Cost-effectiveness of clopidogrel treatment in percutaneous coronary intervention: a European model based on a meta-analysis of the PCI-CURE, CREDO and PCI-CLARITY trials*

Objective: Our objective was to conduct a comprehensive cost-effectiveness analysis of pre-treatment and long-term treatment with clopidogrel in percutaneous coronary intervention (PCI) in three European countries based on a meta-analysis of the PCI-Clopidogrel in Unstable angina to prevent Recurrent Events (CURE), Clopidogrel for the Reduction of Events During Observation (CREDO) and PCI-Clopidogrel as Adjunctive Therapy (CLARITY) trials. This analysis adds to existing knowledge by providing further data on the cost-effectiveness of clopidogrel in PCI across a wide spectrum of patients. Methods: A combined decision tree and Markov model was created. The relative risks of myocardial infarction, cardiovascular death and of major bleedings with clopidogrel were based on a fixed-effects meta-analysis. The risk of ischaemic events in untreated patients and long-term survival were taken from the Swedish hospital and death registers. A societal perspective was used in Sweden and a payer perspective in Germany and France. Costs are stated in €2006 and effectiveness measured in quality-adjusted life-years (QALYs). Results: The pooled effects of clopidogrel on the combined endpoint showed a relative risk of 0.711 (p = 0.003) at 30 days and 0.745 (p = 0.002) at end of follow-up (up to 1 year). Pre-treatment with clopidogrel compared with aspirin alone is a dominant strategy. Long-term treatment with clopidogrel compared with 1-month treatment leads to approximately 0.09 QALYs at an incremental cost of €393 in Sweden, €709 in Germany and €494 in France. The corresponding incremental cost-effectiveness ratios range from €4225/QALY to €7871/QALY. Conclusion: The results of this modelling analysis suggest that pre-treatment and long-term treatment in PCI with clopidogrel for up to 1 year are cost-effective in a range of patient groups and settings given commonly accepted thresholds.