• Marie Tournier, Yola Moride, Ralph Crott, Guillaume Galbaud du Fort, and Thierry Ducruet.
• Center for Clinical Epidemiology and Community Studies, SMBD Jewish General Hospital, Montreal, Quebec, Canada.
• J Affect Disord. 2009 May 1; 115 (1-2): 160-6.

BackgroundIn a real-life setting, differences across newer antidepressants in patterns of use remain poorly explored, particularly in the older patients despite the high prevalence of late-life depression.MethodsAn observational retrospective cohort study was conducted in the community-dwelling elderly population of Quebec using health databases to compare the newer antidepressants with respect to non-persistence, associated health care costs and cost/persistence ratio. A random sample of 12,825 outpatients who initiated an antidepressant treatment in 2000 were followed for 12 months. Non-persistence was defined as treatment duration of less than 180 days. Economic variables included direct costs of prescribed medications, medical services and hospitalizations assessed through RAMQ claims databases and Med-Echo hospitalization database. Cost/persistence ratio and incremental cost/persistence ratio were obtained for each antidepressant product; persistence being considered as an indicator of effectiveness.Results55.6% of antidepressant treatments were non-persistent. Products associated with low antidepressant costs were often associated with high costs of other medications and health care services, and vice versa. Paroxetine was associated with the lowest non-persistence (50.5%; 95%CI 48.5-52.5) and one of the most favourable cost/persistence ratios (CDN$4869 per persistent treatment). Fluoxetine was associated with the most favourable incremental cost/persistence ratio.LimitationsSome services and hospitalizations are not included in the administrative databases. No data on indication for treatment were available. These were likely to be non-differential across newer antidepressants.ConclusionAs found in other populations, non-persistence with antidepressant treatment is very frequent in the Quebec elderly population. Products associated with poor persistence result in increased health care costs. Hence, intervention programs aimed at improving persistence would optimize the use of health care resources and result in economic advantages.

30 keywords...

hide…