Validation of a nomogram used to predict lithium concentration in overdose

Abstract

Introduction Extracorporeal Treatment (ECTR) is an essential component in management of severe lithium toxicity. The Extracorporeal Treatments in Poisoning (EXTRIP) group’s suggested indications for ECTR include “if the expected time to obtain a [Li+] < 1.0mEq/L with optimal management is >36h”. Buckley et al. developed a lithium nomogram which could help predict the fall in lithium concentrations for chronic poisoning. Our aim is to externally validate the lithium nomogram in a cohort of cases with chronic accumulation and acute on chronic lithium poisoning. Methods A retrospective analysis of suspected cases of chronic accumulation and acute on chronic lithium poisoning referred to our Toxicology Unit from May 2013 to 2020 was performed. Results Out of 51 cases, 29 cases of chronic accumulation and eight cases of acute on chronic poisoning were analysed after excluding 14 cases who required haemodialysis. In chronic accumulation cases, the nomogram correctly identified 10 out of 14 patients whose [Li+] failed to drop below 1.0 mmol/L by 36 h (sensitivity 71.4% [95% CI 42 − 92%]), and 8 out of 15 patients whose [Li+] dropped below 1.0 mmol/L by 36 h (specificity 53.3% [95% CI 27 − 78%]), resulting in the positive predictive value (PPV) of 58.8%, negative predictive value (NPV) of 66.7% and accuracy of 62.1%. Conclusions Our study shows that the lithium nomogram is moderately sensitive at identifying patients with chronic lithium accumulation who will have a serum lithium concentration >1 mmol/L at 36 h without ECTR.