Table 3.

Cox regression for risk of first myocardial infarction (MI) in women and men with atrial fibrillation treated with antithrombotic drugs, using those without warfarin treatment as referents. Models shown by “per protocol” analysis (PP) for warfarin and for ASA treatment. Incidence rate per 100 Person-Years at Risk.

	Events/At Risk (n)	Incidence Rate (95% CI)	Warfarin			ASA
			Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
MI
Women	472/5,398	1.587 (1.450–1.737)	0.26 (0.20–0.35)	0.27 (0.17–0.42)	0.25 (0.16–0.40)	0.55 (0.44–0.69)	0.63 (0.41–0.95)	0.60 (0.39–0.92)
Men	528/6,301	1.481 (1.359–1.612)	0.30 (0.23–0.38)	0.30 (0.21–0.41)	0.27 (0.19–0.38)	0.48 (0.39–0.60)	0.47 (0.33–0.67)	0.44 (0.31–0.63)

Prescription of warfarin and ASA was classified as “per-protocol” (“PP”) if present the year before and the year of MI, or present among subjects without MI if present at least 50% of actual years after first recorded year of AF. For warfarin, patients on ASA or clopidogrel were excluded from analyses, and for ASA, patients on warfarin or clopidogrel were excluded.

Model 1 age-adjusted, and Model 2 as Model 1 but also adjusted for socio-economic factors (neighbourhood socio-economic status, educational level and marital status, also including interaction terms between age and marital status), and Model 3 as Model 2 but also adjusted for cardiovascular co-morbidity (hypertension, cerebrovascular disease, CHF and diabetes; also including interaction terms between age and marital status, and for women also between age and diabetes.