legal for patients to pay for and pick out their own lab tests. If they want to follow their lipid panel, every month, without a physician's order, they can do that. And so, I think there's a broader trend here of information becoming more accessible and just that much easier to get.
What are your definitions of a digital biomarker and digital health?
Wow. Those are both really great questions. I'll start with digital health. I look at that very broadly: it's any kind of modality, any kind of space where we're doing something to diagnose, treat, or promote health that involves something that's not traditional. It could involve telehealth, collecting data remotely, wearables, or interacting over the internet, making it easier for patients to get information online. Interaction with the health system to diagnose or treat or promote health in one of these less traditional ways—either doing it at a distance or using technology to facilitate that—I think that’s digital health.
Digital biomarker is a really interesting question. A biomarker itself has a pretty broad definition, but it's typically a blood value or a genetic value, something that's a bit more substantial and has a bit more permanence to it than just a vital sign. In this case, by adding digital to that, we're identifying those things and collecting them facilitated by technology in a way that is less invasive, less time consuming, or less resource consuming for the patient.
You mentioned wearables. What is the biggest step toward wearable data being clinically useful?
I think that the biggest step is going to be the collection of enough volume of wearable data and the ability to easily take that data and link it with outcomes. There are really a lot of steps in that. Collection is the first step, but then you've got to understand the provenance of that data and make sure that it's meaningful, accurate data. It's got to be in a format that you can utilize, and it's got to be in a way where you can identify it.
Once it's collected, you've got to be able to match it up with whatever outcome of interest that you have. We've got some nice little indications that there may be some clinically useful, clinically meaningful data, but we're really waiting for the real, sound, statistically significant, ways in which we match up the data with outcomes. Then the bigger step beyond that is, if you can predict something, can you intervene? Can you actually change the outcome?
What sort of challenges will we have to overcome now? Will the biggest challenge be technical? Legal?
Though there are technical hurdles, and there are legal and regulatory hurdles, I think those things can be overcome given enough resources, and probably not tremendous amounts of resources. I think the biggest hurdle is the fact that that wearable data collection and use of that data really represents a paradigm shift. Data are being collected using devices and a whole ecosystem that is beyond the control of the physician, the hospital, the healthcare institution.
That represents a big change. As physicians and as healthcare organizations, we’re used to patients coming to us, we collect physiologic data using our medical grade equipment that's tested on a regular basis and working properly, and we collect the data in a standardized fashion and under supervision. We feel confident about the provenance of that data: how we got it, how it's stored, what it means.
It's a real paradigm shift now for somebody to hand us their own device or to simply collect data from that device. We aren't sure of even something basic like knowing that the person who says they were wearing the device was in fact the person wearing the device. There are a lot of steps that we have to get used to and find ways of dealing with. It's not our device, and not our system and that's really going to be the biggest challenge.
When do you think consumer wearables will be able to have biological or chemical analytes measured on their Apple Watch or something similar? Do people even want that?
The second part is easy; I think people do want that. I think there is a great appetite for data and information. Everybody seems to be speculating that Apple, for instance, is working on a noninvasive glucose sensor. That presumably would be on something like an Apple Watch. And I think there would be a tremendous appetite for something like that. How long is it going to be? We could have the ability to measure things like that in a matter of months or it could still be years away. I really don't have good insight, but I think it's something that people want. I really think this will be in, sometime in the next decade or two at the most. And it could be much sooner than that.