U.S. flag An official website of the United States government
  1. Home
  2. Animal & Veterinary
  3. Safety & Health
  4. Antimicrobial Resistance
  5. The National Antimicrobial Resistance Monitoring System
  6. NARMS Background
  1. The National Antimicrobial Resistance Monitoring System

NARMS Background

by Dr. David G. White

DR. WHITE: Thanks, Linda. I just want to take a quick second to look at my slides, if I can.


DR. WHITE: Okay. Thank you very much. I think probably too though, Scott, in terms of your question, there is a notebook that is being created right now that will have a list of certain aspects we would like to address throughout the talk. Unfortunately, you don’t have it in front of you right now, but hopefully, most of these talks will kind of steer you in the direction of questions that will probably be forthcoming from all of you folks. And hopefully, we can answer it as best as we can.

DR. McEWEN: Is it an oral exchange of information? We won’t be preparing written answers to the questions?

DR. WHITE: I believe it is oral.

Thank you very much for coming. I appreciate it as well, and I would like to take about maybe 20, 25 minutes to go over the history of this program and try to go back in time to gleam the information from numerous documents about how this program came to be, where it has been going, and hopefully, where we are headed.


We keep seeing more and more articles on antimicrobial resistance. Of course, we only have to go back to 1944, when Penicillin was hailed as a miracle drug and a secret weapon of the allied effort.

Of course, my former mentor’s second edition of his Book, the Antibiotic Paradox. A good reason for those of you who are interested.


In terms of resistance, we all know, I think, that it is a complex phenomenon. The increasing incidence of resistance has serious implications for the future treatment and prevention of infectious diseases in both animals and humans.

Although much scientific information is available on this subject, many ecological aspects of the development and dissemination of antimicrobial resistance remain uncertain.


What is known, however, is that the emergence and dissemination of bacterial antimicrobial resistance is the result of numerous complex interactions among antimicrobials, microorganisms and the surrounding environments. A clearer understanding of these interactions is necessary if a science-based assessment of the relative risks concerning the use of antimicrobials in the animal husbandry environment is to be made.


In terms of our strategy for CVM, it is aimed at assessing relationships between antimicrobial use in agriculture and potential human health consequences. We have adopted a multi-pronged approach that includes a revised safety assessment process. For those of you not familiar with Guidance 152, it is a new process by which antimicrobial drugs are looked at.

We have increased our education outreach activities. In cooperation with AVMA we published a series of prudent use guidelines on antimicrobial use in different food production environments.

Participation in international activities, like Codex and OIE. Expanded research activities and enhanced surveillance activities.


And, of course, the one that is relevant to us today is NARMS, in terms of surveillance activities. This is what I tried to gleam from the history of NARMS. It actually goes back to 1994, 11 years ago, when FDA held a joint Veterinary Medicine and an Anti-Infective Drugs Advisory Committee to address the specific issue of approval of fluoroquinolones for use in poultry.

The committee was united in advising the agency that if the products were to be approved, several restrictions should be placed on the use of these drugs in order to attempt to minimize the public health risks related to the development of resistant bacteria in animals.

The most relevant recommendation to us today was the establishment of a nationally representative surveillance system to monitor resistance trends among both human and animal enteric bacteria. So, that was one of the recommendations from 11 years ago.


So, from that the National Antimicrobial Resistance Monitoring System, Enteric Bacteria or, as we affectionately call it, NARMS, became operational in 1996. Salmonella of animal and human origin was the initial organism accepted for surveillance, and NARMS combines the activities of FDA, CDC and USDA to create a nationwide monitoring system.

It was a tripartite design, development and implementation, and the three founding fathers, so to speak, were Dr. Linda Tollefson, from FDA, Dr. Fred Angulo, from CDC and Dr. Paula Fedorka-Cray, from USDA.

The sources and types of isolates have expanded over the time of the program. Campylobacter was added in 1998 and Enterococcus and E. coli in 2000, and we added the retail arm in 2002.

Each year samples are collected from numerous origins and tested to determine if there have been changes in the susceptibility/resistance of certain enteric bacteria to selected antimicrobial drugs, and the antimicrobial drugs tested annually are selected based on their importance in both human and veterinary medicine.


What NARMS allows the FDA to do is to monitor changes in susceptibility/resistance to drugs used in humans and food animals. Very simply, the human samples are collected from ill people. The animal samples are gathered from healthy farm animals, animal clinical specimens, carcasses of food animals at slaughter and processing plants and ground products at processing plants. The retail meat samples are collected from grocery stores and shops that sell meat and poultry to the public.


In terms of the goals of the NARMS program, it is to provide descriptive data and trends on antimicrobial susceptibility/resistance patterns in zoonotic foodborne pathogens and select commensal organisms. It started out initially with the pathogens, but we have since added Enterococcus and E. coli as commensal organisms.

Also, the goal is to respond to unusual or high levels of bacterial drug resistance in humans, animals and retail meats in order to contain or mitigate resistance dissemination. Also, to design follow-up epidemiology and research studies to better understand the emergence and transfer of antimicrobial resistance and assist the FDA in decision making on approving safe and effective drugs for humans and animals, as well as promote prudent and judicious use of antimicrobial drugs.


There are three testing sites. The human isolates are done at CDC, at the National Center for infectious Diseases, Foodborne and Diarrheal Diseases Laboratory in Atlanta, Georgia. The animal isolates are tested at USDA, ARS in Athens, Georgia in the Bacterial Epidemiology and Antimicrobial Resistance Research Unit. The retail meat isolates are tested at the Center for Veterinary Medicine Office of Research in Laurel, Maryland.


From 1996 to the present, NARMS has been funded through FDA via IAGs, Interagency Agreements with both USDA and CDC, and we also have interagency agreements with APHIS and FSIS, I believe, through ARS.


The animal, human, retail lab testing are comparable in terms of the methodologies used. We have had methods meetings in 2002 and 2003 to standardize the methods between the three arms. In the past we have had an annual NARMS meeting to present the data, but we did not have one this year due to budget limitations. We hope to have one next year. Hopefully, teamed up, possibly, with ICEID.

In terms of the susceptibility testing methods, they are all the same at all three arms. We use the Sensititre System from Trek Diagnostics. It is a broth microdilution panel for Salmonella, E. coli, Enterococcus, and hopefully, as of next week, we will be using the Campylobacter brother microdilution panel, which is going for approval this weekend at CLSI.

So, as of this year we will have all four bugs under surveillance being performed with broth microdilution. In the past Campylobacter has been done by E-test and agar dilution.

We use the same antimicrobial plate formats. We meet generally a couple of times a year to go over what antimicrobials are on the panels. We try to come to some consensus about what should be on there.

In terms of the manufactured lots, Trek manufactures their susceptibility plates in certain lots. We distributed them between the three sites so that each site has the same lots, and we try to use the same isolate handling procedures.

Quality assurance. We have both internal and external programs to monitor quality assurance of susceptibility testing. We follow CLSI/NCCLS standards where appropriate. We do have several antimicrobials which do not have CLSI quality control or interpretive criteria, and we do note that in our reports.

In 2003 we adopted the WHO External Quality Assurance System for susceptibility testing and for Salmonella serology testing I believe.


In terms of the bacteria tested, we do have some similarities among the sites, but each site does have its own unique isolates under surveillance. CDC has Non-Typhi Salmonella, E. coli 0157:H7, Enterococcus, Campylobacter, Salmonella Typhi, Shigella, Listeria and Vibrio.

USDA has Non-Typhi Salmonella, E. coli 0157:H7 when available, Enterococcus and Campylobacter, and the retail meats do Non-Typhi Salmonella, E. coli, Enterococcus and Campylobacter.


As I mentioned before with regards to the methods. The major focus of this program is susceptibility testing, and we have designed this program so that comparable testing methods are used between all three arms. We use the broth microdilution methods using the Sensititre Semi-Automated system, and this is a picture here.

We have all started using the new ARIS systems, which have been a little problematic at the beginning. We have started to use them more, and they have a new software that will allow us to, hopefully, dump the data right into an access program for analysis.

As I mentioned, the participating NARMS laboratories adhere to the appropriate published CLS/NCCLS performance standards. Both the veterinary and the human. The M100-S15 and the M31-A2 I believe. Or, the M31-S1, which is coming out.


So, in terms of the methods for Salmonella and E. coli, it is tested by broth microdilution, Sensititre. The plates have changed over time for the past several years, but there are some core antimicrobials on the plates that have not changed over the years of the surveillance program.

So, the core antimicrobials that have been tested over the length of the program include: amikacin, amoxicillin-clavulanic acid, ampicillin, ceftiofur, ceftriaxone, chloramphenicol, ciprofloxacin, gentamicin, kanamycin, naldixic acid, streptomycin, tetracycline and trimethoprim-sulfamethoxazole.

Now, we did have sulfamethoxazole from 1987 to 2003. We replaced it with sulfisoxazole in 2004, as the QC is only for sulfisoxazole with CLSI/NCCLS, not sulfamethoxazole.

Ticarcillin was on one of the first plates but is geared as an anti pseudomonal. Since we didn’t test Pseudomonas, we moved it off the plate. We had apramycin on there for several years until the manufacture was discontinued by Elanco. Apramycin is no longer being manufactured. So, we removed it from our plate.

Florfenicol, we put on there as a snapshot to see if there was any changes of susceptibility of florfenicol, and we felt that chloramphenicol was suitable as a sentinel for both drugs, because most of the isolates that are Florfenicol resistant -- actually, I think almost every isolate that is florfenicol resistant is chloramphenicol resistant as well.

Imipenim we put on there again as a snapshot, and I don’t think we saw any resistance at all. So, we took it off the plate.

Cephalothin we had on there for six years, but we didn’t feel it was giving us that much information. So we removed it as well. And we added cefoxitin in its place in 2000 to pick up primarily the blaCMY cephalosporinase that we are starting to see in our Salmonella and E. coli isolates.

So, over time the dilutions have improved to include full ranges in the past. Some of our ranges do not include the appropriate quality control or the break points. All of the dilutions on our plates today have both complete quality control ranges in there, as well as the interpretive criteria for SIR.

And we use the appropriate QC organisms for testing. In the case of our Salmonella and E. coli, we used E. coli, S. Aureus, E. Faecalis and Pseudomonas Aeruginosa, depending on the drug and the range.

Enterococcus. Again, it is much like Salmonella and E. coli. It is tested by broth microdilution. We do have several core antimicrobials that have remained the same over the plate, and these include bacitracin, chloramphenicol, ciprofloxacin, erythromycin, flavomycin, gentamicin, kanamycin, lincomycin, linezolid, nitrofurantoin, penicillin, salinomycin, streptomycin, Q/D, tetracycline, tylosin, vancomycin.

A lot of these drugs you may not be familiar with because they are considered animal production drugs. Salinomycin is a coccidiostat*, and we had that on the plate between 2000 and 2003. We removed it from the plate in 2004, and we added Daptomycin once it was approved by FDA. We felt that it was important to put it on our surveillance system.

Again, the dilutions have improved over time, much like the E. coli and Salmonella plates. We have full ranges now for all of these antimicrobials, including the appropriate QC ranges, as well as the interpretative criteria.


So, in terms of the ‘05 Salmonella/E. coli panel, this is probably our best panel we have put together, and like I said, we have worked hard to make sure that everything is quality controlled and we have the appropriate S/I/R in there. We have several beta lactams under surveillance: ampicillin, amoxicillin/clavulanic acid, cefoxitin, ceftiofur and ceftriaxone.

We have nalidixic acid and cipofloxacin. In terms of folic acid inhibitors we have sulfisoxazole and trimethoprim/sulfamethoxazole. In aminoglycosides; we have streptomycin, kanamycin, gentamicin and amikacin. There is no QC for streptomycin by CLSI/NCCLS standards. And then we have chloramphenicol and tetracycline on the plate. And the plate designation by Trek is CMV1AGNF.


This is all going to be in your packet when you get it. So, you don’t need to remember these numbers. This is just to give you the transparency of the ranges that we employ in the testing and the break points used. As I mentioned in streptomycin we don’t have QC. We don’t have interpretative criteria. We do use a break point of 64 µg/mL.


Enterococcus. Just to give you a breakdown, we do have several production drugs, as I mentioned, that are used in veterinary medicine: Bacitracin, flavomycin, lincomycin and tylosin. We do have three aminoglycosides, and these are to measure high level aminoglycoside resistance. These are all at MICs 256 or higher to strep, kan and gentamicin. We have vancomycin on there, as well as the two newer drugs that have been approved for gram positive infections in human medicine, daptomycin and linezolid. We also have penicillin, Q/D, ciprofloxacin, nitrofurantoin, erythromycin, chloramphenicol and tetracycline for the Enterococcus plate.


And again, this will be in your packet when you get it to give you an idea of the ranges used and the break points employed. Again, the red asterisks show that we do not have QC for those drugs, as they are production drugs. CLSI doesn’t do quality control for production drugs, but we hope to change that over time.


Campylobacter surveillance was added in 1998 at USDA and CDC, and from 1998 to 2004 the E-test was employed for antimicrobials, including azithromycin, clindamycin, erythromycin, nalidixic acid, ciprofloxacin, gentamicin, chloramphenicol and tetracycline.

When the NARMS retail arm began in 2002, it tested these isolates with E-test, as well as the agar dilution CLSI approved methodologies to ciprofloxacin, doxycycline, erythromycin, gentamicin and meropenem.


Different methods though between the arms and the lack of reproducibility for some drugs led to the development of a new CLSI/NCCLS approved broth microdilution method for Campy in 2004 using C. jejuni ATCC 33560 as the quality control.

The QC MIC limits of the AST of Campylobacter have been accepted by the CLSI in January of ‘05, and will be published in the upcoming CLSI documents. As I mentioned, the erythromycin was a little bit out of range and is being presented, I believe, this weekend at the VAST and AST meetings. This broth method will be much easier to use than both agar dilution and E-tests and is amenable to semi-automation and has QC ranges for 14 different antimicrobial agents for Campylobacter. So, we are very excited about this and really can’t thank enough the group at OR that was spearheaded by Dr. Pat McDermott, Dr. Bob Walker and Sonya Bodeis for helping to champion this system through. So, as soon as we are ready to go with Campylobacter this year; we will use the broth microdilution method. So, we are ready to jump in there and see what we get.


These are the antimicrobials that are going to be on the new panel, and we have several macrolides and lincosamides: azithormycin, erythromycin and clindamycin. We have nalidixic acid and ciprofloxacin. We added a new Ketolide, telithromycin, as well, and we have others on there: florfenicol, gentamicin and tetracycline. So, we will have nine antimicrobials on the Campy plate.


These are the ranges that are going to be employed, and in terms of the breakpoints, this is all kind of based on populations since there is no interpretative criteria for Campylobacter for CLSI or NCCLS. We have been using these types of breakpoints, and you will notice we don’t have breakpoints for telithromycin or florfenicol at this point.


Other methods used by NARMS. We do have our Salmonella and Campylobacter isolates characterized by Pulsed Field Gel Electrophoresis, and to give you some background, Dr. Zhao will talk about this later on. Genomic DNA is cut with an enzyme and resulting fragments of different sizes can indicate mutations or insertions in the genomes. Whereas PulseNet is an international program -- well, it is national and international now. It assists investigations and improves outbreak detection through rapid linking of cases by comparing bacterial DNA fingerprinting patterns. But a lot of our isolates are indeed now put into PulseNet.


Other capabilities, of course, were warranted. We all have molecular capabilities. All three arms have molecular capabilities to investigate emerging phenotypes and go after the appropriate genotypes. Here are just some examples of some PCRs that we have done looking at resistance genes for high level aminoglycoside resistance and then also we have several programs involved in looking at integrons among bacteria. So, all three arms are able to do this.


So again, just kind of summarizing everything up, this is the way NARMS is designed right now. There are three arms, the USDA, CDC and CVM sides.


We all have annual reports. They are all linked on the CVM home page, but USDA and CDC have their own home pages as well for NARMS, and you can access all three of them on our web link up there. These are just covers of the annual reports for the different sites.


So, in conclusion, we need to remember that NARMS is a public health surveillance system. It will enable informed decision making by FDA, initiated response activities if data warrants. We hope it promotes prudent and judicious use of antimicrobials and will help guide prescription practices. We hope it encourages standardization of laboratory techniques, identify areas for more detailed investigation, promote collaboration between the three different government agencies involved, as well as academia. And ultimately, prolong the efficacy and useful life of antimicrobials. With that, I thank you very much. Are we going to take questions now?

DR. YOUNGMAN: We can take questions.

DR. VOGEL: David, you mentioned that Campylobacter has gone to the broth microdilution. Two questions regarding that. Number one, how is the results of that comparable to the E-test? Is that going to affect your trend analysis or not?

And secondly, how does that affect the ability for your workload, your sample throughput? Will that increase your ability to do additional sampling for Campylobacter?

DR. WHITE: Thanks, Lyle. Those are good questions. I will answer the second one first. The broth dilution method is much easier than the E-test and agar dilution methods. So, if anything, it will shorten the time it takes to do susceptibility testing on those organisms.

Allowing us to do more is all dependent on the budget and everything else. You know, what we can collect for isolates. The numbers we get now through retail meats -- and I am just speaking for myself; I can’t talk for Tom or Paula. We get about now 600 to 800, and it takes a good six to eight months to do that by agar dilution.

This should probably drop us down to maybe two to three months once we are up and running. So, we are really excited about that.

The first question, in terms of the comparison to E-test, E-test is not a CLSI endorsed method. So, if anything, you need to take the E-test data and compare it to the broth, not the other way around.

We have done some studies. We have done both in the past comparing agar dilution and E-test at the same time, and that is probably because of our research laboratory and Dr. Walker’s expertise. So, we have that data to show comparability for some drugs, but that is why we went to broth, because the E-test was not giving reproducible results for several drugs, whereas the broth microdilution method is a reproducible standardized method.

Does that help? Okay. Dan?

DR. SAHM: Thanks for the overview, David. The question that I would like to ask has to do with the determination -- the two key resistance mechanisms to extended cephalosporins that are being seen in enterics are extended spectrum, beta lactams and AMP C mutations or AMP C acquisition.

The selection of cephalosporins you have on your gram negative panel wouldn’t be the most sensitive set of cephalosporins to help detect ESBLs emerging, and I was wondering how that data is looked at, given the cephalosporins that you use as a sentinel for potential ESBL emergence in your study organisms.

DR. WHITE: Thanks, Dan. That is a great question. Actually, we have -- that is a very good point. A lot of ours are just kind of screening. They may be adequate for 10 beta lactams and so forth, but the ESBLs -- we have actually created secondary plates. I didn’t have a chance to talk about it.

We actually have a secondary ESBL plate that has on there cefotaxime*, cefotaxime/clavulanic acid, ceftazidime, ceftazidime/clavulanic acid, cefepime* We have also added Cefquinome*, which is a fourth generation veterinary cephalosporin not approved in the United States. So, we further test anything that has a red flag for Cefoxitin. We kind of take off and put on this ESBL plate and we look for the definite, the three-fold reduction, and then further on we try to characterize from there.

We actually have also a supplementary fluoroquinolone panel as well. So, anything that is nalidixic acid resistant, we then put onto this plate that has on there marbofloxacin*, danofloxacin, enrofloxacin, levofloxacin, ciprofloxacin and so forth.

So, we try to -- you know, these are the second tier plates. So once we get a red flag on a certain phenotype we go to that. I apologize for not showing that. I will try to get that for everybody for tomorrow.

DR. SAHM: Just as a follow onto that, for instance, in the human isolates one of the more sensitive cephalosporins, and the flag for that is any MIC greater than or equal to one or two I believe. I can’t remember.

DR. WHITE: Right.

DR. SAHM: Because of a large number of TEMS that won’t show up with this as well. So, you might want to think about that flagging system.

DR. WHITE: Okay. Thank you. Good point.

DR. ALTEKRUSE: Is that being done in all three arms of NARSM?

DR. WHITE: Those plates are available to all three arms. I think right now CVM has probably done the most of that with that plate, but we had help with CDC. So, all three arms use the plates.

We don’t see much though for ESBLs. At least with the animal isolates and the retail isolates. I think we have seen one ESBL E. coli that fits the classic definition.

DR. KOTARSKI: I have two points to make. The first point is not necessarily a question, and maybe we will come back to this later. The objectives for the program that you have in your first slides don’t quite match up with what we received in our packages, and they are not necessarily the same as what is on the individual websites.

And so, as we go forward in the program, I would like to get some confirmation as to specifically what the objectives are. That might be helpful for the panel in addressing the questions that we have outlined for the afternoon. So, just a comment.

The second is a question. Quickly, what do you use to establish -- what criteria do you use to establish breakpoints for Campylobacter as you work through?

DR. WHITE: That is a great question. I will let Dr. Walker answer that question.

DR. WALKER: CLSI has a working group called, for lack of a better term, Orphan Organism, and they have been meeting -- we have been meeting for the last two years. And one of the things that we are doing in that group is to take infrequently isolated organisms and try to establish testing methods and QC and interpretative criteria with Campylobacter, because we have already gone through CLSI to establish the testing methods and the QC ranges.

What we are going to do, like Dave said, at this meeting come up this weekend and this next week, is establish interpretative criteria for Campylobacter for some drugs. So, CLSI is addressing this. They are doing it off of population distributions for the most part, and I would suspect that come Tuesday afternoon there should be some recommendations as to interpretative criteria for some drugs for Campylobacter.

DR. KOTARSKI: So then basically, it is a screening mechanism for increased MICs relative to the general population, what you would expect? It is not based on therapeutic failures or anything?

DR. WALKER: Yes. And for those who are familiar with CLSI, you know that to -- generally what they do to generate interpretive criteria is to look at population distributions, look at pharmacokinetics, pharmacodynamics and look at the results of clinical trials.

And then those interpretive criteria go into books that are referred to as standards books. For this particular bunch of organisms they are going to look at population distributions, PK/PD where they have it, but obviously the clinical trial is not going to be available.

So, this document will be called the M45, and it will be a guidance document. It will not be a standards document.

DR. WHITE: Great. Thank you very much.

DR. MILLER: One last question. You mentioned C. jejuni for the microdilution new techniques. Are you also moving in that direction for C. coli?

DR. WHITE: It is for both. All Campylobacter.

DR. ANGULO: Dave, just one clarification. CDC started Campylobacter in 1997, and thanks for highlighting the important innovation for having the broth microdilution for Campylobacter because that is really essential for us.

You will hear more, but we have changed the number of samples we are receiving. And that is because we are able to test more samples because we are more efficient at testing what manner. So, thanks for highlighting that.

DR. ALTEKRUSE: You mentioned Listeria being done I think at CDC. Is there any evidence from that that perhaps other arms should also be looking at Listeria?

DR. WHITE: No. I think every once in a while we take a snapshot of something and then we evaluate within our groups and we see. It is kind of like imipenem. We put imipenem on there for a year. We didn’t see any elevated MICs. I don’t think any of us did. And so we said, okay, let’s yank it off.

Remember, we have 96 wells. We are limited to 96 wells. Three of those are control wells. So, we are stuck with 93. So any time we tweak something, we have to take something off.

That is why we got rid of Cephalothin*, primarily because we didn’t think it was giving us information, and with those several wells that we had we were able to improve the QC ranges and cover all the ranges for all of the other drugs on the plate.

DR. ALTEKRUSE: Okay. Thanks.

DR. YOUNGMAN: Before I introduce the next speaker, I just wanted to point out that while our panelists have microphones at their desks, for the rest of you there is a microphone which is live all the time that is here in the center of the room. Please use it. That helps our transcriber in making sure we accurately reflect your questions and comments. Thank you.

If I could, I will introduce Dr. Robert Walker, who is going to be talking about the NARMS budget.

Back to Top