The economic impact of introducing serotonin-noradrenaline reuptake inhibitors into the Brazilian national drug formulary - Cost-effectiveness and budget-impact analyses

Objective: To determine the cost effectiveness, from the Brazilian Ministry of Health viewpoint, of three antidepressant classes for major depressive disorder (MDD), and the budget impact of introducing serotonin-noradrenaline (norepinephrine) reuptake inhibitors (SNRIs) into the current Brazilian national drug formulary, assuming a 6-month treatment duration. Methods: An existing decision-tree model was adapted to Brazil, based on local guidelines. Clinical data were obtained from published meta-analyses. Patients included adults aged >= 18 years with MDD, diagnosed using the Diagnostic and Statistical Manual of Mental Disorders, third and fourth editions (DSM-III/IV), with moderate-to-severe disease (Hamilton Depression Rating Scale [HAMD] >= 15 or Montgomery-angstrom sberg Depression Rating Scale [MADRS] >= 18), without co-morbidities or co-medications, receiving >= 6 weeks of treatment with SNRIs, selective serotonin reuptake inhibitors (SSRIs) and/or tricyclic antidepressants (TCAs). Clinical outcome was remission (HAMD ! 7 or MADRS :512). Direct costs (drugs, physician visits, hospitalisations) were included. Drug costs were obtained from the 2006 Brazilian National Drug Price List, and hospitalisation and physician costs from the 2006 Healthcare System database. Costs were valued in Brazilian Reais ($Brz), year 2006 values ($Brzl = $USO.47). Univariate and Monte Carlo sensitivity analyses tested model robustness. Results: Expected costs per patient treated were SNRIs $Brz4848; SSRIs $Brz5466; and TCAs $Brz5046, and overall success rates (primary plus secondary treatment across all decision tree branches) were SNRls 78. 1 %; SSRIs 74.0%; and TCAs 76.4%. Average costs/success were SNRls $Brz6209; SSRls $Brz7385; and TCAs $Brz6602. SNRIs dominated in incremental cost-effectiveness analyses. Monte Carlo analysis confirmed drug classes' relative positions; however, there was considerable uncertainty. Introducing SNRls into the formulary could generate average savings of 1% of the total budget, with a 52% probability of savings. Conclusions: SNRIs appear to be cost effective against SSRIs and TCAs when 14 prescribed to patients with MDD in Brazil. However, their inclusion into the national drug list would generate minor savings compared with the current formulary of SSRIs and TCAs. Thus, we considered such inclusion as 'cost-neutral', since no major probability of savings or increased expenditures were observed.