Yes, it's worrying - although you need to compare like with like, e.g. young black men with young white men, not to the whole population. Still too high though. And regardless of ethnicity, the "innocent" samples shouldn't be retained at all, IMO.