SpO2 "lag"

[zzyzx]

Can anyone provide an explanation for why the reading on a pulse ox may be a minute or too behind the true value? In other words, the pulse ox shows 90% but the patient may already be profoundly hypoxic?

 

[SicilianMedic07]

Cold fingers, nail polish, CO poisoning? Probe not working correctly?

 

[Achilles]

http://www.nda.ox.ac.uk/wfsa/html/u11/u1104_02.htm

 

[NYMedic828]

The device does not monitor every single passing red blood cell independently. It measures them as sample groups passing through the probe and then reports back its findings on how saturated the sample it measures was. So instead of giving you exact beat for beat measurements on the blood passing the probe at that exact moment it is giving you an averaged measurement that it acquired over a few seconds based on whatever programmed algorithm your device follows. The pleth wave on the other hand should theoretically match up with the radial pulse of the measured limb as that is a visual of what the probe is seeing rather than the number it is giving to you.

 

[Brandon O]

NYMedic points out one important factor -- signal averaging. If you DID see an instantaneous change in SpO2, it would be a good indication that it was artifact; the number calculated is averaged from readings over several seconds. So that's one factor.

Another big one is simply physiology. You mention the sat may read normal while someone is "hypoxic," but I think that's not actually what you mean. What you mean is, there's been a change in the patient's pulmonary/respiratory status, and you want to know about it, but the sat doesn't reveal it yet.

Suppose I stop breathing, for instance. My sat may not drop for a few minutes. Is that a sensor error? No, it's accurately depicting my arterial hemoglobin saturation. But my sat simply hasn't dropped yet due to my circulating (mainly venous) and pulmonary oxygen reserves. No doubt it will eventually, and hopefully by that point you've noticed that I've been apneic. But strictly speaking, that reserve is a good thing. You just have to remember that oxygen saturation is an end-point, and your "early warnings" should never be the end-points, but rather the compensatory mechanisms (respiratory rate and volume, etc.) that maintain them.

There may be a couple other factors, such as hardware and software delay, but in most cases I believe they should be negligible. The above two are the main factors to time delay in an otherwise accurate sat.

 

[VFlutter]

Suppose I stop breathing, for instance. My sat may not drop for a few minutes. Is that a sensor error? No, it's accurately depicting my arterial hemoglobin saturation. But my sat simply hasn't dropped yet due to my circulating (mainly venous) and pulmonary oxygen reserves. No doubt it will eventually, and hopefully by that point you've noticed that I've been apneic. But strictly speaking, that reserve is a good thing. You just have to remember that oxygen saturation is an end-point, and your "early warnings" should never be the end-points, but rather the compensatory mechanisms (respiratory rate and volume, etc.) that maintain them.

A good example of this is watching OR intubations on relatively healthy patients. They might be apnic for a decent amount of time before there is any change in Spo2 and by that time they are tubed. When I was shadowing a CRNA he really took his time and went slow to explain everything and even then there was never a dramatic desaturation.

 

[Brandon O]

A good example of this is watching OR intubations on relatively healthy patients. They might be apnic for a decent amount of time before there is any change in Spo2 and by that time they are tubed. When I was shadowing a CRNA he really took his time and went slow to explain everything and even then there was never a dramatic desaturation.

Although for what it's worth, I think you'd notice a sudden acceleration in the desaturation rate if you let them drop below 90%-ish. That's about when you crest the first "Wile E. Coyote" shoulder of the oxyhemoglobin dissociation curve, and each small drop in PaO2 will yield a greater drop in SpO2.

I suppose letting people desat just to observe this would be a bit inappropriate though...

 

[VFlutter]

Although for what it's worth, I think you'd notice a sudden acceleration in the desaturation rate if you let them drop below 90%-ish. That's about when you crest the first "Wile E. Coyote" shoulder of the oxyhemoglobin dissociation curve, and each small drop in PaO2 will yield a greater drop in SpO2.

I suppose letting people desat just to observe this would be a bit inappropriate though...

Agreed, desaturation is exponential not linear and does increase once below 90.

Another annoying group of people to watch on monitor are patients with sleep apnea. My Spo2 alarms are constantly going off.

We occasionally have terminal patients who die while on monitor and you are able to observe the whole process. The spo2 gradually then abruptly drop as they pass. It may not be the best data since many have multiple system dysfunction or severe cardiopulmonary problems. We had a patient terminally weened off CPAP a few weeks ago. It was interesting to watch from the monitor side of things.


I have also caught a few patients going into PEA because of the Spo2 pleth

 

[MSDeltaFlt]

Can anyone provide an explanation for why the reading on a pulse ox may be a minute or too behind the true value? In other words, the pulse ox shows 90% but the patient may already be profoundly hypoxic?

Simple. Interstitial difusion. The light shines not just through the artery of the finger, but also through tissue/bone/nerves/etc. Depending on the pt, the probe, software, clinical condition it might take a bit for the values to match up

 

[jwk]

The device does not monitor every single passing red blood cell independently. It measures them as sample groups passing through the probe and then reports back its findings on how saturated the sample it measures was. So instead of giving you exact beat for beat measurements on the blood passing the probe at that exact moment it is giving you an averaged measurement that it acquired over a few seconds based on whatever programmed algorithm your device follows. The pleth wave on the other hand should theoretically match up with the radial pulse of the measured limb as that is a visual of what the probe is seeing rather than the number it is giving to you.

Great answer. Depending on the model/manufacturer, all SaO2 monitors generally take 6-12 seconds to start getting a reading. Some may even allow the user to configure the averaging time, the "rolling window" the machine is using to compute that average. If you're using a monitor that shows the pleth waveform, make sure you're getting a good waveform before you believe whatever number might show up initially. A lot of my PACU nurses want to write down the first number they see, even if there is no baseline waveform on the monitor.

 

[sir.shocksalot]

Another big one is simply physiology. You mention the sat may read normal while someone is "hypoxic," but I think that's not actually what you mean. What you mean is, there's been a change in the patient's pulmonary/respiratory status, and you want to know about it, but the sat doesn't reveal it yet.

Suppose I stop breathing, for instance. My sat may not drop for a few minutes. Is that a sensor error? No, it's accurately depicting my arterial hemoglobin saturation. But my sat simply hasn't dropped yet due to my circulating (mainly venous) and pulmonary oxygen reserves. No doubt it will eventually, and hopefully by that point you've noticed that I've been apneic.

Great explanation. This stuff is why preoxygenation is so important before intubation. If your total lung capacity including your residual volume is filled with oxygen you can remain apneic for a longer time period while still having an adequate ratio of oxygen to facilitate alveolar oxygen diffusion.

I'll try not to reiterate what everyone else has said. Keep in mind that the monitor/pulse ox is a computer and the sensor is just providing data, it takes time to turn that data into something the computer can understand as an oxygen saturation and then display it to you. Also a computer doesn't know if it's getting artifact or a genuine reading, it gets bits of data and interprets it the best it can (sometimes why the monitor starts telling you your patient is in V-Fib when really it's just a lot of artifact). I usually recommend watching the pleth to evaluate if you are getting artifact or a good reading.

 

[BLS Systems Limited]

Another factor could be related to the hemoglobin level. When Hb is low, the number could be all over the map. Your question is actually dealt with by the manufacturers when they create the algorithm that goes into their devices. A lot of that stuff is explained when they do the sales material and presentations-things we tune out while we are looking at the display, pretty lights and battery life. Different devices can give you different readings depending on the patient and environmental conditions, based on the above responses.

 

[Brandon O]

Another factor could be related to the hemoglobin level. When Hb is low, the number could be all over the map. Your question is actually dealt with by the manufacturers when they create the algorithm that goes into their devices. A lot of that stuff is explained when they do the sales material and presentations-things we tune out while we are looking at the display, pretty lights and battery life. Different devices can give you different readings depending on the patient and environmental conditions, based on the above responses.

Are you saying that low hemoglobin count can lead to a poor signal? That makes some sense to me (it seems like it would narrow the transmittance differential between arterial and venous blood), but I haven't seen any research that supports it. Surely a low hematocrit, if anything, would have a larger effect? How different really is the "color" of a Hb-rich vs Hb-poor red cell?

 

[BLS Systems Limited]

Are you saying that low hemoglobin count can lead to a poor signal? That makes some sense to me (it seems like it would narrow the transmittance differential between arterial and venous blood), but I haven't seen any research that supports it. Surely a low hematocrit, if anything, would have a larger effect? How different really is the "color" of a Hb-rich vs Hb-poor red cell?

I am basing this on personal experience on several ICU patients (it was a few years ago so the manufacturer may have improved the sensor and software). On several occasions I experienced a vacillating waveform that increased and decreased with remarkable consistency (very high highs and very low lows with each cycle lasting a minute or two). The only common denominator that jumped out was low Hb counts (significantly lower). There may have been other factors, but the Hb stood out as the patients were down a few quarts.

 

[Melclin]

Are you saying that low hemoglobin count can lead to a poor signal? That makes some sense to me (it seems like it would narrow the transmittance differential between arterial and venous blood), but I haven't seen any research that supports it. Surely a low hematocrit, if anything, would have a larger effect? How different really is the "color" of a Hb-rich vs Hb-poor red cell?

A few papers (all rubbish) from the late 80's and early 90's suggest that pulse oximetry "may" be inaccurate with Hct <10 in very low saturations. Other literature disputes this. That number is quoted in several books as well but all based as far as I know on the same "evidence". If it even exists at all, it seems a small error at pretty extreme levels of hypoxia and anaemia. I'd argue its pretty irrelevant. You know they're all kinds of sick either way.

http://www.ncbi.nlm.nih.gov/pubmed/8010546
http://www.ncbi.nlm.nih.gov/pubmed/2352007
http://www.ncbi.nlm.nih.gov/pubmed/2064035

I've never seen newer papers on the topic.

 

[Brandon O]

Interesting, thanks. And of course we should remember that oxygen delivery to tissues is really a factor of both hemoglobin saturation and hemoglobin quantity, so if the latter is inadequate, hypoxia will develop whether or not the lonely remaining hemoglobin are well-saturated. So accurate or inaccurate, we'd need labs to really assess this, which will probably mean both a blood gas and a CBC, thus the pulse ox will be moot.

Or for prehospital purposes without labs, as you say -- they're jes plain sick either way.

 

[VFlutter]

Rant

On a side note.... If your patient has Reynaud's syndrome please do not put the Spo2 sensor on their finger. And if you do, please do not call an RRT because it is reading 70% even though your patient is in no distress.

Also, if the spo2 says the HR is 160 you should probably palpate the pulse yourself before calling an RRT.


Nurses

 

[Melclin]

And of course we should remember that oxygen delivery to tissues is really a factor of both hemoglobin saturation and hemoglobin quantity, so if the latter is inadequate, hypoxia will develop whether or not the lonely remaining hemoglobin are well-saturated. So accurate or inaccurate, we'd need labs to really assess this, which will probably mean both a blood gas and a CBC, thus the pulse ox will be moot.

Or for prehospital purposes without labs, as you say -- they're jes plain sick either way.

Well yes, and cardiac output...in the presence of a circulatory system under enough pressure to maintain adequate perfusion pressures.

Pulse ox never promised to be the lone ranger, then answer to the question of oxygen delivery, but it clearly has its uses. Its not completely useless just because you dont have a gas and a Hb/Hct.

 

[abckidsmom]

Well yes, and cardiac output...in the presence of a circulatory system under enough pressure to maintain adequate perfusion pressures.

Pulse ox never promised to be the lone ranger, then answer to the question of oxygen delivery, but it clearly has its uses. Its not completely useless just because you dont have a gas and a Hb/Hct.

Cause really, a gray, sweaty guy with a pressure of 70 and sats of 100 is not looking any better because of his excellent saturation.

 

[Melclin]

Cause really, a gray, sweaty guy with a pressure of 70 and sats of 100 is not looking any better because of his excellent saturation.

*Peers through tired eyes*...I can tell if you're agreeing with me, disagreeing or saying something different altogether :wacko: