4. Risk Assessment


4.1 Statement of purpose
4.2 Hazard identification
4.3 Exposure assessment

4.3.1 Geographical distribution
4.3.2 Prevalence of L. monocytogenes in fish and fishery products
4.3.3 Growth and survival of L. monocytogenes

4.4 Hazard characterization

4.4.1 Epidemiological patterns of listeriosis
4.4.2 Dose response

4.5 Risk characterization


4.1 Statement of purpose

The purpose of this section is to document current scientific knowledge concerning the risks of listeriosis in relation to fishery products, in order to enable identification of risk contributing and risk mitigating factors for facilitating risk management decisions. While the risk assessment includes some quantitative information, the time frame of the Consultation did not permit a quantification of the risks of listeriosis associated with fishery products.

4.2 Hazard identification

L. monocytogenes is the sole hazard of interest in this assessment.

L. monocytogenes is a bacterial pathogen causing serious human illness such as perinatal infections, septicaemia and meningitis. More recently, a new form of disease caused by L. monocytogenes has been recognized involving mild gastrointestinal symptoms. Although listeriosis occurs infrequently, at an annual incidence rate of 2 to 10 cases per million, the fatality rate is high, usually in the range of 20-30%. Highly susceptible individuals include pregnant women, neonates, elderly people and immunocompromised individuals. This last group includes, in decreasing order of risk, organ-transplant recipients, patients with AIDS, HIV-infected patients and patients with cancer. We are likely to see an increase in the incidence of listeriosis as the numbers of susceptible individuals and vulnerable groups increase over the next decades.

Although other routes of transmission have been described, indistinguishable strains have been isolated from epidemic cases and from the food implicated, clearly identifying the role of food in the epidemiology of listeriosis. Foods associated with transmission are characteristically highly processed, have extended shelf lives at refrigeration temperatures, are capable of supporting the growth of Lmonocytogenes and are consumed without further cooking. Although some strains of L. monocytogenes are more frequently associated with human diseases, at this time, all strains must be considered to be potentially pathogenic.

4.3 Exposure assessment

4.3.1 Geographical distribution

Listeriosis is mainly reported in industrialized countries with few or no reports from Africa, Asia and Latin America. It is not known whether this reflects different exposure rates, consumption patterns, dietary habits, host susceptibility, or lack of testing facilities in different regions.

4.3.2 Prevalence of L. monocytogenes in fish and fishery products

4.3.2.1 Water and raw fish

Unpolluted seawater and ground waters used in aquaculture are generally free from this organism, and fish from these environments are uncontaminated. In temperate regions, the organism has been isolated from surface waters and lakes, and in coastal waters subject to pollution or contamination from industrial, human or animal sources. When L. monocytogenes is present in fish from these environments it is usually present in very low numbers. In tropical regions, the reported incidence of L. monocytogenes is very low and freshly harvested fish from these environments are generally free of this pathogen.

Different strains of L. monocytogenes may be isolated from fish raw material and from final products suggesting that fish can be contaminated at any point between harvest and consumption. In particular, L. monocytogenes can colonize the processing environment and this has been established as a primary mechanism of contamination for some products.

4.3.2.2 Cold smoked fish

Cold smoked fish is produced by filleting raw fish which is subsequently salted (3.0-3.5% NaCl in water phase), dried and smoked at a temperature < 30oC. The fish is then usually sliced and vacuum packaged. It should be stored at a temperature 5oC, and under these conditions has a recommended shelf-life of 3 to 6 weeks. Several surveys reveal that 10-60% of freshly processed cold smoked salmon is contaminated with L. monocytogenes. The contamination level is usually 100 cfu g-1. A similar contamination rate has also been observed on other lightly preserved fish and fishery products that are processed without a listericidal step. During storage for 2 to 3 weeks at 5oC L. monocytogenes may increase in numbers but it is difficult to predict the magnitude of increase. Uncontrollable factors including the physiological state of the organism and the presence and type of microflora on the product will affect the potential for growth of L. monocytogenes.

The processing environment contributes to the presence of L. monocytogenes in cold smoked fish. L. monocytogenes can become established on processing surfaces and may be very difficult to detect and to remove. After production, time and temperature of storage are the most important factors that determine outgrowth. Survey data suggest that levels rarely exceed 103 cfu g-1 at the time of consumption.

4.3.2.3 Cooked fishery products

L. monocytogenes has been recovered from a range of ready-to-eat fishery products, such as cooked shrimps and cooked crabmeat. The rate of contamination of cooked ready-to-eat products varies. For example, less than 1% of ready-to-eat fishery products imported into Canada during 1996/98 were contaminated with L. monocytogenes while pooled data from 6 other published reports (involving 6 countries, 498 samples) on ready-to-eat cooked shrimp indicated a contamination rate of about 4.8%. This indicates post-process contamination, as L. monocytogenes should be reduced to undetectable levels by cooking. When cooked products are stored at chill temperatures (i.e. < 5C) for extended periods, the potential for growth of L. monocytogenes will be less restricted because of the absence of competitive flora.

4.3.3 Growth and survival of L. monocytogenes

4.3.3.1 Growth potential

The major factors controlling the fate of microbial populations in fish and fishery products are temperature, water activity and pH. Specific organic acids, and many preservative compounds, may also play an important role in reducing microbial growth in fishery products. Thus, the potential for growth of Listeria in fishery products is related to the product composition, type and intended end use. Some examples are shown below.

Products with the potential to contain high levels of L. monocytogenes at the time of consumption

Lightly preserved fish products (pH >5.0, NaCl < 8.0%), including:

Heat-treated, before packaging:

Products unlikely to contain high levels of L. monocytogenes at the time of consumption

The growth limits of L. monocytogenes are summarized in Table 1.

The tolerance to a particular environmental constraint is greatest when all other conditions are optimal for growth. For L. monocytogenes, growth is maximal when temperature is in the range 20 – 25C, although the growth rate is fastest at ~37C. When several factors are sub-optimal for growth, the growth permitting ranges of each of those factors will be reduced. Thus, growth at the lowest pH or water activity will only occur when all other conditions are optimal. This behaviour has been embodied in the "Hurdle Concept" and is reflected in the product categories presented earlier in this section.

There are exceptions to this general trend, e.g. while slightly elevated salt concentration may inhibit growth rate, it may at the same time increase tolerance to high temperature of many bacterial species.

Table 1: Growth limits for L. monocytogenes

Environmental Factor Lower Limit Upper Limit
Temperature (C) -0.3 to + 3 ~ 45
Salt (% NaCl, water phase) < 0.5 12-13
Water activity (aw)* 0.91 – 0.93 > 0.997
pH (HCl as acidulant) 4.2 – 4.3 9.4 – 9.5
Lactic acid (water phase) 0 3.8 – 4.6 mM, MIC of undissociated acid

(800 – 1000 mM, MIC of sodium lactate)

* values for NaCl as the humectant

4.3.3.2 Growth rates

The physico-chemical attributes of many fish and fishery products will permit the growth of L. monocytogenes. Some representative growth rates for conditions relevant to these products are shown in Table 2.

Table 2: Representative maximum growth rates of L. monocytogenes predicted by a mathematical model

Generation time (hours)
Temperature (C) pH 7.0, aw 0.990 90 mM total lactate, pH 6.2, aw: 0.990 90 mM total lactate, pH 6.2, aw: 0.965
25 0.8 1.0 1.6
10 6 7 11
7 13 15 24
5 29 33 53
0 620 730 1150

As discussed earlier, it is recognized that other factors may affect potential for growth of L. monocytogenes in foods. Also, L. monocytogenes is known to be sensitive to quaternary ammonium compounds, chlorine and sanitizers containing peracetic acid and peroctanoic acid. Irradiation can effectively reduce L. monocytogenes to undetectable levels in products. Due to all these factors, the presently available predictive models for growth are unable to make accurate predictions. Accurate prediction of growth is a necessary tool for a proper risk assessment.

4.4 Hazard characterization

4.4.1 Epidemiological patterns of listeriosis

The epidemiological pattern of human listeriosis is a background of sporadic cases with occasional outbreaks. A decrease in the number of sporadic cases has been observed during the last ten years in the UK, USA and France. While several reports indicate that fish and fishery products can be frequently contaminated with L. monocytogenes, no major outbreaks associated with these products have been reported. This may be due to inadequate surveillance systems in several countries or because not all factors contributing to both sporadic cases and outbreaks associated with fisheries products have been identified. A low number of cases (Table 3) have been linked to fish associated listeriosis outbreaks when compared to listeriosis outbreaks associated with other foods.

Invasive listeriosis includes perinatal infections, central nervous system infections and bacteremia. Neurologic sequelae have been observed in 30% of patients with central nervous system infections. Ninety per cent of cases occur in populations with impaired immunity. The incubation period ranges from 1 to 90 days, with a mean of 3 to 4 weeks. This variability may reflect variation in the number of cells ingested, differences in host susceptibility, or differences in virulence of the strains. Excepting nosocomial infections, person-to-person transmission has not yet been firmly documented during food-borne community acquired outbreaks. The role of healthy carriers (4 - 6% of the healthy population) in the epidemiology of listeriosis warrants further studies.

Table 3: Main documented cases of foodborne listeriosis associated with fish and fishery products

Location / year

No. cases

Clinical forms1

No. deaths

Food

No. cfu g -1

Strain serovar

USA- 1989 2 2 2 2 0 Shrimp NK 4b
Italy – 1989 1 0/1 0 Fish NK 4b
Australia – 1991 2 0/2 0 Smoked mussels 1 x 107 NK
New-Zealand 1992 4 4/0 0 Smoked mussels NK 1/2b
Canada – 1996 2 0/2 NK Imitation crab meat 2 x 109 1/2b
Sweden – 1994/95 8 3/8 2 "gravad’/cold smoked rainbow trout <100-6200 4b

1: no. of pregnancy related cases / no. of non-pregnancy related cases;
2 : not known if this was the total number of cases
NK: not known

4.4.2 Dose response

The dose response relationship of L. monocytogenes for humans is not known. In general, the infectious dose of a foodborne pathogen depends on a number of variables including the condition of the host, the virulence of the strain, the type and amount of food consumed, and the concentration of the pathogen in the food. Animal studies suggest that reducing levels of exposure will reduce clinical disease. From the reported numbers of Listeria in contaminated food responsible for epidemic and sporadic foodborne cases, there is little evidence that a very low number of L. monocytogenes in food causes listeriosis. In a number of outbreaks, enumeration of units of implicated foods indicated both high (> 1,000 cfu g-1) and low levels (< 100 cfu g-1) of contamination. Some of these data related to fish and fishery products are summarized in Table 3. High levels of L. monocytogenes (103–10cfu gr1) have been detected in soft cheeses involved in 5 outbreaks. These facts, together with data on the recovery of the organism from implicated foods suggest the likelihood that high infective doses are involved in most cases.

4.5 Risk characterization

From the foregoing discussion, it is clear that ready-to-eat fishery products frequently contain L. monocytogenes. The level of contamination at the time of consumption cannot be calculated with confidence. Usually low levels are found and it is known that levels of ~10 – 100 cfu g-1 may occur infrequently at the time of manufacture of cold smoked salmon. In cooked products, lower levels of contamination occur. Growth of L. monocytogenes on these products can be demonstrated, although growth is slow under proper storage conditions. Calculations suggest that per capita human exposure to doses of L. monocytogenes exceeding 1,000 cfu (total ingested dose) is likely to occur several times each year. Despite this exposure, the total incidence of invasive listeriosis is estimated to be 2-10 per million population per annum in countries where data are available.

Listeriosis outbreaks may extend over many months or years, and with low attack rates. The incubation period is usually of the order of 3–4 weeks, but may extend to over three months. These factors combine to make outbreaks difficult to recognize and to investigate. The severity of consequences of listeriosis is high, with 20 - 30% case-fatality rates.

Recognizing the potential for fishery products to be a vehicle for large outbreaks of listeriosis, it is prudent to seek appropriate strategies to minimize human exposure to infectious doses of the organism. Currently, however, there is insufficient data for the dose-response relationship to be determined, but there is growing consensus that the majority of the healthy population is highly resistant to doses in the order of 1,000 to 10,000 cfu L. monocytogenes. There is no consensus on the dose required to cause illness in susceptible individuals. However, using the available epidemiological data, and consumption patterns, crude estimates of the risk of listeriosis from fishery products can be made.

Regarding fish and fishery products, the dose-response has been estimated elsewhere combining data on the incidence of listeriosis in Germany with data on the levels of L. monocytogenes in smoked-fish in that country. It was necessary to make many assumptions, but all were chosen to result in conservative estimates. Epidemiological and food survey data were combined, using a predictive modelling approach, to estimate a dose-response relationship for L. monocytogenes levels and incidence of listeriosis. Two methods were used to model and calculate the dose-response relationship. Both methods gave a similar result. Using that approach, for the estimated immunocompromised sub-population of Germany (20% of the population), the model predicts a 1 in 59 million chance of infection from consumption of a 50g serving of fish containing 100 bacteria per gram.

It must be emphasized that this estimate rests on many assumptions. Nevertheless, when extrapolated across the population and the number of salmon meals consumed per year, the estimate reflects the low incidence of listeriosis in those nations where epidemiological data are available. The above estimate, is of the same order of magnitude as the reported per annum incidence of listeriosis.

Thus, despite the fact that there is considerable potential for fish and fishery products to cause listeriosis, the available data indicate that this potential has not been observed. There is little epidemiological evidence to implicate fishery products in large outbreaks. The reasons for this are unknown.