Winter and spring have been unseasonably wet for central Victoria. With plenty of moisture in the soil, things were looking green and gorgeous so when a friend asked me to farm-sit for him I jumped at it. Looking after horses, a moustachioed cat and a beautiful border collie were a small price to pay for enjoying a landscape of granite boulders and grape vines. That was until the dog ran away.
Dusk arrived and Annie was still missing in action. Terrified by the idea of telling my friend about his lost dog, I decided to chase her down. I had been warned about the risk of getting bogged if I deviated off track…. meh! The ground seemed solid, I had a decent 4wd, and I clearly knew better. Spotlights on and eyes scanning, I started cruising across the paddocks. But pride comes before the fall. It was only when the ute suddenly stopped moving did I appreciate that the ground was less fluid-tolerant than my gestalt told me. Lost car, lost dog, lost for words (other than 4-letter ones), I tried to make meaning of the day’s events and realised it was a teachable moment in the management of sepsis in the emergency department.
When patients arrive in the ED with febrile hypotension, we routinely prescribe in excess of 20-30cc/kg of salt water. The pseudoaxiom is that fluid bolus therapy will improve macrocirculation (arterial blood pressure), and therefore improve microcirculation (tissue perfusion). This is unlikely to be true. Empirically “aggressively resuscitating” people with fluid boluses probably causes harm. Like the fool in the Hilux chasing the missing dog, we chase haemodynamic targets with unproven watery therapies (or disproven if you’re a marginalised African child) that are physiologically suspect and warrant close examination.
The pathophysiology of sepsis is complicated. The basic mechanisms of the disease, however, (at least as we currently understand it) are less complex: vasodilation and glycocalyx (GCX) dysfunction. Organ dysfunction in septic shock can largely be attributed to one or both of these mechanisms. It is not due to hypovolaemia.
Currently there are no treatments for GCX stabilisation (unless you are a Scandinavian neonate having open heart surgery in which high doses methyl-prednisolone seems to reduce concentrations of a plasma syndecan-1; an alleged surrogate for GCX dysfunction). Current treatments for vasodilation include noradrenaline, adrenaline, vasopressin, methylene blue, and angiotensin-2. In sepsis, fluids cause organ dysfunction through worsening interstitial oedema due to GCX dysfunction, and cause vasodilation by stimulating release of naturetic peptides. It is therefore bizarre that it should be used as a first line therapy for septic shock.
Pseudoaxiom one: Fill the tank before you squeeze.
There is no tank to fill in sepsis, and a vasodilated state is probably best managed with vasoconstrictors. Giving a septic patient a fluid bolus will increase cardiac filling pressures, triggering release of naturetic peptides which cause vasodilation. Thus in sepsis, fluids can be considered a vasodilator therapy. If clinicians are concerned that there is inadequate preload, the LV end-diastolic volume should be measured with echo.
Pseudoaxiom two: fluids improve stroke volume
Patients with septic shock have a depressed Starling curve with a reduction in recruitable contractility via increased preload. > 50% of patients with septic shock have diastolic dysfunction which responds poorly to fluid therapy.
Pseudoaxiom three: fluid stays in the circulating volume
In patients with septic shock less than 5% of administered fluid remains in the intravascular space at 1hr. This fluid leaks from the vascular compartment to enter the interstitium, causing organ dysfunction. The Marik-Philips EVLW curve illustrates the respiratory harms of fluid therapy in patients with increasing filling pressures. In the abdomen, increased initerstital oedema causes intra-abdominal hypertension, gut failure and renal failure.
Pseudoaxiom four: albumin is the answer when crystalloid fails
The disrupted GCX enables translocation of albumin into the interstitium, where it continues to exert an osmotic effect causing further interstitial oedema.
The concept of fluid tolerance in pursuit of haemodynamic “stability” in shocked septic patients is as ill-fated as a hunt for a lost dog across a muddy paddock in the dark. Even though there was no surface water visible, the ground swallowed up my ute before I even realised I was on a path to trouble. So too do we drive our septic emergency department patients further into multi-organ dysfunction with iatrogenic salt-water drowning.
So how do I manage septic shock in the ED? After antibiotics are on board I perform a focused haemodynamic assessment using echo to examine preload (LVEDD or LVEDA), contractility (fractional shortening or fractional area change), filling pressures (interatrial septal motion), and diastolic function (E/A, E/e prime). This takes less than 5 minutes. If patients have had no recent oral intake I replace guesstimated deficits (4, 2, 1 method) then commence maintenance fluids (D5W and providing Na, K, Mg as required [N.B. Australian RDI of sodium is 40mmol, not 154mmol]). I concurrently target a MAP of 65-70mmHg using a combination of noradrenaline, vasopressin and adrenaline, depending on the haemodynamic state.
So what happened to the ute you ask? Like the drowned patient needing CRRT and an inpatient bed before they break the NEAT 4-hour rule, I had to phone a critical care colleague who spent 4 hrs helping me dig the car out of the quagmire and haul it to dry ground. Annie came back on her own volition without intervention and I scored a well earned “told you so” from her master.
With thanks to Paul Marik, Thomas Woodcock, Rinaldo Bellomo and John Myburgh for inspiring me to care about fluids.
Special thanks to Caitlin Young for helping to dig me out of my stupidity.