We read with interest the study comparing Paris and EXTRIP criteria of haemodialysis in lithium poisoning, recently published in the journal. 1 The authors used a poison centre cohort of lithium-poisoned patients for whom haemodialysis was recommended or performed.Patients were classified according to both Paris and EXTRIP criteria and to whether they were or not dialysed. The endpoint was worsening neurological status or death, if mentioned in the narrative. The authors concluded that, if discordant, EXTRIP outperformed Paris criteria at identifying potentially ill patients who might benefit from haemodialysis. We would like to raise several methodological concerns limiting the accuracy of the conclusion.