You’re ignore relative prevalence of the subjects getting infected. Sure, African-American women were infected, but that’s 4,000 out of a population of what? 35M? The rate compared to higher risk populations is still incredibly low.