Nevermind, I made a mistake in my original calculations - used variance instead of standard deviation. Actually (from approximating the binomial distribution w

Nevermind, I made a mistake in my original calculations - used variance instead of standard deviation. Actually (from approximating the binomial distribution with n=22k and p=17/22000 with normal one) getting the rate over 30 per 22k has p<0.001.