PCR detetction of Salmonella food contamination not yet ready for prime time.

Implications: PCR is an invaluable tool in the research setting but has only limited usefulness in a clinical context and, in my opinion, even more limited application for broad-based surveillance applications such as this. While PCR done well under laboratory conditions with samples spiked with known quantities of salmonella are able to be done with exquisite sensitivity (picking up all instances of salmonella) and specificity (not falsely picking salmonella when it isn't there), but getting these levels of test performance in an abbatoir or meat-packing facility will be challanging. PCR is notoriously susceptible to cross contamination and is very demanding of proper laboratory technique. Its ability to detect minutely small quantities (e.g. one bacteria or virus) of a pathogen make it prone to false positives. There are also many species of non-pathgenic Salmonella that might be detected (depending on the configuration of the primers used in the PCR) which may also result in false positives. 

Analysis: I don't see this as a metodology that the food industry will find applicable.