Subscribe
Share
Share
Embed
Conversations in Pulmonary, Critical Care and Sleep Medicine by the American Thoracic Society
Eddie: [00:00:00] Hello and welcome. You're listening to the ATS Breathe Easy Podcast with me, your host, Dr. Eddie Qian, also the host of the IC and Podcast. Each Tuesday, the ATS will welcome guests who will share the latest news in pulmonary critical care and sleep medicine.
Whether you're a patient, patient advocate, or healthcare professional, the ATS Breathe Easy podcast is for you. Joining me today is Dr. Ashraf Fawzy, who will be discussing disparities in pulse oximetry readings. Dr. Fawzy is a pulmonologist and critical care physician at John Hopkins University As a clinician, he cares for patients with asthma and COPD as well as critically ill patients in the ICU.
His research aims to understand and improve the real world performance of patient monitoring devices to ensure universal accuracy with a particular focus on pulse oximeter performance in clinical settings, so it makes for a great guest. Welcome, Dr. Fawzy
Fawzy: Thank you so much for having me.
Eddie: No, thank you so much for being here.
I, I think we'll just, we'll just jump right into it and I think let's, let's start a little [00:01:00] bit, kind of at the basics and, and just say, well, why, why is it important for us to have accurate, non-invasive oxygen saturation measurements? That's kind of what we're talking about here.
Fawzy: Yeah, pulse oximetry really revolutionized the way that we practiced medicine all the way back.
It was invented in the seventies, but really wasn't used ubiquitously until the nineties. And you know, before then people just guessed what people's oxygen was, what patient's oxygen was. And after we,
Eddie: my lips blew.
Fawzy: Exactly right. Yeah. Is, are your lips blue? Do I think you, your oxygen is low. In fact, there was a study that looked at whether med students and physicians could tell whether someone's lips were blue and they didn't do a good job, is basically the upshot.
So we really needed something a device to use to be able to tell [00:02:00] whether someone. Was hypoxic what their oxygen level was. And then this really was first implemented and made surgery a lot more safe. And then from there on went, you know, to the emergency department. And then now as you all know, we use it everywhere in the hospital and in the clinics.
So it really is very important to know how much oxygen your patient has in their blood, and that really impacts how you treat them.
Eddie: Yeah. And I think this will play a little bit into the future discussion, but can you step us through at least, maybe not the like exact, all the rigorous science, but how do these pulse oximeters work?
And then is there any difference between hospital machines and like over the counter versions that you could get at your local pharmacy?
Fawzy: So Pulse oximeter is basically a device that shines two types of light. Through the skin to try and estimate how much oxygen is attached to the hemoglobin. I use a red light and an infrared light, and by [00:03:00] doing some calculation between how much is absorbed of those two, we can tell how much oxygenated blood there is and how much deoxygenated blood there is in the in the person's system.
The devices that we use in the hospital are quite different than the ones that are over the counter in that the Food and Drug Administration, the FDA, doesn't regulate the ones that are over the counter. Those are considered to be recreational use. So the. FDA doesn't have anything to say about those.
They have make no guarantees about how they're used, whereas the ones that we use in the hospital are cleared by the FDA, meaning that they. Perform as well as prior pulse oximeters. That's basically the standard and the way that they're tested is on healthy people. So the pulse oximetry manufacturer is going to come out with a new pulse oximeter, for instance.
And to this day, the [00:04:00] standard is that they only have to test it in 10 people, 10 healthy people either and two of which, or 15% have to be, dark pigmented, which they don't define. So essentially that's kind of the difference between the two is the ones in the hospital are tested at least on healthy people, and the ones that are over the counter are not tested and have, we have no idea how good they work.
Eddie: For recreational use. I think it takes a much more imaginative person for me to think of recreation with these pulse oximeters. But I'll leave that, I'll leave that up to the listeners. And then y'all, the so you had mentioned that 15% or two of the 10 people have to have a darker skin pigmentation.
And I think one of the, kind of kind the meat of our discussion today is gonna be talking about the, the kind of disparities in pulse oximetry. And this really, at least for me. And the wakes of COVID came to, came to light in the setting of the [00:05:00] kind of race, ethnicity, and skin color, skin pigmentation, differences in how they can impact his readings.
what do we know about the disparities in race, ethnicity, skin pigmentation as it comes to non-invasive oxygen monitoring?
Fawzy: Yeah. When we talk about this topic I, I definitely have to give a shout out to our, our colleagues at the University of Michigan. Dr. Shing Dr. Wasana, who is now here at Hopkins with me thankfully, and Dr.
Val, who is currently at university of Colorado. But when they were all together at Michigan back in 2020 based on their clinical experience during COVID, they noticed that there was this. Potential inaccuracy in pulse oximetry among the black patients, the darkly pigmented patients that they were taking care of and, and really made this big push to look at it further.
So, so they [00:06:00] published this paper, this letter in the New England Journal in 2020. Essentially showing that at both the University of Michigan and in this much larger cohort, that pulse oximeters were less accurate among those people who were of black race versus white race. And then we were able to follow up on that and kind of take it a step further with our data here at Hopkins that essentially showed again that among black patients, the pulses, oximeters were less accurate.
And the important thing is that that inaccuracy was in the overestimation direction, meaning that. The pulse oximeter told you that the oxygen level among black patients were higher than they actually were, meaning that it made it seem like those patients were less sick. But the step that we took it further is that looking at clinical outcomes and saying, well, does this really matter in [00:07:00] terms of how we treat patients?
What we showed was that it actually did, and, and you know, COVID obviously was ravaging the US and the world at that time. So we had a singular disease that we could look at. So we took patients with COVID-19, and we looked at whether this pulse, oximeter, inaccuracy, delayed care for those individuals.
And, and we defined that by saying. Would they have gotten rem, Desi or dexamethasone, the two treatments for COVID earlier on. And we found that among black patients and Hispanic patients, there was a delay in care to the tune of 23% and 29% among Hispanic and black patients respectively. So that was really a big step forward in our understanding of this problem that.
There isn't just a [00:08:00] little difference in the number that this actually does impact how we make clinical decisions and how patients are treated in the clinical setting.
Eddie: Yeah, it makes sense. And so in preparation for us, us talking about this, you're referencing a paper that you had published, and I believe it was Jam Internal Medicine in 2022, where the cutoff for treatment.
For Dexamethasone resi, I think it was 94% oxygen saturation, 94%. And so you had mentioned that there was an increase in it was the overestimation in patients who have a darker skin pigmentation. And so that would mean that there just a, particularly for myself, that their oxygen levels were lower than the, the pulse oximeter would read.
Is that correct?
Fawzy: Yeah, that's absolutely right. And then. Following up on that jam internal medicine paper that, you know, used modeling to, to try and predict whether there was a delay or not. We actually looked at [00:09:00] actual delay in delivery of care. In a much, much bigger cohort of, of patients in like 186 hospitals.
And found that there was an actual delay in delivery of care that was pretty similar to the one that we found in, in the paper in 2022. And that paper was published in JAMA Network Open in 2024.
Eddie: So, so this might, this might be a naive question, but. If I know that they're systematically lower, why can't I just, you know, set my cut points a little bit higher to kind of account for that?
Fawzy: Yeah, it, it's a great question and, and one that I hear often. There, there's multiple reasons that that may not be a good idea. And I, and I think number one is that the pulse oximeters aren't uniformly inaccurate. Among patients with with darker skin pigmentation. So it's [00:10:00] not like you put a pulse oximeter on someone with darker skin and you get a result that's consistently above their actual, actual OCC oxygen saturation.
There's very clearly a lot of interaction going on here with how thick they are and, and other, other issues that we still haven't really completely fleshed out. So there isn't just one value though that I can say, well, you know, every single black patient you're gonna have a pulse oximeter reading that's X amount higher.
So that's not gonna be a way that we can interpret the, the inaccuracy of pulse oximetry. And then the next thing is that these correction factors have. Been problematic historically. So we've used correction factors based on race and multiple things. And of course the, the two most notable ones, especially [00:11:00] among pulmonologists and intensive care physicians are pulmonary function testing and kidney function.
So we've been trying to roll back using correction factors for those. This is a little different because we are talking about, you know, physiologic skin pigmentation and melanin content in the skin versus the completely social con construct of of race. But at the same time, we're trying to kind of simplify a much more complex problem into saying, well, you know, just subtract this number from, from the pulse oximeter reading that you get that that isn't gonna work.
And then finally, on that point. you may want to consider a higher oxygen level for, for individuals who you suspect might have inaccurate pulse oximetry. [00:12:00] And you know, by. Essentially if you have checked an arterial blood gas and their oxygen level is disparate from, from what their pulse oximetry is showing, that could be one way or one reason to do that.
Or if their pulse oximetry readings are all over the place and you're not really sure whether you're getting good pulse oximetry readings or they're critically ill and you're having trouble picking up a pulse oximetry reading, for instance. So there may be instances where. That could be a good stop gap until we have a better solution.
Eddie: Yeah, that makes sense. That makes sense. We, we, before we started recording you, you had mentioned a little bit about that there might be some other factors that clinicians should be aware of when we're talking about pulse oximetry. Other than just skin pigmentation. What, what are some of those?
Fawzy: Yeah, and there are a bunch of factors that.
We kind of know in the back of our minds probably affect pulse oximetry, but really haven't been studied well or at all. [00:13:00] You know, we all know that having nail polish on, for instance, can impact pulse oximetry. That one we all know. But another one that is very common in the ICU for instance is perfusion.
And hypertension. We have no idea how that really affects pulse oximetry. We just kind of know that it probably does. There have been some studies in. The laboratory setting. So basically on healthy people with pulse oximeters that show that potentially lower perfusion peripherally can impact pulse oximetry, and in fact can make this race difference even worse.
Hmm. But no one has really even studied this in, in the clinical setting. So one project that we're just starting on here at Hopkins is to try and look at that, to try and see [00:14:00] how do vasopressors play in to the problem? How does hypertension play into the problem? How does cardiac output play into the problem?
And. What, how can we correct that problem too? So it's not just about identifying the problem and saying, clinicians go out there. You're on your own. We really do need to take the next step and find a technological solution. At least in my view, it's probably gonna be a technological solution to correct these problems that we're seeing.
Eddie: have you adjusted anything or changed anything in your practice based off of. The data that's out there and some of the data that you are working on,
Fawzy: it's very difficult to, it's really hard. I mean, I've certainly been practicing in the ICU in the many years since this has come out and since we've been talking about it and since I've been researching it.
And still, you know, every day I kind of look at a pulse [00:15:00] oximeter reading, and I, I, I wonder whether it's accurate or not. And you can't really go around doing ABGs on everyone. So. I have definitely been a lot more aware of it. If I am questioning it, if something just doesn't add up, if the patient's lactate, for instance, is rising, even though we can see that everything else looks fine and they're well hydrated, et cetera, then you know, that's one thing that you could potentially go to and say, Hey, is my pulse oximetry reading actually accurate?
Or is that 94%, actually 84%. So. I think being aware of it questioning the, the results you're getting from, from this tool specifically can really help in terms of the way that you interpret and the way you approach patients to make sure that you're providing appropriate care.
Eddie: Yeah, I think that makes a lot of sense.
I've done something similar in my own practice in the ICU where it's the, [00:16:00] if there's something that's unexplained, then you start to question that, the data that you have. And one of the first questions, because of all that we know now about these pulse oximeters is, well, is this reading correct? Is this a patient I need to correlate with an arterial blood gas?
So I think that makes a lot of sense.
Fawzy: Absolutely, and, and it's really something that I certainly wasn't taught in medical school or residency, that these pulse oximeters really were only tested in healthy people and we have no idea how they're actually working in the ICU here. And you know, they may be inaccurate because of skin pigmentation, and it's not for lack of knowledge that we weren't taught this.
So the first paper that reported on this race issue and the skin pigmentation issue was published in 1990. So we've known about this for quite a while, and yet we really haven't done anything about it or really didn't pay attention to it until 2020 when the COVID Pandemic hit. So. It, [00:17:00] it, it really does take a lot of, not just you personally, you know, having that in mind and, and looking at the pulse oximeters and making sure that you're getting accurate readings, but also disseminating that to your trainees, disseminating that to other clinicians who, who are working with you.
Eddie: Yeah. And it makes, it makes so much sense because you, you, you and I both were critical care providers and pulmonologists, and we think about this all the time and it's in our newsletters and and whatnot. But a lot of our tra, like you said, you mentioned at the very beginning, pulse oximeters are used in every aspect of, of medicine, every aspect of clinical care, surgery, clinics, non ICU hospital wards.
And so e everybody needs to be aware of this and know about this. I'm curious, I guess changing this topic just a little bit, you had, you mentioned that, you know, a lot of this came to light during COVID in the wake of COVID. What's, changed? Has anything been done about this in the interim [00:18:00] or has anything been at least like the wheels have been moving on?
Anything since c and we discovered these problems, or at least we brought these problems to a brighter light.
Fawzy: There have definitely been a lot of. People studying this, A lot of studies that have come out. There was a preprint that was published from, from a study that was done at UCSF that was a prospective study.
There have been a bunch of other prospective studies done. We published one in, in the Blue Journal that was a pilot study that looked at objectively measure skin pigmentation and pulse oximeter, AC inaccuracy. And we found that, you know, it really is based on skin pigmentation. we did start to see some discrepancies in, in some of the results that have been coming out.
And, you know, one, one thought that we started to have was have these pulse oximeters have the accuracy of them have. Their performance been kind of changing or are [00:19:00] is the performance of, of Pulse oximeter stable? So that's kind of one question that we've been struggling with because after you get that FDA clearance, you really don't have to give any more data.
You don't have to show that this device is consistently working the same. So, so this is kind of a work in progress that we're, we're doing. And in terms of what you say, what, when you ask what. Has changed. And one thing I'm worried about is that the pulse oximeter performance is actually changing. So that's one thing that we're in the middle of investigating.
Hopefully we'll have a paper out on it soon. Hopefully it'll be in one of the at TS journals. Unclear yet, but we're, we're
Eddie: trying there. But I wish you the best of luck.
Fawzy: Thank you. But we are, we are trying to look at whether that could be an issue going forward in terms of. Impacting the, the results of some of these studies that are [00:20:00] coming out.
And, and in the same time, we're trying to do our own retrospective and prospective studies to try and answer these additional issues of testing in the clinical setting and looking at scenarios that really don't happen in people who are healthy and in whom the these devices were previously tested.
One of them, as I was saying, was perfusion and hypotension and vasopressor use. But another one, and this ties into my other interest of asthma and COPD is hypercarbia so. You know, none of those healthy patients had a CO2 that was elevated, but CO2 could potentially impact the way that the hemoglobin interacts with the oxygen.
So it's never been studied and, and we're hoping to start soon, a prospective study that that will be looking at that.
Eddie: We, we've talked a little bit about kind of like the ATS at S'S [00:21:00] organization are, are there things that you would want patients or patient advocates or on the flip side, large organizations like the a TS as a whole to advocate for, or, how can we people and physicians and organizations not specifically in the space like you are, how, how can we support this?
Fawzy: Yeah, I, I always love the fact that a TS is, is always trying to advocate for. Better patient care, better patient outcomes. And this is no exception. And I've certainly had conversations within the a TS about this. I think one thing that the a TS can do is to bring the issue to the forefront, make sure that it's written in their guidelines.
I mean, the a TS has been certainly at the forefront of the PFT issue in terms of taking race out of that equation. This one is a little bit more complicated probably but. Writing it into guidelines that the a TS puts out, making sure that [00:22:00] there's adequate education out there about this issue, but also at the same time encouraging progress and encouraging change.
And one way to do that, not just the at TS obviously, but large hospital systems should probably say that they're really only going to buy. Pulse oximeter is moving forward that work well in everyone and that don't have this potential RA racial bias to them. And, you know, the FDA has made some progress on this.
There were two advisory committee meetings public advisory committee meetings in the last four years or so. And the FDA came out in January of 2025 with draft guidance for changing things that included this new designation for. Pulse oximeters that work equally well in, in [00:23:00] everyone, regardless of skin pigmentation.
But you know, now we're a year and almost a year and a half out from that, and it hasn't been implemented and nothing has really moved forward after they came out with those. That guidance. So one other thing that the a TS can do is, is really push our regulatory agencies to make the changes and to really follow through on what they started several years back once this issue really came up and, and really move forward with it.
Eddie: And that, that makes a lot of sense. And I think, you know, for individual clinicians, you had already mentioned, it's, it's shocking how many people don't know about this issue. And I, I think it's always, it's always surprising to me when I'm on war as I'm talking with. Residents or medical students and asking a question like, Hey, hey, how does a, how does a pulse oximeter work?
How does, how do these automatic non-invasive blood pressure cuffs work? These things that you use every day, or you're interpreting that data, how do these [00:24:00] actually work and what kind of mistakes could you potentially expect from them? And they, the answer is crickets. People don't really know this.
And so I think. On the individual level, it's, it's also working towards educating providers and through kind of different levels of training to make sure that we're all aware of these, the,
Fawzy: And it, it is an issue, particularly in this modern era where we're being just absolutely bombarded with data, right?
So you don't really have that time to kind of pause and think about the data that you're getting sometimes, and you're just kind of, it's numbers all the time. You're getting labs over and over and over again. You're getting continuous vital signs and all this data kind of just is hitting you all at the same time that you don't have the moment to pause.
So I'm really glad to hear that you're. Talking to your trainees about understanding the limitations of that data that they're getting, and I think that's a very important topic for everyone to, to talk to their trainees about and to their colleagues about.
Eddie: Yeah, absolutely. Thi this is, this has been a really [00:25:00] great conversation.
Do, do you have any other kind of closing thoughts on Pulse oximeters or anything else you wanna leave the listeners with?
Fawzy: I just wonder if. Pulse oximeters are just one of the issues that, that we have to look at. I mean, as we were talking about, you know, there are other devices you mentioned, non-invasive blood pressure monitoring.
There was a very interesting paper that. I think was in jama correct me if I'm wrong, that looked at the non-invasive temperature monitors.
Eddie: Oh yeah, yeah. I think that was jama. Yeah.
Fawzy: Yeah. Those, those temporal monitors, and again, showing that skin pigmentation does have an impact there. It's, you know, another device that uses infrared light to go through the, the skin to, to try and measure something.
So. As far as I've seen that really hasn't been picked up and, and hasn't been researched nowhere near as heavy, heavily as, as the pulse oximeter issue. But you know, there are other devices out there that may be [00:26:00] an issue and may need more scrutiny. And the fact that, you know, we were using pulse oximeters everywhere for 30 years with very little scrutiny about.
Is it really appropriate to use these devices in clinical settings where patients are extremely ill acutely or critically ill? We just didn't really ask those questions until now. And, and then, you know, there are newer and newer devices coming out. You know, those continuous glucose monitors, for instance, are being used very widely now.
And are we really sure that, that those are accurate? I mean, those are just a few examples, but you know, the temporal temperature monitor was definitely one that caught my attention. And you know, when I was talking to a, a pediatrician colleague they mentioned this non-invasive bilirubin monitoring and how that potentially can have issues with skin pigmentation in [00:27:00] terms of its accuracy, again, because it is really kind of measuring color.
So I didn't really know about that because I only take care of adults. But that was a, a very interesting conversation that really. Enlightened me and broadened my, my view of we really need to make sure that all these devices we're using in clinical practice are actually giving us good information.
And this is potentially just a start.
Eddie: That's such a provocative thought to close on how, you know, there's a lot of benefits to having non-invasive monitoring, you know, less sticks, less pokes, less pros, less risks, but we have to make sure that they're right. But this, this really has been one of my, one of my favorite conversations here on the podcast, but.
Thank you everybody for joining us on today's a TS Breathe easy episode. Please subscribe and share this with your episode, with your colleagues. put it on your calendar. Hope to see everybody at the ATS International Conference in Orlando in May. Go to conference.thorasic.org [00:28:00] today.