Researchers call for change of approach to prevent Listeria infections

Robert Roos, Staff Writer May 9, 2003 (CIDRAP News)

A pair of studies on Listeria monocytogenes in ready-to-eat (RTE) foods and associated illness rates suggest that only a small fraction of all Listeria contaminated foods contain enough of the pathogen to cause illness. Therefore, the authors conclude that prevention efforts should focus on limiting the concentration of Listeria in selected foods rather than on the current policy of "zero tolerance," or eliminating the pathogen from all RTE foods.

The studies by researchers with the National Food Processors Association involved the collection and analysis of thousands of samples of RTE foods in Maryland and California over a 2-year period, followed by a comparison of the findings with the incidence of listeriosis in the same areas and time frame. The studies were reported recently in the Journal of Food Protection.

The investigators found L. monocytogenes in 1.82% of more than 31,000 food samples analyzed, but only 21 samples contained more than 100 colony-forming units (CFU) per gram (approximately equivalent to 100 organisms per gram). Seafood salads and smoked seafood were the RTE foods most likely to contain any Listeria, while smoked seafood and luncheon meats constituted most of the samples that had high levels.

In the second study, the authors used mathematical modeling to assess the relationship between the dose of Listeria consumed and the risk of illness. They concluded that nearly all cases of listeriosis are linked with food items that have relatively high levels of Listeria contamination.

"It is clear that the most effective efforts to reduce the risk of listeriosis in RTE foods will involve targeting the food servings that are heavily contaminated, even though the fraction of those servings is very small," the investigators state.

The Centers for Disease Control and Prevention (CDC) has estimated that up to 2,500 cases of listeriosis, with 500 deaths, occur annually in the United States, mostly involving pregnant women, newborn babies, the elderly, and immunocompromised people, the authors note. Given the prevalence of Listeria in RTE foods - up to 5% for certain food types in some studies - it's not clear why the disease isn't more common, the article says.

Since almost no data have been available on the amounts, as opposed to the prevalence, of Listeria in RTE foods, the authors sought to determine both the prevalence and the concentration of the pathogen. They focused on the CDC's Foodnet (Foodborne Diseases Active Surveillance Network) sites in northern California and Maryland, two of nine regions where the CDC actively monitors laboratory-confirmed cases of foodborne disease.

Samples of luncheon meats and deli salads were collected weekly in retail stores over 23 months. In addition, samples of five other products - smoked seafood, seafood salad, blue-veined or soft mold-ripened cheese, bagged salads, and fresh soft cheese - were collected from retail stores for about 50 weeks. Rapid DNA-based assays and culturing were used to assess the presence and levels of L. monocytogenes in the samples.

Of 31,705 samples, 577, or 1.82%, tested positive for Listeria. The prevalence rates ranged from 4.7% for seafood salads and 4.3% for smoked seafood to 0.74% for bagged salads and 0.17% for fresh soft cheese. The rates for deli salads and sliced luncheon meats - 2.4% and 0.89%, respectively - were "considerably lower than expected," the researchers say. By comparison, the US Department of Agriculture (USDA) reported rates of 4.2% to 8.0% in 2,300 luncheon meat samples collected in federally inspected facilities in the 1990s, the article says.

In 402 of the 577 positive samples, the level of Listeria was extremely low—less than 0.3 organisms per gram. As noted, 21 of 577 samples had more than 100 CFU/g. Just five samples contained more than 1,000 CFU/g. Of the samples containing more than 100 CFU/g, nine were smoked seafood and eight were luncheon meats.

Listeria was significantly more prevalent in luncheon meats, deli salads, and seafood salads packaged in stores than in manufacturer-packaged samples of the same foods, the authors found. However, the pathogen levels tended to be higher in the manufacturer-packaged samples; 16 of the 21 samples containing more than 100 CFU/g were manufacturer-packaged.

In the second study, the authors note that when listeriosis emerged in the early 1980s, the federal government established the "zero tolerance" policy whereby food containing any detectable level of L. monocytogenes is considered adulterated. "However, data suggest that L. monocytogenes cannot be eliminated from the environment or from all foodproducts," and it continues to crop up in RTE foods periodically despite extensive control measures, they write.

Because the infectious dose of L. monocytogenes for humans is unknown and cannot be determined experimentally in view of the high mortality risk, the investigators sought to assess the dose-response relationship by using data from the food survey along with listeriosis incidence data from the northern California and Maryland Foodnet sites. The CDC reported 53 listeriosis cases in the two areas during the period of the food survey (2000 and 2001). To compensate for underreporting, the authors assumed that 106 cases actually occurred.

The mathematical model constructed by the authors indicates that essentially all of the 106 listeriosis cases would have been caused by foods containing L. monocytogenes at more than 100 CFU/g. Further, the model suggests that food containing more than 100,000 CFU/g would have caused 67 of the 106 cases. In the food survey, only 2 of the 577 positive samples, both smoked seafood, were contaminated at that level.

Calling for a more targeted approach to preventing listeriosis, the authors write, "The zero tolerance strategy is in essence a prevalence-oriented approach that does not distinguish foods contaminated at high concentrations from those contaminated at low concentrations." Thus, a 50% reduction in prevalence would be likely to yield a 50% reduction in illness cases. According to the authors' model, however, a control strategy that left Listeria prevalence the same while limiting the concentration to 10,000 CFU/g could reduce illness cases by 89%. Limiting the contamination to 100 CFU/g could reduce illness risk by 99.5%, the authors predict.

They further suggest aiming control measures at the RTE food types that contribute the most to listeriosis cases, which their survey implies would be smoked seafood and luncheon meats, rather than targeting all RTE foods. Their study was supported by the USDA and the Joint Institute for Food Safety and Applied Nutrition at the University of Maryland as well as the National Food Processors Association.

Gombas DE, Chen Y, Clavero RS, et al. Survey of Listeria monocytogenes in ready-to-eat foods. J Food Protect 2003;66(4):559- 69
[Abstract http://fidelio.ingentaselect.com/vl=6461888/cl=46/nw=1/rpsv/cw/iafp/0362028x/v66n4/s4/p559]

Chen Y, Ross WH, Scott V, et al. Listeria monocytogenes: low levels equal low risk. J Food Protection 2003;66(4):570-7
[Abstract http://fidelio.ingentaselect.com/vl=6461888/cl=46/nw=1/rpsv/cw/iafp/0362028x/v66n4/s5/p570]