Pathogen Detection: Minimize Contamination Risk

As the risk-based provisions of the Food Safety Modernization Act roll out, risk assessment is being conducted in areas in which it was never before applied and increasing for practices for which it has always been a part. Pathogen detection and sampling is no exception.

As with any risk-based practice, the areas in which the risk of pathogenic contamination are greatest are the areas in which detection, sampling, and prevention should have the greatest focus.

While the most relevant factor in risk assessment is the product being produced, this generally leads to the next question as noted by Stan Bailey, director of scientific affairs for bioMerieux.: Should you sample the product or should you sample the environment?

Many companies rely on finished product testing to ensure that a product will be safe when it goes to the customer. However, Bailey said, finished product testing has significant limitations, because there is a high probability that contamination will be missed in product sampling. For example, he said, referencing the chart (at right), if three samples are taken in a lot of which two percent is contaminated, there is a 94 percent chance that the contamination will not show up; even if 10 samples are tested, there remains an 82 percent chance that the contamination won't be detected.

"End-product testing is not where a food processor can get a comfort level that there's no contamination in a product," Bailey said. "Where you really get your best return and feel for what's going on is environmental testing."
 

Environmental Testing. While every plant is different, there are general practices that all plants can follow for food safety. One such practice for environmental sampling is the four-zone plan, Bailey said, with highest risk and greatest sampling starting with Zone 1.

• Zone 1: Presenting greatest risk is anything that comes in contact with food, such as slicers, conveyors, employee hands, etc.
   
• Zone 2: Anything in close proximity to but not touching foods, such as the exterior of equipment, housing, etc.
   
• Zone 3: Anything not in direct contact with foods that could cause secondary contamination, that is, things that employee hands might touch, such as hand trucks, walls, etc.
   
• Zone 4: Areas away from processing and not likely to contribute to contamination, e.g., cafeteria, loading dock, office area, etc.

The goal of environmental testing should be to reduce the probability of contamination as much as possible prior to end-product testing, Bailey said. In an ideal world, a well-developed, well-implemented environmental plan would make end-product testing unnecessary, but in the real world, it will always be needed as a verification step, he said.

In addition to verifying testing results, the technology should be validated, said Melinda Hayman, principal scientist for Food Safety Net Services. "Environmental testing can be tricky because of sanitizers," she said. Chemicals in the sanitizers can cause a harsh environment that is stressful on the bacterial cells, so they are not as healthy as they may be in the food. This can make them more difficult to detect.

Because of this, she said, "there could be something on a surface that the method doesn't pick up." So it is critical to use a method that is validated, through an in-house or third-party lab.
 

Data Analysis. In addition to collecting data through testing, Hayman said, a complete program includes monitoring and analysis of the data for improvement. "A lot of people collect data, put it in a folder, then put that on a shelf without looking at it again," she said. Instead you need to analyze the data, look for trends, then use it to your advantage.

"It is better to know about a problem early than late, so being proactive is very important," Hayman said.

Creating data for such analyses also means testing up and down the line on a continuous basis, Bailey said. This means knowing what the microbiological profile of your product should be at each step, then testing, measuring, and monitoring against that profile.

Because of the statistical probability of missing contamination, plants should look for appropriate indicator organisms rather than just testing for a pathogen itself. That is, instead of testing for the specific pathogen, test for the category of general bacteria that is likely to be present. This is generally easier to test for, and by knowing what the microbiological level of that bacteria should be at each step, you can base testing on total bacteria count, with an alert and corrective action if the count is higher than specified.

By relating this back to environmental testing, setting it up according to zones, and sampling surfaces at specified frequencies throughout the day, you can establish the baseline profiles against which to conduct testing and set up a routine monitoring program for that. "In a sense," Bailey said, "you are biomapping your plant."

The advantage of this, he said, is that it enables you to see levels of microbiological activity rather than just that a pathogen is or isn't present. "If you are sampling something that is only present one-tenth of the time, you may go a month, or three, where every sample comes back negative. But does that tell you anything about how good your process is?"

If, instead, you are sampling the level of something that you know is there, you are able to tell what stage it is at, and know—in advance—if something is out of control in your process. "If something is too high, then something in your process isn't working right," Bailey said. "You want to know that at any given time you are in control of your process."

When you combine an intensive environmental monitoring program with a proactive supplier qualification program and end-product testing for validation, Bailey said, "You can feel comfortable that the product going out of your plant is as safe as you can technologically make it."

The author is Editor of QA magazine. She can be reached at llupo@gie.net.