Background: Significant differences are known to exist between the United States and Canada with respect to coronary catheterization and intervention. In a post hoc analysis, it was hypothesized that these differences may have the greatest impact on outcome in patients at risk for recurrent events such as those following myocardial infarction (MI).
Patients and methods: The hypothesis was tested in a nonrandomized comparison of the catheterization and revascularization patterns for patients following acute MI in 7029 patients in the United States and 1774 patients in Canada who participated in the Coumadin/Aspirin Reinfarction Study (CARS). CARS tested the effectiveness of low dose warfarin in combination with acetylsalicylic acid (ASA) versus ASA alone in reducing cardiovascular morbidity and mortality.
Results: Before study enrollment (median day 7 to 8), 84.5% of the American patients underwent coronary angiography compared with only 7.7% in Canada (P<0.01). CARS patients in the United States underwent significantly more frequent angioplasty during their hospital admission before study enrollment than their Canadian counterparts (55.8% versus 3.0%, respectively, P<0.01), and there was a more frequent use of intravenous heparin among American CARS patients (96% versus 88%, respectively, P<0.01) but less frequent administration of thrombolytic therapy (44% versus 49%, respectively, P<0.01). For follow-up of up to 32 months, American CARS patients had significantly fewer primary events (cardiovascular deaths, nonfatal MI, nonfatal ischemic stroke) than Canadian patients. Cumulative estimate of a primary end point comparing American with Canadian patients was, respectively, 2.0% versus 3.1% at one month, 8.0% versus 11.3% at one year and 11.6% versus 15.0% at two years. Thus, time to the primary event was significantly longer in American patients (P=0.0001). All-cause mortality estimates at 12 months were 2.2% and 4.0%, respectively, for the American and Canadian CARS subgroups. When management for the index MI (ie, angiography and angioplasty) is included in the model, the risk of death for Canadian versus American subgroups of CARS is not statistically different (0.9, 95% CI 0.6 to 1.2, P=0.40).
Conclusion: Among study participants, American patients experienced a better outcome than Canadian patients, which may be attributable to more aggressive management based on early coronary angiography and angioplasty. However, angioplasty before study enrollment in American patients may have resulted in enrollment of lower risk patients. This selection bias limits the scope of the comparison.