Foreword
In 2021, it's not difficult to keep track of your heart rate.
On June 25, Apple's Apple Watch finally got get get the NDA ECG, atrial fibrillation monitoring medical device listing approval. In fact, in 2018, Apple's Apple Watch Series 4 already had an ECG detection function. Only because the electrocardiogram belongs to the medical device function, Apple failed to pass the certification at first, and had no choice but to "lock" China.
Apple is not the only latecomer to the wearables health space in China. Foreign manufacturers have Fitbit, domestic manufacturers have Huawei, Huami. These manufacturers of smart watches, bracelets products can be heart rate, blood oxygen and other data monitoring, and even in the future may also be "not tied to the needle to measure blood sugar.
Manufacturers can be so strong, but also because of the related areas of the market can be considerable: according to market statistics, by 2025, the medical wearable devices market will be worth more than $ 48 billion.
It sounds like the market for "health bracelets" is booming. However, consumers who wear a bracelet to monitor their heart rate may not have considered the question: who owns the data?
01 Same heart rate, different data
The origin of the problem is a statistical study. The man in charge of the study was JP Onnela, an associate professor at Harvard University's School of Public **** Health.
In academia, he doesn't usually introduce consumer-grade products like the Apple Watch, but instead uses lab equipment to do his research. But he recently collaborated on a study with a hospital that used an Apple Watch to collect data.
That's what got him interested in "heart rate bracelets": everyone knows, both manufacturers and researchers, that there's a problem with the data collected by the devices. He and his team wanted to see how big the problem was.
The team collected heart rate data from the end of 2018 to September 2020, and exported the data via Apple Watch: September 2020 for the first time, and April 2021 for the second. That means that since the raw data didn't change, and the Apple Watch processed the data with no problem, the results of the data processing should have been close between the two times.
However, the results of the experiment were surprising: the overlap between the two data sets was not as high as it could have been if the original heart rate had remained unchanged. The yellow curve and the blue curve "play their own game", and you can't tell it's the same person's heart rate.
If you look at the discrete data, one set of data is more clustered, and it's kind of "connected". The other data is all over the place, "flying off the handle". Together, the two sets of data have little relevance.
According to Onnela's own blog, "the results of the two data sets are probably the most obvious representation of this type of bias".
02 What does the algorithm say about your heart rate?
Why does the output of the same heart rate, from the same Apple Watch, vary so much? The answer is simple: the algorithm.
In traditional heart rate measurements, collecting data is simple: the patient attaches electrodes, and the device exports an ECG. The result is raw, unprocessed data. There are no algorithms, no AI. There are no algorithms, no AI, and the results are what they are.
But when it comes to bracelets, the rules change: after a smart bracelet takes a measurement, it doesn't export it right away, but rather analyzes it and filters it through an algorithm. Researchers are exposed to the "optimized" data, which can be biased when compared to the actual heart rate.
Optimization alone is not the whole story. The algorithms that are analyzed are also "optimized" to be "the same day after day": in the previously mentioned study, Onnela said that wearable device algorithms are "black boxes": Device manufacturers only know how to update the algorithms regularly, but researchers have no idea how the algorithms are counted. The result is a lack of comparable outputs.
The existing results, coupled with possible concerns, have led Onnela to abandon the use of consumer-grade wearables for data collection in subsequent studies. He also implicitly said that the "black box" of algorithms is an "ongoing challenge" for researchers.
Olivia Walch of the University of Michigan was more direct: although she also studies wearables, she lets her research team use raw data directly. Because her research is on sleep monitoring, it requires long-term tracking, and trials are expensive. If you rely on the algorithm of the "smart bracelet" to output the results, then the research will have to start again because of the version change.
From Walch's point of view, even if she was comfortable with an algorithm update, she wouldn't know about it in advance: there's no reason for a company to specifically notify researchers of an algorithm change, but companies often update their algorithms proactively for product updates.
For rigorous research, the data from frequent rule changes is inherently untrustworthy. For health-monitoring apps, the Apple Watch should also provide consistently stable medical data, which it clearly does not.
03 Usage bias, the "bracelet" is not smart
In fact, the Apple Watch has been a "smart health monitoring" device since its official approval. The Apple Watch is a smart health monitoring device that has been officially approved by the government.
In September 2018, Apple announced that the Electrocardiogram (EKG) and heart rate monitoring features of the Apple Watch Series 4 had been cleared by the US Food and Drug Administration (FDA).
The FDA's wording is interesting, however, because the FDA rates new devices on three metrics: public, cleared, and approved.
Products that are publicized do not require professional review by the FDA and have the most lenient standards. Products that require approval have a not insignificant risk of use, require extensive testing and evaluation, and make the category III products that require approval just 10% of the device market.
If you put a "smart watch/bracelet" into this system, it is clear. These devices have a technical threshold, the need to gatekeepers, publicity alone is certainly not enough. But the function of "monitoring heart rate" is not deep into the treatment of disease, the use of risk is very low, the overall view, or license is more appropriate.
The positioning of the "license" also represents the dilemma of wearable monitoring devices: production is not simple, use is not reliable.
Take, for example, the ECG and atrial fibrillation monitoring functions that Apple, Huami, and other manufacturers promote to the public. The reason why manufacturers are promoting this particular area is that under the current technical conditions, "smart bracelets" can only do single-lead ECG monitoring. Compared to the clinical 12-lead, the monitoring method is relatively "rough and ready", and can not give accurate data.
Apple in the Apple Watch related features of the publicity, can only say "data for reference only". It's not like Apple is playing with consumers' expectations by telling them to "master your health" and then saying "if you want to master your health, please find your own doctor".
On top of that, wearable devices remain vague in terms of experience. "The dual attributes of "digital device" and "medical functionality" have led to consumer demand for this type of product, which is both convenient and accurate.
However, consumers who use "smart bracelets" are not professional doctors, and there are differences in their usage habits, leading to a lack of reasonable standards in the product experience: a bracelet that is adjusted too tightly, or going out for a strenuous workout, can lead to "health warnings" on the bracelet. The value of the "FYI" is not even mentioned.
04 Intelligent monitoring, the first rule
Regardless of whether it's "data filtering" or "for reference only", wearable medical devices represented by smart bracelets are facing the same problem.
The current wearable medical device industry, whether traditional medical device manufacturers or emerging digital device companies, want to grow wildly in the early stages of the industry's development, so as to dominate the market. Last year, there were 14 wearable device products approved by the FDA, and 18 devices were also certified by the FDA in China, equivalent to the total number of approved products in the previous three years, and the heat of wearable devices is thus evident. Huawei, Gore, OPPO have been down, the industry heat sharply climbed.
No matter how "digital" the wearable device is, the actual application of the classification is still "medical device". As it relates to healthcare, it is important to manage the product according to the standards of the healthcare sector. However, from the perspective of the relevant product market, whether it is data collection or actual use, wearable medical equipment, the "final interpretation" of too many have run into the hands of the enterprise.
Such a market, just rely on corporate self-discipline to develop industry standards is obviously not enough, but also need the relevant departments of the industry status quo, the introduction of specific industry standards. 2015, the FDA will wearable health equipment into the category of "general health" equipment, and formulated relevant regulations. As a comparison, at the end of 2017, the FDA released the Guiding Principles for Technical Review of Mobile Medical Device Registration to the public, however, the specific implementation of these rules still requires further refinement and clarity.
Consumers who use smart bracelets want the peace of mind that comes from monitoring their health, not the high-intensity competition of the digital industry. If companies are obsessed with updating their products and algorithms, and ignore the real needs, then they need a real "education". Because the consumer is the only one who has a say in the data that represents their health.
Source:
Source: