With permission, of course, Dr. Reed Ferber, PhD, a professor at UCalgary's Human Performance Lab, wants to “tag our athletes like wildlife and follow them out into the real world.” At first, that may sound alarming to people concerned about personal privacy, but to Ferber, who is also a professor in the Faculties of Kinesiology and Nursing as well as the director of The Running Injury Clinic, asking athletes to don wearable sensors is one of the best ways researchers can engineer solutions for health.
Ferber acknowledges that there are huge privacy concerns with collecting health data, but he believes that when it is done properly and securely, there is also huge potential for the data to benefit athletes, health-care patients – and the public.
Ferber and other researchers have already been collecting wearable sensor data from UCalgary marathon groups for the last two years. “So we have a full data set for them,” Ferber says. In 2017, they started collecting data from a new group of marathoners. This year, they began working with and collecting data from marathon training groups at The Running Room, a Canadian retail shop that holds running clinics. Ferber is also partnering with UCalgary Dinos Athletics to do the same for varsity athletes – and, ideally, every Olympic, intramural and recreational athlete that wants to participate, too.
The researchers have started their data collection efforts with athletes because permission to collect the data is only needed from the athlete and the coach – plus, athletes have the extra incentive of wanting to maximize every legitimate tool and piece of data they can to help them improve their performance.
Ferber, who is leading a new Sensor Technology in Monitoring Movement research program, wants to generate a one-of-a-kind data set comprising millions of data points from hundreds of athletes. It will be added to each year – and it will be available to students and researchers for study and analysis.
The program is part of the foundation for a new Specialization in Wearable Technology Program, which is beginning in Fall 2018 in UCalgary’s Faculty of Kinesiology. The interdisciplinary program is the first of its kind in Canada – bringing together biomechanics, data science and analytics, data visualization, knowledge translation and entrepreneurship. The first cohort of students is already recruited, and the new specialization expects to have upward of 100 students over the next decade.
“We're recruiting two types of students,” says Ferber. “Biomechanists – so grad students who understand human movement – and then we're going to train them in data science. And we want to recruit data scientists and train them in biomechanics. Then everyone gets trained in entrepreneurship and we are molding these three together.”
The program, says Ferber, is intended to expand and transform the Human Performance Lab beyond basic and applied research gathered in a laboratory setting. New and expanded research initiatives will incorporate data collected in real-world settings outside the lab using wearable sensors.
How wearable devices are solving health problems
More than 310 million wearable devices – from smartwatches to activity trackers to hearables (just what it sounds like: smart headphones) – were expected to be sold globally in 2017, generating US$30.4 billion in revenue. And that annual number of wearable devices sold is projected to grow to more than 504 million by 2021, according to research firm Gartner.
“Whether we like it or not, we already have wearable technology in our glasses, in our shoes, on our wrists, in our pockets. It's already in our shirts. Soon it will be in our underwear. We will be monitored on a daily basis in many, many ways. This will become part of the health-care system,” says Ferber.
“Let's think about it in terms of your yearly physical,” Ferber continues. “Right now, your yearly physical is a blood draw, it's a physical exam, and it's a snapshot of how you are at one point in time. Wearable technology gives us the ability to look at what is your activity level, what are your physiological measurements, like heart rate, every single day of the year. We want to push that to the forefront so that physicians are not just looking at your chem panel and your lab analysis. We want the physician to understand you as a human being and consider your activity patterns, your other physiological measures, from every day of the year.”
Ferber believes that sensor data’s mainstreaming into health care will begin with efforts to help most at-risk populations first. “It’s probably not geared toward maintaining a healthy lifestyle; It's going to be geared toward attaining a healthy lifestyle,” he says. For example, sensors could be used to monitor patients for three months after a total knee or a total hip replacement, or as patients recover from a triple bypass. And in these types of situations, when patients are eager to recover quickly, they could be more open to give wearable sensors a try.
One of the first wearable sensor companies to aim their products directly at health care is Calgary-based Orpyx, which was co-founded by UCalgary alumna Dr. Breanne Everett (MD ’09, MBA ’13). The company has developed three wearable devices to use with running or walking shoes that work in different ways: essentially, they actively monitor dangerous foot pressure conditions and/or are designed to aid gait and balance. They were created to help patients who can’t properly feel their feet – due to nerve damage resulting from diabetic foot ulcers, or loss of protective sensation, reduced balance and mobility due to multiple sclerosis. The sensors provide the wearer with feedback that, in essence, helps them feel their feet again. If diabetic foot ulcers are left untreated, they can lead to amputations.
“They are solving a real-world problem,” Ferber says of the Orpyx products. “That’s how wearables are going to influence health care. We’re going to take care of the marginalized, high-risk populations by monitoring them on a daily basis.”
Ferber also thinks wearable sensors will help health-care professionals to treat every individual as an individual – because they can provide every patient with their own personal big data set. “We need to look and identify: What is this individual’s typical pattern? How do they typically behave? And then start to identify what is atypical about them. So we don’t need much data. We need a few days’ worth of data to begin with.”
Ultimately, the objective is to predict and prevent injuries. “Arguably, most wearable technology right now does a fantastic job of telling you what you did today and what you did previous days. The devices don’t give you insight into tomorrow, and that’s exactly what we’re trying to bridge,” says Ferber. “We’re trying to train the next generation of sensor-data scientists so they can develop and do the science to tap into the predictive potential of wearable technology.”
How standardizing data benefits patients
To diagnose a patient — and potentially predict outcomes —doctors begin by taking a patient history. Some doctors are now doing their charting digitally by typing into computer terminals or by taking notes with a smart pen and a tablet. Many are still doing it by hand on paper.
Now think about doctors’ handwriting and note-taking vagaries – and imagine, for a moment, the challenges that must be involved in how this raw data is read, digitized, processed, understood, and standardized for comparison with different digital forms of patient-chart data. It’s probably the simplest example among the complex sources that make up the massive amounts of data generated in health care every day.
Perhaps even more so than in other industries, the amount of analysis – and privacy concerns – that go into making health data accurate, safe and useful is truly mindboggling. It’s something researchers at the Methods Hub in the O’Brien Institute for Public Health at the Cumming School of Medicine at UCalgary are working on – to develop the best health care standardization methods that will benefit patients and promote a healthy population.
Health data is messy. “It’s not clean really, the data is not really structured,” says Dr. Hude Quan, PhD, a professor in the Department of Community Health Sciences at the Cumming School of Medicine, who is working with the Methods Hub.
Take hypertension or high blood pressure as an example: “One option is the doctor writes on a chart, ‘This patient has high BP.’ Other doctors may write ‘140/90’ and they don’t say anything about high or low blood pressure. Those are just two examples – there are so many ways of writing this simple condition,” says Quan. “That’s why we need to develop a data cleaning methodology, and ways to manage data and convert the text into an analytical format.”
How data improves global patient care
Quan is also working on solving these kinds of problem on a global scale. As director of the World Health Organization Collaborating Centre in Classification, Terminology and Standards, he and other researchers at the O’Brien Institute are collaborating with other centres around the world and the WHO. Together, they are creating a standardized global tool called the International Classification of Diseases, which lets the world work together to consistently code, classify and track epidemics, look for health trends before they become emergencies, and better distribute health-care resources globally and locally.
The 11th edition of the International Classification of Diseases was released in June with significant upgrades from the 10th edition, which was released in 1990. It includes more user-friendly coding, allows additional combinations of codes, code mapping and more clinical detail, among other tools. The work is widely considered a new global landmark in efforts to collect, analyze and report health status.
“This version really takes advantage of digital technologies,” says Quan. “It’s more efficient and advanced and able to incorporate more details from diagnostic tools to make the classification more precise.” The system is able to track global health trends, such as causes of death for anything from Ebola to heart attacks to pneumonia – and provides the data to health-care workers around the world.
It’s just one example of the need for global standards for data collection and analysis across all kinds of different industries.
How data analysis reduces health-care costs
Quan and members of the Person to Population (P2) Research Collaborative at the Libin Cardiovascular Institute at the Cumming School of Medicine are also working with Alberta Health Services (AHS) on a pilot program to minimize the amount of low-value health-care tests and tasks that get repeated and unnecessarily used. They want to save time, health-care resources and costs.
Through data analysis, the P2 team has already demonstrated saving AHS approximately $145,000 over the course of a year with interventions and new protocols that reduced redundant cardiac risk assessment lab tests. It’s now working on a “Best Care ECG” strategy; the team wants to apply similar data-backed protocols to reduce the number of unnecessary ECGs (electrocardiography) and other tests.
They have already collected baseline data at Calgary’s Foothills Medical Centre cardiology wards where they have been able to demonstrate a 32 per cent decrease in ECGs. The team estimates if these efforts were continued for a year, it could mean more than half a million dollars in annual savings at just this one health facility. As the pilot project progresses, they want to move on to more expensive tests – and generate even more time and cost savings.
Part of the problem, says Quan, is that one doctor will prescribe a test. And then another doctor will do the same because they didn’t see the first test. So then the health system pays for two tests. In Alberta, which has one province-wide health system where health data can be centralized and connected, Quan says there is an opportunity to develop automatic data-cleaning methodologies and then harmonize the health information – while also protecting patient privacy, but in a way that a patient’s information and test results can move from one hospital to the next. It’s a long complicated journey to make that happen, but that’s why the Methods Hub and P2 teams are starting with small pilot projects.
“We need centralized data,” says Quan. “But the process is relatively slow because it is so complex. We need more collaborations between computer science, statistics and data analytics experts working together with doctors and policy makers. We call it micro informatics and we’re working on this at the Cumming School of Medicine.”
When it comes down to it, says Quan, the better the data that doctors have access to, the more precise they can be in using that data to not just diagnose patients, but support patients as they make more informed, timely, health decisions of their own. “That is really the way we can use data to move health care forward,” he says.
– – – – –
Explore our programs
Participate in a research study
– – – – –
ABOUT OUR EXPERTS
Dr. Reed Ferber, PhD, is a professor in the Faculties of Kinesiology and Nursing as well as at the Human Performance Lab at the University of Calgary. He is also director of the Running Injury Clinic, which aims to educate and develop injury prevention and rehabilitation programs for runners and walkers of all ages through world-class research and clinical practice. Read more about Reed
Dr. Hude Quan, PhD, is a professor in the Department of Community Health Sciences at the Cumming School of Medicine at the University of Calgary. He is Director of the World Health Organization Collaborating Centre in Classification, Terminology and Standards at the O’Brien Institute for Public Health; he is also the Lead for Alberta’s Strategies for Patient Oriented Research SUPPORT Unit Methods Support & Development Platform. His research interests include developing novel methods for analyzing big data and improving its quality to enable its optimal use for health research, precision medicine, disease surveillance, and health-care system performance assessment. Read more about Hude