Objectives: Although disease modifying antirheumatic drugs (DMARDs) are the first choice drugs in the treatment of rheumatoid arthritis, many patients still take nonsteroidal anti-inflammatory drugs (NSAIDs) as well. These drugs may cause serious gastric adverse events with continuous usage. Cyclooxygenase-2 (COX2) inhibitors were supposed to have a gastrointestinal (GI) friendly side effect profile. The aim of the study is to compare three therapeutic strategies: conventional NSAIDs, NSAID in combination with proton pump inhibitors (PPIs), and the selective COX2 inhibitor therapy (celecoxib).

Methods: A decision tree model was developed, for 1 year, to simulate cohorts within the three arms (NSAIDs, NSAID + PPI, celecoxib). The efficacy of the different active agents of NSAIDs in therapeutically relevant doses was assumed to be the same, consequently differences can be seen in the side effect profile of the drugs. Medical costs, the costs of the side effects (GI, cardiovascular [CV] events), and quality-adjusted life-years (QALYs) were calculated to gain an incremental cost-effectiveness ratio (ICER). Evaluations were made from a third party payer's perspective. We performed one-way deterministic sensitivity analyses; the results were displayed in tornado diagrams.

Results: Our model indicates that NSAID + PPI offers extra health gain for extra costs compared with conventional NSAIDs (ICER:14,287 euro/QALY), while it dominates celecoxib because of celecoxib's higher costs and lower effectiveness. According to the sensitivity analyses, QALYs had the highest influence on ICER.

Conclusions: Although COX2 inhibitors have elevated GI efficacy compared with NSAIDs, celecoxib seems to be an adequate choice only for a limited group of patients with specific conditions because of the significantly higher price and CV risk profile.