Indications for dialysis in the ICU

Why would you subject somebody to having their entire blood volume sucked out of their body and dragged though a fine filter? This is a favoured subject of the examiners. Previous questions on this topic include Question 16 from the second paper of 2011, Question 8 from the second paper of 2005, Question 6 from the second paper of 2003 and Question 2b from the second paper of 2001. Bellomo, in his chapter for Oh's Manual, lists the following "modern" indications for dialysis in the ICU:

  • Oliguria (less than 200ml in 12 hours)
  • Anuria (0-50ml in 12 hours)
  • Urea over 35 mmol/L
  • Creatinine over 400mmol/L
  • Potassium over 6.5mmol/L
  • Refractory pulmonary oedema
  • Metabolic acidosis with pH less than 7.10
  • Hypernatremia over 160mmol/L
  • Hyponatremia under 110 mmol/L
  • Temperature over 40°C
  • Complications of uraemia: encephalopathy, pericarditis, myopathy or neuropathy
  • Overdose with a dialysable toxin

It is possible to expand upon this list, but it would be impossible to improve upon its brevity, which is the quality of greatest worth to the time-poor exam candidate. One should probably internalise this table, given how likely Bellomo is to have written many of the renal questions in the exam.

ADQI recommendations for the commencement of dialysis

Renal Indications
  • Oliguria with volume overload
  • Oliguria is relative; urine output may be high and still inadequate in clearing the fluid.
  • Uremia with symptoms
  • Hyperkalemia ( K+ over 6.0)
  • Metabolic acidosis due to renal failure (pH < 7.2)
Non-renal Indications
  • Removal of dialysable toxins, i.e. ones which aren’t very lipophilic or protein-bound
    • Pretty much any drug with a volume of distribution less than 0.5L/kg

      If a toxin is equally well cleared by hemodialysis and hemoperfusion, then hemodialysis is preferred, because it will also correct any underlying acid-base disturbance.

    • Valproate
    • Theophylline
    • Salicylates
    • Ethylene glycol
    • Methanol
    • Lithium
  • Control of otherwise uncontrollable electrolytes
    • Hypercalcemia refractory to pamidronate, for one example.
  • Control of body temperature
    • An extracorporeal circuit can help control hypo or hyperthermia which is resistant to other methods of control.
  • Clearance of cytokines to decrease severity of sepsis
    • Still controversial. May be of use in patients with renal failure and sepsis.
    • No evidence that it helps in patients with sepsis who don’t have renal failure.
  • Removal of contrast agent
    • More relevant with old-school high-osmolar contrast

When to start dialysis in chronic renal failure

For people looking after patients longitudinally, the greater skill is knowing when to start the maintenance dialysis early, and knowing when one can safely wait. The timing of RRT is a matter of some debate. The IDEAL trial of 2010 compared early and late starters, and concluded that there was no mortality difference. This influenced some changes to the original 2002 European guidelines.

The indications for maintenance dialysis according to these changes are as follows:

  • Start when the GFR has fallen to below 15ml/min
  • Definitely start before the GFR falls to below 6ml/min
  • Symptomatic uremia
  • Inability to control fluid balance
  • Inability to control blood pressure
  • Progressive deterioration in nutritional status
  • Diabetics may benefit from an earlier start

When to start dialysis in acute renal failure

This is even less clear-cut. In many cases, it is possible to delay dialysis and make a series of medical and biochemical adjustments, with a resulting recovery of renal function.  In a proportion of patients, RRT may be avoided altogether. The patients with the lower severity of illness may get away with just diuretics and medical management. This issues is explored in greater depth in the chaptert on the timing of renal replacement therapy.


Chapter (pp. 540) 48   Renal  replacement  therapy, also by Rinaldo  Bellomo

Interestingly, an article by Bellomo and Ronco from 1999 also contains a list of indications very similar to the one from Oh's, but with slightly different criteria (eg. a temp of 39.5°C, and a urea over 30mmol/L). Also, it contains some lovely black-and-white pictures of 1990s dialysis machines. One may note that in the ensuing decade, they have become subjected to hipster design pressures, developing unnecessarily aerodynamic curves and shiny touchscreens. Not all of us view this evolution of style as an improvement. Some still prefer to be surrounded by equipment which resembles the set of Alien.

Bellomo, R., and C. Ronco. "Renal replacement therapy in the intensive care unit." Intensive Care Med (1999) 25: 781±789