Background: The economic burden of depression is known to be high and was estimated to be USD 83.1 billion in 2000. Serotonin reuptake inhibitors (SRIs), including both selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors (SNRIs), have a superior adverse effect and safety profile relative to traditional agents (e.g. TCAs), and as a result have demonstrated superior cost effectiveness. Although efficacy across the SRIs is similar, the incidence of adverse drug reactions (ADRs) within SRIs remains significant and varies by agent. Patients who experience ADRs from SRIs may seek medical care, require additional treatment, and even discontinue treatment altogether, leading to increased utilisation and cost of therapy.

Objective: This study estimates the direct cost and cost effectiveness, taking into account the impact of treatment-related ADRs, of eight currently marketed SRIs (citalopram, escitalopram, generic fluoxetine, paroxetine, paroxetine controlled release [CR], sertraline, venlafaxine and venlafaxine extended release [XR]) used as initial treatment for depression.

Methods: A decision analytic model with a 6-month treatment goal was used to estimate the direct cost and cost effectiveness of treatment from the managed care/payer perspective. Estimates of SRI-related ADRs, associated treatments and costs were derived from the US FDA-approved prescribing information and published literature. Efficacy was assumed to be similar across all SRIs. Effectiveness was measured using quality-adjusted life years (QALY) based on EuroQol EQ-5D scores derived from the 2000 Medical Expenditure Panel Survey (MEPS). Censored least absolute deviations (CLAD) regression analysis was used to derive age-adjusted estimates of utility for all health states. Univariate and Bayesian second-order multivariate probabilistic sensitivity analyses were conducted to examine the impact of uncertainty in the parameter estimates.

Results: The expected direct cost and cost effectiveness of treatment from least to most expensive were: escitalopram (USD 3891; 0.341), citalopram (USD 3938; 0.340), generic fluoxetine (USD 4034; 0.335), venlafaxine XR (USD 4226; 0.336), sertraline (USD 4250; 0.335), generic paroxetine (USD 4385; 0.332), paroxetine CR (USD 4440; 0.332) and venlafaxine (USD 4613; 0.326). Monte Carlo simulation results suggested that escitalopram was the most likely (77%) to be cost effective for a willingness to pay < or = USD 50,000 per QALY, followed by citalopram (22%), generic fluoxetine (0.3%) and all other SRIs (0%). Sensitivity analyses indicated that the results of the study were robust to the assumptions underpinning the model.

Conclusions: SRI-related ADRs have a significant impact on the direct cost and cost effectiveness of treatment. Escitalopram, with the lowest ADR rate of the SRIs, had the lowest expected treatment cost and greatest effectiveness when compared with citalopram, generic fluoxetine, generic paroxetine, paroxetine CR, sertraline, venlafaxine and venlafaxine XR.