Abstract 253: Comparative Effectiveness of Clopidogrel in Medically Managed Patients with Unstable Angina and non-ST Segment Myocardial Infarction

Abstract

Background: The CURE trial demonstrated improved outcomes with the addition of clopidogrel to aspirin in unstable angina (UA) or NSTEMI, leading to guideline adoption and widespread use. Although the role of clopidogrel after PCI is well studied, 40-50% of patients with UA/NSTEMI are medically managed, i.e., without revascularization during the index hospitalization. The effectiveness and optimal duration in real-world medically managed patients is unknown. Methods: We conducted a retrospective cohort study of all adult members of Kaiser Permanente of Northern California, a large, integrated health care delivery system, without known CAD or prior clopidogrel use who presented with UA/NSTEMI from 2003-2008 and did not receive revascularization (PCI or CABG) during the index hospitalization or within 7 days post-discharge (i.e., “medically managed”). We measured the association between use of clopidogrel within 7 days of discharge with subsequent all-cause mortality and MI at 2 years of follow up. Outcomes were examined in unmatched and propensity-matched multivariable Cox regression analyses, adjusted for demographics, comorbidities, longitudinal medication use and secular trends. Results: We identified 18,771 patients (mean age 69.3 years; 42.4% women) with incident UA (34.7%) or NSTEMI (65.3%) followed for a mean 2.5 years after the initial event. Clopidogrel prescription within 7 days of discharge was observed for 34.5% of the sample. Among these, the mean duration of continuous clopidogrel use was 255 days. An additional 1.6% filled a clopidogrel prescription between 7 and 30 days after discharge, and 5.0% filled a prescription after 30 days. During the first 2-years of follow up, patients prescribed clopidogrel within the first 7 days after discharge had lower unadjusted rates of all-cause mortality (7.5% vs 19.2%, p < 0.001; HR = 0.42, 95% CI [0.38-0.45]) and subsequent MI (6.5% vs 8.2%, p = 0.002; HR = 0.85 [0.76-0.95]). In multivariable Cox regression models, clopidogrel users had lower risk of 2-year all-cause mortality (HR = 0.65, [0.60-0.71]) but similar rates of MI (HR = 1.07, [0.96-1.20]). Propensity-matched models demonstrated similar results. A sensitivity analysis in which clopidogrel use was defined by initiation within 30-days post-discharge also demonstrated similar results. Conclusions: In a large community-based sample of patients with incident medically managed UA/NSTEMI, receipt of early post-discharge clopidogrel was independently associated with lower rates of death but similar rates of MI. Future research should identify the subset of patients who experience the greatest net clinical benefit with clopidogrel after UA/NSTEMI as well as the optimal duration of treatment.