All right. So an additional thought is one of the questions at the end of the meeting was, you know, why go with a clinical trial ready wearable versus a consumer-facing wearable? And there are some interesting things that look the same on the outside, but are usually different on the inside. And I think that’s important to start educating people about. So for example, both the consumer-facing and the clinical trial-facing will tell you that they’ll give you a pulse rate...
All right. So an additional thought is one of the questions at the end of the meeting was, you know, why go with a clinical trial ready wearable versus a consumer-facing wearable? And there are some interesting things that look the same on the outside, but are usually different on the inside. And I think that’s important to start educating people about. So for example, both the consumer-facing and the clinical trial-facing will tell you that they’ll give you a pulse rate. Sometimes they’ll say heart rate, but if they’ve gone through FDA, they’ll say pulse rate because, you know, you can still have a heartbeat and constriction here and cut off the pulse, right? So if it’s measured on the wrist and it’s clinical trial ready, it’s probably going to say pulse rate. And then you want to look at, are they just saying they provide pulse rate or have they gone through clinical validation and ideally the clearance of an organization like FDA? Because if they, so any consumer device can claim that they provide it. And, you know, now people have shown, for example, putting that consumer device on a potato and it shows the potato has a pulse rate of 60 to 70 beats per minute, right? So these are clearly not the quality that you want for medical. Now, some of the better-known consumer companies have done nice validation of their heart rate. But when you ask them for their raw data, they won’t give it. And one of the problems with that is most of them won’t give it, as was illustrated recently in a trial by Harvard University that JP Onella, there the PI on the trial, blogged about. He was using a very famous, most advertised brand smartwatch heart rate data and heart rate variability data, and it looked really good the first six months they downloaded, and then six months later they had 12 months of data, and they decided to not just download the most recent six months, but just to, with the same button, re-download the whole 12 months of data. To their shock, the first six months of data no longer matched, even though it was the same period of time, the same data. Like, huh? We downloaded it before. We just re-downloaded it. It should be the same. Why did it change? Well, when they dug around, it turned out that the consumer company, which is not in the business of providing data, they just happened to be, you know, allowing that at the time as a side thing, they had changed their algorithm and they no longer offered the same quality of interval data, which meant that one of the key measures JP Onella and team were using, the pulse rate variability, was now unable to be computed accurately for their study. So it literally sabotaged one of their endpoints because they were using a consumer device and the consumer company is not in their first and foremost business is selling devices, not data. So if you are doing especially a long-term study where consumer companies are constantly changing things, it’s important to do an arrangement up front with your provider of the data that they will give you the raw data or maintain, you know, provide the same exact computational biomarker over the duration of your trial.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
