A REVIEW OF ECONOMIC EVALUATIONS IN ANKYLOSING SPONDYLITIS: CONSIDERATIONS AND PROPOSAL FOR AN OMERACT REFERENCE CASE

Background: Since healthcare resources are scarce, choices have to be made on how they will be allocated. The use of cost effectiveness analyses has increased rapidly as policy makers have realised its value in maximising the populations benefits (in terms of length of life and health status) within a given budget. The recent increase in attention of the inflammatory disease Ankylosing Spondylitis (AS), has come from the findings that the disease is nearly as prevalent and equally disabling as Rheumatoid Arthritis, and the discovery that anti-tumour necrosis factor (TNF) therapy is highly effective in improving the manifestations of the disease.Objectives: We reviewed the current literature on economic evaluations with the aim of determining important considerations in the development of economic analyses. Following on from efforts by OMERACT to create reference case definitions for the conduct of economic evaluation in rheumatoid arthritis, osteoporosis and osteoarthritis, we sought to identify various methodological areas and research decisions that could benefit from a consensus between researchers, clinicians and drug developers in terms of an AS reference case.Methods: We searched MEDLINE and major rheumatology conference abstracts to identify full economic evaluations specific to patients with predominantly axial AS, published from 1980 to December 2005. Each was reviewed to highlight common areas deemed worthy of further consideration.Results: Eight evaluations were found. One analysis focused on spa therapy, another on COX-2 selective inhibitors and 6 on TNF antagonists. Only the analysis on spa therapy was based on a randomised controlled trial, while the others used a synthesis of evidence using decision analytic modelling. Some 5 of these were only available in abstract form and over half were published in the last year. After discussion and cross referencing with previous reference cases, ten specific issues were determined.1. Time horizon2. Duration of treatment3. Extrapolation of effects beyond trial duration/or trial duration4. Choice and synthesis of comparisons and treatment strategies where clinical trials do not exist5. Outcome measures6. Valuation of health in terms of Quality of Life or health utilities7. Measuring and valuing resource utilisation8. Population risk stratification9. Uncertainty Analysis10. Model development and reportingMany of the issues are generic to economic evaluations, but have specific pertinence in AS. For example the time horizon of an economic model is normally determined by the plausible duration of the effect of the intervention of interest. If TNF inhibitors can, for example, delay the structural progression of AS, this would make a lifetime horizon appropriate. However, there is limited data on the long-term of effect of TNF antagonists in AS, and the longitudinal course in AS disease is relatively unknown. However, new methods of data synthesis, and sensitivity analysis can be used to help decision makers make informed decisions when such uncertainty exists. There was also disparity in how studies were developed and reported, since some of the assumptions, and data analyses reviewed were not transparent, or developed in a manor usually deemed acceptable.Conclusion: The number of economic evaluations in AS is growing. We identified ten considerations which will be important for future development of evaluations. We believe further discussion through the OMERACT process will enable consensus on a number of these, aiding the policymaking of interventions for AS in the future.Citation: Ann Rheum Dis, volume 65, supplement II, year 2006, page 275Session: Health service and outcome research