Apple Watch yields heart data nearly as good as clinical tests, report claims

New research could point to a significant step forward in AI-driven remote health monitoring.

Apple has been playing a long game in digital healthcare. The main focus – where Apple is right now – CEO Tim Cook  recently said, is “...to empower people to own their health.” And as the tools to enable such ownership develop, more opportunity knocks.

Protecting the human

We know Apple’s devices can track fitness levels. We know the heart health monitoring tools it makes are already saving lives. Now, fresh research shows that data gathered by an Apple Watch can enable remote monitoring of heart conditions. This is an important step in the company's smart plan for digital health.

Underpinning that plan is data. What Apple has proven across the last few years is that the data its watch gathers can make a difference. But what’s different about the latest Stanford research is that it shows not only that Apple Watch can detect health problems, but also that it enables pretty much the same quality of heart health data you’ll pick up when visiting a clinic.

It suggests the device can gather sufficiently robust data remote medical staff can use to make accurate decisions concerning heart health. This matters, because more than 17 million people die each year globally from cardiovascular disease, an estimated 31% of all premature deaths worldwide.

What the research states

The latest research saw 110 patients asked to use an iPhone and Apple Watch to monitor heart health. They were also required to attend the clinic for tests.

The study found that:

  • In an in-clinic setting, the Apple devices used a VascTrac app and were able to accurately assess "frailty" with sensitivity of 90% and specificity of 85%.
  • Outside the clinic in an unsupervised setting, the test was 83% sensitive and 60% specific in assessing frailty.

In other words, passive data collected at home was nearly as accurate as a clinic-based test (a six-minute walking test), the researchers claimed. The test suggests that remote patient monitoring of heart health conditions is possible.

The researchers said:

“While the benefits of telemedicine and remote monitoring—convenience, low cost, improved data quality—have been postulated for some time, the COVID-19 pandemic has made accelerated implementation a safety imperative. In this study, we showed that smart device-based measurements, including both a 6MWT (Six Minute Walk Test) and passively collected activity data, provide clinically accurate and meaningful insights about functional capacity in patients with CVD.”

And, I think, because it’s data-based the monitoring itself can be automated.

The burden of proof

This is a significant step forward toward building remote health monitoring systems, which I’m increasingly certain forms part of Apple’s plans in the digital health sector. Essentially, it’s a space in which AI meets human performance data and accurate analytics to deliver remote automated health monitoring services.

The problem with remote patient monitoring is trust.

We’re accustomed to using our watches and other devices to track personal data, such as steps walked, locatio,n or even books read. When it comes to personal medical data, however, we remain reticent to share this information with vast corporations that appear beyond our control.

Apple’s stance on privacy helps, but others in the chain are less committed to this, or so it appears. There are also challenges around access to these tools, the commodification of privacy, and the danger that increasing reliance on digital devices for healthcare poses the danger of exacerbating existing health inequalities. (The University of Cambridge has published an insightful policy briefing concerning this.)

Coming to a pet near you?

This latest Apple Watch research forms an able demonstration of how citizen-generated data can provide valuable personal oversight while also enabling research into existing conditions. Apple’s Research Kit is an excellent illustration.

It’s possible we’ll see the first truly significant work in terms of AI-supported analysis of patient data in the animal kingdom. Pet-focused firms such as Vet AI use data gathered by smart connected wearables and AI to monitor pet health. They then provide on-the-spot veterinary advice if an anomaly is spotted through their Joii app.

What’s good for the animal kingdom will likely become part of human healthcare, which means the key data gathered by smart devices such as Apple Watch will end up being used by remote patient health systems, likely tied to health insurers.

The problem, of course, is accuracy.

Where to find verifiable data

No one wants to rely on systems that don’t work, and in order to rigorously test such remote digital health solutions vast stacks of data – combining actual diagnosis with a vast quantity of patient data – is required. After all, you need strong, accurate and verifiable data to train the machine intelligence required to monitor the vital signs gathered by connected wearable devices.

Researching that data — even with ResearchKit — is expensive.

That need is why I think there has been a race to get hold of information from the UK NHS. It seems inevitable someone wants to mine this information for insights that may inform remote patient monitoring systems.

That we now know data gathered in real time by a wearable device can yield valuable insight that can itself be automated means we now know for sure that wearable devices will form part of future healthcare.

I’m expect Apple will play a part.

Please follow me on Twitter, or join me in the AppleHolic’s bar & grill and Apple Discussions groups on MeWe.

Copyright © 2021 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
 
Shop Tech Products at Amazon