Why wrist vs. chest heart rate monitoring makes me worried when pharma r&d and digital health mate
The Polar heart rate monitor was invented in about 1977 and was released to the retail market in the early 1980’s. Let’s call it 1984. Fast forward 35 years and I suffer from terrible FOMO from looking at all the super trendy wrist worn “fitness” trackers that have eradicated fine watches. However, wrist worn fitness trackers, while they have improved, are not as accurate as the chest strap introduced more than 35 years ago. Yes, I still rely on that uncomfortable strap, well, because its very accurate and that matters if you are actually using heart rate training techniques. When a wrist heart monitor is worn in an inactive state, it is pretty accurate. The more active you are, or the more complex your movement, the less accurate they become; up to 30% inaccurate.
The conclusion by most observers is wrist worn heart monitors are sufficient for a recreational user; or by their own count, the purpose of your smart watch is to help you establish healthier habits and as motivational tools. That’s great. You don’t need very detailed or accurate data to drive behavior change. Wrist worn heart rate monitors represent a now $50B market that is growing fast. The jury is still out on whether they drive any kind of lasting behavior change.
However, every athlete I know wouldn’t use a wrist heart rate monitor to drive athletic performance…let alone make health care decisions. Not yet. I love data and technology but coming from the finance world, data integrity is king and a requirement to any profit making trading system. The need for it has driven many of the major innovations of the last decade. Similarly, one of the biggest problems that remains for health data analytics to show value in health care is data accuracy. This is where those of us coming from tech need to understand that their data will not be used for just motivational or recreational purposes. Its difficult for me to see how wrist worn retail smart watches are ready for precision health when they are not ready for precision heart rate training. A few such as BioBeat are making headway but they go through much more rigorous testing, the FDA process and offer a patch or other way to truly act as a monitoring system. Health care providers see the potential but remain concerned about both primary and secondary reliance on aggregated data. Finally, additional data inaccuracy is driven by provider- digital health interaction. Doctor frustration with EHR systems for example causes them to click through screens to simply get through the system which drives compliance and their legal departments crazy with nightmares of risk exposure or a provider might click on procedures, advice or warnings that are never given. I think the greatest challenge is when a group aggregates enormous amounts of bad data, sells it to a pharma company who wants to honestly leverage the power of large data sets and technology back into research…and then bad data yields bad results.
To solve that problem, pharma will have to start with good questions, and demanding data integrity; what is the exact question or problem they are trying to solve, what is the exact data to be analyzed, how will you know it is accurate or inaccurate. We urgently need digital health and pharma to marry; but they must marry well.