Jun 302019

Jon-Emile S. Kenny MD [@heart_lung]

“I’ve looked at clouds from both sides now ...”

-Joni Mitchell


In a recent physiological study – brought to my attention by the erudite and kinetic [@iceman_ex] – the detailed physiological effects of intravenous fluids, norepinephrine and dobutamine were sequentially evaluated in patients with sepsis.  Sepsis was defined as a known or presumed infection coupled with hypotension and elevated lactate.

With recent, though imperfect, data on early norepinephrine in septic shock and the anticipated CLOVERS study years away from completion, we rely upon pointed assays of cardiovascular physiology during early and evolving sepsis as guideposts.  What are the specific cardiovascular effects of volume expansion, norepinephrine and dobutamine administration in un-resuscitated sepsis?  Professor Pinsky and colleagues offer us some captivating data that, potentially, contest conventions concerning the pillars of hemodynamic sepsis management.

What They Did

Fifty-five septic patients – defined above – were enrolled and had multiple hemodynamic measurements made during the sequential provision of intravenous fluids [30 mL/kg], norepinephrine and dobutamine per the Surviving Sepsis Guidelines.  Importantly, the patients enrolled had received no more than 250 – 500 mL of intravenous fluids.  Norepinephrine was given to maintain a MAP above 65 mmHg, if needed, following volume expansion; dobutamine was given following norepinephrine only if there was persistent organ dysfunction/hypotension.

Patients had full transthoracic or transesophageal echocardiography performed along with right atrial pressure monitoring and auto-calibrated pulse-contour analysis cardiac output assessments.  An array of hemodynamic metrics was obtained including arterial elastance [Ea – as a marker of total arterial impedance], left ventricular end-systolic-elastance [Ees – an index of LV contractility] and their ratio [Ea/Ees] as an indicator of ventriculoarterial coupling.  Additionally, left ventricular efficiency, total cardiac efficiency [Eh], mean systemic filling pressure, the gradient for venous return and dynamic arterial elastance were measured before and after the provision of volume expansion, norepinephrine and dobutamine – if needed.

What They Found

The findings of this investigation are manifold and extend far beyond what this simple summary can comprise.  Nevertheless, the salient findings are as follows.

The 30-day mortality rate for all 55 patients enrolled was 47%, indicating that the population was quite ill.  These patients had baseline right atrial pressure and LV ejection fraction values between 7 and 8 mmHg, and ~42%, respectively.  Because the Surviving Sepsis Recommendation is to give 30 mL/kg of volume expansion [VE] immediately, those enrolled did not have ‘volume responsiveness’ checked beforehand.  Despite this, 49 of 55 patients [89%] had an increase in cardiac index by at least 15% following initial volume expansion.

Both those patients who had a significant increase in cardiac index [n = 49] and those who did not [n = 6] increased left ventricular end-systolic elastance following volume expansion; Ees is a marker of contractility.  By contrast, those who increased cardiac index had a fall in arterial elastance [Ea] – a marker of arterial impedance, or ‘afterload.’  Thus, those patients who had an increase in cardiac index had a fall in the Ea/Ees index – or improved ventriculoarterial coupling.

While nearly all patients had an increase in cardiac index in response to volume expansion, 20 of 55 continued to have a MAP of less than 65 mmHg after intravenous fluids.  These patients therefore received norepinephrine infusions – 12 of them increased their MAP while 8 did not.  In totality, norepinephrine increased arterial elastance [Ea] more so than it increased Ees such that ventriculoarterial coupling worsened to its baseline state prior to fluids.  In those patients who had either an increase in cardiac index or MAP in response to norepinephrine, there was a higher pre-norepinephrine Ees – suggesting that a higher pre-norepinephrine contractility buffered the effect of an increase in arterial elastance Ea [i.e. afterload].

Dobutamine was given to 6 of the 8 patients who were hypotensive following both volume expansion and norepinephrine.  On average, and predictably, dobutamine increased Ees and restored ventriculoarterial coupling, increased stroke volume, and MAP.  Also as expected cardiac performance increased and right atrial pressure fell.  There was a larger increase in stroke volume with greater increase in Ees in response to dobutamine.


While I am no proponent of blindly giving intravenous volume expansion to patients with sepsis, this small study deserves our communal contemplation.  The large proportion of patients who were fluid-responsive in early sepsis echoes data from the recent, and much larger, ANDROMEDA-SHOCK study.  It is probable that ‘fluid unresponsiveness’ evolves over the course of sepsis and that this can be dramatic even in the first 6-8 hours.  One wonders how the capillary refill time of these 55 patients changed during this study – in response to intravenous fluids, norepinephrine and dobutamine?

Also of great consequence is that – in this study of septic patients with only mediocre baseline cardiac function [i.e. right atrial pressure of ~ 8 mmHg and LVEF in the low 40% range] – volume expansion increased cardiac output in the vast majority and did so, partly, by boosting cardiac contractility as measured by the LV end-systolic elastance.  How might this occur?  The authors themselves are uncertain but hypothesize that following volume expansion increased diastolic blood pressure raises the pressure head for coronary artery perfusion.  Satiating coronary oxygen demand in early sepsis, in theory, enhances cardiac squeeze.  Additionally, because fluids decreased arterial elastance [this has been previously reported], a diminished 'afterload' should decrease myocardial oxygen demand and, consequently, nurture cardiac myoenergetics.

Next, the provision of norepinephrine – somewhat predictably – increased arterial elastance, a marker of afterload.  In totality, the effect was to drop LV performance despite raising MAP above 65 mmHg in about 50% of those who received the drug.  While norepinephrine was given only after receiving 30 mL/kg of intravenous crystalloid in this study, it should still give pause; there is no flawless hemodynamic therapy in sepsis.  Vasoactive infusions, like intravenous fluids, have beneficial and harmful consequences; alas, everything is a titration.

In summary, this study does not provide decisive recommendations on the management of sepsis, but it does provide an enthralling, descriptive overview of the physiological consequences of common supportive therapies in sepsis and septic shock.  The findings also highlight the individualized nature of therapy.  While 89% of patients had a favourable response to an immediate 30 mL/kg bolus of crystalloid, this implies that – on a population level – tens-of-thousands of patients each year are receiving this intervention without circulatory benefit.  Similarly, in ANDROMEDA-SHOCK, 30% of patients were unresponsive to intravenous fluids in very early sepsis – arguably, causing patient harm.  We must also all remind ourselves that even if a patient's left ventricle is responsive to intravenous fluids, giving them may still be harmful and that other vital organs should be assessed when providing rational fluid therapy.  Nevertheless, as Joni beautifully reminds us, if we look at ‘both sides of the cloud,’ we see that intravenous fluids in early sepsis can be salubrious; arguably with more inotropic activity than norepinephrine!

Sweltering in Gordes, France,


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

Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


Intravenous Fluids Act as Inotropes in Recent Sepsis Study