Wearable medical devices: Getting into tech, talk and the ‘rehabilitation mindset’

Wearable technology may be a novel concept for some patients, but there are plenty already using fitness trackers and various “smart” watches, glasses, clothing and rings.

Analyst Market.us Scoop reported a survey that found 41% of Americans already own at least one health-related wearable device. Globally, the nation is on the low end, with India, China, the United Kingdom, Sweden, Netherlands and Germany all reporting even greater ownership rates.

Matthew Smuck, MD, is chief of physical medicine and rehabilitation (PM&R) and professor and vice-chair of orthopaedic surgery at Stanford University. He’s also founder and director of the school’s Wearable Health Lab, dedicated to studying health care technology that patients can put on as part of their treatment regimens. This article continues his discussion with Medical Economics about the lab’s research and how wearables are a growing part of medicine.

This transcript has been edited for length and clarity.

Medical Economics: At the 2025 American Academy of Physical Medicine and Rhabilitation (AAPM&R) Annual Assembly, you presented a program on applying digital health and AI when monitoring physical activity and low back pain. You described three parts, starting with a rehabilitation mindset. Can you describe what that is?

Matthew Smuck, MD: Yes. A lot of my work stems from my subspecialty focus as a PM&R physician. Human function is a quantifiable part of existence, yet we tend to measure the functional impact of disease through questionnaires and not through precise measurements. Understanding what function is and how to measure it accurately requires using an appropriate rubric, because what I mean by "function" and what an engineer developing an app or wearable device thinks of as "function" could be two totally different things. If we're not talking about the same thing or measuring the same thing, then we won't be successful.

People in my subspecialty have deep knowledge on human function, and one of the reference tools we use for the assessment of human function and the thinking around human function is a book called the International Classification of Functioning [ICF], developed out of the WHO [World Health Organization]. It essentially outlines the modern thinking of what human function is, what are the components, and how do you think about it. As a quick example: It describes two major components of human function, one being capacity, and the other being performance. Capacity is defined as what a person is capable of doing in a controlled environment. In medicine, that might be measuring somebody's tandem gait in the clinic or muscle strength in the clinic. Performance is how they physically function in the real world under real circumstances. Capacity informs performance, but the two are distinct from one another.

People with musculoskeletal problems, when they come to clinic, are concerned about the impact of their condition on their performance. Yet the only things we tend to measure are capacity measurements. So there's a mismatch between what we're doing in the clinic to assess the condition and what's actually happening to the person in the real world. If we want to do a better job of caring for people and understanding their condition and how treatments impact it, we need more accurate measurement of their performance — and that's one thing wearables can do for us. That's what I mean by having a rehab mindset: understanding what we mean by how people are functioning and why we're measuring what we measure. We need to start there to make sure we're approaching diseases with the appropriate measurements.

Medical Economics: In your presentation, you did talk about the importance of measurement in understanding a disease and communicating with patients. Can you elaborate on that point, especially about communicating with patients?

Matthew Smuck, MD: It's an interesting story, actually, because it wasn't part of my design of this sort of approach to research, and it's something I discovered along the way, if you will, and somewhat to my surprise. But I think it's a really important insight. Like I mentioned earlier, patients come to clinic concerned about their performance. If I have somebody with a back problem, they'll say things like, "Dr. Smuck, I'm just having trouble getting on the floor and playing with my kids. Is this treatment you're recommending going to help me with that?" Or, "I have trouble getting through long meetings at work." These are the kinds of concerns patients have.

My knowledge about how any treatment I prescribe for low back pain improves their function is generally based on questionnaires — the most commonly used ones are called the ODI [Oswestry Disability Index] or the Roland Morris Disability Questionnaire. I can tell people what their percent likelihood of improvement is on one of those scales, but I can't tell them it they can get on the floor and play with their kids or if they can get through a two-hour meeting at work. The only way I can answer that question is through my experience and anecdotal information about it, which is not scientific.

However, if we could start to measure people's real-world performance and use those measurements to determine clinical outcomes, then I now have insight into those things and I can start to have a discussion with somebody about the implications of their disease and the pros and cons of treatments and expectations around those treatments that are in the context of something that’s meaningful to them in their life. A lot of the frustration that exists in the current health care system around diseases where things are measured through questionnaires and other imprecise tools — and I don't want to disparage these tools too much, because they've gotten us to where we are now and they are important and there’s some utility in them — but they don't answer certain questions. Those questions that they don’t answer are pain points for our patients, and they're frustration points for us because they create this inability to communicate: the patient is asking questions we don't have answers to, and the answers we do have don't really mean anything to the patients.

Medical Economics: We've talked about measurement and communication — let's bring it back to the technology. How are wearables going to help improve that situation?

Matthew Smuck, MD: Wearables can improve that specific situation by providing continuous performance measurement. I could know what a person's day looks like, how frequently they're involved in certain activities, and how they will improve in response to treatment. For example, if a person has a herniated disc and sciatica from that herniated disc, the conventional knowledge under current medical research standards is that 90% of the time that people who have surgery get better. But "get better" is defined through these questionnaires, and the threshold for saying a person has improved is defined through a system called the minimum clinically important change [MCIC]. It's a whole bunch of statistical wrangling that tries to determine what threshold of improvement actually matters. For pain, which is one of the important measurements for sciatica, that threshold for improvement that’s been determined to be the minimum clinically important change is two points on a 10-point scale. So a person's pain might change from an eight to a six, and we call that meaningful and successful treatment, whereas the patient might say, "That doesn't really mean that much to me." Or a change in their function is based on those functional questionnaires I mentioned, and is even more nebulous than the pain change I just described and there's no way to put that into context for a patient to truly understand it. If I say your surgery has a 90% chance of success and they say, "What does success mean?" and I say, "A two-point improvement in your pain and a 15% improvement in your ODI score" — they say, "Well, that doesn't really mean much to me. Can I get on the floor and play with my kids?"

The exception that proves the rule is professional athletes. Their performance is measured on a weekly basis, and their on-field performance we know. If you are an NFL (National Football League) player and you herniate a disc and have surgery, you have a 74% success rate, not 90% like I described for the general population. But success is defined differently: It's defined as returning to play and having the same on-field performance after you return to play as you had before you were injured. That performance measurement in the real world allows for a more precise assessment of the success of treatment, and it puts things in terms that are meaningful to the patient. So wearables will do that for the average person, because as far as I can tell, there's no way for me to get all my patients to play in the NFL.

Medical Economics: You gave real-world examples — for instance, getting through a long meeting at work, or getting onto the floor to play with your kids. Can you give some other examples of treatments you've worked with, and the conditions involved?

Matthew Smuck, MD: I'll give you two examples, because they are the prototype examples we use in the lab to help understand how physical activity monitors can provide insight into common diseases. The two disease states we've studied more heavily than any others are knee osteoarthritis and lumbar spinal stenosis. I chose those two in part because they're very common, they affect a similar age group, and while they both limit mobility, they do so through different mechanisms. Knee osteoarthritis is a mechanical problem with the joint that causes inflammation and pain, while lumbar spinal stenosis is a neurologic condition that affects the nerves of the spine and causes pain and reduced mobility. So both are painful conditions that limit mobility, but through completely different mechanisms.

We've been able to find, for example, that the way these two conditions affect walking, differs. While that has been measured to some extent in gait labs, the ability to do it at scale matters here, because if we're going to use this as a precise way of measuring the impact of disease, you can't do that through a gait lab — they're big and costly, and they take a lot of time for the patients and the researchers. Being able to do the same thing through a wearable makes it scalable. And not only can you do it, you can find measurements that fit the disease process. For example, lumbar spinal stenosis gait tends to be impacted more in gait variability, which has to do with neurologic control and it is a neurologic condition. Knee osteoarthritis doesn't have that same variability impact on gait.

Medical Economics: How do you envision wearable technologies improving primary care practice for both physicians and patients?

Matthew Smuck, MD: That's a great question, and it's definitely a concern of mine. As a practicing physician I understand that the average physician is not going to nerd out on wearable information, and, in fact, I don't do that even in my own clinic. What I do have time for is a quick summary. So data from wearables, especially those that have utility in the medical world, will ultimately present itself in the medical arena much like MRIs and other imaging technologies have. I'm old enough to remember the earlier days of MRI in medicine. In the early days of MRI, the MRI reports would include all sorts of detail about the physics of how the scanner was set up, physics involved in the different sequences, and all sorts of information that radiologists were very interested in but that clinicians had no ability to read, let alone consume. It was wasted information and made it harder to find the important details. Over time, we've learned from MRIs of the spine what the common findings are within certain age ranges, what's expected and what's unexpected, and the reports have evolved to focus on the clinically meaningful, unexpected things. You get a distilled bit of information that has nothing about scanner physics, just the small amount of clinically useful information. That's what's going to happen with wearables. Once we figure out which tidbits of information are clinically meaningful, we can deliver those tidbits at the appropriate time points.

Medical Economics: What did I not ask about that you would like our audience to know?

Matthew Smuck, MD: A lot of these things are still under development. The things I'm working on in the musculoskeletal realm and neurologic realm have proven in some ways to be more challenging than I expected early on, but other parts have been easier. I think the examples of health situations where wearables are already being used really point to the fact that this is part of the future of medicine, and it's going to expand over the coming decades.

Lastly, I'll point out that some of my remarks may have sounded disparaging toward patient-reported outcomes or questionnaires. But ultimately, what makes for the best medical care is when we understand both things: when we understand objective information about a person's disease, and we understand how that person feels about their disease and their care. Having both together will drive the best systems of care. So wearables aren't going to replace what we already do, but it’s going to supplement it and make it more precise.