Using forecasting models to estimate the effects of changes in the composition of claims for selective serotonin reuptake inhibitors on expenditures

Background: The use of selective serotonin reuptake inhibitors (SSRIs) as antidepressant therapy has increased considerably since the introduction of fluoxetine in 1989. By 1999, 3 of the 4 available SSRIs were among the top 10 most frequently used drugs in the United States. In addition, SSRIs were one of the major contributors to the growth in psychotropic medication expenditures during the past 5 years. Objective: The purpose of this article was to examine the utilization patterns of the 4 most commonly used SSRIs and their contribution to rising antidepressant medication expenditures among claimants in a publicly funded drug program. Using the results of forecasting models, we explored possible ways to control these growing expenditures. Methods: Cross-sectional antidepressant claims and expenditure data from the Ontario Drug Benefits program for 1992 to 1998 were examined. Five scenarios were modeled in which future SSRI expenditures and claims were predicted using exponential smoothing models. Results: If the historical patterns of use continued, a 20% increase in the 1998 level of expenditures was expected to occur by the year 2000. Predicted expenditures are sensitive to the composition of the SSRI claims. Exclusive use of 1 of the 4 major SSRIs (fluvoxamine, fluoxetine, paroxetine, and sertraline) could decrease projected expenditures by 30% or increase them by 11%. An "equal shares" approach, in which each of the 4 SSRIs are used in equal proportions in the population, may reduce expenditures by similar to8%. Conclusions: The current trends in the utilization data suggest that sertraline and paroxetine are being used as first-line treatments. The results of the forecasting models suggest that growing expenditures could be curbed if these 2 antidepressants were not used in that manner. Short of limiting the drugs available on benefit formularies, there may be a way to control costs through the use of a prescribing algorithm. Although our results support the use of fluoxetine for first-line SSRI treatment as a cost-control measure, we do not definitively recommend its adoption. These findings contribute to the discussion about using fixed versus flexible formularies as a potential cost-control mechanism.