Systematic review and quality assessment of economic evaluations and quality-of-life studies related to generalized anxiety disorder

Abstract

Background: The objectives of this article were to systematically review, summarize the results of, and assess the quality of economic evaluations and humanistic studies related to patients with generalized anxiety disorder (GAD).

Methods: EMBASE, EBM Reviews, MEDLINE, and HealthSTAR databases were searched (from the time of inception through April 2008). Full-text publications describing full economic evaluations (cost-benefit, cost-minimization, cost-effectiveness, and cost-utility analyses), partial economic evaluations (cost, burden-of-illness, and resource-utilization analyses), and humanistic outcomes (utilities, preferences, and willingness-to-pay analyses) were included. GAD diagnoses per official publications (eg, Diagnostic and Statistical Manual of Mental Disorders) and associated comorbid conditions were included; anxiety-related symptoms without a diagnosis of GAD were excluded. Study quality was assessed with a 38-point checklist of criteria previously developed by the Panel on Cost-Effectiveness in Health and Medicine.

Results: Thirty-six articles were included. Full economic evaluations (n = 5) were based on conventional decision-making modeling or population-summary data, using time horizons < or =12 months. Cognitive-behavioral therapy by a public-salaried psychologist and evidence-based care generated savings compared with current care. Pharmacotherapy with extended-release venlafaxine treatment was cost-effective compared with diazepam; escitalopram was cost-effective compared with paroxetine because of productivity gains. Full economic evaluations addressed 55.3% to 68.4% of the 38 items on the quality-assessment checklist. Partial evaluations were reported; GAD incurred larger mean marginal health care costs compared with other anxiety disorders (a difference of US $2138 in year-1999 values). GAD patients with severe pain interference incurred significantly higher costs than did patients with pain but no GAD. Furthermore, GAD patients used more services from a primary care provider or specialist than did patients with other psychiatric disorders. Comorbidities were associated with greater absenteeism than was having a diagnosis of GAD alone. Mean (SE) utility scores for quality-of-life assessments among patients with GAD (15D, 0.783 [0.019]; EuroQoL EQ-5D, 0.589 [0.038]) were similar to those for patients who were 20 years older and reported somatic conditions such as Parkinson's disease or heart failure.

Conclusions: Current evidence suggests that GAD is associated with substantial economic and humanistic impact on patients and health care systems. Future research should address economic evaluations from the private-payer perspective, studies related to the cost of underdiagnosed or untreated GAD, and full economic evaluations that incorporate longer clinical courses of the disorder.