TY - JOUR T1 - Advantages and limitations of utility assessment methods in rheumatoid arthritis. JF - The Journal of Rheumatology JO - J Rheumatol SP - 2193 LP - 2200 VL - 34 IS - 11 AU - Ariel Beresniak AU - Anthony S Russell AU - Boulos Haraoui AU - Louis Bessette AU - Claire Bombardier AU - Gerard Duru Y1 - 2007/11/01 UR - http://www.jrheum.org/content/34/11/2193.abstract N2 - Utility assessment and cost-utility analyses such as costs/quality-adjusted life-years (QALY) are frequently presented to demonstrate the value of new treatment options in rheumatoid arthritis (RA). However, utility indicators require various methods that introduce significant methodological challenges, which directly influence the results and ensuing reimbursement decisions. Our objective was to review and discuss these challenges and the validity of frequently used utility assessment techniques in the context of RA. Coding the intensity of preferences or variations in patient satisfaction in order to assess utility implies extreme mathematical assumptions about a patient's rationality regarding his/her preferences towards different given health states. The construction and assumptions of commonly used "direct approaches" (standard gamble, time tradeoff, visual analog scale) and indirect approaches (EQ5D, HUI, SF6D) are presented. Other approaches such as transformation in utility of data from clinical (Health Assessment Questionnaire) or quality of life instruments ("mapping technique") are analyzed as they appear to generate uncertainty and a wide variation in estimated utility values in the context of RA. Utility assessment and cost-utility analyses in RA, which form the basis of the QALY, are frequently published and often requested by health technology assessment agencies to assist -reimbursement decisions. However, when interpreting the results, the medical community must take into consideration the limitations and significant uncertainty of these approaches.In light of these findings, real cost-effectiveness analyses based on observed clinical outcomes appear to be more robust and reliable to assist decision-making, particularly in the context of RA. ER -