Pharmacoeconomic and health outcome comparison of lithium and divalproex in a VA geriatric nursing home population: influence of drug-related morbidity on total cost of treatment

Abstract

Objective: Clinicians use mood stabilizers for treating agitation in older patients, but limited information is available regarding side effects and costs in clinical practice. Total costs of treatment were assessed for a subset of geriatric patients receiving either lithium carbonate or divalproex sodium for agitation.

Study design: Retrospective cohort examination of the medical records of 72 patients, 55 years of age or older, in a Veterans Administration long-term, skilled nursing care facility, with a diagnosis of dementia or bipolar affective disorder or both.

Patients and methods: Patients treated with lithium or divalproex during the previous 4 years (1994-1997) were evaluated. Quantitative information was collected and compared regarding routine care, including cost of treatment and laboratory monitoring; and occurrence of adverse events and associated diagnostic and treatment measurements.

Results: Routine care costs for the 2 groups were similar. The lower annual acquisition cost per patient-year for lithium ($15 vs $339 for divalproex) was offset by higher laboratory monitoring costs associated with its administration ($278 vs $53 for divalproex). Examining the adverse events showed that the lithium group had more medication-related adverse events (32 total) than the divalproex group (10 total) and more severe occurrences, including 6 cases requiring medical intensive care unit (MICU) hospitalization. The total mean cost of treating drug-related mild-to-moderate morbidity was $3472 for lithium and $672 for divalproex. An additional cost per admission of $12,910 ($77,462 for all 6 cases) increased total morbidity-related expenditures in the lithium group to $80,934.

Conclusions: Treating geriatric patients with lithium requires careful monitoring because of side effects. Staffing and resource limitations of a skilled nursing care facility may compromise optimal lithium monitoring in elderly patients. The collected data indicated that divalproex does not result in as many as or as severe adverse events and is, therefore, a safer treatment. The use of lithium was not only more expensive (on average $2875 more per patient) than treatment with divalproex but, more importantly, it was associated with poorer patient outcomes.