Accelerated Goal Directed Therapy for Septic Shock

The Surviving Sepsis Campaign has raised awareness that septic shock is a medical emergency.  However, these guidelines recommend a stepwise approach to resuscitation, which commonly results in a gradual escalation of treatment intensity.   Additional therapies are added over several hours if the patient fails to reach treatment goals.  For some patients, this approach may not be rapid enough to get ahead of the disease process.  

Accelerated goal directed therapy is a streamlined approach designed to escalate resuscitation more rapidly and achieve stabilization more quickly.  This is primarily designed with the sickest patients in mind.  However, when in doubt, it may be safer to err on the side of aggressive stabilization followed by prompt de-escalation once the patient is recovering. 

A typical approach to septic shock is shown above.   Initial therapy typically consists of antibiotics and 4-6 liters of fluid resuscitation, often over several hours.  If this fails, a series of vasopressors are added sequentially.   Finally, steroid is initiated for patients with vasopressor-refractory shock.   The time between initiation of treatment and maximally aggressive therapy is often 6-12 hours.  The IMPRESS trial, a multinational survey of sepsis care released this month, revealed only 66% compliance with the use of vasopressors for hypotension within six hours among the sickest cohort of patients.  

The sequencing of this approach is suboptimal for the sickest patients.  It used to be believed that vasopressors only increased afterload and contractility.   In that case, it would make sense to provide fluid first to "fill the tank," before starting vasopressors.   However, we now understand that norepinephrine causes venoconstriction as well as arterial constriction, thereby increasing the preload and "filling the tank" by itself.  By increasing preload, afterload, and contractility simultaneously, norepinephrine stabilizes circulation immediately and reverses the physiologic derangements of early sepsis (which is a vasodilatory shock state).  In contrast, severe septic shock responds poorly to fluid, so a fluid-first approach often delays stabilization.

Early initiation of bactericidal antibiotics is universally recommended.   As discussed previously here, evidence supporting this is correlational.  In some cases bactericidal antibiotics release inflammatory bacterial products into circulation  (e.g. lipopolysaccharide from gram negative rods).  Releasing this inflammatory material without providing other therapies to stabilize the patient may cause deterioration.

Thus, a strategy of starting with antibiotics and fluid alone may cause temporary improvement while setting the stage for subsequent hemodynamic collapse.  Inflammation combined with intravascular volume overload may damage the endothelial glycocalyx, causing capillary leak.  Although providing fluid may temporarily improve hemodynamics, this fluid may be rapidly lost from leaky capillaries with rebound deterioration.  This may explain the results of the FEAST trial, a study of septic children in Africa wherein administering fluid boluses seemed to cause initial improvement but actually led to delayed hemodynamic death (Maitland 2011).

Our current management of sepsis is based on the concept of Early Goal-Directed Therapy (EGDT), wherein therapies are escalated depending on the patient response.  Ideally this would work as shown below, with therapies carefully titrated to match the intensity of the disease.  

However, some very sick patients may deteriorate while resuscitation is being escalated (figure below).  This may result in a delay or inability to stabilize the patient.   By the time resuscitation has escalated, the disease has already spiraled out of control: 

With accelerated goal directed therapy, escalation is not a finely titrated process.  Instead it is accepted that the initial resuscitative efforts will exceed the minimum level required to stabilize the patient.  Subsequently, as the patient improves, the intensity of resuscitation is carefully reduced.  

The surviving sepsis guidelines include a number of complex resuscitation goals which may delay stabilization.   For example, mixed venous oxygen saturation requires placement of a central line, obtaining a blood sample, and sending it to the lab.   Lactate levels requires two sequential blood samples to be drawn and sent to the lab to reveal the change of the lactate level.  Any resuscitation strategy built around these goals will inevitably be delayed. 

Goals for accelerated resuscitation must be easily and rapidly measurable at the bedside.  A reasonable set of goals may be: 

(#1) MAP goal:  Achieving a mean arterial pressure (MAP) capable of perfusing the vital organs is essential.  Usually a target of 65mm is selected initially.  As discussed further below, this goal should ideally be achieved almost immediately (e.g. within 10-15 minutes) using peripheral vasopressors.  An arterial catheter is desirable for most patients, but achieving the MAP goal should not be delayed while awaiting an invasive blood pressure measurement. 

(#2) Perfusion goal:  Urine output may be the most clinically relevant measurement of organ perfusion.  Unfortunately in some cases urine output cannot be evaluated (for example, in patients with acute tubular necrosis or chronic renal failure).  In such cases, extremity perfusion (e.g. mottling, temperature, capillary refill) might provide a rough estimate of perfusion.

One example of how accelerated goal directed therapy might be applied is shown above, with the components discussed below as follows:


The concept of starting norepinephrine immediately to stabilize hemodynamics and defend organ perfusion was discussed in detail previously here.   In short, for a patient with severe shock and hypotension, there is nothing to be gained by delaying a norepinephrine infusion.  The safety of initiating a peripheral norepinephrine infusion is increasingly supported in the literature (e.g., Cardenas-Garcia 2015). 

Fluid resuscitation may be started simultaneously with norepinephrine.  However, fluid overload correlates with mortality, renal failure, and may perpetuate a state of chronic septic shock as discussed previously here.  Thus, the ideal strategy may combine norepinephrine with a moderate amount of fluid, for example 2-3 liters.   Norepinephrine may increase the preload due to venoconstriction, maintaining an adequate preload while avoiding volume overload. 

Vasopressin is often conceived as a treatment for catecholamine-refractory shock.   However, this use has not been borne out by the evidence.  Retrospective subgroup analysis of the VASST trial suggested instead that vasopressin might be more effective when used in patients with mild shock (patients on 5-15 mcg/min norepinephrine).  Other RCTs of vasopressin have similarly found improved renal outcomes when vasopressin is initiated early in the course of sepsis (studies are reviewed here).  Therefore, if there is a role for vasopressin in sepsis, it should probably be started early.

Thus, my approach is usually to add a fixed, low-dose vasopressin infusion of 0.03 units/minute when the norepinephrine is running at a low rate (i.e. ~10 mcg/min).  The goal of the vasopressin isn't necessarily to increase the blood pressure but rather to improve renal function. 

There are many reasonable approaches to the use of vasopressin, and indeed it would be reasonable not to use it at all.  One approach which is probably unhelpful is to wait until the patient is refractory to high-dose norepinephrine and then see if adding vasopressin will stabilize the patient.  Adding low-dose vasopressin to high-dose norepinephrine probably won't make a big impact.  Instead, ordering a vasopressin infusion and waiting to see if it will work may delay initiation of a more effective agent (i.e. epinephrine).  So, if you're going to use vasopressin, it may be best to start it early in the resuscitation, without delaying other therapies. 

Patients refractory to norepinephrine and vasopressin often have adequate afterload but may benefit from increased inotropy.   Epinephrine functions as an inotrope at low doses (0-10 mcg/min), with additional vasoconstrictive activity at higher doses.  Epinephrine titration therefore is a simple way to first provide inotropy, and then provide additional vasoconstriction if necessary.   This may achieve hemodynamic stabilization faster than, for example, performing a complex titration involving norepinephrine, vasopressin, dobutamine, and phenylephrine.   The benefits of using epinephrine as a second-line inopressor were explored further in a prior post here(1). 

The use of steroids in septic shock was explored in detail last week.  In short, steroids are neither as beneficial nor as dangerous as often thought:  they do not decrease mortality, but neither do they increase the risk of superinfection.  Steroids have been consistently shown to improve hemodynamic stability and reduce the duration of shock.   It remains unknown which patients might benefit from steroids, perhaps patients who are the sickest and lack contraindications.    

Usual practices regarding the timing of steroid initiation are paradoxical.  A common misperception is that the benefit of steroids is restricted to patients with vasopressor-refractory shock.   This leads to the practice of waiting until the patient is refractory to vasopressors and on the verge of death before starting steroids.   However, it seems likely that steroids would be more effective if started earlierin the disease process. 

The ideal timing and patient selection for steroids is unknown.  However, in a patient with severe shock who is responding poorly to initial therapies it seems reasonable to start steroids sooner rather than later.

Conflicts of Interest:  None.