Objective: This study evaluated risk factors for utilization of acute care services (ACS) (hospitalization or emergency department or urgent care visit) for lithium toxicity and the prevalence of lithium toxicity in a large, ambulatory population.
Methods: A nested case-control study compared lithium users with ACS utilization for lithium toxicity (case group) to lithium users without toxicity (control group) by using data from Kaiser Permanente Colorado for patients with at least one lithium prescription purchase. Patients in the case group were matched 1:5 with patients in the control group who had purchased lithium within 39 days of the ACS encounter. Possible lithium toxicity, identified by lithium level or diagnosis, was confirmed by chart review. Multivariable, conditional logistic regression analysis was used to identify patient and prescription characteristics associated with ACS utilization for lithium toxicity. The prevalence of lithium toxicity was determined.
Results: Of 3,115 individuals who took lithium, 70 experienced lithium toxicity, with or without ACS utilization, for a prevalence of 2.2%. Identified risk factors for ACS utilization for lithium toxicity included a newly initiated potentially interacting medication (odds ratio [OR]=30.30, 95% confidence interval [CI]=2.32-394.95), a higher number of treated chronic diseases (OR=1.28, CI=1.12-1.45), older age (OR=1.05, CI=1.02-1.09), and higher total daily lithium dose (OR=1.00, CI=1.00-1.00).
Conclusions: Newly initiated, potentially interacting medications are a major preventable driver of ACS use for lithium toxicity, whereas age, chronic disease, and total daily lithium dose are small but significant factors. Clinicians should use extra caution when initiating a potentially interacting medication.
Keywords: Lithium, Toxic states/psychopharmacology, Epidemiology, Utilization patterns & review, Drug interactions.