• 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

Animal Isolate Sampling

by Dr. Neena Anandaraman

DR. ANANDARAMAN: I am going to be talking about FSIS’ contributions to NARMS in just a little bit more detail than Paula went through. First, I will just give you a brief introduction to the agency and its lab and then I will talk about the Salmonella HACCP verification testing program. A
lot of you are probably already very well familiar with it, and then I am just going to briefly go over some other programs within FSIS that contribute isolates to NARMS.

(Slide)

As you all probably know, FSIS regulates nearly 6,000 meat, poultry and egg processing plants nationally. Samples from these establishments are sent for microbiological, chemical and pathological analyses. Of course, today we will be concentrating on the microbiological part.

We have got three labs. One is located in Athens, Georgia, and that is the Eastern Lab, and it is housed in the same building as Dr. Cray’s lab. Then we have the Midwestern Lab in St. Louis, Missouri, and finally, we have the Western Lab in Alameda, California.

(Slide)

In FSIS we have two categories of testing that we do. One is baseline testing, and this testing is actually weighted by production volume and it is statistically designed to estimate national prevalence levels. The other is our HACCP verification testing program, and this is a compliance testing program and it is done to verify that establishments are meeting regulatory performance standards.

(Slide)

Our HACCP verification testing program does provide the bulk of FSIS derived isolates to NARMS, and the same year that NARMS was established, in 1996, coincidentally FSIS also published its PR/HACCP Systems Final Rule. One of the provisions of that rule is that plants would meet prescribed performance standards for Salmonella and that FSIS would conduct testing to verify that these standards were being met.

The performance standards were based on prevalence estimates from baseline testing that was conducted in the early to mid 1990s, and if you are interested in more on the baseline studies, they can be found on our website.

(Slide)

So, prior to implementing our HACCP rule, which was published in 1996 and implemented in 1998, we conducted what is called pre-implementation testing. With pre-implementation testing we were trying to get an idea, both for FSIS and for the establishments, where they stood prior to implementation.

This testing was not statistically designed to estimate prevalence levels, and isolates from this program were sent to NARMS.

(Slide)

We have more information on the pre-implementation testing program in an article by Dr. Schlosser and also on our website on several documents listed.

(Slide)

So then, HACCP was implemented in stages, depending on plant size. We define large plants as having 500 or more employees. HACCP was implemented in 1998 in those. And then in 1999 in small plants, and small plants we define as having 10 to 499 employees.

For very small plants which have less than 10 employees or less than $2.5 million in sales we implemented in January of 2000. So, as of 2000 all sizes of establishments were included.

(Slide)

So, sampling for HACCP is done on an ongoing basis after the establishment’s implementation date on forward. Product class, whether carcasses or raw ground product, determines the number of samples that are collected per set and it also determines the performance standard.

Once an establishment is targeted for testing, the inspection personnel at that establishment collect a daily sample until the set is completed. Then the samples are shipped chilled the same day of collection.

(Slide)

Then there are two different categories of testing that we do. Routine testing, the ongoing testing, is what we call "A" sets, and for "A" sets establishments are scheduled approximately every six to 12 months and this is ongoing and routine. Sometimes you will see it referred to as random.

The "B", "C" and "D" sets are what we call targeted testing. These are conducted if an establishment fails a set. So, if an "A" set fails, a "B" set is done. If a "B" set fails, a "C" set is completed.

(Slide)

So, just for each of the product classes I am just going to go through how much comprise a sample set and some details of those. Fifty-one samples comprise a broiler carcass set. So, for the broilers we randomly sample after immersion chilling and we do a whole bird carcass rinse and then save 30 milliliters of the rinse fluid, which is then sent to the labs for analysis.

(Slide)

For raw ground products, for chicken, turkey and beef, we have a 53 sample set. We collect our 25 grams after final grinding and also send these on to the lab for analysis.

(Slide)

For the cattle and swine carcasses we have 82 samples for young cattle, 58 samples for mature cattle and 55 samples for swine. These are chilled carcasses which are randomly selected after 12 hours of cooling and then we swab three different sites on the carcasses. We then put the swabs in a sterile bag and send these on for further analyses.

(Slide)

This just summarizes the seven product classes that we test, the number of samples comprising a set and the maximum number of positives that are allowed. Again, this goes back to our baseline prevalence estimates.

So, for example, for young chickens they are allowed to have 12 positives out of a 51 set sample before the sample fails. If they get 13 positives, then a "B" set would be collected.

(Slide)

Our laboratory analysis can be found in our Microbiology Laboratory Guidebook that is on the web. Also, if you have more detailed questions on micro methods, Dr. Bonnie Rose is here. We confirm culturally all of our presumptive positives and then we send the isolates to APHIS’ National Veterinary Services Laboratory in Ames, Iowa. Once they have serotyped it, they send the serotype information back to the originating lab and then the isolate, along with the serotype information, is sent to Dr. Cray’s lab for antimicrobial susceptibility testing for NARMS.

(Slide)

So, this is just to give you an idea of the number of isolates that we have sent through the HACCP Compliance Testing Program over the years. Since 2000, when we have had all plants on board, we have collected approximately 40,000 to 50,000 samples a year. These are the numbers of Salmonella isolates that we then forwarded on to ARS.

(Slide)

Just some of the limitations to interpreting data derived from the program are that it is not statistically designed to estimate national prevalence levels, like our baseline studies are. We did have staggered implementation. All of the plants were not included until the year 2000.

And then, for the first several years of sampling we really heavily concentrated testing in the first half of the calendar year and we are not really taking into account any seasonal variations that may occur.

(Slide)

And then we have a couple of publications, and also, our yearly web report that talks in more detail about our testing program.

(Slide)

And just a little bit about some of the other programs that have contributed isolates to NARMS. For Salmonella we have had several egg baseline studies that we have been sending to Dr. Cray for analysis. Also, we have been sending isolates from our Ready-to-Eat Program.

(Slide)

For Campylobacter we have had a couple of programs in the past that have contributed isolates to Dr. Cray. From 1999 to October 2000 we had our young chicken microbiological baseline data collection program, and that baseline study is going to undergo review by our advisory committee, the National Advisory Committee on Microbiological Criteria of Foods.

And then since that study ended in October of 2000, in October of 2001, as Dr. Cray mentioned previously, we have been giving her spent rinse samples from our broiler rinses in the Eastern Lab, which is housed in the same building as her lab, so that she can isolate Campylobacter, generic E. coli and Enterococci.

(Slide)

Some of the descriptive information that we have provided since July 2002 with our isolates are product class information, the date of collection, region, the plant size, the set and serotype.

(Slide)

So then in summary, our verification testing program does provide the bulk of the isolates that we submit for NARMS, and though it is not a statistically designed program for looking at national prevalence levels, we feel like it provides very good information. It is quite robust.

And we do plan on forwarding any isolates from future baseline studies to NARMS for susceptibility testing. Any questions?

DR. KOTARSKI: You stated -- I would just like to compliment first on a nice presentation and quite a robust program. You said the program does not provide national prevalence data on Salmonella in meats, the different meats? It does not provide that information?

DR. ANANDARAMAN: Well, I’m saying it is not designed specifically to provide that information. It is not designed to estimate national prevalence. It is for regulatory compliance. So we are doing the program to track plant performance, rather than trying to get a national prevalence estimate.

By extension, it may be providing us with that type of information, but we just want to make sure everybody understands the limitations; that it is not specifically designed for that purpose.

DR. KOTARSKI: For clarification then, by extension it doesn’t provide us national prevalence of resistance in Salmonella, if that is the source of the samples?

DR. ANANDARAMAN: I can say that it is not designed to do that. Whether it is doing that? It may be, but I can’t say for sure.

DR. ALTEKRUSE: It is not an unbiased sample. That is the issue. On a periodic basis establishments are tested. Sometimes those tests take longer to complete because of a whole range of issues that are involved. So, you couldn’t take this percentage and say that is the percentage of resistance to this strain in these Salmonella or even that these are the percentage necessarily by serotype 100 percent. It is not an unbiased sample. It is a regulatory driven program.

One aspect of that is that we have already indicated that we are going to, in likelihood, go towards a more risk-based approach to Salmonella. So we know already that some plants have high rates of positivity and some have very low rates of positivity. And so, based on those performance standards that you saw, we are going to increase the frequency of sampling in plants that have high prevalence and decrease it in plants that have low prevalence.

So, in fact, it is going to become even more biased in its approach in the future. But one of the things that we want to try to do is to establish a parallel of sort of ongoing surveillance activity. But then, as everyone has pointed out today, money is the issue. So the design of that parallel program to obtain national representative data is going to be driven by how much funding we have.

DR. WALKER: I have two questions. Number one, on these broiler rinses that go to the Eastern Lab, are those only from the eastern United States? Do the Midwestern Labs receive isolates from -- or samples from the Midwestern and the Western states from the western states?

DR. ANANDARAMAN: Well, what happens is when a set is scheduled, it is randomly assigned to a lab. So there is a variety of states going to the Eastern Lab. It is not only eastern.

DR. WALKER: So, the eastern lab could end up with samples from Oregon or Washington?

DR. ANANDARAMAN: Sure. Absolutely.

DR. WALKER: And when you talk about shipping samples to the lab, what is the conditions and the transport time?

DR. ANANDARAMAN: Are you talking about the isolates themselves or --

DR. WALKER: The samples.

DR. ANANDARAMAN: The rinse samples?

DR. WALKER: Yes. Or the ground beef or --

DR. ANANDARAMAN: Are you talking about the spent rinse samples for Campylobacter, E. coli? Or for Salmonella isolates?

DR. WALKER: Well, for the boiler carcasses you have the rinse samples and you say shipped to the FSIS lab or the ground beef you ship to an FSIS lab.

DR. ANANDARAMAN: Okay. For the rinses that we are supplying after the Campylobacter baseline testing was completed, those aren’t getting shipped because they are only getting sent from the Eastern Lab. So the Eastern Lab -- they are walked down to Dr. Cray’s lab.

DR. WALKER: From the point of collection.

DR. ANANDARAMAN: Right. What happens with the --

DR. WALKER: Like the Oregon sample. If it were going to the Eastern Lab, how long would it take it to get there?

DR. ANANDARAMAN: Well, it is shipped overnight. It is Fed Ex’ed overnight.

DR. FEDORKA-CRAY: They are shipped overnight and processed the next morning, and as soon as they are done being processed, we get it. And they are kept refrigerated the entire time that they are --

DR. ALTEKRUSE: They are rejected -- Bonnie would know this. But they are rejected if they are above a certain temperature.

DR. WALKER: But have you ever taken like Campylobacter and ran it through it that scenario to see how well it survives?

DR. ANANDARAMAN: We are only doing that with Salmonella isolates. With the Campylobacter we only get those out of the rinse samples that are in the Eastern Lab. In the same building. We don’t ship that.

DR. WALKER: But those rinsates could come from Oregon to the Eastern Lab?

DR. ANANDARAMAN: No. No.

DR. WALKER: Just the chicken, non the rinsate.

DR. ANANDARAMAN: Oh, you’re talking about the rinsates for FSIS.

DR. WALKER: Right.

DR. ANANDARAMAN: What happens is the rinse -- once the rinse sample is collected it is refrigerated and it is supposed to be sent --

DR. WALKER: In Oregon?

DR. ANANDARAMAN: In Oregon. And it is supposed to be sent by noon that day Fed Ex’ed to our lab. So, it is Fed Ex’ed overnight to our lab so it doesn’t get there on a weekend.

DR. WALKER: So my question is have you ever done any studies to see how well Campylobacter survives that type of --

DR. ANANDARAMAN: Have we, Bonnie, that you know of? I’m not sure.

DR. ROSE: They are shipped -- all the samples, the rinses, the ground product and the sponge samples from the cattle and swine carcasses are shipped in insulated shippers with frozen gel packs. They are delivered overnight by Fed Ex to the FSIS labs.

DR. WALKER: Right. I appreciate that, and E. coli and Salmonella is probably not a problem. Or Enterococcus. They will survive, you know, standing on their heads. But Campylobacter is more fastidious and doesn’t survive as well.

DR. ALTEKRUSE: Well, one thing we do know -- we have done work occasionally with isolates or with rinses that were above the temperature that they were supposed to be brought in at, and Campylobacter doesn’t grow well in those.

We are able to isolate Campylobacter and enumerate it from rinses that are shipped overnight and under eight degrees ‘c’. Otherwise, they are rejected. What that means is that the -- an inspector in the plant is given another sample request form, and that one doesn’t count against the performance set.

DR. ANANDARAMAN: As far as specific studies done on Campylobacter survivability, we would have to check with the labs. I am not sure if we have or not.

DR. FEDORKA-CRAY: bob, we know that when we initially started to get the samples from FSIS we had a low, what we considered a low, recovery rate of Campylobacter compared to what we would have expected. We were getting about 11 percent.

What we ended up doing was we took the 10 mils of rinsate and we concentrated it by sonerfigation* step. Then we took the pellet and enriched that, and we brought our isolation up to 20, 30 percent. That is the current isolation rate that they are getting out of plants now. So we know that we are comparable to what they are seeing and it depends upon which plant you are getting these from.

Are these the ideal samples for Campylobacter? Probably not. Are they only thing that we have that would even give us a chance of getting Campylobacter? Right now? Yes.

DR. ALTEKRUSE: One additional item, if I could, and that is that keep in mind these are from birds that have just come from an online reprocessing system where they are going through some pretty harsh chemicals. So, in effect, I believe that they are not ideal samples. Under the best of circumstances, that you are dealing with injured cells as the nature of the organism. And perhaps by the time you get to retail it has some opportunity of recuperation. That is a big question right now; is what to do about damaged cells on poultry rinse carcasses.

DR. FEDORKA-CRAY: Right. And we are doing several studies with that looking at taking, in particular, those rinsates and what happens when you put them under different gas conditions or different growth conditions to see. And we have also taken a number of these rinsates and concentrated and just done PCR on them to see if we can get any Campylobacter DNA out of them at all, and all of these data are being put together and we hope to be able to have some story that we can tell from this.

The fact that we are getting any Campylobacter at all I think is, like Sean said, somewhat fairly remarkable, considering the condition that they are taken off right out of the chill tank.

DR. VOGEL: I guess to me that raises the question that if these poultry carcasses have been exposed to TSPs and other type of treatments, antimicrobial treatments, what we are getting may be preselected for resistance, --

DR. ANANDARAMAN: That is a good point.

DR. VOGEL: -- because there can be cross resistant between these antimicrobials that are used in plants as sanitizing agents and the antimicrobial drugs that we are using. So we may be biasing it towards a resistant Campylobacter.

DR. ANANDARAMAN: Unless we had some more information that we could look at those sorts of things with the isolates we wouldn’t be able to tell.

DR. ANGULO: Is every HACCP isolate sent from NVSL? Do the FSIS labs send all HACCP isolates to be serotyped to NVSL?

DR. ANANDARAMAN: Yes, they do.

DR. ANGULO: And does NVSL send every isolate to ARS for NARMS?

DR. ANANDARAMAN: Right. NVSL actually doesn’t send the isolate. They just send the serotype information back to the FSIS lab and then FSIS sends an isolate with the serotype information to ARS.

DR. ANGULO: What proportion of the HACCP isolates are "A" set isolates?

DR. ANANDARAMAN: About 90 percent.

DR. ANGULO: Have you looked to see if the other than "A" set isolates have a remarkably different resistance pattern than the "A" set isolates?

DR. ANANDARAMAN: Not yet. No.

DR. ANGULO: Because one of the easy reductions in sampling could be that just only -- in terms of NARMS. Because of this bias. It is the same plant being retested in the B and C and D sets. You can consider just submitting only the "A" set isolates into NARMS. That would reduce it by 10 percent, about 320 isolates per year, if there is an inherent bias of that.

DR. ANANDARAMAN: That is a possibility.

DR. ANGULO: Another possibility is to not sample -- not send every isolate. I mean, we do one in 20 Salmonella isolates from ill people. Has anybody done a power calculation to see what extra information is created; how much precision is created by testing 3,200 isolates this calendar year versus 1,400 or versus 1,200 isolates?

DR. ANANDARAMAN: No. I don’t think we have done that calculation.

DR. ANANDARAMAN: No.

DR. ANGULO: And the last extension of that question is how much cost savings -- how much cost saving would there be by reducing the sampling? If the shipping of every single isolate is $10.00 or something and every plate is $10.00, there could be important cost savings from reducing the number of samples tested.

DR. ANANDARAMAN: Those are costs that FSIS absorbs though. It is not coming out of the NARMS budget.

DR. ANGULO: But the plates. With the plate, NARMS purchasing the plate, and the technician time. I just question how much value is there testing. This year was a low year. That is because the positivity rates in everything except broilers is declining. I guess it was only 3,500 this year?

DR. FEDORKA-CRAY: It was only 2,500 this year, and I think part of it -- the other thing that you have to look at is this is split between -- I mean, there is chicken and then there is turkey and there is beef and there is pork in there. So it is not just 2,500 chicken isolates.

And then the other thing is that with what we are seeing now with our PFGE is a lot of diversity amongst our isolates, and that would be lost if we were cutting down. So, how would we know exactly which one to cut down? We have to run the PFGE, I would think, for at least two years to be able to take a look at that to make sure that we weren’t losing something there.

DR. ANANDARAMAN: I just wanted to add one more thing though on these numbers of isolates being different. Sometimes what happens too is that we may start a few sets at the end of one year and then a bunch the end of the next year, so it may appear that there are higher numbers of samples in the year that most of the set at the end of the year was collected.

DR. ANGULO: But this is a great discussion. It seems to me that the isolates from beef are so valuable because you get so few of them. Maybe the selection should be every one of the cattle isolates from HACCP.

But since you get so many isolates from broilers, I don’t know that -- I wonder if we -- maybe you could have a selection -- it seems there could be a relatively easily imposed sampling scheme at the FSIS labs. Tell them to send every other chicken isolate instead of every chicken isolate.

DR. ALTEKRUSE: If I could comment on that?

DR. ANANDARAMAN: Sure. Go ahead.

DR. ALTEKRUSE: Actually, we are very committed to serotyping all of our isolates. You don’t know what you have until you have serotyped it, and we want to provide that information back to the establishments.

DR. ANGULO: That makes sense.

DR. ALTEKRUSE: So, if you only take one tenth of isolates and give them information on it, you are giving them much diminished information.

DR. ANGULO: But this is a great discussion. It seems to me that the isolates from beef are so valuable because you get so few of them. Maybe the selection should be every one of the cattle isolates from HACCP.

But since you get so many isolates from broilers, I don’t know that -- I wonder if we -- maybe you could have a selection -- it seems there could be a relatively easily imposed sampling scheme at the FSIS labs. Tell them to send every other chicken isolate instead of every chicken isolate.

DR. ALTEKRUSE: If I could comment on that?

DR. ANANDARAMAN: Sure. Go ahead.

DR. ALTEKRUSE: Actually, we are very committed to serotyping all of our isolates. You don’t know what you have until you have serotyped it, and we want to provide that information back to the establishments.

DR. ANGULO: That makes sense.

DR. ALTEKRUSE: So, if you only take one tenth of isolates and give them information on it, you are giving them much diminished information.

DR. ANGULO: But I wasn’t talking about serotyping. I agree completely. Serotype all of the isolates. I was talking about the -- the burden is on the resistance testing and the cost inherent to that.

DR. ALTEKRUSE: Right. Well, there is no cost of sending them because the two laboratories are housed right there together.

DR. ANGULO: No. There is three regional FSIS sites.

DR. ALTEKRUSE: Okay. Fair enough.

DR. ANGULO: I am just raising these points about economies of scale.

DR. ALTEKRUSE: And let me follow up on it, because I am interested in Paula’s thought on this, because we know that a lot of our isolates are not the same isolates that are found to be causing human infections. And so, the question is do they -- should they have the same priority for NARMS that other isolates that are from the top 20 serotypes that cause 65 percent of the human illnesses we see?

For example, we keep on talking about Kentucky and Derby as examples of that.

DR. FEDORKA-CRAY: I think maybe we can crunch some numbers. But if you look at the -- one of the ways I look at the bang for the buck, if you will, is you look at the number of isolates that go through the animal arm of NARMS per year and you look at the Salmonella isolates, the Campy, the Enterococci and E. coli, look at the $1.5 million IAG, you take out $150,000 and that is $1.35 million that actually goes to the laboratory.

You look at about 5,000 to 6,000 Salmonella isolates, 1,000 Campylobacter, 2,000 E. coli and 1,500 to 2,000 Enterococci and we are up to 10,000 isolates. So you take those 10,000 isolates, you look at the amount of money that you are getting from it and you look at all of the other programs and they have all of the other associated costs with sending money to labs and how many isolates that you do at the end of the year, and I think that right now the savings is not going to outweigh the information that we would generate for the program.

We have 2,500 isolates that we are getting in for the slaughter samples. The other thing is that we would have to look and make sure that they are regionally representative, because in addition to just taking every 10th, you would have to make sure that you were getting an even distribution amongst regions.

I think that we talked biasing, and if you are going to have higher plants that are -- if you are going to have some plants that are going to have repeated problems and you only take an "x" number of isolates, then you are going to have the probability that you might exclude even more of the other plants that you could be testing.

I am not -- just in looking at it -- not just because I’m part of the animal arm of NARMS. But just in looking at it I don’t think that reducing the number of slaughter isolates that we do now, until we get some of these other programs up where we have the robustness going from farm to plant, is going to save us any information.

DR. ANGULO: It will save us money. That is the point. You need 10 percent less money. How are you going to cut it? You have got to cut sw. That is my point. Where would you cut? It seems to me some sampling scheme could be imposed. I am just trying to find savings somewhere.

DR. FEDORKA-CRAY: And I agree that I think we have to find savings, but I think that this would include us sitting down and looking at some of the raw numbers and asking exactly where some of that money and information is going to. We have all had a reduction over the years in the amount of monies that we have had to expend. So, it is not something that is being unequally borne by any one arm right now.

DR. KOTARSKI: As a follow up to Dr. Angulo’s and Dr. Cray’s discussion, I come to a question I asked earlier. What is the objective for the slaughter isolation monitoring component of the NARMS?

Regardless if you want to cut 10 percent arbitrarily or if you say I am going to cut the repeated trustings, I think it will help you if you go to your objective. Do you want to reflect what is actually seen at the slaughter house? Do you want to reflect your prevalence in the slaughter house? Do you want to reflect prevalence on a national basis?

What is it exactly that you are looking for and that will help guide where your cuts will come, if they need to come.

DR. ANGULO: I agree completely. The problem is though -- and I think it would be great to be explicit about that. The problem is though you have to take advantage of the isolates that are available. In this instance FSIS isolates are free. There is a system to sample them already.

So, while I agree completely with your point, we would love to have a random representative sample nationwide, but we can’t create a whole new sampling scheme. We have to use available isolates.

DR. KOTARSKI: And in no way am I trying to suggest that we recreate a sampling scheme. But since all of the samples right now, if I understand correctly, are going to Georgia, there is a subset that are actually on a randomized basis.

If you just looked at those sample sets "A", do those reflect any information about prevalence at the slaughter plant? What does that sample set represent?

DR. ANANDARAMAN: They are less bias definitely.

DR. ALTEKRUSE: But they are going to become with risk based verification, and that is the reason why there is an interest in FSIS in adding another layer, which would be a baseline, an ongoing baseline to try to get at that information, which would be nationally representative.

DR. FEDORKA-CRAY: And we are going to be involved in this upcoming baseline, as in subsequent baselines, and actually getting all of those isolates and testing them. It would be along the same lines of the NAHMS and CAHFSE program where you have nationally representative studies and sets of isolates coming in so that we can move away from bias and move to report a more national representation.

DR. ANANDARAMAN: But I agree with Paula, in that until we get those programs going, this is all we really have to look at over time. And if we start changing now, it may not be as comparable as to what we have seen in the past.

DR. YOUNGMAN: Thank you very much. We are going to be talking more about sampling after the break. What I would like to propose now is that we take a break now and continue talks after the break.

I would like us to come back at 3:30 if we could. We are also proposing to move the last talk for today to tomorrow morning. Dr. Zhao has agreed to that in the interest of time.

So, we will take a break now until 3:30, and then we will have two talks after the break. Thank you.

(Whereupon, at 3:16 p.m., a recess was taken.)

DR. YOUNGMAN: If we can start again, we are trying very hard to keep things on time. We have a few changes to the agenda as written. One is that the last talk that was planned for today we are now proposing to do tomorrow morning. Dr. Zhao will be talking about the molecular characterization tomorrow morning. That means that after this break we will only have two talks today, and hopefully, finish around 4:30, which is when we had originally planned to finish today.

What I would like to do before Terry Proescholdt comes up to talk about the retail meat sampling is to just remind the panelists in particular of the six questions that we had posed to you before this meeting. The latest version of those questions has a heading that says discussion points, and these are just a brief rundown of what those six questions are.

So if, as people are talking and as you think about things tonight and tomorrow morning, if you could be thinking about those six questions and what your individual responses would be, that would be really helpful.

Also, at the suggestion of one of the panelists, Scott McEwen, I thought what might be helpful to each of you is that tomorrow to make a bit of time at the end of the day for each of you individually to give us some of your responses to the six questions and other comments that you had for us. So, we will give you time tomorrow at the end of the day to each individually give your responses.

But just briefly, the six questions that we hoped we could get your individual responses to are: Is the current sampling process by FSIS adequate and effective for Salmonella surveillance? We have already heard some comments about that.

Are these samples adequate and effective for Campylobacter? Campy dies more quickly. So, how do we address that?

What other pathogens should or should not be isolated and susceptibility tested from the poultry rinsates? What makes sense? Again, these are brief versions of those questions that are on your pages, those six questions.

If additional sampling is suggested, please consider what we should stop doing in NARMS since no additional funding is available. It really comes down to some hard choices. If we are going to add new things, if we are going to do a better job of a certain type of testing, what do we stop doing?

Are the current sampling strategies for the retail meat arm of NARMS adequate as currently conducted? You will be hearing more about that in a minute.

Is reporting sufficient to be able to use data from the three arms effectively for public health surveillance? We have had a lot of discussions about trying to make the databases more uniform so you can more easily use data from humans, from animals, from retail meat.

How would change the reports to accomplish this so that the three arms can be used more effectively?

What should be retained and/or omitted in future reports? There is a lot in all of the reports. What are the pieces that work? What are the pieces that don’t work? How can we do a better job?

What are the top three elements of a well coordinated collaboration for the molecular characterization of isolates from all three arms? You are going to be hearing more about that tomorrow. You heard some about that today.

Molecular characterization from all three arms so we can demonstrate or refute a continuum from animal, food and human origins for specific pathogens and/or resistant phenotypes.

The fifth question: How should NARMS be involved in international efforts? We are going to hear about that tomorrow.

How do you suggest that we enhance and sustain funding for NARMS? We know that our appropriated dollars are very likely going to be going down. What can we do about that? Are there other ways that we can raise funding or other programs that we can combine somehow?

So, those are the six questions that we would really like to hear your responses.

Now, if I can turn this over to Dr. Terry Proescholdt, he is going to be talking about the retail meat sampling scheme and particularly focusing on the Iowa Pilot Study.