Please help explain this Blood Pressure Monitor 'calibration' system

[NTL2009]

I've wanted to check the accuracy of my Blood Pressure Monitor (BPM) (an automated Omron unit), and finally found this old (2016) thread on another site (linked below). Prior to this, I only found the less helpful 'compare the reading to your Doctor's office' (BP varies with every reading, not so great), or 'send it to the mfg' (probably costs more than a new one!).

The author's goal was to rely on gravity and water, rather than another meter (with questionable accuracy). So this made some sense to me. He adjusts the water depth in the column on the left as a 'regulator' so that the BPM maintains a constant pressure so you have time to get a steady reading, rather than trying to catch it as it rises as it fills the cuff. Pretty clever, I think.

But then, why does he need the column on the right? Wouldn't the depth of water above the tube on the left indicate the actual pressure in inches water column? Just compare that to the reading on the BPM? I suppose the plastic bottle could be used (if needed) as a reservoir to smooth the pulses from what I think is a diaphragm pump (the air volume in the cuff might provide this in actual use)?

https://www.uglyhedgehog.com/t-378334-1.html

TIA - NTL2009

 

[Tom.G]

That is almost clever!

I would put the BP Monitor above the top of the tubing exiting the "Pressure Regulator" though. That reduces the risk of water entering the Monitor.

The basic theory is the pressure generated by Monitor increases the pressure in the bottle, thereby forcing water up the Manometer arm.

The tubing in the Pressure Regulator acts as a relief valve to limit the maximum pressure in the system. The pressure limit is set by the distance the end of the tubing is below the water level in the Pressure Regulator.

[Edit:]
I just realized that the Manometer arm and T-fitting are not needed.
The Pressure Regulator also serves as a Manometer.

• Start with the tubing to the BP Monitor disconnected
• Insert the tubing in the Pressure Regulator as shown in the drawing
• Connect and start the BP Monitor
• Measure the distance from the water surface of the Pressure Regulator to the water surface in the tubing that is in the Pressure Regulator

[/Edit:]

Hope this helps.

Cheers,
Tom

 

[Lnewqban]

I believe that the reason for the right section is to eliminate the error introduced by the bubbles.
In aquariums, bubbles in a water column are used to pump water from an undergravel filter up to slightly above the surface of the aquarium.
The bubbling column of water has less density than the all-liquid column, reaching a higher level when in pressure balance.

Please, see:
https://animals.howstuffworks.com/pets/choosing-aquarium-equipment7.htm

 

[NTL2009]

Thanks for those replies. Yes, keeping the BPM above the level in the 'regulator' makes good sense, I didn't think of that!

Ahhh, the old bubble pump! I didn't think of that effect either. Hmmm, he uses a 3" diameter PVC and 1/2" hose, the 3" PVC would seem to reduce the magnitude of the effect.

The use of the 1/2" hose for the regulator (vs 1/8" for the manometer) he does mention is to reduce resistance, so pressure is about equal at the hose end and at the monitor, but I don't see why this matters - the tee to the manometer is near the BPM. It seems (within reason), that is all that matters if you aren't measuring from the regulator, but only using it to adjust to the target point. It also seems a simple adjustable valve would suffice - just set it to bleed off air until the BPM reads the target number, then read the manometer. Maybe a little more finicky to get set, but it eliminates a large PVC tube filled with water, which is kind of awkward.

 

[sophiecentaur]

The author's goal was to rely on gravity and water, rather than another meter (with questionable accuracy).

I also have a BP monitor and would like to calibrate it. However, I think the pressure measurement is easier to calibrate than actual determining when the pulses are detected by the machine. My daughter is a midwife and she reckons her ear is better than what the machine does. And she is always right, guys.

I'm thinking that she would probably be a good source of calibration, with her (regularly calibrated) manual aneroid job. And yes, the calibration service is expensive but, of course it would be as it's a safety of life matter.

 

[NTL2009]

Yes, I was also thinking about the accuracy/repeatability of detecting the start/stop of the pulses. I was expecting a microphone in the cuff, but it is just an air hose, so they are picking up the sound through that air hose and a microphone in the machine, which might be just fine. With the proper filtering and threshold detection, I'd think the machine *could* be (but not necessarily *is*) more consistent than a random sample of humans who do this task. But that's just a guess on my part.

 

[sophiecentaur]

I would put the BP Monitor above the top of the tubing exiting the "Pressure Regulator" though. That reduces the risk of water entering the Monitor.

I was expecting a microphone in the cuff,

The 'signal' the machine really wants is subsonic (despite the use of a stethoscope by human ears) so I think the word "microphone" should be replaced by "pressure transducer" and having the transducer inside the machine would isolate it better from nearby sources of interference.

 

[NTL2009]

The 'signal' the machine really wants is subsonic (despite the use of a stethoscope by human ears) so I think the word "microphone" should be replaced by "pressure transducer" and having the transducer inside the machine would isolate it better from nearby sources of interference.

Good point, the 'beats' really are a 'bump' in pressure that would be riding on top of the downward slope as the air in the cuff is slowly released. Humans can't hear that, they have to listen to the 'swishes' that accompany the beats. But those are sort of an artifact of what we really want to measure.

And I almost wrote 'sensor' instead of microphone in that post, I was sort of thinking along those lines, but you made it clearer in my head just what the machine is likely looking for.

 

[Lnewqban]

Can those machines be calibrated via an internal device?
Which pressure is being calibrated, the max or the min shown in the display?
My wife's machine results at home seem to be off (around 8 inches lower) respect to measurements taken at her doctor's office.
Could that be due to alterated heart beats due to the driving, walking, talking, etc. prior to the nurse's measurement?

 

[NTL2009]

I'm not sure if there is a calibration pot or anything. Most likely it is an offset stored in memory, so no easy way to update that w/o their code.

In my career, we performed calibrations like that, and "it all depends". If the response is linear enough for the application, a single point offset can be stored. If it isn't linear enough, then you measure and store as many points as needed - maybe even across temperature, voltage, altitude - whatever affects the reading. I'd guess a one or two point calibration would be sufficient for these, but you'd need to characterize the sensor they use and the accuracy goals to know for sure.

Yes, most instructions I've seen call for you to sit still for 5 minutes, feet on the floor. Take reading with your arm at heart level. I would not be surprised at all that a reading would be 8 or more points higher w/o that prep. Ahhh, this link says to repeat after 1 minute.

https://www.heart.org/en/news/2020/05/22/how-to-accurately-measure-blood-pressure-at-home

 

[sophiecentaur]

But those are sort of an artifact of what we really want to measure.

I agree but the machine has to match the human version for BP. As they are both probably measuring different parameters then there could well be an offset between the two if the machine relies on a purely objective measured value.
The comparison between a static value of pressure and the reading of the machine under static conditions may not actually need to give a zero difference for the machine to be 'correct' for dynamic conditions. You'd have to find out what the official calibration system uses.
Personally, I'd have reservations using myself for calibration because my BP wanders from minute to minute and that may or may not be normal. Heart rhythm should ideally be pretty regular (apparently) but I've always noticed that mine is all over the place. (So does my Omron machine). So much so that I thought that the Apps for measuring it were all rubbish. Then the time came for surgery and I was told they detected Atrial Fibrillation and I instantly became someone with a heart condition. That's in spite of regular running for eight to ten miles and hiking in the Pyrenees every year (with the same silly pulse).

 

[NTL2009]

I agree but the machine has to match the human version for BP. As they are both probably measuring different parameters then there could well be an offset between the two if the machine relies on a purely objective measured value. ...

That reminds me of something my old boss would say: "Sometimes it's better to be 'wrong' together, than to be right alone!".

Can those machines be calibrated via an internal device?

To add to my previous comment - in that old thread, he didn't actually calibrate the maching in any way, he just made a chart of the offsets that he could manually apply (or with a spreadsheet) to the reading.

 

[sophiecentaur]

Could that be due to alterated heart beats due to the driving, walking, talking, etc. prior to the nurse's measurement?

Easy to test that. She gets into the car, walks, gets a bit cross with someone (easy to arrange if you know her well ), then takes the home measurement. That could easily be worth +8.
Or just take regular tests over a typical day.

 

[Tom.G]

My wife's machine results at home seem to be off (around 8 inches lower) respect to measurements taken at her doctor's office.

The most frequent reason my wife and I have found of different readings is different operators adjust the cuff differently. A tight cuff is already slightly compressing the blood vessels, requiring less air pressure to collapse them.

For instance there is a new nurse at our doctor's office that regularly gets readings that are 6 to 8 mmHg lower than expected. I finally asserted myself, loosened the cuff, and told her to take a reading. Sure enough, it was within my expected range.

The cuff should be LOOSE around your arm; about 1.5 inches extra diameter than my arm works well. Put the cuff on so you slip two fingers, oriented in the radial direction of your arm, between your arm and the cuff.

If using a manually inflating device, one with a squeeze bulb, you should get a zero pressure reading for the first several squeezes.

[edit] deleted "several"
Cheers,
Tom

 

[sophiecentaur]

A tight cuff is already slightly compressing the blood vessels, requiring less air pressure to collapse them.

That doesn't necessarily follow; it would depend on the actual arrangement. In an ideal static case of pistons in series in a cylinder (with a central section of the cylinder representing the pressure due to the BPM), the final output pressure would be the same as the pressure in the intermediate cylinder section.
But if the sleeve is restricting blood flow then the dynamic pressure where the BPM is sensing, could be lower. In a doubtful situation like this, it would be clearly be better to have a loose sleeve - moreover, they (Omron) recommend positioning the sleeve well above the elbow joint.

I'm actually surprised that such a popular measurement system works at all. But what alternative could there be, that is as portable, cheap and robust?

 

[NTL2009]

... The cuff should be LOOSE around your arm; about 1.5 inches extra diameter than my arm works well. Put the cuff on so you slip two fingers, oriented in the radial direction of your arm, between your arm and the cuff. ...

I let the cuff be slightly loose this AM, and got readings ~ 9 pts (SYS) and 2 pts (DIA) lower than the average of the past 6 days (under as close to the same conditions as I could manage- early AM, 5 min resting).

But I downloaded the manual and it says the cuff should be wrapped firmly.

https://omronhealthcare.com/wp-content/uploads/BP742N-EN.pdf

3.1 Applying the Arm Cuff
Remove tight-fitting clothing or tight rolled up sleeve from your left upper arm.
Do not place the arm cuff over thick clothes.

1. Insert the air plug into the air jack securely.
2. Wrap the arm cuff firmly in place around your left upper arm.

 

[sophiecentaur]

But I downloaded the manual and it says the cuff should be wrapped firmly.

My home BPM has a smart, semi rigid cuff that holds itself around the arm, leaving finite gaps all round. It would be much harder to cut off the blood supply with this cuff than the traditional fabric strip which can act like a tourniquet (see comments higher up). That's a great invention and removes potentially massive experimental errors from home measurements. Probably also 'professional' measurements - if they could bring themselves to use them.

 

[Dullard]

It certainly can't hurt to calibrate the BP monitor, but it may be equivalent to:
"Measure with a micrometer, mark with chalk, cut with an axe."

It has been clearly demonstrated in numerous studies (and makes intuitive sense) that the size, shape, and composition of the arm in question significantly impacts the 'performance' of the cuff. Interestingly, (to me, anyway) a common problem is the use of a too-small cuff on (increasingly) obese patients - the measured BP is overstated (when compared to a 'properly sized' cuff). It's too bad that there isn't an expensive pharmaceutical intervention for normal blood pressure.

 

[Lnewqban]

Could somebody explain to me in what principle that measurement (machine or manually) is based on?

 

[NTL2009]

I was wondering about one aspect of his test set-up, and did a very crude test this AM. As I suspected, when I try to bleeed off some air, the machine is smart enough to detect that the pressure is not rising and throws an "E1" error.

With a more controlled set up, you might be able to slow it down and get a stable reading before it errors out. But I think it is also looking for heartbeats, and I think that will also throw an error, so I guess you'd need to be wearing the cuff while doing the test.

 

[sophiecentaur]

Could somebody explain to me in what principle that measurement (machine or manually) is based on?

Afaik, it's based on finding what pressure over the band around your arm is enough to stop the circulation, at various parts of the heart's cycle. If you ran the air pump at 200, your arm would die, eventually (tourniquet). The pressure in the band is reduced until it's just less than the peak of the Diastolic pressure at that point, the band pressure has blips at your heart rate. The machine detects the blips and the pump pressure is the high number. Then the pump pressure is reduced until the intermediate (lower) Systolic peaks are just detected and that's the low number.

I can often tell what reading I'll get by the feeling inside my arm as the test is running. Good for +/- two or three. Imo, that's a good way of settling myself down for the procedure and likely to give a suitably low value.

 

[Lnewqban]

How can the peak be detected?
Zero flow or little flow should have about the same pumping pressure?

 

[NTL2009]

I think that when the cuff pressure falls just below the peaks of the heart pumping, some blood starts to flow, and you hear that pulse of flow (or the machine detects the little 'bump' in pressure). Above that pressure, there is no observable flow (pulse) at all.

On the low side, I think that when the cuff pressure is no longer providing much resistance to the flow, there is very little sound created, it just flows through unobstructed.

I was just thinking, how does that outside pressure end up applying itself directly to the artery ? Wouldn't it just get pushed out of the way? But I guess that our bodies are ~ 80% water, so that our arm is essentially in-compressible. So the pressure from the cuff around the arm ends up being pretty equal pressure all around every side of that artery.

 

[Dullard]

"But I guess that our bodies are ~ 80% water, so that our arm is essentially in-compressible. So the pressure from the cuff around the arm ends up being pretty equal pressure all around every side of that artery."

I think that is the simplifying assumption upon which the technique is based. But:

We aren't bags of water. There are 'fringe-effects' due to interconnection - tissue outside (at the 'ends') of the cuff are subject to some part of the applied force (tension). The larger the length/diameter ratio of the cuff, the less significant (I expect) the fringe effect. I would also expect that adipose vs muscle would (for 'fringe' purposes) behave differently. This doesn't invalidate the "pretty equal pressure all around every side of that artery" part of the assumption, but it could definitely alter the magnitude of that pressure WRT the 'applied' pressure.

 

[NTL2009]

"But I guess that our bodies are ~ 80% water, so that our arm is essentially in-compressible. So the pressure from the cuff around the arm ends up being pretty equal pressure all around every side of that artery."

I think that is the simplifying assumption upon which the technique is based. But:

We aren't bags of water. There are 'fringe-effects' due to interconnection - tissue outside (at the 'ends') of the cuff are subject to some part of the applied force (tension). The larger the length/diameter ratio of the cuff, the less significant (I expect) the fringe effect. I would also expect that adipose vs muscle would (for 'fringe' purposes) behave differently. This doesn't invalidate the "pretty equal pressure all around every side of that artery" part of the assumption, but it could definitely alter the magnitude of that pressure WRT the 'applied' pressure.

That makes sense. I suppose the effects are consistent enough that the numbers are still good in a comparative sense, even though they may be lacking in absolute accuracy. It does seem that body fat would be more compliant as you mention, and result in more pressure required to close down the artery - so maybe some of the usual "lose weight to lower blood pressure" advise is really only lowering the reading (though I certainly wouldn't discourage anyone from being fit)?

 

[sophiecentaur]

so maybe some of the usual "lose weight to lower blood pressure" advise is really only lowering the reading

Isn't that the point, though? For a given flow of blood, the heart needs to do less work if you have low BP. That is surely a good thing for you (using a simple model) but there are reasons for great panic when a patient's BP becomes low. The norm is the thing to aim for, I guess.

I think we may be trying to impose "Engineers' attitudes" on Medics here. Our body systems are far too complex to associate simple causes and effects and the attitudes of Medics has been to accept this and to apply a more wooly logic to treating many ailments / conditions. There may be a number of well known relationships between ailments and treatments that can be relied on and 'everyone' goes along with those. But Medical Consultants have a much bigger range of opinions about diseases and treatments than a set of competent Research Engineers on design of a motor car.
I have a regular discussion with my very competent daughter in law, who's a Consultant Physician (haha - name dropping again), about dosages and testing policies. She basically says that, in the present environment, there is not time (or available money) to venture outside the normal gamut of medicine and the system tries to benefit the maximum number of people with the available money.
So BP is God and can be measured anywhere and any time.

 

[NTL2009]

(quoting got messed up), sophiecentaur said:

so maybe some of the usual "lose weight to lower blood pressure" advise is really only lowering the reading
Isn't that the point, though? For a given flow of blood, the heart needs to do less work if you have low BP. That is surely a good thing for you (using a simple model) but there are reasons for great panic when a patient's BP becomes low. The norm is the thing to aim for, I guess.

I think we may be trying to impose "Engineers' attitudes" on Medics here. ...

Yes, but what I was getting at is that the fat might alter the reading in a way that doesn't correlate to what's happening in real life.

Well, I am trying to impose an "Engineers' attitude" on this, but only in the sense of understanding how it works. With that, we are better positioned to understand what to make of the readings we get.

Getting back to some earlier posts, I'm frustrated to find that 99,99% of the web info refers to "calibrating" a BPM as checking it against what you get at the doctor/nurse reading. And since the measurement typically does change a few points at least with each reading, that's a pretty poor system.

 

[NTL2009]

so maybe some of the usual "lose weight to lower blood pressure" advise is really only lowering the reading

Isn't that the point, though? For a given flow of blood, the heart needs to do less work if you have low BP. That is surely a good thing for you (using a simple model) but there are reasons for great panic when a patient's BP becomes low. The norm is the thing to aim for, I guess.

I think we may be trying to impose "Engineers' attitudes" on Medics here. ...

Yes, but what I was getting at is that the fat might alter the reading in a way that doesn't correlate to what's happening in real life.

Well, I am trying to impose an "Engineers' attitude" on this, but only in the sense of understanding how it works. With that, we are better positioned to understand what to make of the readings we get.

Getting back to some earlier posts, I'm frustrated to find that 99,99% of the web info refers to "calibrating" a BPM as checking it against what you get at the doctor/nurse reading. And since the measurement typically does change a few points at least with each reading, that's a pretty poor system.

(had to go back and fix quotes)

 

[sophiecentaur]

Ok about the quotes. I do it too.
But the fat is one of the things that high BP indicates. How is that a problem and how else would you measure it with a passive device that a medic can carry. The Sphygno sounds like a good machine, to me.