I think it's important to understand that scientists don't work in a perfect world. Science is all about uncertainty and the unknown and moving forward to reveal information that improves confidence, but scientists don't deal in absolutes. As Ray has pointed out, with any hit to the data bank, there is a statistical analysis that is conducted to try to address that theoretical possibility. So no matter how many advances we make in terms of increasing the numbers of tests we do, there will always be a theoretic possibility that a hit has occurred as a result of a coincidence, but that theoretical possibility, even today, is infinitesimally small.
On April 28th, 2009. See this statement in context.