The Oligo-Anuric Patient

A common question asked of ICU doctors every shift is “the patient’s urine output is low, what do you want to do? Give fluids?” What follows is a basic approach to oligo/anuria. Hopefully, by the end it will be clear why the solution to oliguria is most often not IV fluids!


The patient is passing less than 0.5ml/kg/hr urine. Often this is simplified to <30ml/hr.

Why does oliguria matter?

Oliguria suggests that there has been either a decrease in glomerular filtration rate or there is mechanical obstruction to urine flow. This may be due to:

  • decreased renal blood flow (from low cardiac output, distributive shock or hypovolaemia)
  • acute tubular necrosis from a period of hypoperfusion to the kidneys
  • mechanical urinary tract obstruction
  • hormonally mediated (as a stress response more ADH is secreted and increased activation of the RAAS system occurs).

So, we care because low urine output may signify that the patient is shocked, and we need to do something about it, or their urine flow is obstructed and we need to do something about it. We do not treat urine outputs however, we treat patients.


If a patient is hypervolaemic with an acute kidney injury (i.e. has peripheral oedema and/or is requiring oxygen) do not administer IV fluids for a low urine output. If they are requiring oxygen and they receive more IV fluids they almost invariably will have an increase in their FiO2 requirement. These patients may have inadequate renal blood flow due to distributive or cardiogenic shock or they may just have ATN and be unable to make urine.

If a patient is hypervolaemic and has an oxygen requirement they will benefit from diuretic therapy to decrease their pulmonary oedema. Frusemide is not a nephrotoxic drug. Frusemide is a vasodilatory agent and may actually increase renal blood flow! (This is also the mechanism by which it may cause hypotension). Hypovolaemia will result in worsening renal function, try to avoid giving frusemide to hypovolaemic patients.

A diagnostic and therapeutic approach to oliguria

Mechanical obstruction

  • Is the catheter blocked? Unblock it.
  • Have they had a CT KUB or renal tract ultrasound? All new acute kidney injuries should be investigated at the earliest possible time for obstruction to avoid potentially permanent injury.


  • Are they genuinely hypovolaemic? Based on history (thirsty, history of fluid losses [vomiting, diarrhoea, bleeding], clinical exam (absent oedema, dry mucous membranes, concentrated urine) and biochemically (hypernatraemia, disproportionately high urea).
  • If you’re sure they’re hypovolaemic then give them a fluid bolus (or blood if they’re bleeding). If their urine output doesn’t increase and their MAP is already >65mmHg then they may not be hypovolaemic and you should consider other causes. 250 – 500mls of balanced crystalloid in this context (i.e. they are hypovolaemic) is unlikely to have hurt anyone. If you’re given more than this consider carefully your diagnosis if there is no response.

Decreased renal blood flow

  • Is this cardiogenic shock? Are they already on an inotrope like dobutamine, milrinone or adrenaline and should it be increased? Indicators of this would be poor perfusion to skin (slow capillary refill time, weak radial pulses), or a rising lactate.
  • Is this distributive shock (i.e. sepsis)? Aim for a MAP >65mmHg, but if patients are hypotensive nursing staff won’t be asking you what to do about their low urine output. In chronic hypertension patients’ kidneys may be used to a higher MAP, consider aiming MAP >70mmHg if you have a clear history of poorly managed hypertension.
    • Renal Perfusion Pressure = MAP – Right Atrial Pressure (or ~CVP)
      • A reduction in MAP (outside the auto-regulatory zone) or an increase in atrial pressure (or CVP) will reduce renal perfusion pressure.
  • Is there obstruction to renal blood flow? Abdominal compartment syndrome may decrease urine output and the cause needs to be addressed. A simple cause is ascites, a more complex cause is catastrophic abdominal sepsis or bleeding which requires surgical intervention.

Acute tubular necrosis

  • Has the patient had a period of hypotension or a cardiac arrest? Their kidneys will have suffered a period of ischaemia and this may result in acute tubular necrosis. If you suspect this to be the case and feel that they may be hypovolaemic (see above) then try a fluid bolus. If their urine output and haemodynamic markers do not improve then STOP.

Hormonally mediated

  • Do none of the above really fit? Then it may just be the patient is retaining fluid like all physiologically stressed patients do due to increased circulating catecholamines, increased RAAS (renin-angiotensin-aldosterone system) activation and increased ADH secretion. Don’t give them fluids.

Using diuretics to manage volume state

Fluid administration to a shocked or hypotensive patient is like taking out a loan. If you give a person who is euvolaemic or hypervolaemic IV fluids for their haemodynamic instability they will need to urinate it out at some point. If you give a patient 3 litres in 24 hours and they only urinate 300mls, that 2.7L will need to come out at some point. In the meantime it sits in their lungs, liver, kidneys, bowel or subcutaneous tissues. Waterlogged organs do not work well.

Additionally, large volume fluid administration as part of haemodynamic resuscitation, will often manifest itself a few days later with a clinically hypervolaemic patient. The term “de-resuscitation” has been coined to describe the period in a patient’s critical illness journey when haemodynamic stability has been achieved and the goal is to transition to a negative fluid balance.

So you have a hypervolaemic patient, and you want to make them less hypervolaemic:

  • Give frusemide, if their renal function is good (GFR >60) then 40mg is a reasonable starting dose. If their renal function is not good (GFR<60) then you’ll need more, 80mg is reasonable. If they do not respond to multiple doses then consider a frusemide infusion.
  • A frusemide infusion can be increased to up to 100mg/hr, titrate to the desired urine output and required negative fluid balance.

After days of frusemide therapy you will find your patients becoming hypernatraemic, hypokalaemic and developing a metabolic alkalosis. This biochemical derangement may become prohibitive in the ongoing diuresis of your hypovolaemic patient. Some useful drugs to consider:

  • Hypernatraemia your problem? Hydrochlorothiazide is a solution. This is a Na/Cl transporter that acts on the proximal convoluted tubule of the nephron. It will result in a higher urine sodium and will decrease your patients’ serum sodium. 25mg BD is a reasonable starting dose.
  • Hypokalaemia your problem? Spironolactone is a solution. This works on the Na/K transported in the distal convoluted tubule and will result in less potassium being excreted in your patients’ urine. 25mg BD is a reasonable starting dose.
  • Metabolic alkalosis your problem? Acetazolamide is a solution. This is a carbonic anhydrase inhibitor, which will result in more HCO3 secreted in your urine, less H+, resulting in a decreased pH and decreased serum HCO3. 500mg BD is a reasonable starting dose.

When diuretics fail, it’s time to consider renal replacement therapy

Author: Nick Ryan

Leave a Reply