I've read people recommending pulse oximeters here. I got one based on somebody's recommendation.

I got the STAT newsletter this morning by email and they have a warning about accuracy. Here's a copy/paste, followed by links to 2 of the resources that they link to:

FDA issues alert on ‘limitations’ of pulse oximeters, but omits racial differences
Without explicitly mentioning racial disparities, the FDA issued a public warning on Friday about potentially inaccurate readings from pulse oximeter devices. “Be aware that multiple factors can affect the accuracy of a pulse oximeter reading, such as poor circulation, skin pigmentation, skin thickness, skin temperature, current tobacco use, and use of fingernail polish,” the FDA’s alert reads. A December 2020 letter in the New England Journal of Medicine said the oximeters were nearly three times as likely to miss hypoxemia — below-normal blood oxygen levels — in Black patients versus white patients. “I am hopeful that the FDA will … reconsider the standards they use for determining whether these devices work the same for people of all skin tones,” Michael Sjoding, an author of that NEJM letter, told STAT.

Link to FDA announcement: [www.fda.gov]

Link to Black patients NEJM letter: [www.nejm.org]

Also note that the FDA says that OTC pulse oximeters aren't reviewed and so, "...should not be used for medical purposes." I presume that's essentially boilerplate; I'm going to continue to use ours.

FYI. There are phone apps that work great.

Never tried one on a black person, but the phone, a $19 one from Ebay and the wired one in a hospital all read the same O2 levels. Tried it multiple times on several people.

---

"he is going to start WW3!!!!"

Michael,

Thank you for the heads up. I have a model from iHealth at home and at my office. The readings both offer nearly identical to each other. I took them to the doctor's office with me and checked their readings against the pulse oximeter they use and the results were pretty similar. Did this a couple of times with the same result. Good enough for me!

Robert

FYI. There are phone apps that work great.


With any PulsOx device, a specific one, or what device?

I have my Watch and an iHome that give the exact same reading.

So they're both similarly accurate or inaccurate.

I have no lab grade comparison.

---

I am that Masked Man.

All you can do, is all you can do.

There’s trouble — it's time to play the sound of my people.

Your boos mean nothing to me, I've seen what you cheer for.

Insisting on your rights without acknowledging your responsibilities isn’t freedom, it’s adolescence.

I've been to the edge of the map, and there be monsters.

We are a government of laws, not men.

Everybody counts or nobody counts.

When a good man is hurt,
all who would be called good
must suffer with him.

You and I have memories longer than the road that stretches out ahead.

There is no safety for honest men except
by believing all possible evil of evil men.

We don’t do focus groups. They just ensure that you don’t offend anyone, and produce bland inoffensive products. —Sir Jonathan Ive

An armed society is a polite society.
And hope is a lousy defense.

You make me pull, I'll put you down.

I *love* SIGs. It's Glocks I hate.



Edited 1 time(s). Last edit at 02/22/2021 08:19AM by RAMd®d.

Quote
Mr645
FYI. There are phone apps that work great.

Never tried one on a black person, but the phone, a $19 one from Ebay and the wired one in a hospital all read the same O2 levels. Tried it multiple times on several people.

That's neat, considering that those phone apps are frauds.

[www.theverge.com]

...

The thing about the pulse oximeters is that you don't need to have a perfectly accurate reading.

Digital thermometers aren't accurate either, especially the no-contact thermometers.

Nevertheless, both are helpful when you're trying to figure out whether something is wrong.

Understand that they usually show a slightly higher O2 level than you've actually got and if it's dropped a bit then you may need to take a reading multiple times to get a good ballpark, so if the reading is low a few times in a row, you've got a problem. If you test low and you're feeling sh!tty, get your ass to the doctor, cause symptoms plus low O2 on a pulse oximeter together are a fairly reliable indication that you're ill.

---

Yeah, design bias. Remember the HP facial recognition kerfluffle ?

I’m severely melanin deficient and my cheap Bezos pulse ox meters present results very close to my doctors devices.

I think the exact reading would be somewhat less important than the direction of change over time (in things like Covid).

We have several laying around our house and my Mother's house.
Many varieties.

We have a saying "Pick a finger".

Moving it to another finger can make all the difference. Why ask why.

Battery life and giving them a spin (oxidation on terminals) are important.

Go for one that is the least confusing and easy to read.

---

Staunton, Virginia

Quote
pdq
I think the exact reading would be somewhat less important than the direction of change over time (in things like Covid).

Exactly. The trend is what to watch for.

That's neat, considering that those phone apps are frauds.

Thank you for the link, as it also explains how the apps work with the phone's camera.


Digital thermometers aren't accurate either, especially the no-contact thermometers.

Nevertheless, both are helpful when you're trying to figure out whether something is wrong.


Monitoring any trend is how I use both, given that neither is lab-grade accurate.

I only have one thermometer, digital, so I can't speak to it's accuracy,

But it consistently gives me the same temperatures when taken at the same times each day, so If I feel off, I'm confident I can rely on it for conformational bias.

---

I am that Masked Man.

All you can do, is all you can do.

There’s trouble — it's time to play the sound of my people.

Your boos mean nothing to me, I've seen what you cheer for.

Insisting on your rights without acknowledging your responsibilities isn’t freedom, it’s adolescence.

I've been to the edge of the map, and there be monsters.

We are a government of laws, not men.

Everybody counts or nobody counts.

When a good man is hurt,
all who would be called good
must suffer with him.

You and I have memories longer than the road that stretches out ahead.

There is no safety for honest men except
by believing all possible evil of evil men.

We don’t do focus groups. They just ensure that you don’t offend anyone, and produce bland inoffensive products. —Sir Jonathan Ive

An armed society is a polite society.
And hope is a lousy defense.

You make me pull, I'll put you down.

I *love* SIGs. It's Glocks I hate.

I think this is good practical advice that could apply to a number of over the counter medical devices.

Get familiar with the device before you need it. Make sure you can get a good reading when you feel well so that when you need it you know it works. Otherwise you could have a chest cold, try using a mostly useless pulse ox and cause yourself a great deal of concern.

---

A good article on the Apple Watch Series 6 oximeter:

[tidbits.com]

Sarcany,

I take issue with your statement about digital thermometers. The implication is that their accuracy is insufficient for obtaining a valid reading. I don't buy that. I've got three different types. Oral. Forehead. No-touch. The readings I get from them aren't always identical but they are always very similar. They might not be as accurate as an old-school mercury model but they are definitely sufficiently accurate enough to gauge whether or not my wife, I or Little M have an elevated temperature.

Robert

I got one off Bezosworld that had a comment from a nurse who bought it who said that it was indistinguishable in function or accuracy from one that they use in the clinic. I'm pretty sure it IS accurate because I get essentially the same readings whatever finger I use it on, compared to another (cheaper) model I'd gotten previously -- where different fingers would give quite different results.

---

Freya says, 'Hello from NOLA, baby!' (Laissez bon temps rouler!)

Peter,

Exactly. That's why I brought the iHealth pulse oximeters I have with me when I visited the doctor once just to see how they compared to the model(s) the "pros" use in their office. I figure as long as you buy one from a reputable company, then it'll be more than sufficient for its job.

Robert

Quote
PeterB
I got one off Bezosworld that had a comment from a nurse who bought it who said that it was indistinguishable in function or accuracy from one that they use in the clinic. I'm pretty sure it IS accurate because I get essentially the same readings whatever finger I use it on, compared to another (cheaper) model I'd gotten previously -- where different fingers would give quite different results.

Which one is that?

again, those $20 pulse oximeters are +/- 6% accuracy...medical grade are +/- 2%, but start around $200.



Edited 1 time(s). Last edit at 02/22/2021 04:29PM by Bill in NC.

I don't think absolute accuracy is all that important.

Just take readings when you're feeling fine. The key point is readings, not reading. Take enough of them so you can get a decent average, and calculate the standard deviation for your unit.

Use that calculated information to interpret readings when you're not feeling good.

Another factor is what kind of shape your finger joints are in. Mine are all somewhat distorted by arthritis, so results vary, based on the amount of contact that is possible. But maybe I should try a different meter for comparison.

---

My life goes smoothly and in regular intervals

Just got an Apple watch 6 and the blood ox sensor reads 90 to 94. Test with the doctors finger reads 96 to 98% Unimpressed with the watch so far

Bill,

You sure your post is accurate? I checked some of the cheapies on Amazon.com. Many advertise and claim accuracy +/- 2% and FDA approval.

Robert

Quote
Robert M
Sarcany,

I take issue with your statement about digital thermometers. The implication is that their accuracy is insufficient for obtaining a valid reading. I don't buy that. I've got three different types. Oral. Forehead. No-touch. The readings I get from them aren't always identical but they are always very similar. They might not be as accurate as an old-school mercury model but they are definitely sufficiently accurate enough to gauge whether or not my wife, I or Little M have an elevated temperature.

...Which is pretty much what I said.

---

Quote
Dennis S

Quote
PeterB
I got one off Bezosworld that had a comment from a nurse who bought it who said that it was indistinguishable in function or accuracy from one that they use in the clinic. I'm pretty sure it IS accurate because I get essentially the same readings whatever finger I use it on, compared to another (cheaper) model I'd gotten previously -- where different fingers would give quite different results.

Which one is that?

This is the one -- it's actually come down in price since I bought it, which was towards the peak of corona-phobia with respect to sat O2. [www.amazon.com]

---

Freya says, 'Hello from NOLA, baby!' (Laissez bon temps rouler!)

Sarcany,

Not at all. You and I said _very_ different things. You said digital thermometers - especially no-touch models - aren't accurate. That is definitely not the same thing that I said in my post. 'course, now that you've said it, I'd love to see the basis of that claim.

Got some tests that demonstrate digital and no-touch models are inaccurate? Or, maybe, I should say it as inaccurate enough that they ought not be used to check someone's temperature for signs of illness?

Robert

Quote
Robert M
Got some tests that demonstrate digital and no-touch models are inaccurate? Or, maybe, I should say it as inaccurate enough that they ought not be used to check someone's temperature for signs of illness?

I didn't say that they weren't accurate enough to check for signs of illness.

In fact, I'd go so far as to say they're almost as accurate at detecting signs of illness as the back of my mom's hand.

I guess I need to clarify: There's nothing wrong with using a digital thermometer to check your temp, so long as you're only using it for rough guidance and not relying on the fine-tuned results from it.

Analog/mercury thermometers are the gold-standard for taking accurate temperature measurements. Digital devices are handy because they're fast and cheap to mass-produce without fear of mercury spills. It can take 10 minutes or more to get a proper result from a "real" thermometer as opposed to a thermistor-based device which can get a ballpark reading in under a minute.

If you have $800 to spend on a good temporal thermometer, one of those is probably an adequate substitute for a mercury thermometer if you take multiple readings and average them.

Or you can use whatever digital thermometer you have and just be wary that all you're getting is at best a highly-variable approximation of your temp. It's useful. It's not authoritative.

Edit: You want a study? Here:
[www.nursingtimes.net]

(Top Google result. Next time, do your own Googling.)

---

Edited 1 time(s). Last edit at 02/22/2021 10:16PM by Sarcany.

Sarcany,

I stand by my assessment. And, thank you for the link. it confirmed my thoughts on the differences. I could’ve found the link but why would I? I didn’t make the statement. You did. i wanted to see if you’d be willing to back it up. Thank you for doing it.

Robert

Quote
Robert M
Bill,

You sure your post is accurate? I checked some of the cheapies on Amazon.com. Many advertise and claim accuracy +/- 2% and FDA approval.

Robert

I'm sure they do...not like there's any real consequences...to them, at lest.

Bill,

It just shows your post isn't necessarily true. This assumes the companies are providing accurate information. I'm inclined to give them the benefit of the doubt, especially if it's a reputable company. If you really wanted the details, you could probably obtain it from them since they have to have it on record somewhere.

Robert

Quote
Robert M
Bill,

It just shows your post isn't necessarily true. This assumes the companies are providing accurate information. I'm inclined to give them the benefit of the doubt, especially if it's a reputable company. If you really wanted the details, you could probably obtain it from them since they have to have it on record somewhere.

Robert

It's not like they're advertising the $20 versions as a certified medical device.

It's more common to see language like on this product:

[www.amazon.com]

"NOTE: This device is for Sports and Aviation use only and not intended for medical use."

Also:

"Some pulse oximeters are sold OTC as general wellness, sporting or aviation products.
OTC oximeters are not intended for medical use and do not undergo FDA review.
Other pulse oximeters are cleared via the 510(k) pathway and are available with prescription."

[www.medtechdive.com]

This review found most of the inexpensive ones they tested could be off, unsurprisingly, as much as ~6%:

[journals.lww.com]



Edited 4 time(s). Last edit at 02/23/2021 05:03PM by Bill in NC.

Bill,

Nothing you've said changes my assessment. It's on the buyer to make sure that it a suitable devices for its intended purpose.

Robert

Quote
Robert M
Bill,

Nothing you've said changes my assessment. It's on the buyer to make sure that it a suitable devices for its intended purpose.

Robert

As long as the buyer understands accuracy claims of +/- 2% on those $20 meters are...questionable, sure.

+/- 6% may be just fine for paying that little, though.



Edited 2 time(s). Last edit at 02/25/2021 03:40PM by Bill in NC.