The folly of chasing urine output with fluid in sepsis

Through medical school and house office years it is easy to develop many ‘reflex’ responses to certain conditions. Treating low urine output with fluid is one of these.

This makes sense in certain conditions. Hypovolemia leads to reduced renal perfusion. Correcting hypovolemia is therefore a good thing. Aggressive fluid resuscitation to restore renal perfusion makes sense in conditions where the patient is significantly fluid deplete, like enteritis or DKA, or where diuresis is helpful to prevent nephrotoxicity such as rhabdomyolysis or tumour lysis syndrome.

Unfortunately this has been extrapolated to every condition associated with AKI, resulting in massive fluid volumes being given to patients in the hope that the fluid will somehow drive the kidneys to work better. This is entirely devoid of physiological sense. This is most apparent in septic conditions. I particularly recall patients on the surgical ward with pancreatitis, who were given fluids for their oliguria and renal failure until they were swollen like the Michelin Man.

Chest Journal has published an article this month (1) addressing fluid management in acute kidney injury. This narrative is supported by a 2017 article of the same name. There are a few things to note.

Firstly, there is really no scientific evidence that macrovascular renal blood flow is routinely compromised in sepsis. Septic patients may be hypovolemic due to fluid shifts into the extracellular space but generally the problem is one of vasodilatation. The pathophysiology of acute kidney injury in sepsis is complex and involves tubular apoptosis and dysfunction at the cellular level.


Pathophysiological data do not show any increase in urine output in sepsis following a fluid bolus, despite an increase in blood pressure of 8mmHg that was sustained for an hour. A large animal model of sepsis showed no change in renal perfusion pressure, global renal blood flow or renal cortical flow following fluid bolusing (1).

There is also literature that excessive fluids make septic AKI worse. One RCT looking at restrictive fluids in sepsis versus standard care showed more AKI in the standard care group. Various observational studies have associated larger positive fluids balances in AKI with worse outcome (2).

There are several potential mechanisms. Firstly, the kidney is an encapsulated organ, and any oedema in the renal capsule will compromise the net pressure driving filtration. Secondly, just like the heart the kidney has a supply-demand balance and it seems likely that making the kidney do the work of filtering and reabsorbing sodium when all it wants to do is rest is not a good idea.

It is of course reasonable to rule out hypovolemia as a cause of renal dysfunction. The logical way to tackle this conundrum is to give a measured dose of fluid, and attempt to observe a response. Pounding three litres of fluid into a patient regardless of the response you get is not logical. Pounding three litres in to get a response of 20ml/hr extra urine is not logical. As Einstein said, ‘only a fool does the same thing over and over expecting a different result’.

When the primary problem is occurring on a cellular level repeatedly administering fluid without evidence of volume responsiveness will not help, may make things worse, and will impair the function of other organs, especially the lungs.


  1. Marlies Ostermann PhD, Kathleen Liu MD, Kianoush Kashani. Fluid Management in Acute Kidney Injury. Chest April 2019
  2. Anders Perner1* , John Prowle2 , Michael Joannidis3 , Paul Young4 , Peter B. Hjortrup1 and Ville Pettilä. Fluid management in acute kidney injury. Intensive Care Med (2017) 43:807–815 DOI 10.1007/s00134-017-4817-x

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s