Economic impact of nonpersistence with antidepressant treatment in the adult population of Quebec: a comparative cost-effectiveness approach.

Background: Although for the great majority of indications, practice guidelines recommend that antidepressants (ADs) be used for at least 6 months, premature discontinuation is very frequent in a "real-life" setting. Previous studies have assessed the economic impact of such nonpersistence, but differences across antidepressant products remain inadequately explored.

Objective: To compare treatment persistence and incremental cost/persistence ratios (ICPRs) across individual new ADs (selective serotonin reuptake inhibitors and atypical ADs) as well as the associated direct health-care costs in the adult population covered by the public drug program of Quebec.

Methods: A retrospective cohort study was conducted in 13,936 adults aged 18 to 64 years who started an AD treatment in 2003. Persistence was defined as treatment duration of at least 6 months regardless of whether a product switch had occurred. Economic impact was assessed over the first year of treatment through drug, medical services, hospitalization, and total health-care costs. Comparisons across products were conducted using the ICPR.

Results: Adjusting for confounders, treatment nonpersistence ranged from 60.4% (paroxetine) to 65.1% (citalopram). The product associated with the highest total health-care costs was citalopram (CDN$2653) and the lowest was venlafaxine (CDN$2168). Fluvoxamine had the lowest mean AD costs (CDN$215) and venlafaxine (CDN$309) the highest.

Conclusions: Total health-care costs were similar across products except for citalopram, which was more costly. Comparisons based on the ICPR revealed that paroxetine, fluoxetine, and venlafaxine were more favorable than the other AD alternatives.