• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Animal & Veterinary

  • Print
  • Share
  • E-mail

Assessment of Risk: Microbiological Risks

Dick Whiting, Ph.D.

DR. WHITING: Thank you very much for the introduction there, Keith. And it is my pleasure to be here and talk a little bit about microbial risk assessment. And microbial risk assessments are really a new area. I think there has probably been less than a dozen, say, full microbial risk assessments that have been done anywhere in the world at this point.

Within the U.S. Government, we did the Salmonella enteritidis in eggs risk assessment a year ago. That was the first one. Being presented today downtown, the USDA is talking about their E. coli 0157:H7 risk assessment in ground beef. Within the Food and Drug Administration, we have a risk assessment on Listeria monocytogenes and one on Vibrio parahaemolyticus ongoing right now. So this is a new area and we are sort of inventing microbial risk assessment as we go.

I see the risk assessment as really a pre-regulatory process. In other words, we see the risk assessment as basically an information-gathering and evaluation process. And in that, we follow the recommendations that have come out that risk assessment and risk management should be kept sort of separate.

We see the microbial risk assessment of trying to follow the paradigm that has been developed by the chemical people, that is risk assessment, risk management, risk communication. And within the risk assessment area, we talk about hazard identification, exposure assessment, dose response and risk characterization.

And in short, you know, the risk assessment determines, you know, what can happen, how likely is it to happen, what are the consequences. Or you can say the risk assessment determines what do we know and how certain are we of what we know.

Now, when it comes to trying to do a microbiological risk assessment, we have had some real problems doing this. It is a new area. The data gaps are quite large in the field of microbiology. I suspect that is because most people who have had an inclination for science decided to go into microbiology because they didn't like math and statistics. So this is now coming around to haunt us when we try to do risk assessments.

You know, microbiologists just don't develop models when they publish papers. They don't characterize the variations and the standard deviations. And they love to do that presence-absence type of analysis which, you know, just doesn't get us too far when it comes to risk assessments.

But we do have despite saying we are following the paradigms of -- that have been pioneered by the chemical people, we do see some real differences in microbial risk assessment versus some of the others. In microbiology, we are generally concerned about acute situations and single doses. The statistics say you are likely to get a food-borne illness about once every ten years. So, you know, the chances of having two in a day are rather low.

And we also think about acute illness. But even as I say this, I realize there are some exceptions. We are beginning to talk about long-term sequelae to some of the microorganisms, Guillain Barre syndrome, HUS from E. coli or reactive arthritis from Salmonella.

And we also realize that perhaps chronic exposure to low levels of certain microbial pathogens may affect your susceptibility to when you are exposed to a large dose or is there maybe some sort of immune type response going on here. But at this point, we just really don't know enough to do much in terms of risk assessment or modeling of this.

One big difference with microorganisms is we can't just keep diluting them. Eventually, we get to one bacteria. And at that point, we then have to start talking about probabilities of occurrence. In other words, if you have one big tank, we can talk about one bacteria surviving a pasteurization process in 40,000 gallons perhaps.

But then you begin to put it into an individual carton for retail sale. And we eventually get down to one bacteria which is now in, say, one package out of 100 or one package out of 1,000. So we have to now switch from sort of a quantitative level to more of a probablistic type of hazard assessment.

But perhaps the biggest difference with microbiology is bacteria can grow. And if there is an abuse period with a food, it is not unreasonable to see 100,000-fold growth. Certainly, a 1,000-fold growth is very likely. So -- and also, we can see a similar sort of decrease. If we do a pasteurization step, we can see a million-fold or more decrease in the levels of pathogens within a few seconds.

So what we then have is trying to put together a food process model we call it or a process risk assessment in which we take the food from the raw materials and go through the various processing steps including pasteurization, but also storage, transportation, you know, all the way to the consumer and try to model the changes in bacterial numbers as they go up and down through this whole process.

And this then becomes a very major part just in terms of size and complexity of the microbial risk assessment. But we are I think very close now to being able to do a process risk assessment like this. And despite Keith's comment on our pathogen modeling program, I would encourage you all to take a look at it. I really do think it is quite user-friendly.

And, you know, I think we are there to where this type of risk assessment can be done. And we would really like to see a PC in a program like this on every food microbiologist's desk, particularly in industry, so that people in the food industry can look at their particular food processes and do this type of calculation.

And then that kind of becomes the underpinning for a HACCP Program. Now, I don't know if you people are familiar with food industry and the HACCP, Hazard Analysis Critical Control Point Program. But I see the current efforts in this area are basically sort of qualitative. That is, when they develop a HACCP, they look at each step separately. You look at the raw materials and you put into play certain standards and reactions to things out of specs. for that step. And then you look at the pasteurization step. And then you look at the storage step.

But now that we can do this process risk assessment and actually model the whole flow from raw materials through, you can put all of this together. And we can compare one step in the process versus another. And maybe two processes are slightly different. But we can then evaluate at the end and say are they equivalent.

One step might rely on -- or one process might rely on good quality raw ingredients where another process might have a pasteurization step. I think about, say, fresh orange juice right now. Some people do not want to pasteurize orange juice. Can we evaluate one process that uses it versus one process that does not And I think we are beginning to be able to do that.

This then leads us to a calculation of the number of pathogens that might be in the food at the time of consumption. So we have, say, 2,300 Listeria in a serving. So what Is this a hazard or is this not And this then leads us into the dose response section of the risk assessment. And I would say this is probably one of the weaker links at the moment. But, you know, we do have some idea, certainly compared to some of the chemical hazards like radon which they are trying to argue over what is a serious level.

We do know for E. coli 0157:H7 that from ten to 100 organisms is enough to be a serious threat to a child. We do know a little bit about food matrix and that that affects the effective dose. We know that there is a lot of variation between one strain of a pathogen versus another. And we also know something about the wide variation in human susceptibility to these different bacteria.

I would say these food-borne bacteria are generally opportunistic organisms. That is, they like to strike children, elderly, various immunocompromised people and pregnant women. So we are making progress in research in this area. And I think most of the gains will probably come from improved epidemiological investigations. You can't really run experiments particularly on the susceptible population that we are most interested in. But with careful investigations of outbreaks that do occur, we can get much better information.

An example of this is there was an outbreak of Listeria in Finland last spring that occurred in a hospital with severely immunocompromised patients. It was due to relatively low levels of Listeria in the butter. But we can analyze the butter. We know how many organisms are there. We know how many people consumed the butter. We know how many got sick. We can really begin to characterize the dose response for this one outbreak.

So we are at the point now we can calculate the amount of bacteria in the food. And then we can look at the dose response. And that then leads us to the question then of what kind of standards are we now going to set on this. How do we set the standards Who sets the standards And what sort of process do we have

And I would say for food microbiology right now, we really have not gotten to the point of really addressing these questions yet. You know, we have talked about food, I think both from the public and certainly the government side, as saying your food is safe. And, you know, as a risk assessor, that word, "safe", is really one I don't like. I mean, safety, as your previous speaker said, is an absence of risk, an absence of something.

And, you know, to say one food is safer than another, I mean, a food is safe or not safe. I mean, in a certain sense, the word, "safer", is not really a logical term. What we prefer in risk assessment is to talk about a hazard which could be Salmonella. That is something specific. And then we can talk about the risk of that hazard, so many Salmonella per gram or a certain probability of illness from consuming a certain number. So we can talk about the hazard and the risk.

But what is an acceptable or tolerable risk from the various food-borne pathogens is a question that we really have not begun to face yet. And I don't pretend to come up here and say I really know the answers to what that should be. I think there is a consensus we want to do better from where we currently are.

And perhaps just for discussion, I would throw out the figure that CDC statistics say that we get a food-borne illness about once every ten years. There is about 1,000 meals a year. That means your chances of getting illness from lunch today is about one in 10,000. Now, is that high or low This is a decision, something we have to think about.

Should the risk be the same for all of the various food-borne organisms Should Salmonella and E.coli be considered the same I would say probably not because some of the organisms have much more severe consequences than others. E. coli, for instance, causes hemolytic uremic syndrome. It can cause death and severe kidney failure in children where Salmonella for the most part just makes people sick for three days.

Should the risk be the same for all foods Again, I would say probably not. But nobody has really discussed this. Should we have the same risk for different populations, different sub-populations Should we have the same standards for children Should we have the same standards for nursing homes and other institutions

What sort of choice should we as consumers have

If I like my eggs sunny-side up, if I happen to like raw oysters, should I have the choice to consume those foods or not What is the acceptable level of risk Should it be based on current standard practice Is that a good place to start Perhaps it is.

But then again, we have found in recent years that many of our traditional foods are not quite as safe as we thought they were. We thought eggs were safe until a few years ago. We have seen problems with fermented meat products. We have seen problems now with fresh orange juice. Foods that we had considered safe, we are suddenly finding there are some problems.

Should we base our level of standard on what is considered the best feasible technology But then, of course, that brings the cost factor into what is feasible technology. For example, on eggs, there is about one egg in every 20,000 which is contaminated with Salmonella enteritidis.

There is a process that you can pasteurize in-shell, whole eggs with a hot water treatment. And that will inactivate any Salmonella and it costs about 24 cents a dozen. Should we mandate this for protection or not

I really don't know the answers to any of these. But I think we must begin to, you know, face these questions and begin to discuss them. And the answers to these are really a public and societal or political decision. This is not a scientific decision.

So, therefore, in conclusion, I would say what I am most certain of, that there is a lot of communicating that we have to do over the issues of food microbiology. Thank you.

(Applause.)

DR. STERNER: Any questions for Dr. Whiting Yes.

MS. : Dr. Whiting --

DR. STERNER: Could you go to the microphone

DR. WHITING: I can't hear you.

MS. : Okay.

DR. STERNER: We are fixing that.

MS. : In the risk assessments that you mentioned as having been done recently, the E. coli, the Vibrio, Listeria, the S. e. in eggs, did you take that to the human health impact like we did in the Campylobacter risk assessment In other words, did you use the FoodNet data from CDC to look at the ill humans and try to associate that with the dose that you calculated in the product

DR. WHITING: Yes. All four of those have tried to do that. You can find the Salmonella enteritidis on the internet if you go into the USDA FSIS, Food Safety and Inspection Service, and then Office of Public Health and Safety. And it is available there. And it has a series of modules. And one module is called the Public Health Module.

But the Listeria one, trying to determine what we know about the dose response is one of the major parts of that risk assessment. So, yes.

DR. STERNER: Well, we have talked about food and now we are up to water, or down to water depending on how you want to look at it. And I think that the room temperature is moderating a bit. And my glass is not ice yet, but there are times where it feels a bit like it.

Our next speaker, Dr. Steven Shaub, is a microbiologist. He received his bachelor's degree from Washington State University and his Ph.D. from West Texas, University of Texas at Austin. Excuse me, a Longhorn.

And from 1992 to the present, he has been with the United States Environmental Protection Agency's Office of Water in the Office of Science and Technology. He is a Senior Microbiologist there. And he heads up the pathogen risk assessment methodology development. He supports the drinking water and recreational water regulation development. Dr. Shaub.