A watch is no longer a watch. It is a wrist-mounted health and fitness monitor for a large and increasing population. Indeed, one in six consumers in the United States now uses a smartwatch or fitness band.
The accuracy of these devices is variable, depending on a variety of factors including model and brand, exercise intensity, type of activity and heart rate. Much of the testing on the main models has been conducted in controlled laboratory settings, and it is unclear how this data relates to real-world use.
In addition, nearly all wearables are governed by proprietary algorithms that are undisclosed and therefore unevaluable.
Take footsteps, for example, which are the most commonly captured variable in the wearable world. Most programs reward 10,000 steps a day. The evidence for the cardiovascular benefit of exactly 10,000 steps is unclear, but it is a nice round number. After all, it is a pain to remember to set different step measures for different problems: for example, 8165 steps to achieve significant weight loss when used with a diet, 6034 to achieve moderate weight loss when used with a diet, or 1000 steps on the first day postoperatively if you don’t want to be readmitted. A mere 6000 steps a day will get you 0.2 frequent flyer points with an Australian airline.
What we do know from available practical evidence is that brands tend to underestimate step count during light exercise, such as comfortable walking speeds, by at least 25 percent.
More worrying are normal variations in bodily functions that can temporarily cause high readings that may unnecessarily cause distress to the wearer. For example, heart rate may temporarily increase for a variety of reasons, such as climbing stairs, having sex, or even hiccupping. While it’s helpful to know that one’s heart rate is elevated for a sustained time period, and one should consult a doctor in that case, it can be disconcerting to see a heart rate of 180, even if it’s fleeting. Moreover, the accuracy of these devices for measuring heart rate diminishes during physical activity due to body motion and changes in pressure/contact between the skin surface area and equipment.
For new applications, the errors are of the same margins. In aqua exercise, swimming stroke measurements were observed to be inaccurate during light intensity exercise.
It’s not all bad news for gadgetry. Bits are a boon for the obsessive and those who rally with continual motivation. In healthcare, measurements are important only if they help achieve an outcome such asincreasing motivation to exercise and providing safe boundaries for those with various physical limitations. Continuous monitoring of individuals over long periods of time is new, and the benefits and pitfalls are yet to be comprehensively evaluated.
Watches can now do much more than watching. Individual health problems are unique, complex and often involve multiple morbidities, such as depression, substance abuse, obesity and cancer. Watches could assist individual patients and improve public health if they used their enormous computational power to help manage the multiple conditions most patients confront.