Pulse Ox Allday: Covid-19

No.

any information is better than none  .

Umm not in this case if the inaccuracy that is inherent in these watches leads people to believing they have a problem and then add more stress to overwhelmed health providers.

If you want to accurately measure you oxygen saturation levels go buy a proper pulse oximeter. Plenty to choose from:

https://www.walmart.com/browse/health/pulse-oximeters/976760_1005860_1161882_2477346

Wow - I always figured I would feel it if the oxygen saturation levels dropped, but that article suggests people aren't even short of breath. Scary. I'll still stick with the night schedule for Pulse Ox as it sounds like the levels drop over several days. Still - it'd be cool to have an alarm for this like we have for high/low pulse.

I have not.

So, it's a random number generator?

You’re welcome to believe what you want to believe. My point is that since it has not been produced or registered as a medical grade device it should not be used to evaluate the presence or otherwise of respiratory dysfunction.

I turned on All-day PulseOX a couple weeks ago (usually, I have pulseOX totally off, and only use the widget for manual measurements).

Definitely is killing my battery life, I only get about 10 days between charges now (normally get 21-30 days). But I figured more data can't hurt anything. It's not like I'm going to check myself into the ER if my O2% drops, I just figure that if it seems unusually low then I'll stay at home more and isolate more than I have been.

My 2 cents:

first of all: the device has no medical approval/licence.

A lower than usual PulseOX value doesn't have to mean anything: bad day; measurement error, bad (beta) fw update...

But it can be an indicator, if you have other symptomes, too. In that case, I would call a doctor.

Don't futz around.  You're saying that the device reports data that is so unreliable, it shouldn't even be used as a basis to make a phone call to a doctor, right?  Own your position, man.

That's exactly my thinking on how you might use it.  I get the "not designed as a medical device" thing.  But, to paraphrase Gene Kranz in the Apollo 13 movie, "I don't want to know what it was designed to do, I want to know what it can do."  BTW, by way of comparison, my Oura Ring app sent out a notice inviting users to participate in a Covid-19 study to see if that device could predict the onset.   It's not a "medical grade" device either ...

On average I get around 94/95 all-day percentage and 93/94 at night.  

When I've run it in the doctor's office vs their device, which is hard, because it doesn't always take, it's been 2 to 3 points lower: 97 vs 94.

The thing is, I once took 3 consecutive reads and got 89, 92 and 97.

I guess, if I ran it all-day, felt lousy, and saw an 88 for the day, I'd probably make a phone call.  

Running it one-time is a dice-roll and that's is, and is not, a problem.  For example, if a scale is off, in one direction, I can still use it to see if I'm gaining or losing weight.  There's no problem except being 10 pounds heavier gets you a lecture about your BMI being too high.  A scale that's random, up or down 5 pounds, can still work because you can average it and work trends, but, it can be serious because sudden weight drops can be indicator's of problems.  Sorry, going back to heart rates, but when the watch spikes, do I have a heart problem, or do I have a watch problem, because a spike could also be serious.

I guess what I'm saying is, maybe it doesn't matter but maybe it does, and, regardless, it's almost certainly not reliable.  So, spend the $30 and get a dedicated device, if you can find one.