Early digestive tract decontamination in acute-on-chronic lithium poisoning does not call conventional therapy into question

To the Editor:

In our preceding paper we investigated early digestive tract decontamination in acute-on-chronic lithium poisoning. Our study does not imply any reconsideration of conventional therapy by haemodialysis or saline infusion. In our study, 54/59 patients received normal saline infusion and the number of patients infused was not statistically different in both groups (p = 0.6). The exact amount of saline serum was unknown but in most cases the volume used was between 1 and 2 L a day, to maintain normal blood volume.

Given the retrospective design of the study, it was not possible to assess renal function of all patients in the first 12 h after ingestion. However, the patients’ serum creatinine level on hospital admission was known in 49/59 cases (12/15 in the early decontaminated group and 37/44 in the comparison group). Patients’ serum creatinine level on hospital admission (therefore, before rehydration) did not differ statistically between the two groups (median 75.5 vs 82.25 μmol/L, p = 0.18). Of the 11 patients who received dialysis (all in the comparison group), only seven had an acute renal failure.

In order to describe clinical severity of poisoning, we chose the Poisoning Severity Score (PSS) rather than Hansen and Amdisen's criteria¹ for several reasons. Hansen and Amdisen's criteria only assess neurological features, while the PSS assesses the overall clinical and biological severity. Moreover in a case series involving 212 patients, the Hansen and Amdisen classification does not appear to be a useful clinical tool to predict either morbidity or mortality.² We cannot agree that difference in neurological features on hospital presentation may have influenced the decision of the physicians in charge to perform GI decontamination because the PSS during the first 12 h was not statistically different in the two groups.

As you point out regarding the need for mechanical ventilation or haemodialysis, a statistically significant difference between the two groups could not be demonstrated, probably because of a lack of statistical power. But despite this lack of power we were able to demonstrate a difference in other relevant criteria such as admission to the ICU, the Glasgow coma score and the overall PSS. We believe that a gap PSS score of almost one point on a scale of 0–4 is a relevant difference. The PSS score difference of 0.7 – observed in our study population with a pooled standard deviation equal to 0.92 – corresponds to an effect size equal to 0.76, that is a medium to large effect according to the Cohen classification.^3–4 As you know, there is no consensus for dialysis criteria. As already expressed in our paper, these criteria were very heterogeneous and dependent on the ICU physician in charge. Even with sufficient statistical power, it would have been difficult to highlight a difference between the two groups. Finally the length of hospital stay does not seem to us to be an appropriate end point. Indeed, in a retrospective study, the length of stay has not been previously defined and often includes psychiatric care.

We are not questioning the use of conventional therapy by saline infusion and dialysis in lithium poisoning. Even if, as digestive tract decontamination, no prospective randomized clinical trials have been performed to demonstrate their benefit on clinical outcome. Early digestive tract decontamination may be a further treatment option for acute lithium poisoning.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.