Abstract

Objective: To estimate the incremental cost-effectiveness ratio (ICER) of antibodies against cyclic citrullinated peptides (aCCP) in the early diagnosis of rheumatoid arthritis (RA).

Methods: A Markov model was used to model 10-year progression of RA in patients first diagnosed with undifferentiated arthritis (UA) and to estimate the incremental costs and quality-adjusted life years (QALYs) of using aCCP additionally to American College of Rheumatology (ACR) criteria. The impact of later diagnosis and treatment due to non-use of aCCP was modelled as increased Health Assessment Questionnaire (HAQ) progression. Utilities were assigned to HAQ states for calculating QALYs. Uncertainty was analysed using univariate and probabilistic sensitivity analyses (Monte Carlo simulation).

Results: Baseline ICER was €930/QALY. Univariate sensitivity analyses identified the impact of later diagnosis on HAQ progression as a major source of uncertainty, resulting in an ICER range from “dominance” to €153 092/QALY, compared with a maximum ICER of €4870/QALY for other variables. Monte Carlo simulation resulted in a 95% uncertainty interval from –€3537/QALY (dominance) to €5429/QALY; when indirect costs were considered, Monte Carlo simulation resulted in a 95% uncertainty interval from –€78 115/QALY (dominance) to –€23 444/QALY (dominance).

Conclusions: Using aCCP in the diagnosis of RA in patients with UA is likely to be cost effective compared with using ACR criteria alone. When indirect costs are incorporated, aCCP seems to save costs. Clearly, more research is needed relating the effects of diagnosis and treatment on the long-term course and the resulting functional impairment of RA as measured by the HAQ.