Hemodialysis in metformin poisoning

May 8, 2015, 12:45 am


Extracorporeal Treatment fo Metformin Poisoning: Systematic Review and Recommendations From the Extracorporeal Treatments in Poisoning Workshop. Calello DP et al. Crit Care Med 2015 Apr 9 [Epub ahead of print]



Metformin is now the oral drug most commonly prescribed to treat non-insulin-dependent diabetes in the United States. The drug is large eliminated by the kidneys. Toxicity presents with severe lactic acidosis, and can occur when decreasing renal function causes accumulation of therapeutic doses, or in the case of acute deliberate overdose. According to some reports metformin poisoning, though rare, is associated with a mortality rate of 30%.

Metformin is a small molecule that is not protein-bound, making it amenable to dialysis despite having an appreciable volume of distribution (1-5 L/kg.) This paper, from the Extracorporeal Treatments in Poisoning (ExTRIP) Workgroup, aimed to review systematically literature relevant to the use of extracorporeal methods to treat metformin poisoning, and present recommendations for using such methods.

The authors identified 175 relevant articles, including case reports and observation studies. Unfortunately, as in this group’s previous papers regarding different poisons, the quality of the evidence was so poor that all of their recommendations were rated 1D (1=strong recommendation, D=very low level of evidence). As we’ve pointed out in discussing the group’s earlier papers, this makes their conclusions opinions rather than evidence-based recommendations. However, given the expertise and experience of members of the group, their opinions must be reckoned with.

The paper concludes that metformin is dialyzable and that extracorporeal treatment (ECTR) is recommended in “severe” metformin poisoning. Specific suggested indications are as follows:

  • ECTR is recommended if:
    • Lactate > 20 mmol/L
    • Blood pH < 7.0
    • Treatment with supportive care and bicarbonate fails

The group’s “suggestions” are less strong than their “recommendations,” reflecting less of a consensus among the members:

  • ECTR is suggested if:
    • Lactate > 15-20 mmol/L
    • Blood pH < 7.0-7.1

The group also suggests that presence of the following conditions should lower threshold for starting ECTR:

  • shock
  • renal insufficiency
  • liver failure
  • decreased level of consciousness

The group recommends that intermittent hemodialysis should be the initial ECTR method of choice if available, but that continuous renal replacement therapy (CRRT) may be helpful if it is not. Subsequent ECTR can involve either hemodialysis or CCRT. The end point should be lactate < 3 mmol/L and pH > 7.35. Since the metformin level can rebound and increase after ECTR is stopped, patients’ acid-base status should be followed closely until it is clearly stable.

Again, all of these recommendations and suggestions are based on very poor evidence — so poor, in fact, that the authors describe their analysis as somewhat of a “guess.” But the paper is well worth reading and represents a good starting point for thinking about ECTR in metformin poisoning. Recommended.

Related posts:

Hemodialysis in acute methanol poisoning: is there really good evidence?

Hemodialysis in lithium poisoning: what is the evidence?

Must-read: consider hemodialysis in cases of massive acetaminophen overdose

What enhanced elimination techniques are useful in critical toxicology patients?

Hemodialysis and other extracorporeal modalities in toxicology cases



Comments are closed.