Drowning in data

July 13, 2020
Jordan Rosenfeld

Medical Economics Journal, Medical Economics July 2020 Edition, Volume 97, Issue 11

How doctors can make data from wearable devices usable

The variety of devices and apps that measure health data, known as wearables, are growing in popularity with patients. Despite their ubiquity, they are technologically in their infancy, according to experts, leaving physicians to figure out what to do with the data produced. While wearables may offer patients useful ways to engage in their own health, these devices are still developing the sophistication and clinical validity to be of optimal use to physicians.

So how can physicians comb through the noise of so much data for useable information that can help their patients? Experts say it takes a combination of creativity, discernment and patience.

How valid are wearables?

Physicians should realize that health technology is still developing, says Crystal Riley, Pharm.D., a lecturer of health administration at New York University’s Robert F. Wagner Graduate School of Public Service. While there may be some quality difference between an app that a patient can download to a smart device, such as a watch or Fitbit, and a glucose monitor, very few wearable devices have been clinically validated.

“The devices, from a technological perspective, are not all created equal,” Riley says. “There haven’t been extensive studies done across all of these different wearables to make sure that the data they’re collecting (are) accurate.”

If inaccurate, she surmises, how meaningful can the data be?

Even if the data were all valid, Riley points out that for patients who wear their devices frequently or regularly, no physician would be able to sort through the amount of data that would accumulate.

Indeed, everyone is drowning in data, according to Christian Habermann, MBA, co-founder of MayaMD, an artificial intelligence (AI)-based health care company in Boston.

“Patients today have more health data at their disposal than ever before and so do doctors,” he says. “This is the information age after all, and it’s very accessible. The problem is that doctors today are already inundated and somewhat overwhelmed with their current workload, not to mention the endless hours of having to enter their patients data into EHRs.”

More data are not the solution and simply contribute to physician burnout, he says.

“Until there’s a way to really integrate data in a meaningful way, for the moment, I think it’s a really good patient empowerment tool and a mechanism for physicians to start the conversation about recognizing health patterns,” Riley says.

Snapshot of patient health

The most useful data to come out of wearables may actually be qualitative, gleaned from a conversation between physician and patient.

“Once you can figure out how to interact with this data without drowning in it, it’s worth the challenge because our patients see us three or four times a year and we make huge decisions about their (health),” says Betty Rabinowitz, M.D., FACP, chief medical officer of NextGen Healthcare, a health care technology company in New York that supports ambulatory care.

Rabinowitz says there’s a great deal of evidence that the data a patient brings from wearables can help physicians get a better snapshot of that patient’s health during a clinical visit or to the root of complex problems.

Additionally, Riley suggests that wearables also encourage patients to engage more with their own health care. “So the patient can look at their data, and the physician can ask if there are any spikes and give them markers to help evaluate their own health care.”

And, as Haberman points out, research shows engaged patients lead to better health outcomes.

These conversations around the data patients are gleaning from their own wearables can also help physicians personalize patient care, says Kyna Fong, Ph.D., CEO and co-founder of Elation Health, a health care technology company in San Francisco.

“That conversation about what (patients) are doing outside of your clinic is an important part of coming up with a treatment or care plan that works for that patient and can be really valuable for building relationships,” Fong says.

The question of integration

For wearable data to be truly useful, however, they have to go beyond the qualitative and offer actionable information that can integrate into a physician’s EHR. However, many medical groups and hospitals are reluctant to allow wearable data into their EHRs due to the lack of validity, says health care data expert, Theresa Hush, co-founder and CEO of Roji Health Intelligence, a health care technology and consulting company based in Chicago.

“We’re in the wild west of biometric data,” Hush says.

What’s more, she says, as AI-based algorithms continue to improve, the technology will begin to grow in capability — which also means growth in the amount of data.

Haberman says the data will begin to become useful, and even essential, when more sophisticated AI begins to pull trends and insights out of the streams of data.

“This is when the data can be used for real precision medicine, population health and preventative care,” Haberman says, citing the prediction and detection of most chronic diseases as one of its potentially powerful uses. “A lot of this new patient data could be pulled from wearables and biometric devices and then reside behind the scenes, if you will, in databases and warehouses that can collect and synthesize it. If certain associations can be made with the data and (that information) meets certain criteria that makes it imperative for a doctor to know, then it can be alerted in EHRs.”

Analyzing wearable data

A few vendors are already offering software and technology to help aggregate this data coming from multiple sources, such as wearables, health information exchanges and insurance claims, says Rabinowitz.

So how do physicians and practice managers decide whether or not these products are right for their practice? Rabinowitz offers four questions to ask of a software that might help a physician decide:

1 Can it ingest and interpret data from multiple sources?

2 Was the system designed for clinical use?

3 Can the system separate out inconsequential data from actionable data?

4 Do the insights integrate back into the EHR workflow?

Rabinowitz believes that without a comprehensive solution for collecting and analyzing data, physicians will eventually find that patients must use multiple wearables/apps to collect a number of different data points from blood pressure to sleep quality.

“That is absolutely time prohibitive and not scalable,” Rabinowitz says. The more comprehensive, the better.

Become part of the
design conversation

As these technologies continue to evolve, Hush feels that it’s crucial for physicians to find their way into the design conversation.

“I think the worst thing to do with physicians is tell them what you want them to do at the last stage of the process,” says Hush. “They’re really needed in mapping out the clinical pathways of patients (who) are presenting with wearable data.”

She suggests that physicians speak to administrators and IT folks to be sure they’re a part of the design conversation whenever possible. “There shouldn’t be any black boxes deciding what a patient is at risk for that physicians are not validating through their own learning and education,” Hush says.

And, either way, she says, “I definitely advise physicians to start ramping up their understanding of this technology and how it can be applied. Look beyond their own EHRs and into other ways of aggregating data and putting (that data) to use, even experimentally.”

Embrace with care

Though physicians should begin to embrace these technologies, they should do so with some care so they don’t wind up assuming responsibility for reviewing health data they did not solicit, advises Fong.

“If you don’t review all the pieces of the data or something is found later, you could potentially have liability.”

Furthermore, Fong adds, to have value, “(the data need) to have purpose, timeliness and a feedback loop where physicians can flag and correct issues. And health data systems need to start with patient benefit. When we strategically use physicians as care coordinators and data curators, and we listen to them when they say something isn’t valuable or correct — instead of dumping millions of raw records at their clinic door — that’s when data will become the most valuable for patient care.”

Hush believes that it won’t actually be up to the individual physician to sort through these individual streams of data, but rather the organizations or group practice administrators who will synthesize data and pass the information on to physicians.

“(The data) will be queued up for them by industries or population health solutions that show them the patients who appear to be at risk from a variety of factors and suggest the next step.”

In short, Haberman says, “For doctors to even think about accepting additional patient data from today’s wearables, it needs to be very concise and needed. So providers need to understand their patient population and what data from certain cohorts would really help to improve outcomes.”

download issueDownload Issue : Medical Economics July 2020 Edition

Related Content:

News | Technology