Abstract
Abstract only
Background:
Anticoagulants are effective in stroke prevention in patients with atrial fibrillation (AF). Warfarin’s comparative effectiveness in AF Medicare beneficiaries by race/ethnicity and gender is not well described.
Methods:
Medicare claims data for years 2000-2010 were used to calculate: 1) AF per annum prevalence (N > 1.9 million); and 2) rates of new AF cases (first Medicare reimbursement for AF), hospitalization, and mortality. Three 20 % samples of AF beneficiaries for years 2000 (n=266K), 2005 (n=316K), and 2007 (n=311K) were used to calculate warfarin use and stroke (ischemic) rates. Consistent with prior study methods, INR test claims were used as a surrogate for warfarin use with a greater than 95 % precision rate and specificity. New AF cases were extracted from each of the three sample cohorts (mean 72K cases each) and continuous survival analyses conducted to assess warfarin’s relationship to stroke, mortality, and hospitalization after adjustment for CHADS2 score, age, gender, and race/ethnicity.
Results:
AF prevalence and warfarin use increased while stroke and mortality rates declined across gender and race/ethnicity from 2000 to 2010. Survival analysis comparing Blacks, Hispanics, and Asians to Whites showed: 1) even when adjusted for warfarin use, Blacks were 40 % (p<0.0001) more likely to have a stroke; 2) in 2007 Hispanics had a 35 % (p<0.01) higher prevalence of stroke and warfarin use did not reduce their stroke risk; and 3) Asians had better outcomes. Warfarin did not reduce stroke risk in women as well as in men, but women using warfarin had a lower risk of death and hospitalization than men. While there was a >70% (p<0.0001) reduction in mortality for warfarin users (Blacks still had a 25 % (p<0.0001) higher mortality risk than Whites after adjusting for warfarin use). 5-year survival rate for all AF beneficiaries was about 50 %.
Conclusions:
Significant differences in key metrics between race/ethnicity and gender exist. Across all metrics, Blacks had worse outcomes with less warfarin use and comparatively worse outcomes even when on warfarin. Patient diversity should be a focus for future trials in AF-related cardiovascular outcomes.