In episode 130 of the podcast and in many subsequent discussions, I have advocated for femoral-artery-monitoring during cardiac arrest to allow titrated use of vasopressors. Many of my colleagues in the advanced resuscitation community have a similar practice. In 2013, the AHA has released an advanced practice guideline recommending the same practice.1
But…
What if the diastolic blood pressure number shown on the monitor is the WRONG PRESSURE. That is the contention of our guest today:
[su_box title=”Today's Guest”]
Per-Olav Berve is a Norwegian anaesthesiologist who works for the Oslo air ambulance and in-hospital at the Oslo University Hospital. He is currently wrapping up a PhD on CPR physiology, focusing on multimodal monitoring. His main project is a OHCA study on mechanical active compression-decompression CPR. (Bio from scanFOAM).[/su_box]
It seems the machine picks the DBP that is the lowest point between systoles. This works well with waveforms generated by the normal cardiac compression. During CPR, however the lowest point (decompression phase nadir) actually represents the reformation of the aortic outflow tract after it has been compressed. This generates a brief negative pressure which can give extremely low pressures. If you use this as your DBP you will be steered in exactly the wrong direction–the better the cpr and the more pronounced the vascular tone, the lower this number will be.
In the arterial wave form image above, Point A is a DBP of zero. This is the number that will appear on the monitor; this number can actually be negative. In the past, I would be rezeroing the transducer thinking it must be in error. And when I spoke about refractory vasoplegia, it was because of this number. What we need to be looking at is Point B, the point just before the systolic upstroke–this is the true DBP to titrate your vasopressors according to PO. The better the CPR and the higher the vasomotor tone, the lower Point A and the higher Point B will be. So as we are doing better, we will see lower and lower DBPs if we go by the displayed numbers.
Here is a normal waveform from an actual beating heart as contrast:
Solutions
You must visualize the waveform rather than looking at the computer-generated DBP
- Change displayed waveforms to 10 sec. if possible
- Auto-Wave (match the top and bottom of the scale to the actual waveform parameters) to make the waveforms as big as possible
- If your machine has a line that can be set to a pressure, put it at 40 mm Hg and then just look to see if the beginning of the systolic upstroke is above this line
The Full Lecture from the Big Sick
ScanFOAM has an amazing post with PO's full lecture, slides, and all the references from the lecture.
The Paradis et al. Papers
- 1989 Paper2
- 1990 Paper3
Superimposed Aortic and Right Atrial Waveforms
Note that the CVP and arterial waveforms look almost identical–you cannot differentiate arterial versus venous placement with waveforms alone.
The period of diastolic filling in the late decompression phase is shown below:
Identical Waves with the AutoPulse
Additional Information
- Review Article on where to measure DBP during CPR4
- Understanding Arterial Pressure5
- Experimental Simultaneous Waveforms on Swine6
- Case Report7
We screw up ETCO2 as well
- More from PO on this topic coming soon…
Now on to the Podcast…
Additional New Information
More on EMCrit
- Podcast 191 – Cardiac Arrest Update(Opens in a new browser tab)
- EMCrit 342 – More Cardiac Arrest Mastery with PO Berve – Pre-Tidal CO2
Additional Resources
You Need an EMCrit Membership to see this content. Login here if you already have one.
Professor
Nassau University Medical Center
No conflicts of interest (coi).
- EMCrit 385 – Eye Trauma I – Retrobulbar Hematoma, Orbital Compartment Syndrome, and Paracanthal Decompressions - October 5, 2024
- EMCrit Wee – Did this Really Just HAPPEN? – The HAPPEN Trial Hot Take – NIPPV for COPD - September 29, 2024
- EMCrit 384 – The Vascular Guy on Vascular Access - September 23, 2024
Scott, Thanks for the thought provoking podcast. Really we should be looking at the area under the curve, not diastolic nadir. Coronary perfusion is proportional to the pressure difference between that in the aorta during diastole and ventricular diastolic pressure. This applies for all of diastole, not just the nadir. The nadir is a convenient place to look, but oversimplifies this physiologic relationship. What we need is a monitor that splits the mean arterial pressure (MAP) measurement (which is calculated based on area under the curve) into two parameters: systolic mean pressure and diastolic mean pressure. With the diastolic mean… Read more »
Hi Ari, PO here jumping in on this one:) Thanks for this interesting comment. I just need to point out that the term “nadir” in the text above refers to the lowest point on the curve – the one chosen by all monitors I know of to represent diastolic pressure – and our main issue is that this does not reflect changes in peripheral resistance properly. I guess you are referring to the end decompression point in your comment. In regards to your main argument – the area between the curves theory : the Oslo group (Sunde et al (including… Read more »
To elaborate a bit further on the issue with the two MAP strategy( aka: compression phase MAP and decompression phase MAP). This is a doable (semi-manual) job in the post hoc data analysis of most animal trials (with typically 10-20 animals and treatment times of 10-20 minutes). Real-time analysis and / or validation in bigger human studies would demand an automatic algorithm like the one you suggests. This is similar to i.e the PICCO or arterial-line based CO tools that are used for spontaneous circulation. However, the dynamic curve-forms during CPR creates som challenges. The difficult parts are 1) to… Read more »
After a few moments of listening I was biting my thumbs for not seeing this sooner. I had a few cases with arrest fem lines – which have become routine at our shop – where I had negative diastolics and clearly mistakenly blamed the line setup. This is really important stuff, and setting the line at somewhere between 30 and 40 to make sure the pre-systolic plateau can be estimated at a glance is the way I’ll be going from now. Of course I think it is important not to consider this as an “artefact” as it isn’t at all.… Read more »
yes, super curious to see this as well. Felipe is already v. involved in this stuff and has been working with Paradis, so I think we will have these answers.
Hi! I agree – it is not an artifact – these spikes occur even with membrane catheters used in the research setting. Therefore, the assumption made in the Paradis ’90 paper that the dynamic trace in the decompression phase (they use the term relaxation phase) is due to movement of the fluid filled long catheters is likely not correct. Importantly, the spikes tend to be more evident with good peripheral resistance and be more dynamic or even disappear if a solid ROSC is established during compressions. Thus, they may be telling us important stuff about how things are going before… Read more »
Very interesting, everything make sense.
I would like someone to point out on the “Here is a normal waveform from an actual beating heart as contrast:” and label about where we are supposed to be reading the DBP. While the labeled picture above this picture is a little helpful. . .that’s not what we are using in practice. On the real monitor wave provided, it looks like a smooth ascending curve, straight from the lowest point, so the DBP and the “true DBP” that you are wanting us to use seem to be one in the same.
Gabe
it is the same point on the normal waveform above–the systolic upstroke. the difference is that the machine will correctly use that point, because it is the lowest point between systoles.
Hello PO Hello Scott Thanks for this enlightening Podcast! > The difficult parts are 1) to make the algorithm choose the correct upstrokes/downstrokes to work on and 2) to make the algorithm able to handle all cannulation sites Comparing to other algorithmic challenges in medical engineering it is not such a difficult task. I think I could even program a phone app that you coul hold in front of an arterial Signal to calculate the appropriate diastolic value and/or an area under the curve. But ist not worth it, I think, since as scott mentioned a lot of machines allow… Read more »
Hi Alexander! You are right – when we analyse pressures on each side of the coronaries – the readings will often indicate a possibility for perfusion during (part of) the compression phase. However, we actually do not know if it is the case as no study I know of has looked into this specific question in humans. There are animal studies suggesting even retrograde flow – and the mechanism would be that the external squeeze on the heart muscle hinders forward flow and even push blood backwards. I would guess that this is a very dynamic situation, but have anecdotally… Read more »
Thanks for your reply. Indeed, very dynamic situation for sure. So probably for now best is to use the estimated ‘real’ diastolic measurement..
Best wishes,
Alex
Super interesting!. The arterical curves from Larsen (2010), optained under LUCAS-1 massage, do not display the same deep downward deflection towards zero-pressure. Why is that? Less suction form LUCAS-1? Due to values measured at the aortic root in the Larsen paper?
Hi Jesper! The example in Larsen is collected at the coronary ostia (thus very proximal aorta) – these are the lowest lines presented in the figure. They actually show such a negative spike on both atrial and arterial side. the have marked out how they calculated perfusion pressures and of note is the simultaneous end decompression phase pressures that correlate well with perfusion. for further reading I recommend the excellent PhD thesis by Henrik Wagner – showing flow patterns as well. I am uncertain wether active decompression actually affects this as I have seen very low readings with standard LUCAS… Read more »
Incredible. FINALLY this makes sense. And I feel like a complete moron. Honestly I never even looked at the waveform…just that diastolic pressure number. Like Scott, I just assumed that number was going down due to vasoplegia and nearing death. This is a HUGE thing for those of us who are doing ECPR because it appears I’m overdosing my vasopressors. Now, once I decide I’m putting someone on ECMO during CPR I do not give any more pressor until after they are on-pump. But based on this, I’ve been giving too much vasopressor early on, during traditional CPR. Gosh I… Read more »
Hi Joe! Great to hear this from you! I do of course agree with you that having the correct decompression phase pressure to work with is of immense value when you are trying to decide if cerebral perfusion pressures are adequate for brain survival. To be more precise: as you know, for cerebral perfusion it is MAP that drives blood up to the brain. However, the monitors MAP calculations by means of whatever formula it use (i.e. SBP+2xDBP/3) can be horribly wrong depending on which value the monitor use as DBP. So – when I calculate MAP in this situation… Read more »
Hi Scott & Per
I hope I followed this. If so, if you put your line across & your systolic inflection point is above 40mmHg, am I right in thinking that there is no additional gain in pushing more epinephrine for your patient?
Dean, if the DIASTOLIC takeooff is above the line, not the systolic.
This was super helpful to know. I had been wondering why all my arrests that I’ve inserted fem lines have been having pressures of like 140/10 or 120/0, basically no diastolic pressure at all.