![[ Visit Client Website ]](images/banner.gif)
Purpose: Data supports that early treatment for rheumatoid arthritis (RA) improves outcomes, but this approach requires accurate risk stratification to identify those most likely to benefit and avoid over-treatment. MRI has been suggested as a potential tool for risk stratifying early RA patients. Our objective was to explore the cost-effectiveness of adding MRI to standard early RA prognostic markers.
Methods: Using TreeAgeŽ decision analysis software (v. 1.0.2), we created a simulation model of early RA management comparing the addition of MRI to standard assessment (i.e., RF status, CCP antibodies, baseline disease activity) in a population with disease duration <2 years and no erosions on plain radiographs. The model estimated total costs (in 2007 USD) and quality adjusted life years (QALYs). Clinical assessments were performed at 3 month intervals to determine the need for treatment escalation based on disease activity (DAS28) and change over time (ACR50). Adverse events (AEs) were classified as mild (with no associated cost or mortality) or moderate to severe (e.g., requiring treatment or hospitalization and resulting in cost and mortality increases). Cost, quality of life, treatment response, and AE estimates represent the range of data published in randomized trial and observational studies. The model was constructed from the societal perspective with 3% discounting. Sensitivity analyses examined the impact of varying all input estimates, in addition to best and worst case scenarios.
Results: In the base analysis, adding MRI to standard testing increased average per person costs for the first year of treatment from $7,079 to $7,835 and quality of life from 0.810 to 0.815 QALYs, yielding an incremental cost-effectiveness ratio (ICER) of $151,200/QALY gained. Increasing the proportion of individuals at risk for severe disease or increasing the proportion of individuals receiving more aggressive treatment (e.g., by decreasing MRI specificity or improving sensitivity) improved the cost-effectiveness of MRI (Figure 1).
Conclusion: This model provides a formal framework for
evaluating the cost-effectiveness of MRI in early RA
management. Short-term analyses suggest that, despite conferring small quality
of life gains in the first year of treatment, MRI is
unlikely to be cost-effective in the short-term at currently acceptable
thresholds among individuals at low risk for severe disease (e.g., very early
RA or undifferentiated inflammatory arthritis). Longer-term analyses are needed
to better define the populations, treatment regimens and MRI
characteristics for which MRI is cost-effective. 