[Lab Management] The Challenges of Pathogen Detection

If the food safety system of the United States is among the world’s best, why are recalls always in the news?

Recalls may seem to indicate that U.S. food has become less safe. But, said BioMerieux Director of Scientific Affairs Stan Bailey, "I would argue the opposite—that because we are having recalls, our food is safer."

"I think we are having more recalls because we have better tools to epidemiologically link more cases," Bailey said, citing programs such as CDC’s PulseNet, which enables analysis of genetic typing of outbreak-causing bacteria to better detect outbreaks. When the data is fed into a common database, "we are able to detect and see linkages that we never saw before."

Wendy Warren, chief science officer of Food Safety Net Services, agrees that a key reason behind U.S. recalls is the nation’s increased sophistication in detecting the root cause of illness. "Our ability to detect illness and assign it back to a pathogen is sophisticated," Warren said. "And it’s continuing to become more sophisticated."

Even noting the advances and sophistication of U.S. food safety, both scientists agreed that the industry continues to face challenges in pathogen detection that make it difficult, if not impossible, to have pathogen-free foods.

"These are the same challenges there have always been," Bailey said. The bad bacteria are generally present in small numbers, alongside good bacteria in larger numbers, so they can be very difficult to detect. "You’re searching for a needle in a haystack.

"Bacteria are everywhere," he said. "And only a small portion of them are bad; we couldn’t live without the good bacteria."

The question is often asked: Why does it take so long to detect pathogens in foods? Because the pathogens are usually present in small numbers, we have to grow them to the level that our current technologies can detect them, Bailey said. "All of the current commercial technologies need between one thousand and 10 thousand organisms per milliliter of enrichment broth for detection."

This is true, at least, in the current state of technology which does not always allow for real-time analysis, Bailey said. "Because we have to take that small number of pathogens and grow them up to a number that can be detected."

Sampling Challenges. A key challenge in detection is getting a representative sample, Warren said. "Pathogens tend to be very much unevenly distributed in food products that are not produced in a homogenous way."

Warren cited an example of the differences in sampling of milk and beef trim. Because the milk would be mixed during processing, pathogens would be evenly distributed and a sample more likely representative. In a solid, such as beef trim, or a food with large chunks, she said, "you can have pathogens that are scattered inconsistently, and generally they are at low levels."

So, she asked, "How do you properly sample a batch so you get a good sample for the entire batch?"

Sample methods are, in fact, often based on mathematical and statistical models that assume even distribution, she said. Only recently have sampling plan publications begun to add sub-caveats noting that organisms can be unevenly distributed. In these, she said, they will often refer to homogenous foods, such as liquids which have a high degree of mixing; intermediate homogenous foods, e.g., mixed solids, such as ground beef; and inhomogenous foods, those which are completely unmixed solids.

The question in sampling can then become a matter of balancing risk between the consumer and the producer. That is, Warren said, "You don’t want to release a batch that is potentially risky, but you don’t want your plan to be so terribly stringent that you are rejecting everything and have no product to sell."

*****

The Line in the Sand

"The best quality program in the world will find and eliminate systemic errors. But there’s always the chance of the single-time whoops," said Mike Wolf, lab director of JL Analytical Services.

Whether you work on a bench as a chemist or in an office as an accountant, it is important to double check your work before you call it complete. "There seems to be a line in the sand," that has to do with whether a person learned to calculate math problems with a slide rule or with a calculator, Wolf said. For those too young to remember, a slide rule is defined by The American Heritage Dictionary as "a device consisting of two logarithmically scaled rules mounted to slide along each other so that multiplication, division, and other more complex computations are reduced to the mechanical equivalent of addition or subtraction." Or, if you prefer, a simpler definition from Dictionary.com: "a device for performing mathematical calculations, consisting essentially of a ruler having a sliding piece moving along it, both marked with graduated, usually logarithmic, scales: now largely replaced by the electronic calculator."

It is precisely its being replaced by the calculator that Wolf sees as the line in the sand. When calculations are made with a slide rule, tracking has to be done in one’s head. Because of this, he said, "Once you finish the calculation, the next thing you do is ask yourself: Does this make sense?"

To relay it to lab work, Wolf said that when a test is finished, the results should be put in numerical order, then the values reviewed with the thought in mind: Does this make sense?

Students who learned to calculate problems with slide rules were taught to check themselves; to be self doubting. Those who learned on an electronic calculator tend to not question if it could be wrong. Because, of course, a calculator can’t be wrong. Unfortunately this does not take human error into consideration—that the "calculator" who is pushing the numbers could be wrong.

*****

Method Validation. A second key challenge is that of selecting a sampling method that can adequately detect the specified level for a specific product, Warren said. That is, has the method been validated according to the conditions in which it is to be applied? Even if the method was validated for the product, it may be at a different defect level or portion size.

Thus, the challenge is, she said, "making sure the test method is appropriate for how it is being used."

Over the last year a great deal of forward movement has been seen in test method validation and laboratory accreditation, Warren said, noting that she sees it continuing. "I expect to see fruition of that this year."

This is partly due to the fact that the regulatory agencies have heightened their focus on what’s happening in the labs, she said. In addition, when there is a lack of clarity and weakness in lab operations, "that could be misleading food manufacturers in terms of data." Thus, it is important to ensure that there is validation data on the labs’ testing methods and all applications are verified, she said, "so that all the wheels are turning in the right direction."

"I think that’s an area that has just been assumed," she said. "If there’s no oversight, you can’t get verification. It’s a good time for transparency."

Emerging Organisms. The question was also posed to the scientists as to whether new disease-causing organisms are emerging. Although neither knew of any specific "new" organisms, they did not discount the possibility.

"There is always the opportunity for new disease-causing organisms," Warren said, adding, there could be something occur of which we are not aware.

Bailey answered the question with examples of pathogens discovered in recent decades, then countered with a question of his own.

In the mid- to late 80s, Salmonella enteritidis was discovered in poultry, he said. At the time, the industry had a "pretty good" system for Salmonella tracing, but "all of a sudden we saw Salmonella enteritidis primarily linked to eggs around the world." Soon after, he said, the pathogen became one of the top two serotypes causing illness. Because the detection methodology had not changed, this was probably an emerging pathogen.

Another example Bailey gave was Campylobacter. The apparent prevalence of Campylobacter significantly increased in the late 1980’s, he said, adding that, in his opinion, "this increase was due to improved methods of detection and was not necessarily a true increase in prevalence."

Thus, the question, Bailey posed regarding newly "emerging" pathogens: "Is it an emerging pathogen or is it an emerging technology able to detect the pathogen?"

Incremental Improvements. In the similar way, Bailey sees the technology of pathogen detection as gradually evolving rather than having sudden emergences—though he keeps an open mind to this as well. "There’s always a possibility of some dramatic breakthroughs, but it seems to me the diagnostic industry is making incremental improvements."

Rapid detection methods, for example use two or three basic platforms, but the specific pathogen tests on these platforms are improving, so they are more able to be performed rapidly with less cross reactivity, he explained.

And can rapid detection keep evolving or does it reach a point at which results are sacrificed? "You can only get so rapid at this point without sacrificing the probability of detecting at low concentrations," Warren said.

However, there is a working group, established by the International Association for Food Protection (IAFP), that is currently looking at sample prep methods for detection. Charged to investigate new technologies, the group is seeking prep methods that could potentially provide an alternative to the time-consuming nature of enrichment, such as techniques added into the process to help concentrate it and increase the reliability of detecting the target if it’s in there, she said.

Because today’s tests rely on culture-based detection, they require that bacteria grow to certain levels before they can be detected. "And they require a certain amount of time to grow," Warren said. "There is a fine balance between getting reliable detection versus a potential for a false negative if you shorten the time."

Accountability. In the new economy, with growing variance between small business and mega-corporations, questions of scale and accountability also come into play. "Do you treat a small farmer the same as a big company?" Bailey posed.

Having grown up on a small farm and now working in pathogen management, Bailey can see both sides of the equation. "My heart reaches out to the small farmer," he said. "But from a food safety aspect, I can’t say you have to treat them different. Maybe a little, but the ultimate outcome has to be the same.

Foodborne illness can originate on even the smallest spinach or pepper farm, Bailey said. "It’s the weakest link that pulls the whole chain apart."

*****