The Physician and the Fitbit

IMG_4643About six weeks ago, I lost my Fitbit Flex.  I’d been wearing it for two years, almost to the day.

This apparently makes me an anomaly. Consumer surveys suggest that many people — perhaps even most — lose interest in their devices within a few months.

I’m not surprised that many people would abandon using their tracking devices. People have a lot going on in their lives. It never seemed very plausible to me that vast numbers of “ordinary” people would adopt technology to change their health habits.

But does that mean wearables will be useless in healthcare (the organized part of the system) or “health care” (the way people take actions to improve their health or the health of someone else)?

It is really too soon to say. So much depends on whether a company like Fitbit can better understand what different types of users want and need from a device. I personally believe that the people who are most likely to benefit health-wise from the data-gathering capabilities of a device are also the ones who are most likely to be involved with healthcare professionals. People like Joan, for instance.

This is actually why I purchased a Fitbit in the first place. In the fall of 2013, it was a popular wearable device and I wanted to see whether it might be able to help address some common issues that we often help people with in primary care.

In this post, I’m going to share my own experience with my Fitbit, including how I can envision it being useful in the context of primary care. But if you are interested in wearables for older adults, I highly recommend reading this AARP report (July 2015), in which they studied the experience of 92 older adults using sleep and activity trackers. (Lots to pick apart in that report; perhaps in a future post.)

There is also some extremely interesting information on the state of the digital health industry here, summarizing Rock Health’s 2015 report on consumer adoption trends in digital health.

Could a Fitbit Help With These Two Primary Care Issues?

Here are the two primary care issues I had in mind, when I decided to try a Fitbit:

  • Complaints of sleep difficulties. This is a pain point often reported by patients to their doctors. Evaluation and management always require some sleep tracking, which historically we’ve asked people to do via paper and pen. I wanted to see how a device such as the Fitbit might help with this.
  • Trying to help patients increase their walking. This is not something that patients usually complain to doctors about. Instead, this is something that we clinicians are often interested in, because we generally propose it as part of a strategy to help patients reach their health goals, such as weight loss, improved mood, better glucose control, better balance and strength….the list of ways walking can improve health goes on and on!

For both the above issues, getting information from the patient is usually a barrier. Here’s how it often goes with sleep problems:

Patient: Doctor I’ve been having trouble sleeping. Is there anything you can give me, or what should I do?

Me: Hm. Well, tell me more about the sleep problem. Is the problem falling asleep or staying asleep? Is it every night, or just occasionally? [and so on. There is actually a long list of questions and information that we are supposed to gather, to determine the likely cause of the sleep problems, to figure out whether it’s secondary to a medical condition, to figure out if it’s a sleep-related breathing disorder, etc.  But getting data on when people go to bed, fall asleep, how long they stay asleep, and so forth is always important.]

Me: Tell you what. What we really need, in order to help you with your sleep problem, is more information about how your sleep is right now. So you’ll need to keep a sleep diary and bring it back.

Unsurprisingly, it’s often a pain for us as providers to give people a sleep diary, and they often have difficulty completing it. So we stumble forward, with clinicians trying to dig the data they need out of the patients’ memory, or worse yet, prescribing sleeping pills or some other form of management without decent data guiding us, because the data is so hard to come by.

Sleep is very important to health and sleeping pills are actually quite risky for older adults — more than most older patients and probably most doctors realize. So I’ve long been quite interested in how we can make it easier for the average clinician to more effectively help people with sleep.

How about the walking issue? There, the problem is that in so many situations, I find myself asking an older patient “How much do you walk every day? How much exercise do you get every week?” Here again, the answers are often vague or inaccurate.

I can’t blame people for this. I can tell you how often I go to a scheduled exercise event, but otherwise it’s hard for me to say how much I walk. Clearly, if sensors and devices are becoming cheaper, here is an opportunity to make the work of helping people with their health much easier. What gets measured can be managed.

How Well Did My Fitbit Work for Tracking Sleep and Exercise?

The first time I got a Fitbit I bought the One, and I promptly lost it within 3 days. I replaced it with a Flex and wore it day and night for two years. I did have to replace the band a few times, and the charging got wonky during the past several months (turns out that cleaning the Fitbit’s charging terminals with an alcohol wipe helps a lot). But I found it an easy device to wear.

Tracking sleep

A few years ago I went through a period of really bad insomnia, likely related to chronic stress and being a little too wound up by life as a working mother. This had gotten better by the time I received my Fitbit, but I still felt my sleep wasn’t ideal.

Prior to getting the Fitbit, I’d tried a few smartphone apps for sleep tracking. Most of these leverage the accelerometer in the phone, to detect your movements during the night. (Presumably you can also log your sleep manually, but there are obvious advantages in not requiring a user to perform manual data entry.) I am not sure how accurate the smartphone apps were, but the thing I really hated about them was having a phone in my bed. YUCK. (I’ve always kept my phone charging in my office at night.)

So it was nice that the Fitbit Flex did not require a mobile device anywhere nearby while sleeping, although you do either have to tell it when you are going to bed, or record the approximate “in bed” hours in the app the next day.

On the whole, I found it quite helpful to see my sleeping patterns. In general, I haven’t looked back at them long-term, to see if my sleep has gotten better over the past year or no. I mostly used it as a nudge to try to get at least seven hours most nights.

Is the Flex accurate? I never worried it about it a ton, as it didn’t seem wildly inaccurate. (Which I can’t say about another device I recently tried, which reported on several occasions that I’d slept 30 minutes for the whole night; wrong!!) I did come across this interesting HuffPo article recently, in which a sleep researcher tried several devices and compared them to a research-grade actigraphy device. He seemed to find the Flex’s performance acceptable.

However, in checking the literature today, I did come across this October 2015 research letter “Fitbit Flex: an unreliable device for longitudinal sleep measures in a non-clinical population.” (Sorry, not freely available to all.) In this letter, researchers report their experience trying to train 107 college students to use Fitbit Flexes for an IRB-approved pilot of a sleep study. The researchers found the devices to be extremely unreliable, with only 14% of devices able to record 6-7 nights of sleep. Positing user error, the researchers retrained the students and only those demonstrating correct use of the device for sleep tracking were allowed to repeat the pilot. Despite this, 70% of the devices could not capture sleep data. The researchers contacted Fitbit regarding the problem but state that the company was unresponsive and did not seem very interested in rectifying the problem.

I was surprised that the research team had such a negative experience, because the two Flexes I’ve used do gather data (yes, I replaced mine and I explain why below), and as I said, it seems roughly accurate.

Tracking steps

Although I appreciate having the ability to track my sleep, the step counter has been more important.

When I first started using the Fitbit, I was working from my home office (as I still do). Initially, I took a look at my step count and it was almost always over 10,000 steps, so I lost interest in that quite quickly.

However, the following summer we decided to move our small dog to my mother’s house (she has a yard and we don’t). My step count immediately dropped by 60%, as I discovered that without a dog obligating me to walk around the block several times daily, and to the park regularly, I was prone to just sit at my computer for large parts of the day.

(I should admit here that I am not one of those people who feels compelled to work up a sweat every day; I don’t run, I don’t go to gyms, I don’t even do yoga. This is not great for my health but, on the other hand, it makes me not terribly dissimilar from many of my middle-aged patients. I do also play Ultimate frisbee with my city’s recreational league once a week in the summer and fall, and otherwise I try to get to a dance class once a week but I don’t always make it.)

I decided my very sedentary days were problematic. You’ve surely all heard that “sitting is the new smoking.” Well here was evidence that I was doing a lot of sitting and not enough walking. I don’t tend to focus a lot on optimizing my own health, but this still seemed like an issue worth addressing.

So, since our dog has moved away, here is what I’ve done. Every evening, I take a look at the Fitbit’s step count. If it’s low, I force myself to take a 30-minute walk after my kids have gone to bed.

The step counter has also shifted the way I view most opportunities to walk. Instead of driving to my kids’ school for a meeting, I’m more likely to walk. If I’m forced to park several blocks from the desired destination, I think “well at least I’ll get some more steps.”

“But you could use your phone to do that!” This is what several of my non-wearable using friends have told me. True, but I don’t like being tethered to my phone. I do take it when I leave the house, but at home I tend to leave it parked on my desk.

Why I Bought Another Fitbit Flex

It was interesting to lose my Flex after two years. I actually went several weeks without a device, as I tried to figure out whether to buy a different one, or just do without. The lack of step count, in particular, bothered me, so I decided I had to get another device. I also missed having a record of my sleep, to my mild surprise.

I briefly considered the possibility of using a smartwatch, but dismissed it quickly. I won’t go into it now, but although I use my smartphone daily and it has some apps I consider extremely useful, I’m a bit ambivalent about my phone. Many apps don’t work well, the whole device gets stuck occasionally, it can be annoying when something updates and it works differently than before, etc. Suffice to say, I have some sympathy for those who’d prefer to use mobile devices less often rather than more often. So I will not be a smartwatch early adopter…I will probably only try it when I see people who look like me (meaning, tech open but not particularly tech enthusiastic) using them and finding them fantastic, easy, and reliable.

This meant I had to decide on a wearable. And after briefly considering the options, including a newer fancier Fitbit device, I decided to just get another Flex. The main reason is that overall I found it usable and useful, and I didn’t want to spend time and money working out the quirks of another device or discovering that another device didn’t work as well for me. Also, I like that the Flex is slim and feels unobtrusive to me, whereas some of the other devices would be like wearing a second watch. (Why would I wear a Misfit Shine on my wrist along with my own watch?)

Would I Recommend a Fitbit to a Patient?

Right now I have a small consultative practice, and the small number of patients I follow are quite elderly, and either quite physically disabled or quite cognitively disabled. So wearable activity or sleep trackers are not really an option for them. Furthermore, they have so much going on that neither the patient nor the hands-on caregivers have the bandwidth to take on a new device.

But I do find myself thinking about my previous practice at Over 60 Health Center, where I had a more “normal” primary care practice seeing people aged 60+, all of whom had multiple chronic conditions.

Having used my tracker for two years, I do think a similar device could be really helpful, either in evaluating sleep issues or in assessing and encouraging more walking.

Would the patient be likely to abandon it within a few months? Maybe. According to the AARP report, many older adults found the trackers to be hard to set up, hard to sync, and sometimes uncomfortable. If I were suggesting a tracker to my patients, it would obviously be best for me or my team to be prepared to provide some support in getting started. At a minimum, we should be ready to point people to a place where they can get such support. (Young people volunteering to help seniors with their tech might be nice.)

On the other hand, we know that older patients often take their doctors’ recommendations quite seriously. A doctor’s recommendation, and interest in the data a tracker generates, will surely have an influence on patients, especially if this leads to noticeable progress on a health issue that the patient is extremely interested in.

And would I recommend a Fitbit Flex specifically? Maybe. It’s not a perfect device, and I think the data reporting and summarizing functions as seen on the web need a lot of improvement. But my personal familiarity with the device counts for a lot.

So yes. I might suggest it, or something similar.

Ezekiel Emanuel recently told the crowd at the Connected Health Symposium “I think you can forget about wearables for the masses.” I’m not sure what to think about that particular statement, as it depends on who one has in mind when one says “masses.”

But I’m not going to forget about wearables for patients. There’s a lot to be done, to improve usability and design and usefulness. This is especially true to make them usable for older adults. But I think they have a lot of potential, and I’d be surprised if eventually we aren’t making more use of them in healthcare.

Have you tried a sleep or activity tracker? Tell me about it below. I’d especially like to know if you are using it on your own, and whether you’ve discussed the data with a clinician.

[Note: I have no financial ties with Fitbit, or with any technology company for that matter. (You can read my financial disclosure policy in full on this about page.) Although I’m occasionally given a device or service to try for free, in this case I purchased my Fitbits on my own.]


  1. Alice Hulbert MD says:

    The Fitbit has been invaluable in helping coordinate the large number of caregivers looking after an adult nonverbal autistic patient of mine. Communication between caregivers doesn’t convey nearly as much information about his physical status as reviewing the Fitbit information over a week or two- extremely helpful.

    • Leslie Kernisan, MD MPH says:

      Fascinating! What exactly do you look at, in terms of his Fitbit data? Just the steps or also the sleep? And if you are tracking sleep: who puts it into sleep mode?

  2. genie deutsch says:

    Exercise is something I do for myself not for an MD…I make it a point of walking, swimming or use exercise machines…no digital app will make me do it if I don’t choose to. I am aware it is the best thing to maintain brain health and possibly physical health. That is the only incentive that Ineed.

    I am aware if I have slept well or not…I try to do activities that contribute to it…no digital app will help me sleep better unless it involved CBT which I understand improve sleelp. I do not know of that kind of digital app.

    • Leslie Kernisan, MD MPH says:

      Agree that apps generally will not be able to make anyone do something they don’t want to. But many of us are having difficulty doing something we want to do, or feel we should do (e.g. getting enough steps in the day, in my own case). As I’ve found my device helpful to my own efforts, I do think about how it might help patients too.

      Historically, most of my patients have not been as motivated and able to execute — regarding self-healthcare behaviors — as it sounds like you are.

  3. john trenouth says:

    I too have been a FitBit user for about 3 years not. Had the One, Flex and now Charge HR. I have a few others but like the FitBit best. I have lost a few and broken even more. When broken FitBit has always been great about sending a replacement no questions asked. I appreciate that.

    I can understand the high customer churn rate. Honestly the device has no material impact on my activity. I do not do any more or any less because of it. My body provides all the feedback I need. I wear it, but can go days without looking at it. I may have used the dashboard and handful of times over the years. The email reports are nice but only in an anecdotal way.

    However where people are perhaps less in tune with their own body’s feedback, or where continuous activity monitoring provides some insight into otherwise hidden behaviour patterns, a passive tracker can be extremely useful. Even if its is highly inaccurate, unless you are training to reach particular objective goals, relative fluctuations can still reveal a great deal.

    Leslie, perhaps you should post an email address for reader to add you to their FitBit friend list — this way you can test the psychological and motivational impact a social network has, and how that impact changes as the network scales.

    • Leslie Kernisan, MD MPH says:

      Thanks for this interesting comment. Not sure I quite understood the last part: are you saying I should try to add my friends to my Fitbit service, so that I can see how the involvement of a social network affects things? I’ve wondered about this but in general err away from social media…fear of getting too many messages and notifications. Plus I don’t really like tech companies knowing a lot about my friends and social circles. (Which may be silly, as Google presumably knows quite a lot about my social patterns since I use their services a lot.)

Speak Your Mind