Comment by kiltedpanda

2000 people is a pretty good sample size. It is true though that I'm hazarding a guess that the background rate of chest pain in any given timeframe is slightly lower than ~5%.