In putting together the article above, Christina Farr was gracious enough to listen to my theory about Apple’s not-so-secret plan that explains why there’s suddenly an ECG in the Apple Watch 4. What surprised me the most is that everyone focused on it as an ECG, including all of the tech and digital health press, and missed Apple’s brilliant, extremely broad patent from 2014.
If you go looking, the name for what Apple patented (again, brilliantly) is a way of measuring something called pulse transit time (PTT) from wrist-worn devices alone. PTT is a proxy for cardiovascular health, in the same way measuring blood pressure with a cuff is a proxy for cardiovascular health. What you’ll find is that as of a little more than a decade ago, pulse transit time in laboratory settings was considered unreliable as a proxy for a blood pressure cuff. But then… that wasn’t being done by Apple.
Here’s how I think this will go: first, people will use the Apple watch ECG. There may even be a few cool ECG apps and stories about the ECG feature saving lives. Then, Apple will announce that the Apple Watch 5 (or the next Watch OS in 2019) includes PTT. People will start asking whether it really is a reliable proxy for a blood pressure cuff, and not long after, someone will point out that actually blood pressure measured via cuff is kind of not all that reliable anyway, because it’s just the easiest proxy measure we’ve got. And though the Apple Watch’s PTT will correspond pretty nicely to a blood pressure cuff, there’s no reason why PTT couldn’t just replace blood pressure as a more reliable measure of cardiovascular health.
Why can Apple do this where others have failed? For the same reason that the iPhone’s camera has taken better pictures than competitors for so many years: when you control both the hardware and the software, you can make your system do things others can’t. Your calibrations are consistent, you can rely on your data quality, and your measures are standardized to a precision at least as good as in any certified medical device. Your users share this data with you because they trust your reputation (as Apple) for exceptional data security.
It’s under conditions like these that machine learning becomes truly powerful, and turns an integrated platform of hardware, software and data into a highly reliable and useful product.
Add this cardiovascular health strategy to Apple’s investment in user privacy and in secure digital health records, and the world’s biggest company has likely found a way to bring tens of millions of people onto their health platform, to keep them there through continuous improvements and feature additions, and to collect monthly recurring revenues from every one of those people for many, many years. Some people (or their insurers, or the companies making their medications) will likely pay a lot more per month for easy, wearable, AI-powered cardiovascular monitoring than the bulk of users pay for an iTunes subscription, because an iTunes subscription is arguably a less life-or-death proposition. This is a very good business model.
