How often should you take vitals on a critical patient?

[JPINFV]

Talking about NIBP, I ended up having to call shenanigans on a BP today of 66/24. Got 90/50 which, while not necessarily great, is much more realistic in an asymptomatic patient.

 

[Anonymous]

Talking about NIBP, I ended up having to call shenanigans on a BP today of 66/24. Got 90/50 which, while not necessarily great, is much more realistic in an asymptomatic patient.

this was an automated BP?

 

[JPINFV]

this was an automated BP?

Yep... in the ED...

 

[mycrofft]

It was your machine, it was a bilateral machine. Take it on each arm, then add them together.
:rofl:

 

[Brandon O]

My main problem with automated BP is that it gives you no indication of data quality. Even an HR calculated via pulse ox often displays a pleth waveform, which tells you a lot about pulse strength and regularity, as well as reliability of the number. But the BP just spits out a reading, and you don't know whether it's quite certain or borderline, something you know intimately if you obtained it yourself.

I suppose when it takes forever to produce a reading, that tells you something, and if it can't get one at all that says something too, but in the end you're still trying to "read through" the machine's work, which is not very transparent.

 

[JPINFV]

My main problem with automated BP is that it gives you no indication of data quality. Even an HR calculated via pulse ox often displays a pleth waveform, which tells you a lot about pulse strength and regularity, as well as reliability of the number. But the BP just spits out a reading, and you don't know whether it's quite certain or borderline, something you know intimately if you obtained it yourself.

I suppose when it takes forever to produce a reading, that tells you something, and if it can't get one at all that says something too, but in the end you're still trying to "read through" the machine's work, which is not very transparent.

Then there's clinical correlation. The A/Ox4, generally asymptomatic patient who's talking to you probably doesn't have a BP of 66/24.

 

[Brandon O]

Well, of course. But my point is, particularly at the BLS level, we have few enough insights into the patient's condition, so when I take vital signs I don't just want bare numbers -- I want to wring out every drop of physical information that's available. The idea that (using the other example) a pulse is reducible to a mere number throws a truckload of traditional but extremely useful clinical medicine out the window, and once you start doing that there's not much left.

 

[xrsm002]

The department I'm doing my ride outs on have NIBP on them but apparently none of the crew trust them. My IFT I worked at we had NIBP and that's pretty much all I used. Unless I got a really high or low number

 

[xrsm002]

How do you take a manual pulse ox?

 

[Medic Tim]

How do you take a manual pulse ox?

the same way you get a palp diastolic pressure

 

[Brandon O]

the same way you get a palp diastolic pressure

Not to get off topic here, but you just gave me a thought.

I've obtained a BP in a very tough-to-assess patient by placing the sat probe on their finger, inflating the cuff proximally, and "palpating" the systolic by watching for the appearance of a waveform while I deflated the cuff. You just made me realize that, in principle, one could probably get a diastolic in the same way -- the waveform should steadily grow as you continue to deflate, and the diastolic pressure will be the point at which it returns to the original, baseline amplitude.

In theory. Must try this...

 

[Melclin]

Not to get off topic here, but you just gave me a thought.

I've obtained a BP in a very tough-to-assess patient by placing the sat probe on their finger, inflating the cuff proximally, and "palpating" the systolic by watching for the appearance of a waveform while I deflated the cuff. You just made me realize that, in principle, one could probably get a diastolic in the same way -- the waveform should steadily grow as you continue to deflate, and the diastolic pressure will be the point at which it returns to the original, baseline amplitude.

In theory. Must try this...

I think there is a delay in the readings provided by pulses oxs such that its basically real time but not quite real time enough to be accurate unless you were deflating the cuff super slowly and pausing incrementally for however long the pulse ox takes to generate new info.

I think.

 

[JakeEMTP]

I think there is a delay in the readings provided by pulses oxs such that its basically real time but not quite real time enough to be accurate unless you were deflating the cuff super slowly and pausing incrementally for however long the pulse ox takes to generate new info.

I think.

Definitely agree. Masimo and Nellcor both have stated there is a delay in their literature. The interface and the software used will make a difference.

 

[Brandon O]

I think there is a delay in the readings provided by pulses oxs such that its basically real time but not quite real time enough to be accurate unless you were deflating the cuff super slowly and pausing incrementally for however long the pulse ox takes to generate new info.

I think.

The quantitative saturation figure is certainly delayed -- usually up to a minute or two for the blood to reflect changes at the lungs, and even then the measured sat is averaged over a few seconds. But I don't think that the visible pleth waveform (when displayed) or even the little bouncy signal bar should be delayed more than a moment or two for hardware and software latency. If anyone has evidence to suggest otherwise, please do share.

 

[NYMedic828]

As stated there is a delay.

Theoretically if it is in real time there should be an exact correlation when feeling a radial pulse as to when you see a pleth spike generated.


I think from an EMS standpoint pulse oximetry is another tool we just dumped onto people without proper education.

People do not understand almost every aspect of pulse oximetry outside the simple "it measures oxygen in the blood." (Which is only partially true in simplest terms)

My class taught nothing outside of giving oxygen if a low day presented. (People seem to think 95% is low and needs O2 in an elderly smoker)

The Bohr curve, the long list of potential factors resulting in inaccuracies of SpO2 measurement, hemoglobin binding and releasing oxygen. All had to be acquired via self education.

I am sick people wanting more toys but not being willing to learn to use them.

 

[JakeEMTP]

The quantitative saturation figure is certainly delayed -- usually up to a minute or two for the blood to reflect changes at the lungs, and even then the measured sat is averaged over a few seconds. But I don't think that the visible pleth waveform (when displayed) or even the little bouncy signal bar should be delayed more than a moment or two for hardware and software latency. If anyone has evidence to suggest otherwise, please do share.

The quantitative value does not take a minute but for perfusion accuracy the number you are looking for will. Usually the value will be a 3 second average much like the ECGs.

Sometimes it is easy to see the delay on the pleth in the ICU by palpating arrhythmias and watching the pleth. Some machines have a solid 1 - 2 second delay. But, if you go to the Masimo website for their literature on this when they established their specs you can read about it.

 

[VFlutter]

Sometimes it is easy to see the delay on the pleth in the ICU by palpating arrhythmias and watching the pleth.

Or even better correlating the pleth with the atrial waveform

 

[Brandon O]

The quantitative value does not take a minute but for perfusion accuracy the number you are looking for will. Usually the value will be a 3 second average much like the ECGs.

Not sure what you mean -- can you clarify? My understanding is that the displayed oxygen saturation is an average of the past several seconds in order to stabilize the data, since there's no physiologically plausible mechanism for significant second-to-second fluctuations in sat. However, I'd expect the qualitative plethysmography wave to be close to realtime, subject only to the latency of processing and filtering.

Sometimes it is easy to see the delay on the pleth in the ICU by palpating arrhythmias and watching the pleth. Some machines have a solid 1 - 2 second delay. But, if you go to the Masimo website for their literature on this when they established their specs you can read about it.

A second or two wouldn't surprise me too much. I'll poke around for some literature when I get home.

 

[JakeEMTP]

Not sure what you mean -- can you clarify? My understanding is that the displayed oxygen saturation is an average of the past several seconds in order to stabilize the data, since there's no physiologically plausible mechanism for significant second-to-second fluctuations in sat. However, I'd expect the qualitative plethysmography wave to be close to realtime, subject only to the latency of processing and filtering.
.

The machine does not care if the patient data is stabilizing. It is going to give you a number based on the algorithm in the machine. The number you get is an average of usually 3 second. If perfusion is low, it may take a little longer to gather the data set by the algorithm parameters. This is not much different than the ECG which might read a HR of 100 and then one PVC might drop the numeric value to 50. "Stabilizing data" is an odd term. Machines collect data by looking for what is within their parameters with assumptions and use this information to produce a result.

 

[Brandon O]

The machine does not care if the patient data is stabilizing. It is going to give you a number based on the algorithm in the machine. The number you get is an average of usually 3 second. If perfusion is low, it may take a little longer to gather the data set by the algorithm parameters. This is not much different than the ECG which might read a HR of 100 and then one PVC might drop the numeric value to 50. "Stabilizing data" is an odd term. Machines collect data by looking for what is within their parameters with assumptions and use this information to produce a result.

Okay, I think we're describing the same thing.