That's the point. 91 people confirmed by test to have it after recovering out of 376,000 people altogether. The odds there are just as likely to be coincidence due to faulty tests or any other number of reasons that wouldn't apply to a broader population. Get back to me if/when they find more.
I'm not sure that's right. The point is that very few people later get retested after recovery. So, to get 91 is probably a disturbingly large number of the retests...