Two Advances in Health Informatics: New Devices and Collecting SDOH Data

The future of healthcare will rely on an ever-evolving landscape of health data standards.

At the March 2021 annual meeting of the Office of the National Coordinator (ONC) for Health Information Technology, the information-focused initiative of the Department of Health and Human Services, speakers focused on how to incorporate new but important types of data into patient care.

ONC’s Advancing Standards for Precision Medicine Project has focused on two areas of standards of current importance: mobile health, sensor, and wearable data; and social determinants of health (SDOH) data. Presenters shared progress on two different use cases in these areas, walking attendees through what it takes to put a new data standard into practice for the collection of patient data outside of the clinic, for inclusion in patients’ electronic health records (EHRs).

How Mobile and Home Devices Facilitate Health Data Collection

Mobile health, which also includes the burgeoning extension of tech-powered healthcare to home settings, is already making a difference in the lives of patients, said Larry Garber, MD, medical director for informatics and associate medical director of research of Reliant Medical Group in Massachusetts, group medical practice made up of more than 2,000 people.

Dr. Garber recounted the story of a 90-year-old woman patient who had congestive heart failure: “She was going to the emergency room every one or two months for congestive heart failure exacerbations,” Dr. Garber said. “We set her up with some wireless home devices—a blood pressure monitor, a wireless scale, wireless pulse oximeter, wireless blood pressure, as well.” Notably, these devices are available for consumers to purchase online.

Each morning, when the woman got out of bed, she would step on her scale, then check her blood pressure and oxygen levels via the pulse oximeter on her finger. All of these readings were wirelessly transmitted to the internet and right into her electronic medical record at Reliant.

Clinicians were only alerted when the readings represented actionable information, such as a critically high or low value, or a significant change in the woman’s weight. In this case, the woman gained six pounds over the course of two days.

“A message showed up in my nurse’s in-basket,” Dr. Garber said. “She contacted my nurse practitioner, who contacted the patient, adjusted her diuretic, wrote a prescription, sent it to the pharmacy, did the documentation in our EHR, and did all of that within 30 minutes of that patient getting up in the morning, and stepping on her scale.”

Six months later, the patient continues to use these devices, and has not required any trips to the emergency room, Dr. Garber said.

These are far from the only devices going home with patients. Band-Aid-like devices, stuck to a patient’s chest, can listen to their lungs, tell their respiratory rate. “They can tell if you’re coughing, if you’re vomiting, whether you’re laying down or standing up, or how active you are,” Dr. Garber said. “They can sense your temperature.”

Much of this technology originated on battlefields with U.S. armed forces but has now begun making its way into broader societal use. Similar technology has been woven into T-shirts for babies that monitor the newborn’s breathing at night to reduce the incidence of sudden infant death syndrome (SIDS).

Glucometers can measure glucose. Perspiration sensors can monitor electrolytes and even blood alcohol levels. During the Covid-19 pandemic, fetal monitors were sent home with 200 expectant mothers to provide remote monitoring of fetal heart rates, Dr. Garber said.

These less-common use cases join the vast number of people carrying or wearing activity trackers such as Fitbits or Apple Watches.

Some activity trackers also can detect the stability of the wearer while they are walking, and thus offer important clues about the future risk of those wearers experiencing a debilitating fall.

“These devices generate tons of data,” Dr. Garber says. “Each sensor can generate over 100 measurements every second. For researchers, that’s great. But I’m an internist. I don’t want to see that.”

Fortunately, the device manufacturers and certain data aggregators will summarize this flood of data for physicians, potentially to be uploaded into EHRs.

The world of sleep is similarly opening up to being measured and evaluated. Sensors worn under a mattress of fewer than 15 inches can monitor sleep patterns, heart rates, and respiratory rates. “They’re all wireless, and you can buy these for $80 online,” Dr. Garber says.

Contact sensors on elderly patients’ refrigerator and cabinet doors can help give clinicians an idea of how often patients are actually eating. Motion sensors also can detect the frequency of bathroom use.

The Usefulness & Challenges of Mass Health Data Collection

Amidst all this technological change, the National Institutes of Health launched its All of Us research program in 2016, inviting one million people across the U.S. to help build one of the most diverse health databases ever built. Researchers will use the data to learn how human biology, lifestyle, and environment affect health, to generate knowledge that may find ways to treat and prevent disease.

Already, the program contains more than a quarter of a million patients’ EHR data, deidentified, in its database, as well as the activity of 8,000 Fitbit users, Dr. Garber says. By the end of 2021, the program expects double that number of Fitbit users.

The time savings for electronically reporting patient data can be considerable. “We’ve got diabetics who every week had to call in and report 21 sugar measurements plus how much insulin they were using,” Dr. Garber says. “If you can do that all electronically, it saves time for the patient, and for my nurse, who is entering all of this information manually.”

At this point, work is still involved in integrating the data flowing out of these devices with the EHR. Reliant, for example, turned to Open mHealth, an open-source software project, which has developed schemas representing the data from various consumer health devices.

With the help of the Georgia Tech Research Institute, these interfaces, known as Shimmer and Granola, were translated from the Open mHealth standard into the Fast Healthcare Interoperability Resources (FHIR) standard format, developed by the HL7 international standards-developing organization for healthcare.

In this way, clients who can talk to servers holding FHIR information can query the data and laid it into EHRs, Dr. Garber said.

“It does take a fair amount of work to set this up,” Dr. Garber says. “I do hope someday this will be a little bit more plug and play.”

In the workflow eventually set up by Reliant, patients download an app onto their mobile device, enter an access code and the last four digits of their social security number for validation in Reliant’s Epic EHR. Patients then choose their device’s apps for each type of measurement, then log into the Reliant app to authorize access.

“Every five minutes, my EHR is checking and updating the data or files silently,” Dr. Garber says. Again, as in the weight gain example, only abnormal readings generate any messages to nurses.

So far, Reliant has equipped 15 patients with these sensors, and looks forward to doing a much larger study to show benefits at a level of statistical significance, Dr. Garber says.

Collecting Data on the Social Determinants of Health

ONC has also funded demonstration projects where healthcare organizations are bringing in data from the social determinants of health (SDOH). Boston-based federally-qualified health center Fenway Health received ONC funding to collect SDOH data and put it to effective use in its EHR.

SDOH includes areas such as income, educational attainment, transportation options, employment status, food security, and housing.

ONC’s SDOH Project is intended to expand the standards and data types that can be expressed in an EHR to give a more complex picture of the patient’s reality, and to give health care providers data and tools needed to provide patients with individualized treatment and achieve better outcomes.

Project collaborators include health technology company Audacious Inquiry; University of Washington cPRO, an open-source project to computerize patient-reported outcomes; and EHR provider athenahealth.

Patients complete surveys on the cPRO platform, and then these data are sent in the FHIR data format to the athenahealth EHR, says Chris Grasso, associate vice president for informatics and data at Fenway Health.

Depending on the answers patients give, Fenway’s system sends them messages containing appropriate resources that may be helpful in overcoming their SDOH challenges, Grasso said.

The system is also able to register patients into resource activities and send them reminders via text message as appointments approach.

“Pushing the limits around technology is always fun and exciting, but we also want to understand how it works from the patient’s experience,” Grasso said.

It is important to start considering this data to be as important, in its own way, as vital signs are.

“We know housing instability, transportation needs, or employment can change,” Grasso said. “It’s so important for us to be able to ask these questions, collect these data and make them ultimately actionable.”

Because the SDOH question responses were new, EHR vendor athenahealth had to add some custom questionnaire responses into the Fenway EHR.

Fenway will continue to explore the variety of screening tools it uses. The ultimate goal is to ensure that “the recommendations that a clinical team is seeing is really based on the person in front of them, and not on a population,” Grasso said.

The sensitive nature of some of the data being collected in these projects gives rise to any number of privacy considerations. “One of the things we seeing, as far as the technology, is putting more of the processing of data actually right at the source of the data,” Dr. Garber said. “So it’s only the summarized data that leaves the home.”

Ultimately, legislation will emerge that further protects the rights of patients, as these important but sensitive bits of their life find themselves into life-saving and life-extending health systems.

Scott Mace
Scott Mace

Scott Mace has been writing about technology for 40 years, and about technology's role in healthcare for the past 20 years. He served as a senior editor at InfoWorld in the 1980s and 1990s, and as technology editor for HealthLeaders Media from 2012 to 2017.

He studied medical writing at UC San Diego Extension and won a Neal Award for his healthcare journalism work in 2015. (Twitter: @scottmace)