|Cost-effectiveness of pregabalin versus venlafaxine in the treatment of generalized anxiety disorder: findings from a Spanish perspective
|Vera-Llonch M, Dukes E, Rejas J, Sofrygin O, Mychaskiw M, Oster G
This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.
This study compared oral pregabalin, an anticonvulsant medication, with venlafaxine extended release, for patients with generalised anxiety disorder. The authors concluded that pregabalin appeared to be cost-effective, in the Spanish health care setting, for these patients with moderate-to-severe anxiety. The reporting was adequate, but the assumptions underpinning the effectiveness and benefit outcomes mean that the results should be considered to be highly uncertain.
Type of economic evaluation
Cost-effectiveness analysis, cost-utility analysis
The aim was to assess the cost and health effects of pregabalin, an anticonvulsant agent, compared with venlafaxine extended release, for the treatment of patients with generalised anxiety disorder. A hypothetical cohort of 1,000 patients with the disorder and with an average Hamilton Anxiety Scale (HAM-A) score, before treatment, of 16 or more, indicating moderate-to-severe chronic anxiety, was assessed.
Oral pregabalin once daily (300 to 600mg) was compared with venlafaxine extended release once daily (75 to 225mg), for patients with moderate-to-severe chronic anxiety.
A patient micro-simulation decision-analytic model was constructed to extrapolate the clinical findings to a longer period. The model synthesised published data from various sources, mainly relying on a key European randomised controlled trial, a trial by the pharmaceutical manufacturer, and national cost schedules. The authors stated that the study perspective was that of the third-party payer and the analysis covered a maximum of one year.
The main clinical endpoints were the expected mean HAM-A score with treatment and the expected number of weeks with no, minimal, mild, moderate, or severe anxiety. The clinical treatment efficacy data were from the Pregabalin Efficacy in Anxiety Clinical Evaluation (PEACE) trial (Baldinetti, et al. 2009, see ‘Other Publications of Related Interest’ below for bibliographic details) and other published studies. The mean HAM-A score changes for pregabalin and venlafaxine were obtained for the eight-week trial duration and extended to one year. The simulation assumed that the treatment benefit at eight weeks was maintained up to one year. It was also assumed, in the base case, that no patients would switch or discontinue treatment due to lack of efficacy or side effects; this was tested in the sensitivity analyses.
Monetary benefit and utility valuations:
The health-state values, for the HAM-A score categories, were based on those from a previous cross-sectional study of 456 patients with generalised anxiety disorder. In this study, participants were selected from 134 centres across Spain and the European Quality of life (EQ-5D) instrument was used to value the health states (Rovira, et al. 2007, see ‘Other Publications of Related Interest’ below for bibliographic details).
Measure of benefit:
The measures of benefit were the additional weeks with minimal or no anxiety and quality-adjusted life-years (QALYs).
The direct medical costs included primary and specialist care visits, blood tests, electrocardiograms, thyroid function tests, in-patient days, and medications. The resource estimates were from a previous study (Rovira, et al. 2007) and pharmaceutical manufacturer data. The values were from published national sources (Catalogo del Consejo General de Colegios Farmaceuticos. 2008, see ‘Other Publications of Related Interest’ below for bibliographic details) and a consulting firm. All costs were reported in 2007 Euros (EUR).
Analysis of uncertainty:
The deterministic uncertainty was measured using one-way sensitivity analyses on the key parameters and the results were presented in tables. A probabilistic sensitivity analysis was undertaken, with left-truncated normal distributions assigned for the weekly expected change in HAM-A scores. One-thousand Monte Carlo simulations were performed and 95% confidence intervals were generated. The results were illustrated in cost-effectiveness acceptability curves.
In the base case, the one-year expected mean number of weeks with no or minimal anxiety (HAM-A score of nine or less) was 13.5 for pregabalin and 4.3 for venlafaxine. Treatment with pregabalin produced fewer weeks with mild (HAM-A score of 10 to 15), moderate (HAM-A score of 16 to 24), and severe (HAM-A score of 25 or more) anxiety.
The estimated mean QALYs were 0.740 for pregabalin and 0.713 for venlafaxine extended release.
Total costs over one year for pregabalin therapy were EUR 3,871 compared with EUR 3,234 for venlafaxine; an increment of EUR 637 with pregabalin. When all costs were considered, the incremental cost per additional week with no or minimal anxiety was EUR 70 (95% CI 61 to 80) for pregabalin over venlafaxine. The incremental costs per fewer weeks at other anxiety states were not reported. The incremental cost per QALY with pregabalin was EUR 23,909 (95% CI 20,820 to 27,006).
One-way sensitivity analyses indicated that the incremental cost per QALY ratios, with pregabalin, ranged from EUR 19,829 to EUR 67,928 and were most sensitive to varying the the time horizon (to eight weeks), the 25% quartile utility values, and the rate of switching to paroxetine when therapy was discontinued. The probabilistic sensitivity analysis indicated that there was full likelihood that the incremental cost per QALY would be below a willingness-to-pay threshold of EUR 30,000.
The authors concluded that pregabalin appeared to be cost-effective compared with venlafaxine extended release, in the Spanish health care setting, for patients with moderate-to-severe generalised anxiety disorder.
The authors provided clear descriptions of the two treatments, but did not discuss other options that might have been relevant comparators. It was suggested that these drugs might be given sequentially, but the full sequence was not reported, and no treatment might have been a valid comparator.
The data for the model were from the PEACE trial, which was an eight-week, multi-centre, three-arm randomised controlled trial that compared each of the interventions against placebo. This report did not provide the direct effectiveness estimates for the two interventions, and the original publication should be consulted for an assessment of the overall quality and validity of the data. The trial lasted eight weeks and a number of assumptions were required for the model. The main assumptions were the maintenance of the treatment benefit from eight weeks to one year, and the lack of withdrawals due to side-effects or lack of efficacy. Withdrawals were considered in the sensitivity analysis and the rates were presented in the paper, but it was not clear why they were not considered in the base case. The utilities were a key driver of the findings and they were derived using a recognised measure, from a representative sample of 456 patients from 134 primary health centres in Spain (Rovira, et al. 2007). These utility values were estimated for various severities of anxiety, as measured by the patient's HAM-A score. This seems to have been appropriate, but the HAM-A scores were based on the assumption that the pregabalin efficacy remained constant from eight weeks to one year. This was a favourable assumption and its validity is critical to the findings.
The direct medical costs were analysed and they appear to have been appropriate to the perspective. Thorough details of the resource types, how these were measured and valued, and their unit costs were provided. The cost analysis seems to have been appropriate and was well reported.
Analysis and results:
The model structure was explicitly described and illustrated. All the inputs and data sources were provided and justified. The analyses were comprehensive and the results were clearly reported. The authors gave a number of study limitations including the one-year time frame, the assumption that the short-term efficacy held constant up to one year, and the omission of treatment discontinuation due to side-effects or treatment failure, in the base analysis. Sensitivity analyses were conducted and presented, including cost-effectiveness acceptability curves from a probabilistic analysis of the base case. The reporting was generally sufficient.
The reporting was adequate, but the assumptions underpinning the effectiveness and benefit outcomes mean that the results should be considered to be highly uncertain.
Supported by Pfizer, Inc.
Vera-Llonch M, Dukes E, Rejas J, Sofrygin O, Mychaskiw M, Oster G. Cost-effectiveness of pregabalin versus venlafaxine in the treatment of generalized anxiety disorder: findings from a Spanish perspective. European Journal of Health Economics 2010; 11(1): 35-44
Other publications of related interest
Baldinetti F, Bandelow B, Kasper S, Herman B, Nivoli G, Van Ameringen M, Petralia A, Mandel FS. Efficacy of pregabalin and venlafaxine-XR in generalized anxiety disorder: results of a double-blind, placebo-controlled 8-week trial. International Clinical Psychopharmacology 2009; 24: 87-96.
Rovira J, Albarracin G, Salvador-Carulla L, Rejas J. Cost and burden of illness of generalized anxiety disorder (GAD): a Spanish perspective from the primary health-care setting (The ANCORA Study). European Journal of Neuropsychopharmacology 2007; 17(Supplement 4): S507-S508.
Catalogo del Consejo General de Colegios Farmaceuticos. Base de datos de medicamentos: Consejo General de Colegios Oficiales de Farmaceuticos. 2008. Available at http://www.portalfarma.com/home.nsf [Accessed March 2008].
Subject indexing assigned by NLM
Anticonvulsants /economics /therapeutic use; Antidepressive Agents, Second-Generation /economics /therapeutic use; Anxiety Disorders /drug therapy /economics; Confidence Intervals; Cost-Benefit Analysis; Cyclohexanols /economics /therapeutic use; Delayed-Action Preparations; Economics, Pharmaceutical; Humans; Models, Economic; Models, Statistical; Monte Carlo Method; Pregabalin; Psychiatric Status Rating Scales; Psychometrics; Quality-Adjusted Life Years; Spain; Treatment Outcome; Venlafaxine Hydrochloride; derivatives /economics /therapeutic use; gamma-Aminobutyric Acid /analogs &
Date bibliographic record published
Date abstract record published