Does early GI decontamination improve outcomes in acute-on-chronic lithium toxicity

April 19, 2013, 5:55 pm


Lithium poisoning: the value of early digestive tract decontamination. Bretaudeau Deguigne MB et al. Clin Toxicol 2013 May;51:243-248.


Although whole bowel irrigation is being used less and less as a means of gastrointestinal decontamination in cases of overdose, one possible remaining indication is a potentially significant ingestion of a poison that is not well absorbed to activated charcoal. Practically, this applies to ingestion of heavy metals, including lithium, iron, and the occasional lead chip. The objective of this French study was to . . . well, I’m not sure I can tease out a convincing objective.

The authors state that:

The objective of this study was to evaluate the effects of early digestive tract decontamination on the severity of acute-on-chronic lithium poisoning (acute poisoning in patients under long-term therapy).

Without specifying specific outcome measures, the authors open up the possibility for all sorts of data dredging and post hoc analysis. But this is just the start of the problems with the paper.

The authors retrospectively searched the records of the Angers Poisons and Toxicovigilance Centre for cases of acute-on-chronic lithium overdose. They identified 59 patients. Fifteen patients had received early GI decontamination and 44 had not. (“Early GI decontamination” was defined as whole bowel irrigation (WBI) and/or administration of sodium polystyrene sulphonate (SPS) within 12 h of ingestion).

The two groups were similar as to mean age, estimated dose ingested and documented poisoning severity score (PSS) during the first 12 h after ingestion. A higher percentage of “early decon” patients had ingested a sustained-release preparation.

Although the authors claim their statistical analysis found that early decon was significantly associated with a lower risk of severe poisoning, their criteria for determining this were so vague that I thought the claim unconvincing. It would be a good journal club exercise to list all the design flaws in this paper. One could start with these:

  • Being an observational study, there was no standardization as to which patients received early GI decontamination, who received WBI or SPS, or who underwent hemodialysis.
  • The study population was small, and was not analyzed with regard to what type of early devon — WBI, SPS, or both — was administered.
  • No information is provided as to how long after ingestion the lithium level was drawn, or PSS determined.
  • While the definition of severe poisoning seems to have been based on a PSS > 2, a PSS of 2 is actually considered moderate, and I wonder if this definition was decided on post hoc.

The authors conclude that:

Our results, nevertheless, highlighted the possible usefulness of early digestive tract decontamination [in acute-on-chronic lithium toxicity]. . . . Randomized, controlled and prospective studies will now be necessary to enable a separate assessment of these two techniques for digestive tract decontamination.

Look, one might argue, despite the flaws inherent in a retrospective observational study this paper is valuable to generative a hypothesis that, as the authors suggest, can be subject for further study. Nonsense. It is clear to me that a good randomized control trial of this question will never be carried out — neither the will nor the funds to support such a study exists. While there are theoretical reasons to think that early WBI in these cases improves outcome, and the intervention is recommended in many texts, in my opinion we will never demonstrate this scientifically. It will continue to be a matter of faith.

Comments are closed.