So I’m in the process of putting together my resus handbook, and the really good thing about writing something up is that it forces one to beef up the entire mental database and fill in blanks that may sometimes be filled by belief, habit, culture or leaps of faith. So part of my process will involve discussing stuff with the brightest guys I know. Who happen to be pretty bright. So I figured it might be stuff worth sharing!
Here, Segun and I discuss the possible uses of Pmsa, of resuscitation philosophy, and touch on the issue of blood pressure vs perfusion. (please skip to 0:30 – sorry can’t cut out!)
Love to hear some additions to our discussion!
Here is the paper I was referring to, with the graph on page 2:
cheers
Philippe
thinking critical care 
A very thoughtful and interesting discussion!
I have a few thoughts to add.
I never had heard of the device mentioned that measures mean systemic pressure, however MSP can be extapolated in ventilated patients if inspiratory holds are done at increasing airway pressures (PMID 222344243). With each increase in airway pressure, the CVP will rise and the cardiac output (essentially the venous return) will drop. If, as on a Guyton diagram, the cardiac output/venous return is plotted on the Y axis and CVP on the X axis, and each point from each airway pressure is connected, a descending line can be plotted that eventually comes to a point of zero venous return. This is the mean systemic pressure, since VR is zero when CVP equals MSP as there is no longer a driving pressure. What was interesting was that in this study, arterial BP was also plotted against CO and the pressure at which the CO was zero was higher than the point of the mean systemic pressure. This indicates that not only is there a vascular waterfall in the venous return circulation (as CVP becomes zero and no longer inhibits venous return), but there is a second waterfall somewhere between the capillaries and the small arterioles. Also, this indicates the point that is intuitive (but not often consciously recognized), that the arterial system has its own intrinsic resistance that is different from the venous system’s resistance. These points surely have some practical implications to resuscitation and hemodynamics.
With regard to the discussion you guys had about the patient on high dose vasopressors with mottled cold extremities, I think there is something to be said about where the BP comes from. I think that in a sick patient, what is generating the MAP is just as important as what the MAP is. Of course the three factors that generate the MAP are the CO, SVR, and CVP. Patients in sepsis have high CO and low SVR. The CVP is low as well for a variety of reasons. Partially because the hyperdynamic heart is pumping the blood like crazy and sucking the blood from the SVC/IVC, but also because of an increase venous (and arterial) compliance, which takes otherwise stressed volume that drives venous return and converts in to unstressed volume. The conventional wisdom that we’ve been taught is to give crystalloids to “fill the tank” but the real problem is lack of SVR and the heart tries to compensate by increasing the CO to maintain the MAP. I think the rational approach is to start vasopressors very early, rather than to give a lot of crystalloid, then start the vasopressors when they don’t work. I have even started to start my patients NE gtts with normal BPs if they look sick and have a high CO/low SVR/CVP syndrome. Maybe this is overkill, but they really look and feels better in my experience. Conversely, a patient with low CO, high SVR, and high CVP needs an inotrope to fix the problem. A patient with a MAP of 80, CO of 2.5, SVR of 2500, and a CVP of 20 probably looks very sick, but if you give them dobutamine, you may increase the CO to 4.5, decrease the SVR to 1200, and the CVP to 15, and they look MUCH better. What I’m saying is if you aim to normalize the CO, SVR, and CVP you will fix the MAP anyway, and you’ll have a good balance of flow (CO) and capillary perfusion pressure (SVR). Too much or too little of either one, there’s probably a significant hemodynamic problem. This also speaks to Phil’s point about the insanity of chasing some extra stroke volume (because they are volume responsive) when they already have a CO of 9 L/min.
Another point is that we can measure a lot of pressures with a pulmonary artery catheter and we can aim to get these pressures to physiologic variables, but this misses the fact that pressure does not equal preload due to differences in LV and RV ventricular compliance. Normal LVEDP/left atrial pressure is about 6-12 mmHg and most healthy people with normal LV compliance will have optimal LV preloading at 15 mmHg. Pulmonary edema doesn’t usually develop in people without leaky pulmonary capillaries until the LV filling pressures exceed 18 mmHg. Someone with sepsis induced diastolic dysfunction will need considerably higher LVEDPs to get the same LV preload. Furthermore, the pulmonary capillaries are likely leaky in bad sepsis, so you will get extravascular lung water at much lower LVEDPs than if you didn’t have sepsis. Because of this, POCUS/ECHO skills, I think, are way more valuable than a pulmonary artery catheter that just gives you numbers. With POCUS I can evaluate both the systolic (EF) and diastolic (E/e’ ratios) function, as well as monitor for B-lines, indicating extravascular lung water. As an aside, perhaps if you see the B-lines, you’ve already lost the resuscitation game as you already over volume resuscitated the patient. Perhaps this is where transpulmonary thermodilution devices have a role, but who knows? For what it’s worth, I’ve used them, and found them helpful in conjunction with ECHO. Back to the main point, ballpark values of LVEDPs can be calculated with ECHO anyway, as well as very accurate numbers for CO, which enable you to calculate SVR. For me a pulmonary artery catheter has nothing on ECHO for resuscitation.
Next, there’s the fact that the RV and LV are different animals, and they handle volume and pressure loads differently. They interact both in series and in parallel (because they share the inter-ventricular septum). Any complete understanding of what’s going on with your sick patient’s hemodynamics require some idea about what each ventricle is up to, how they are interacting, how relatively compliant they are, and what their contractility is.
With regard to overly aggressive crystalloid therapy, as mentioned, I think this is rampant. You guys touched on this already, but with regard to the LV, too much fluids puts the lungs and RV at risk for failure. As far as the RV goes, there’s interstitial renal edema (AKI) and iatrogenic portal hypertension/gut edema (continues the sepsis syndrome due to bacterial translocation and bad cytokine release) that occurs as a result of a reckless fluid administration. I have my almost hard stop points for fluids. E/e’ ratios and calculation of estimated LVEDPs, consideration of pulmonary vascular permeability, dilated IVC or hepatic vein Doppler with S'<D', portal vein pulsatility, or a congested intra-renal venous Doppler pattern all makes me think really hard about giving any more fluids.
So for me, unfortunately, there's no hemodynamic number or combination of numbers that can tell you what to do for your shocked patient. It would be nice for your resident to look at a machine and give you some numbers and you tell them what to do, but it's not reality. It's all about taking these cases one by one, personally examining and assessing your patient and tailoring the therapy after consideration of everything that's going on. This is hard to learn and takes a lot of work to master, but that's why we're resuscitationists and are good at what we do.