The prevalence question isn't really answered in the article. If it is 500 cases in an area in which 1m people live then it might not really be a cluster. But if it is 500 cases out of a town of 40000 then it is quite something different.