Food Engineering

Rapid detection methods for common foodborne illnesses

July 5, 2011
New assay methods shorten the time food is held up in quarantine waiting for an OK to ship, while automation is taking the human error out of test results.


Stan Bailey. Source: bioMérieux.
The speed at which E. coli O157:H7 or Salmonella can be detected in a food sample is typically takes hours to a couple of days-certainly not as fast as the couple of milliseconds it takes to find metal or other foreign bodies in food. However, for the most common bacteria-such as E. coli O157:H7 and Salmonella-same-day or next-day assay results are becoming a reality, and in the future, new technologies such as biosensors could reduce assay times dramatically, says Dr. Stan Bailey, director of scientific affairs for the food industry at bioMérieux and former senior research scientist at the USDA Agricultural Research Service.

For processors that must quarantine products until assay results are known, a shorter results time means products quickly enter the supply chain, thus having a longer shelf life, occupying less warehouse space and saving money.

Bailey, in an exclusive FE interview (PDF), said today’s rapid detection methods are much faster than the four to seven days normally required for cultured tests 20 or more years ago. “Now we’re moving into generations of next-day (24-hr) tests, with a few selected cases, primarily for E. coli O157:H7 and maybe some Salmonella in raw meat products in an eight- to 16-hr. range,” says Bailey. For most food products, the Salmonella test requires 16-18 hrs. of incubation, whereas E. coli can run 8-10 hrs., depending on sample size, says Bailey. Bailey’s company has just released a rapid test for Salmonella, which uses bacteriophage technology to detect the Salmonella. A rapid test of E. coli O157:H7 has been available for a year or so.

There has been much discussion on whether the non-shiga toxin E. Coli (STEC) serotypes are as dangerous as O157:H7. “Unfortunately, the non-O157s don’t have the same phenotypic trait. They look like every other run-of-the-mill E. coli on earth, which is about the most common bacteria there is,” says Bailey. “So being able to pull these non-STECs out of the rest of the E. coli population in a sample is much more problematic than finding O157. The non-STECs are only important if they’re present and they possess certain pathogenic virulence factors.”

So the question is an economic one. Typically, 90 percent of E. coli cases in the US have been the result of O157:H7, and that leaves the other 10 percent to non-STEC bacteria, says Bailey. Tests, of course, are available for the non-STECs, but how many samples should be taken and tests run, and how many tests can processors afford to run, holding product in limbo while waiting for results? These are tough questions-especially when unknown bacterium may potentially wreak havoc as was the case with STEC 104:H4 in Germany.

To read the complete FE interview with Stan Bailey, follow this link to the PDF.