Escitalopram and Duloxetine in Major Depressive Disorder A Pharmacoeconomic Comparison Using UK Cost Data

Background: Selective serotonin reuptake inhibitors (SSRIs) and serotonin-noradrenaline reuptake inhibitors (SNRls) are approved for the treatment of major depressive disorder (MDD). The allosteric SSRI escitalopram has been shown to be at least as clinically effective as the SNRls venlafaxine and duloxetine in MDD, with a better tolerability profile. In addition, escitalopram has been shown to be cost saving compared with venlafaxine. Objective: To evaluate the cost effectiveness of escitalopram versus duloxetine in the treatment of MDD, and to identify key cost drivers. Methods: The pharmacoeconomic evaluation was conducted alongside a 24-week, double-blind, multinational randomized study (escitalopram 20 mg/day and duloxetine 60 mg/day) in outpatients with MDD, aged 18-65 years, with Montgomery-Asberg Depression Rating Scale (MADRS) score >= 26 and Clinical Global Impression Severity (CGI-S) score A, and baseline duration of the current depressive episode of 12 weeks to 1 year. The analysis was conducted on the full analysis set (FAS), which included all patients with >= 1 valid post-baseline health economic assessment. Effectiveness outcomes of the cost-effectiveness analyses (CEA) included the change in Sheehan Disability Scale (SDS) score (primary CEA), treatment response (MADRS score decrease >= 50%) and remission (MADRS score ! 12) rates at week 24. Cost outcomes were assessed from the societal perspective. Healthcare resource use and sick leave were evaluated using a health economic assessment questionnaire. Unit costs of healthcare services were obtained from standard UK sources ( pound, year 2006 values). Results: Over the total 24-week study period, escitalopram was associated with significant cost savings compared with duloxetine (total per-patient monthly cost pound 188 vs 334 pound, respectively). In the primary CEA, escitalopram dominated duloxetine (i.e. was more effective on the disability scale and less costly). Treatment with escitalopram resulted in significantly lower mean sick leave duration per patient over 24 weeks than duloxetine (30.7 days vs 62.2 days). In multivariate analyses, escitalopram as a treatment choice was associated with a 54% reduction in sick leave duration (p < 0.001). Treatment with escitalopram also resulted in 49% lower total costs than treatment with duloxetine (p = 0.002). Absenteeism accounted for about two-thirds of the overall cost. Early clinical improvement (mean change in MADRS total score, response and remission) had an independent significant impact on the sick leave duration, after controlling for key co-variates. Conclusions: Escitalopram was associated with significantly lower duration of sick leave and significant savings in the total cost compared with duloxetine; it dominated duloxetine when effectiveness was assessed on the SDS scale. Indirect costs due to sick leave accounted for the most substantial portion of the total cost and should, therefore, be an important consideration when pharmacoeconomic comparisons between treatments are made from the societal perspective. The link between decrease in absenteeism and early (8-week) clinical improvement suggested in the additional analyses may explain the reduced sick leave observed with escitalopram, given its superior short-term efficacy compared with duloxetine (demonstrated in the underlying clinical trial).