Echocardiography Does Not Correlate with Volume Status says CoDE-MiN Study
Jon-Emile S. Kenny MD [@heart_lung]
“Nolite te bastardes carborundorum”
-Margaret Atwood
Case
An 86 year old woman presents with two days of decreased intake by mouth and 3 days of melena and vomiting. She is followed closely by her internist and cardiologist for hypertension, severe mitral regurgitation, pulmonary venous hypertension and right ventricular dysfunction with co-morbid atrial fibrillation. She normally takes an ACE inhibitor, beta-blocker, warfarin and a statin as an outpatient. One week prior to her presentation, she experienced generalized muscle aches after a long day of gardening and treated herself with acetaminophen; her presenting INR is 5.7. In the ED, the patient is in rapid atrial fibrillation at 113 beats per minute and has a blood pressure of 115/40. The admitting resident is worried about transfusing the patient given her known cardiac co-morbidities so he performs point of care ultrasound. On examination, the patient’s inferior vena cava is large and unvarying. He withholds blood transfusion and calls his ICU fellow for assistance.
Background
One of the most important published figures in the history of critical care literature was divulged in 1984 by Shippy and colleagues. This very simple, yet – seemingly – ubiquitously overlooked illustration plotted volume status on the x-axis and central venous pressure [CVP] on the y-axis. The dotted diagram can best be described as an amplified view of a pointillistic painting in the style of Signac or Seurat. At the moment when I first genuinely contemplated Shippy’s figure as a resident – in the heyday of Early Goal Directed Therapy – the absurdity of everything I had been taught crumbled into the little points before me. I was reminded of the scene in Ferris Bueller’s Day Off when Cameron gazed at “A Sunday Afternoon on the Island of La Grande Jatte.” At this moment – his transition from adolescence to adulthood – he appears to realize the grand preposterousness of what he heretofore held real.
While Shippy and colleagues evaluated the CVP – a pressure – the volume of a cardiac chamber or great vessel follows a similar physical argument; pressure and volume relate by compliance but both cardiac pressures and volumes are physiologically removed from a patient’s overall ‘volume status.’ I have previously argued that the use of echocardiography as an estimator of volume status is akin to using the volume of water in the hull of a leaking ship to predict the size of its surrounding ocean. With this, I was interested to read the recently-published CoDE-MiN study.
What They Did
As an objective measure of ‘volume status,’ the authors used the equilibrium-pressure in the radial artery [Parm] following rapid [i.e. less than one cardiac cycle] inflation of a specialized tourniquet cuff proximal to a radial arterial line. The equilibrium-pressure in the radial artery during no-flow approximates the mean systemic filling pressure [Pmsf], itself directly-determined by the stressed blood volume and inversely-related to venous compliance; this physiology has been previously-explored.
Twenty-seven, elective neurosurgical patients were studied prior to general anesthesia; Parm was obtained 30 seconds following rapid cuff occlusion, as well, echocardiography was performed. Basic measurements such as left ventricular end-diastolic volume, area, E wave velocity, e’, the E/e’ ratio as well as end-expiratory inferior vena cava [IVC] diameter were performed. All patients then received a 500 mL, room-temperature, saline bolus over 10 minutes. Following crystalloid expansion, both Parm and echocardiography were repeated.
What They Found
Nineteen patients had an increase in Parm following the crystalloid infusion – on average from 21 mmHg to 26 mmHg. Of note, however, 8 patients displayed a decrease in Parm following the infusion – from 25 mmHg to 21 mmHg. No consistent correlations were noted between the Parm and echocardiography before and after fluid. In the multiple regression analysis, the changes in Parm only weakly correlated with IVC and IVC indexed to body surface area; further, no significant correlations were found between change in Parm and LVEDA/LVEDAi, LVEDV/LVEDVi or E/e′ changes.
Thoughts
The authors of the CoDE-MiN study should be commended for their effort to objectively study a marker of total volume status in conjunction with echocardiography. Nevertheless, the fact that Parm fell in such a large portion of patients following a crystalloid infusion should raise substantial concern over the use of Parm as a surrogate for total blood volume, or ‘volume status.’
Firstly, even if Parm were a perfect proxy for mean systemic filling pressure [Pmsf] – the latter is imperfectly related to total volume status. As demonstrated decades ago, and reviewed here, Pmsf can vary by as much as 7 – 10 mmHg without any change in blood volume – because of changes in venous tone and, therefore, venous capacitance. The converse is also true, blood volume can rise without any change in pressure. How? If blood volume increases, but there is a simultaneous fall in stressed venous volume [increased venous capacitance], then the recoil pressure [Pmsf] of the veins will not change.
The authors perceptively discern that the group of 8 patients in whom the Parm fell following volume infusion appeared to have higher baseline adrenergic tone [higher heart rate, higher diastolic blood pressure]. If crystalloid volume were to reduce adrenergic tone, it is quite plausible for Pmsf to also fall – raising concern, generally, for employing pressure as a measure of volume.
Secondly, the Parm may not perfectly track Pmsf nor may Pmsf completely follow mean circulatory filling pressure. In a previous study by Dr. Pinsky’s group, changes in Parm and Pmsf tracked each other very well. Importantly, however, these changes were detected in heavily-sedated, post-cardiac surgery patients receiving mechanical ventilation. It is possible that adrenergic tone is less variable in a sedated cohort. This is important because it is physiologically plausible for an increase in adrenergic tone to decrease Pmsf. How might this be physiologically realized?
The secret may be the difference between mean circulatory filling pressure [Pmcf] and mean systemic filling pressure. The mean circulatory filling pressure reflects both the systemic and pulmonary vasculature. Thus, the Pmcf mirrors the total blood volume more so than the mean systemic filling pressure when blood selectively partitions into the pulmonary circulation and ‘hides’ between the right and left heart. One could imagine a situation where provision of relatively cold [i.e. room temperature] crystalloid in anxious, awake, pre-operative patients could preferentially increase mesenteric arterial resistance, recruit unstressed splanchnic venous volume and decrease splanchnic venous resistance. As elegantly described by Gelman over a decade ago, the aforementioned changes in the splanchnic circulation could raise pulmonary blood volume at the expense of the systemic circulation as blood is jettisoned into the hepatic veins, IVC and then pulmonary tree. In other words, total circulatory blood volume rises with saline infusion, but there is an increase and decrease in the pulmonary and systemic blood volumes, respectively. This is yet another mechanism whereby estimates of pressure confuse volumetric reality.
Finally, and regardless of my mental meandering above, my previous concerns still hold. Even if radio-labeled albumin was used to accurately quantify total blood volume, inspecting vascular structures to know a patient’s total blood volume is fraught with physiological failings; the volume of any vascular structure is determined by its time-averaged inflow relative to outflow. ‘Volume status’ is but one fickle variable affecting cardiac inflow – high inflow met with equally high outflow will produce a small chamber, and vice versa.
As an analogy – in the immeasurable Pacific Ocean, a leaking ship with an excellent bilge pump will have a fairly dry hull; conversely, in diminutive Lake Ontario an overloaded hull will develop should the unfortunate boat have impaired pump function.
Return to Case
The ICU fellow arrives; she takes a history and performs a focused physical examination. The patient is short of breath, but has no crackles on examination and no B-lines on lung ultrasound. Her hemoglobin is 5.7 g/dL. The fellow explains the underlying physiology of the patient in the Guytonian sense and the patient is given both packed red blood cells and more fresh frozen plasma. The patient is brought to the step-down unit for endoscopy. After receiving all ordered blood products, the patient’s heart rate is 88 bpm with a blood pressure of 135/72 mmHg. Notably, reassessment of the patient’s IVC following volume expansion shows that it is decreased in size and that her E/e’ fell from 20 pre-transfusion to 12. The patient’s dyspnea is now gone and she remains without crackles on lung examination.
Best,
JE
Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he is also the creator and author of a free hemodynamic curriculum at heart-lung.org