Microbiological testing is no longer the method of choice for ensuring food safety. Manufacturers now rely on HACCP-based food safety management systems to make sure foods don’t contain pathogens rather than asking the laboratory to test for them in finished products. Nevertheless, testing remains an important backup, both for verifying that the all-important HACCP system is actually working and for tracking pathogens in food supply chains.
So any suggestion that food microbiology labs regularly report false negative and false positive results is a concern. But that is exactly what research conducted in the USA by the American Proficiency Institute (API) discovered. Most labs carry out proficiency testing to check that their methods and procedures work properly and API looked at the results of those tests going back 14 years. They found that on average labs reported more than 9% false negative results for Campylobacter and nearly 5% for Salmonella. These figures are a worry because they indicate that labs may be missing a significant number of contaminated samples, and that may lead to cases of foodborne disease that are actually preventable.
In fact, I suspect that the rate of false negative reports in real food samples, as opposed to artificially spiked proficiency test samples, could be even higher in some cases. I am reminded of a survey of Campylobacter in retail poultry meat about ten years ago, which seemed to suggest that the rate of contamination in one half of the UK was almost twice that in the other half. This appeared inexplicable until one looked in detail at the survey report, when it became apparent that two labs were used to test the samples, one receiving samples from one half of the country, and the second from the other half. The clear implication was that the differences were a result of inter-laboratory variation. Quite simply, lab 1 was nearly twice as good at finding Campylobacter as lab 2, and I’m guessing that both were enrolled in proficiency testing schemes. I’m certainly not suggesting that proficiency testing is not useful – quite the contrary – but I do think that food safety professionals sometimes forget that even the best run and equipped lab is not infallible. Pathogen detection methods get better all the time, but none can guarantee to find every needle in every haystack.