• 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

NARMS 2004 Executive Report - Methods

A. Sampling Methodology

Sample collection is an integral part of many public health surveillance systems, including NARMS. Sampling strategies necessarily differ among the three components of NARMS and are described below.

1. Human Component

Sampling for the human pathogens depends on public health laboratory-based surveillance and is driven by the occurrence of laboratory-confirmed cases. NARMS testing of non-Typhi Salmonella began in 1996 with isolates from 14 states. Subsequently, additional states joined the program. Since 2003, Salmonella isolates have been collected from clinical laboratories by state and local health departments in all 50 states and sent to CDC for susceptibility testing. From 1996 through 2002, participating sites submitted every tenth non-Typhi Salmonella they received to CDC for antimicrobial susceptibility testing. In 2003 and 2004, participating sites submitted every 20th isolate. Salmonella serotyping was performed by the participating state and local public health laboratories prior to shipping.

Surveillance for Campylobacter began in 1997 with five FoodNet sites submitting one isolate each week. This was expanded through the years, and in 2004, Campylobacter isolates were submitted from 10 FoodNet sites (California, Colorodo, Connecticut, Georgia, Maryland, Minnesota, New Mexico, New York, Oregon, and Tennessee); one isolate per week was tested from each site.1

2. Retail Meat Component

In 2002, retail meat sampling began in January with FoodNet laboratories in Connecticut, Georgia, Maryland, Minnesota, and Tennessee; Oregon joined in September. FoodNet laboratories in California and New York joined in 2003, while FoodNet laboratories in Colorado and New Mexico joined in 2004.

Each FoodNet site attempted to purchase a total of 40 meat samples per month, including 10 samples each of chicken breasts, ground turkey, ground beef, and pork chops. An attempt was made by each site to sample as many different stores as possible each month and to purchase as many different brands of fresh (not frozen) meat and poultry as possible. All sites cultured the retail meats for Salmonella and Campylobacter. Four sites (Georgia, Maryland, Oregon, and Tennessee) also cultured the meats for E. coli and Enterococcus. Once identified, bacterial isolates were sent to CVM’s Office of Research for further characterization, including serotype (Salmonella) or species (Campylobacter and Enterococcus) confirmation and antimicrobial susceptibility testing.

3. Animal Component

The animal component of NARMS was launched in 1997 and initially included monitoring of antimicrobial susceptibility among Salmonella. Data for Salmonella isolates from food animals at slaughter are included in this report. The isolates were recovered from carcass rinsates (chicken), carcass swabs (turkey, cattle, and swine), and ground products (chicken, turkey, and beef) collected at federally inspected slaughter and processing plants throughout the U.S. as part of the FSIS PR/HACCP verification testing program and pre-implementation testing. Serotyping was performed by the National Veterinary Services Laboratories (NVSL).

USDA began testing Campylobacter isolates in 1998. From 1998 to 2000, Campylobacter isolates from chickens were obtained from a variety of FSIS programs for inclusion in NARMS. In 1998, Campylobacter isolates were submitted only from the Eastern FSIS laboratory, whereas in 1999 and 2000, isolates were obtained from all three FSIS laboratories (Eastern, Midwestern, and Western laboratories). FSIS cultured samples for Campylobacter using the most probable number method described in the FSIS Microbiology Laboratory Guidebook. 2 Nalidixic acid susceptibility and cephalothin resistance were initially used as identification criteria for Campylobacter jejuni/coli. This likely resulted in an underreporting of quinolone resistant Campylobacter until 2001, when use of this method was discontinued. From January through June 2001, various isolation methods were compared and a new ARS method was adopted in July of 2001. Since that time, Campylobacter reported in the NARMS animal component have been isolated by ARS from spent chicken carcass rinsates submitted by the Eastern FSIS laboratory. The rinsates were collected as part of the Salmonella PR/HACCP verification testing program. In addition to testing the Campylobacter isolates for antimicrobial susceptibility, the ARS laboratory speciated the isolates. This Executive Report contains data on Campylobacter recovered from chicken carcass rinsates for the period July 2001 through December 2004, when the new ARS isolation method was used.

USDA began testing E. coli isolates for antimicrobial susceptibility in 2000. ARS isolated the E. coli from spent chicken carcass rinsates submitted by the Eastern FSIS laboratory. The rinsates were collected as part of the Salmonella PR/HACCP verification testing program.

B. Antimicrobial Susceptibility Testing Methods

The dilution schemes and antimicrobial content of the susceptibility testing panels have undergone several design iterations as the NARMS program has matured. This has resulted in testing arrays that now meet international standards for quality control. The content of the panels was amended, as appropriate, to accommodate new antimicrobial agents entering the market, to omit those no longer available or used, or to adjust dilution ranges. The susceptibility testing panel formats undergo annual review for possible improvements. Customized testing panels have also been designed, and are available for use in phenotypic assessment of extended spectrum beta-lactam and fluoroquinolone resistance. Antimicrobial minimum inhibitory concentrations (MICs) for Salmonella and E. coli were determined according to manufacturer instructions using the Sensititre® semi-automated antimicrobial susceptibility system (Trek Diagnostic Systems, Westlake, Ohio). In 2004, Salmonella and E. coli isolates were tested using a custom plate developed for Gram negative bacteria (catalog # CMV1AGNF). Clinical and Laboratory Standards Institute (CLSI) 3 recommended quality control organisms were used each time that antimicrobial susceptibility testing was performed. The quality control organisms included Escherichia coli ATCC 25922, Enterococcus faecalis ATCC 29212, Staphylococcus aureus ATCC 29213, and Pseudomonas aeruginosa ATCC 27853. 4,5 Two content changes were made in the Salmonella/E. coli plate format in 2004: cephalothin was omitted from the testing panel and sulfamethoxazole was replaced with sulfisoxazole.

Methods used to determine antimicrobial MICs for Campylobacter have changed over time. Through 2004, the human and animal components of NARMS used Etest® (AB Biodisk, Solna, Sweden). The antimicrobial agents tested using Etest® included: azithromycin, chloramphenicol, ciprofloxacin, clindamycin, erythromycin, gentamicin, nalidixic acid, and tetracycline. Based on Etest® manufacturer recommendations, MIC results that fell between the two-fold dilutions described in CLSI documents were rounded up to next two-fold dilution before interpretation. 6 The retail component used the CLSI-approved agar dilution method in 2002 and 2003.  The antimicrobial agents tested using agar dilution included: ciprofloxacin, doxycycline, erythromycin, and gentamicin.  Recognizing the need for a standardized semi-automated method, CVM developed a broth microdilution method. The retail component began using this method in 2004 and the human and food animal components adopted the method in 2005. This standardized method was approved and published by CLSI in 2006. 7 Testing was done using the Sensititre® semiautomated antimicrobial susceptibility system (Trek Diagnostic Systems, Westlake, Ohio) and a custom plate developed for Campylobacter (catalog # CAMPY). The antimicrobial agents included in the broth microdilution testing were: azithromycin, ciprofloxacin, clindamycin, erythromycin, florfenicol, gentamicin, nalidixic acid, telithromycin, and tetracycline. Campylobacter jejuni ATCC 33560 was used as the quality control organism.

C. Breakpoints

The breakpoints used in this report are shown in Tables 1 and 2. CLSI-approved breakpoints were used when available. For Salmonella and E. coli, CLSI breakpoints were available for all antimicrobials tested except streptomycin. 8 For Campylobacter, CLSI breakpoints were available only for ciprofloxacin, doxycycline, erythromycin, and tetracycline. 7 NARMS breakpoints were used when CLSI breakpoints were not available. NARMS breakpoints were established based on the MIC distributions of NARMS isolates. Since the NARMS 2003 Executive Report was published in 2006, there were changes in the breakpoints for Campylobacter for the following antimicrobial agents: azithromycin, clindamycin, doxycycline, erythromycin, gentamicin, and nalidixic acid. Resistance data for all years were recalculated using the new breakpoints.

Table 1. Breakpoints Used for Susceptibility Testing of Salmonella and E. coli 9

Antimicrobial ClassAntimicrobial AgentMIC Breakpoints (µg/ml)
AminoglycosidesAmikacin≤ 1632≥ 64
Gentamicin≤ 48≥ 16
Kanamycin≤ 1632≥ 64
Streptomycin≤ 32N/A≥ 64
AminopenicillinsAmpicillin≤ 816≥ 32
Inhibitor Combinations
Amoxicillin–Clavulanic Acid≤ 8 / 416 / 8≥ 32 / 16
CephalosporinsCeftiofur≤ 24≥ 8
Ceftriaxone≤ 816 - 32≥ 64
Cephalothin 10≤ 816≥ 32
CephamycinsCefoxitin≤ 816≥ 32
Folate Pathway InhibitorsSulfamethoxazole/Sulfisoxazole 11≤ 256N/A≥ 512
Trimethoprim–Sulfamethoxazole≤ 2 / 38N/A≥ 4 / 76
PhenicolsChloramphenicol≤ 816≥ 32
QuinolonesCiprofloxacin≤ 12≥ 4
Nalidixic acid≤ 16N/A≥ 32
TetracyclinesTetracycline≤ 48≥ 16

Table 2. Breakpoints Used for Susceptibility Testing of Campylobacter 12

Antimicrobial ClassAntimicrobial AgentMIC Breakpoints (µg/ml)
AminoglycosidesGentamicin≤ 24≥ 8
KetolidesTelithromycin≤ 48≥ 16
LincosamidesClindamycin≤ 24≥8
MacrolidesAzithromycin≤ 24≥ 8
Erythromycin≤ 816≥ 32
PhenicolsChloramphenicol≤ 816≥ 32
Florfenicol 13≤ 4N/AN/A
QuinolonesCiprofloxacin≤ 12≥ 4
Nalidixic acid≤ 1632≥ 64
TetracyclinesDoxycycline≤ 24≥ 8
Tetracycline≤ 48≥ 16

D. Reporting Methods

The remaining three sections of this report contain NARMS surveillance data for Salmonella, Campylobacter, and E. coli. Section III contains data for Salmonella isolates recovered from food animals at slaughter, retail meats, and humans. Antimicrobial susceptibility data are presented for all non-Typhi Salmonella, as well as the following four Salmonella serotypes: Typhimurium, Enteritidis, Newport, and Heidelberg. Section IV contains data for Campylobacter isolates recovered from humans, retail meats, and chicken carcass rinsates. Antimicrobial susceptibility data for C. jejuni and C. coli are presented separately. Section V contains susceptibility data for E. coli isolates from retail meats and chicken carcass rinsates.

Each section begins with a table that shows the number of isolates tested by source and year. This is followed by a table and two figures that show the percentages of retail meats that tested positive. Data are also provided on the distribution of Salmonella serotypes and Campylobacter species isolated from humans, retail meats, and food animals.

Data on antimicrobial susceptibility follows. MIC distribution tables include MIC distributions, the percentages of isolates displaying intermediate susceptibility and resistance, and 95% confidence intervals for the percent resistant, by source for 2004. Confidence intervals were calculated using the Clopper-Pearson exact method.14 The unshaded areas in the MIC tables indicate the range of concentrations tested for each antimicrobial.15 Single vertical bars indicate antimicrobial susceptibility breakpoints, while double vertical bars indicate antimicrobial resistance breakpoints.

The MIC distributions are followed by tables that show the numbers and percentages of isolates that were resistant, by year, through 2004. 16 The total number of isolates tested per year for each source is listed at the top of each table. An empty cell in this area indicates that surveillance was not conducted for that particular source, whereas a zero indicates that surveillance was conducted, but no isolates were available for testing. Below the section containing the number of isolates tested, empty shaded boxes indicate that there are no data to report because surveillance was not conducted or isolates were not available for testing.

Resistance to ceftiofur and nalidixic acid among Salmonella isolates is highlighted in several pie charts and graphs (Figures 6-14). 17,18 Third-generation cephalosporins (such as ceftriaxone) and fluoroquinolones (such as ciprofloxacin) are antimicrobial agents commonly used for the treatment of severe Salmonella infections in humans. Ceftiofur is the only third-generation cephalosporin approved for use in food animals in the U.S. and elevated MICs (≥ 8 µg/ml) correlate with decreased susceptibility to ceftriaxone (MIC ≥ 2 µg/ml). Similarly, resistance to the quinolone nalidixic acid (MIC ≥ 32 µg/ml) correlates with decreased susceptibility to ciprofloxacin (MIC ≥ 0.125 µg/ml). Finally, for Salmonella and E. coli isolates, data on multidrug resistance (MDR) phenotypes of public health importance are also presented (Tables 13-17, 21, 24, 28, 32, and 47).

The data contained in this report may, in a few cases, differ from those previously reported. These differences may be due to changes in breakpoints (Campylobacter) and/or the dynamic nature of the data, which are updated if new information is obtained about the bacterial isolates under surveillance or specific isolates are retested. In a few cases, differences may be due to other reasons. Etest® MIC results for Campylobacter that fell between the two-fold dilutions described in CLSI documents were rounded up to next two-fold dilution prior to interpretation in this report, but not in USDA’s annual reports.

1 Further details about sampling of Campylobacter isolates can be found in the NARMS Human Isolates Final Report for 2004.
2 http://www.fsis.usda.gov/Science/Microbiological_Lab_Guidebook/index.asp
3 The organization was known as the National Committee on Clinical Laboratory Standards (NCCLS) until the name
was changed in January of 2005 to CLSI. In this report, the name CLSI is used.
4 NCCLS. 2002. Performance Standards for Antimicrobial Disk and Dilution Susceptibility Tests for Bacteria Isolated from Animals; Approved Standard—Second Edition. NCCLS document M31-A2. NCCLS, Wayne, PA.
5 NCCLS. 2003. Performance Standards for Antimicrobial Susceptibility Testing; Thirteenth Informational Supplement. NCCLS document M100-S13. NCCLS, Wayne, PA.
6 In USDA’s NARMS annual reports, MIC values were not rounded up prior to interpretation.
7 CLSI. 2006. Methods for Antimicrobial Dilution and Disk Susceptibility Testing of Infrequently Isolated or Fastidious Bacteria; Approved Guideline. CLSI document M45-A. CLSI, Wayne, PA.
8 CLSI. 2008. Performance Standards for Antimicrobial Susceptibility Testing; Eighteenth Informational Supplement. CLSI document M100-S18. CLSI, Wayne, PA. 
9 Breakpoints were adopted from CLSI (Clinical and Laboratory Standards Institute), except for streptomycin, which has no CLSI breakpoints.
10 Cephalothin was tested through 2003.
11 Sulfamethoxazole was tested from 1996 through 2003 and was replaced by sulfisoxazole in 2004. 
12 Breakpoints were adopted from CLSI (Clinical and Laboratory Standards Institute), when available. CLSI breakpoints were available for erythromycin, ciprofloxacin, doxycyline, and tetracycline.
13 For florfenicol, only a susceptible breakpoint ( ≤ 4 µg/ml) has been established. In this report, isolates with an MIC ≥ 8 µg/ml are categorized as resistant. 
14 Newcombe RG. Two-sided confidence intervals for the single proportion: comparison of seven methods. Statistics in Medicine 1998; 17(8): 857-872.
15 The concentration ranges are also listed in the Appendix.
16 Data on Campylobacter recovered from chickens is presented only for the period of July 2001 through December 2004, as described in Section IIA.
17 Note that the scales vary from figure to figure, based on the maximum percent resistance.
18 Below each graph is a table that shows the number of isolates tested. Empty grey boxes indicate that surveillance was not conducted, while boxes with zeros indicate that there were no isolates available for testing.