Sep 14 2018

New Apple Watch and Health Screening

Apple recently announced two new health features of their Apple Watch – fall detection and heart monitoring. These are being sold as useful health measures, but there are concerns that this new technology may be more gimmicky than useful, and might actually be counterproductive.

The first feature detects possibly dangerous falls. If the motion detector senses that the wearer has fallen to the ground, and then they don’t move for one minute, the watch will automatically call 911. This superficially sounds reasonable, but the concern is that the tech company has not adequately tested this feature in coordination with medical professionals.

Specifically – what is the sensitivity and specificity of this algorithm? How often will it detect an event that reasonably requires a call to 911, and how many false positives will it generate? How much will this overwhelm emergency services?

What is naive about this feature is the lack of appreciation for the fact that medical interventions need to be evaluated. Why, for example, one minute? That is a round number, but is that really the threshold where the true-positive to false-positive ratio is optimal, while avoiding dangerous delays if someone is really injured?

Also, what will be the net effect of this? Will the drain on resources actually cause more harm than benefit to the users? How will 911 call centers handle these calls? Will they be able to interpret them?

It’s possible this feature will work out wonderfully – the problem is, Apple is shooting from the hip here. It’s great marketing, it remains to be seen if it’s good medicine.

The watch can also be programmed to call chosen contacts in the case of a possible fall. It would have been better, in my opinion, to test this new feature by only calling contacts, not 911. Then collect data on true vs false positives. If the feature proves reliable enough, then the 911 feature can be activated.

This brings up the greater question – for big technology and social media companies, what rights do they have to plug their technology into common services? Apple is making decisions that can have a great impact on our health care system – do they really have the right to make these decisions on their own?

The second feature is a heart monitor, that monitors the rate and rhythm of the pulse. This feature only provides information back to the user, and doesn’t contact any aspect of the healthcare system. Apple is touting that it has FDA approval for the device – however, this only means the FDA approves that the technology does what it says it does, and will not kill people. When it comes to devices, the FDA does not evaluate or approve how they should be used and how effective that use is.

This is an important distinction. The FDA did not evaluate the net health effect of using this function, only that the monitor accurately records the pulse. Again, Apple is assuming (or at least marketing) the notion that this feature will be useful.

I have written previously about the fact that in medicine more information is not always better, as counter-intuitive as that may sound. Here the main issue is not so much false positives or negatives, but what is the proper threshold for reacting to health information?

Let me give you an example. It used to be common practice to monitor fetal lifesigns during delivery. This makes sense – if the fetus starts to exhibit signs of severe stress, then the delivery can be moved along, or a C-section can be done. But it was found that the monitoring was a net negative, because doctors were overreacting to fluctuations in vital signs that did not really require any intervention.

Now we are going to be this kind of monitoring information into the hands of the general public, probably selecting for individuals who are more neurotic about their health. What is going to be the net effect of this? Sure, there will likely be cases where useful information will alert the user to a problem who will then seek out appropriate health care. But it seems far more likely that many individuals will overreact to normal fluctuations in their heart rate or benign anomalies in their rhythm.

Again – what burden will this place on primary care doctors who now have to deal with the fallout? How much anxiety with this generate among users?

Even worse, this can drive a cottage industry of dubious health care providers who will happily treat whatever nonexistent problem these anxiety-provoking devices convince you that you have. That is the context that worries me the most – we are weakening regulations, allowing for the proliferation of unscientific or ethically-challenged providers, while we democratize health information. And now we are going to bypass health professionals altogether and encourage people to monitor their own biological parameters.

This kind of feature could be benign, if implemented properly, with adequate feedback on what to do, and not to do, with the information. At the very least, this should be a conversation with your primary care doctor or cardiologist. And of course we should just fix the gaping problems with pseudoscience in the health care system.

Apple should also fund independent research on the net effects of using these features, and tweaking them if necessary.

To be clear – I think that giving individuals the ability to use advanced technology to monitor health parameters can be a very useful thing. This information can be very useful to alert to genuine problems, and provide information to health care providers. I just think that this should be carefully studied and coordinated with science-based experts, and not rolled out as a marketing gimmick.

No responses yet