Clinical impact of early clopidogrel discontinuation following acute myocardial infarction hospitalization or stent implantation: analysis in a nationally representative managed-care population.

OBJECTIVES

To evaluate the association between discontinuation of clopidogrel therapy and risk of acute myocardial infarction (AMI) hospitalization or cardiac revascularization in a nationally-representative patient population following hospitalization for an AMI or coronary stent insertion.

RESEARCH DESIGN AND METHODS

This observational cohort study was performed using data on patients from the PharMetrics Anonymous Patient-Centric Database who were hospitalized for an AMI or coronary stent insertion and subsequently treated with clopidogrel. Cox proportional hazard modeling was used to evaluate the association between clopidogrel discontinuation prior to 1 year post-initial AMI hospitalization and the primary endpoint of repeat AMI hospitalization or coronary intervention defined as percutaneous coronary intervention (PCI) with or without stent, or coronary artery bypass graft (CABG).

MAIN OUTCOME MEASURES

The main outcome for this study was AMI hospitalization or coronary intervention defined as PCI with or without stent placement or CABG.

RESULTS

A total of 31 835 patients were included in the analyses. Patients were predominantly male and the average patient age was approximately 60 years. After controlling for baseline patient characteristics and follow-up time, discontinuation of clopidogrel was associated with a significantly higher rate of hospitalization for AMI or coronary intervention (HR 1.34, 95% CI 1.22-1.44).

CONCLUSION

Within a population of ACS patients drawn from a database of 85 US health plans, clopidogrel discontinuation within 1 year following hospitalization for AMI or stent placement is associated with an increased risk of AMI hospitalization or coronary intervention. The results of this study should be interpreted within the context of observational research, which does not address cause and effect relationships.