Comparison of automated microbroth dilution and agar dilution for antimicrobial susceptibility of Campylobacter jejuni isolated from dairy sources

Transcription

1 Journal of Antimicrobial Chemotherapy Advance Access published August 30, 2005 Journal of Antimicrobial Chemotherapy doi: /jac/dki309 Comparison of automated microbroth dilution and agar dilution for antimicrobial susceptibility of Campylobacter jejuni isolated from dairy sources Lisa W. Halbert 1, John B. Kaneene 1 *, Linda S. Mansfield 1, Pamela L. Ruegg 2, Lorin D. Warnick 3, Scott J. Wells 4, Chuck P. Fossler 5, Amy M. Campbell 1 and Angela M. Geiger-Zwald 2 1 Population Medicine Center, College of Veterinary Medicine, Michigan State University, East Lansing, MI 48824, USA; 2 Department of Dairy Science, College of Agriculture and Life Sciences, University of Wisconsin-Madison, Madison, WI 53706, USA; 3 Department of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA; 4 Department of Clinical and Population Science, College of Veterinary Medicine, University of Minnesota, St Paul, MN 55108, USA; 5 Centers for Epidemiology and Animal Health, USDA-APHIS-VS, Fort Collins, CO , USA Received 21 April 2005; returned 21 June 2005; revised 25 July 2005; accepted 7 August 2005 Objectives: To compare the agreement between microbroth dilution and agar dilution for antimicrobial susceptibility testing of Campylobacter jejuni. Methods: Utilizing commercially prepared antimicrobial panels, microbroth dilution was compared with agar dilution for determining antimicrobial susceptibility in C. jejuni isolates. To assess the performance of both techniques for ampicillin, 190 C. jejuni isolates from dairy cattle were utilized. A group of 172 C. jejuni isolates from dairy sources were used to compare the susceptibility to ciprofloxacin, erythromycin, nalidixic acid and tetracycline. Results: Our results indicate that microbroth dilution and agar dilution agree within 1 log 2 dilution for 86.7% of the isolates tested. Ciprofloxacin had the highest level of agreement for isolates tested by both techniques, resulting in a kappa of and 97.1% agreement 1 log 2 dilution. The least agreement was observed in determining the susceptibility of isolates to ampicillin and erythromycin (82.1 and 79.7% agreement 1 log 2 dilution). However, kappa statistics were considered to have good agreement for these antimicrobials. There were no significant differences in the summary statistics for any of the five antimicrobials evaluated for the isolates analysed by the percentage of resistant isolates, MIC 50, MIC 75 or MIC 90 beyond 1 log 2 dilution. There was no association in the classification of resistance by the testing methods employed. We also demonstrated that the quality control strain of C. jejuni ATCC performed in a consistent manner for both agar dilution and microbroth dilution. Conclusions: Microbroth dilution may be an acceptable alternative to agar dilution for determining susceptibility of C. jejuni in research or surveillance where flow of samples, labour efficiency and cost may restrict the use of agar dilution. Keywords: antibiotic resistance, epidemiological surveillance, C. jejuni Introduction Campylobacter jejuni is among the most frequently identified causes of bacterial gastroenteritis in the United States. 1 Throughout many months in any given year, C. jejuni infections have often outnumbered other infectious causes of food-borne illness in the United States, such as Salmonella, Escherichia coli O157:h7 and Shigella, based on CDC reports of annual disease incidence. Based on these data, it has been estimated that up to two million cases of illness may be caused by this organism annually. 2 Most C. jejuni... *Corresponding author. Tel: ; Fax: ; Kaneene@cvm.msu.edu... Page 1 of 6 Ó The Author Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please journals.permissions@oupjournals.org

2 Halbert et al. enteritis cases are mild, self-limiting episodes of vomiting, cramping and diarrhoea. However, a more serious form of the illness occurs in infants, geriatric patients and immune suppressed individuals. In these cases, bloody stools, dehydration, septicaemia and long-term sequelae can occur. 3 Secondary effects of C. jejuni gastroenteritis can include the demyelinating neurological disorder of Guillian Barre syndrome or intermittent arthritis. 4,5 C. jejuni can colonize the gastrointestinal tracts of mammals and birds without causing disease. 6,7 Thus, faecal material from apparently healthy animals may contaminate meat, raw milk or crosscontaminate fresh produce, resulting in human illness. 8,9 Another primary concern with C. jejuni is that this organism has demonstrated the ability to develop resistance to antimicrobials. 10,11 Human isolates are displaying increased resistance to many classes of the drugs throughout time and introduction of new pharmaceuticals. 12,13 This is a global problem. 14 In countries with lax restrictions on human and animal treatment, very high levels of antimicrobial resistance are observed. 15,16 However, in developed countries, there is ongoing debate regarding the role of antibiotics in animal husbandry and veterinary therapeutics in the food-borne dissemination of bacterial resistance mechanisms. 17,18 There is documentation of increased fluoroquinolone resistance in C. jejuni and other bacteria once these antimicrobials were approved in some food animal species. 19 In the laboratory setting, C. jejuni is a fastidious organism requiring very specific growth conditions such as a microaerophilic environment and selective media. 20 The NCCLS in the United States has recently standardized agar dilution for antimicrobial testing in this organism. 21,22 However, many other methods have been employed both for surveillance and in research laboratories including disc diffusion, microbroth dilution and Etest, which has made it difficult to track or compare global trends in C. jejuni resistance Luber et al. 26 have compared the results of C. jejuni isolates tested by agar dilution, Etest and microbroth dilution. These authors found microbroth dilution performed using the SensiTitre Ò system with commercially prepared antimicrobial reliable, efficient and provided results with a high degree of agreement in classifying the susceptibility of isolates when compared with Etest or agar dilution. However, the C. jejuni used in this German study consisted of isolates from clinical submissions from humans with gastroenteritis and poultry. 26 Agar dilution requires the preparation of serial dilutions of antimicrobials incorporated into a blood agar plate, which is both expensive and time consuming. The prepared plates also have a very short shelf life that therefore requires frequent preparation with ongoing testing such as may occur in longitudinal study designs or clinical laboratory settings. Another consideration is that the NCCLS guidelines for C. jejuni do not incorporate antimicrobials of interest for some research questions. Currently, ciprofloxacin, doxycycline, erythromycin, gentamicin and meropenem are the antimicrobials that are part of the standardized protocol. 21 However, the National Antimicrobial Resistance Monitoring 25 surveillance programme in the United States monitors C. jejuni resistance in humans and animals to azithromycin, chloramphenicol, clindamycin, erythromycin, gentamicin, ciprofloxacin, nalidixic acid and tetracycline using the Etest. In 2000, Wallace and Fedorka-Cray found comparable results between microbroth dilution, Etest and agar dilution for C. jejuni isolates when assessing the antimicrobials of interest for NARMS surveillance. Therefore, the primary objective of this study was to evaluate the agreement of MICs between automated microbroth dilution using a SensiTitre Ò system and agar dilution for determining antimicrobial susceptibility of C. jejuni isolated from dairy cattle in the United States as part of a large, longitudinal research project that would require efficient processing of thousands of C. jejuni isolates. Materials and methods Isolates: cattle samples C. jejuni isolates that had been isolated from dairy cattle were utilized for this study. In order to assess the agreement between ciprofloxacin, erythromycin, nalidixic acid and tetracycline, 172 C. jejuni with diverse MIC outcomes for these four antimicrobials as determined by microbroth dilution technique were used. In order to assess the agreement of ampicillin, 190 isolates with a broad distribution of MICs as determined by microbroth distribution were used. These five antimicrobials were chosen for comparison purposes by the ranges of MICs demonstrated by our isolates, the on-farm exposure to these classes of drugs and/or interest in human C. jejuni resistance surveillance. Farms were enrolled in the study under confidentiality agreements, and cattle sampling for faecal collection was performed under each respective university animal use guidelines (Michigan State University, Cornell University, University of Wisconsin and University of Minnesota). Cattle samples were collected by placing 10 g of faecal material obtained by rectal retrieval into Whirl-Pak Ò bags. A separate glove was used for the collection of each sample. Animal faecal samples and milk samples were suspended in PBS solution. Although the study design included sampling the farm environment using enrichment, recovery of C. jejuni from farm environments resulted in very few isolates. Therefore, C. jejuni from environmental sources were not included in this study. PBS suspended biological samples were streaked on selective C. jejuni Blaser plates (BD Diagnostics) and incubated at 42 C in microaerophilic conditions (5 10% CO 2 )in water-jacketed CO 2 incubators (NuAire, Inc., Plymouth, MN, USA, model NU-4500 and VWR, Cornelius, OR, USA, model 2400) for 48 h. Typical colonies were selected and streaked on sheep blood agar (Cleveland Scientific) and incubated at 42 C in microaerophilic conditions (5 10% CO 2 ) for 48 h. C. jejuni identification was performed from isolated colonies by Gram s staining, oxidase testing and motility testing. Hippurate hydrolysis was used to speciate C. jejuni. Over 97% of our isolates were classified as C. jejuni. 27 Taqman PCR was used on a subset of isolates to verify the performance of hippurate testing to adequately identify C. jejuni for 1430 isolates. Among those hippuratepositive isolates tested, 1311 were classified as C. jejuni. Only isolates demonstrating agreement between hippurate hydrolysis and PCR were used in this study. In vitro susceptibility testing: microbroth dilution Initial in vitro susceptibility testing was performed using the microbroth dilution method, following guidelines provided by the NCCLS. 21 Bacterial isolates from frozen stock were grown on Brucella agar supplemented with 5% defibrinated sheep blood (BASB) for 48 h at 42 C under microaerophilic conditions (5 10% CO 2 ). Individual colonies from each plate were subcultured on BASB under similar growth conditions. Bacteria were swabbed from the BASB and suspended in 5mLofH 2 O and the turbidity was adjusted to that of a 0.5 McFarland standard. This suspension was used to make a final bacterial inoculum concentration of cfu/ml using Haemophilus testing medium (Trek Diagnostics). Customized microbroth dilution plates (CMV1USDA) were purchased pre-made from TREK Diagnostic Systems, Inc., with a prepared range of drug concentrations of azithromycin, chloramphenicol, Page 2 of 6

3 Comparison of microbroth dilution to agar dilution ciprofloxacin, clindamycin, erythromycin, gentamicin, nalidixic acid and tetracycline. C. jejuni ATCC was used as a quality control strain. Each plate was inoculated by adding 100 ml of the bacterial suspension using a SensiTitre Ò autoinoculator, covered with a gaspermeable seal and incubated at 42 Cat5%CO 2 for 48 h. The MIC was determined as the minimum antimicrobial dilution at which no bacterial growth occurred. Following the observation that dairy isolates did not demonstrate resistance patterns similar to those from humans, another customized antimicrobial panel (CMV2DMSU) was developed with Trek Diagnostics to address drug exposures that are common to dairy cattle management. This antimicrobial panel included 17 drugs encompassing drug classes used on our study farms such as b-lactams and cephalosporins. 28 In vitro susceptibility testing: agar dilution Bacterial isolates from frozen stock were grown on Brucella agar supplemented with BASB for 48 h at 42 C under microaerophilic conditions (5 10% CO 2 ). Individual colonies from each plate were subcultured on BASB under similar growth conditions. Agar dilution was performed according to the NCCLS VAST guidelines. Mueller Hinton agar (Michigan State University Media Services) was supplemented with 5% sheep blood (Cleveland Scientific, Cleveland, OH, USA) and each antimicrobial agent according to the concentration ranges in Table 1 in order to compare agreement of the antimicrobial susceptibility results obtained using microbroth dilution. Bacterial cultures were suspended in Mueller Hinton broth (BD Biosciences) and adjusted to a 0.5 McFarland standard and inoculated onto the prepared plates using 1 mm pins with a Cathra Replicator (Oxoid, Inc., Ogdensburg, NY, USA). Agar dilution plates were incubated in microaerophilic conditions (5 10% CO 2 )at42 C for 24 h. MICs for each isolate was recorded as the lowest antimicrobial concentration for each drug that inhibited growth of each C. jejuni isolate. Data analysis Agreement was defined as 1 log 2 dilution between the two tests for each isolate MIC value in order to be consistent with other authors 26,29 and to account for the amount of error considered inherent in reading the test results. 30 Using the breakpoints defined by NARMS for C. jejuni and general enteric breakpoint for ampicillin (Table 1), isolates were classified as resistant or susceptible for descriptive purposes. SAS Version 8.2 was used to generate c 2 tests to determine whether there were significant differences in the determination of resistance associated to the testing methodology. The kappa statistic was used to determine the agreement for isolate classification using both agar dilution and microbroth dilution methodology. The kappa, the proportion of agreement between two observers above what would be expected through random chance, was calculated as follows: level of agreement observed level of agreement k ¼ due to random chance 1 level of agreement due to random chance where the level of agreement observed is the proportion of isolates with the same MIC value by both agar dilution and microbroth dilution, and the level of agreement due to random chance is the proportion that would be matched owing to simple random chance. Results It should be noted that the isolates chosen for this study do not represent those of all C. jejuni isolated during the longitudinal study. Isolates chosen represented the broadest range of MICs with which to compare agar dilution and microbroth dilution, and were therefore more resistant than the overall susceptibility demonstrated by the C. jejuni isolated from the participating 132 dairy farms enrolled. Also, only isolates from cattle samples that were confirmed to be C. jejuni by PCR were included. Quality control strain of C. jejuni ATCC was used for each batch of susceptibility testing. In Table 2, it is demonstrated that the quality control organism performed within the quality control ranges for those antimicrobials for which performance ranges have been established for agar dilution methods. Overall, it is apparent that the quality control organisms performed with similar consistency for both agar dilution and microbroth dilution. Ampicillin Table 1. Dilution ranges for the antimicrobial agents used and interpretative breakpoints While exact agreement as determined by the same MIC for each isolate by agar and microbroth dilution was 18.9%, agreement as defined within the 1 dilution error for test interpretation was 82.1% for this antimicrobial (Table 3). SensiTitre Ò tended to define isolate MICs at one dilution higher when there was disagreement between the two tests, as demonstrated by the difference in MIC 50, MIC 75 and MIC 90. The overall level of agreement between the two tests was determined to be good with a k of While the overall percentage of isolates determined to be resistant was also higher by the microbroth dilution (6.3 versus 3.2%). This difference was not significant (P = 0.14)., Antimicrobial agent Microbroth dilution test ranges (mg/l) Agar dilution test ranges (mg/l) NCCLS quality control agar ranges for C. jejuni ATCC NARMS interpretative breakpoints S I R Ampicillin a NA b Ciprofloxacin c d Eythromycin c d Nalidixic acid c NA 32 d Tetracycline c d S, susceptible; I, intermediate; R, resistant. a Trek Diagnostics SensiTitre Ò Panel CM2DMSU. b General enteric breakpoint. c Trek Diagnostics SensiTitre Ò Panel CMV1USDA. d Campylobacter breakpoint. Page 3 of 6

4 Halbert et al. Table 2. Performance of microbroth dilution and agar dilution for determination of MIC quality control levels for C. jejuni 33560: paired comparison of MICs generated by both methods NCCLS quality control agar ranges for C. jejuni ATCC Ciprofloxacin MIC (mg/l) Replicate microbroth agar Ampicillin a NA Ciprofloxacin b mg/l Erythromycin b 1 8 mg/l Nalidixic acid b 8 32 mg/l Tetracycline b 1 4 mg/l a Trek Diagnostics SensiTitre Ò Panel CM2DMSU. b Trek Diagnostics SensiTitre Ò Panel CMV1USDA. Ciprofloxacin is often used to represent the fluoroquinolones drug class when antimicrobial susceptibility or surveillance of foodborne pathogens is being studied. Therefore, it is very encouraging to note the performance of this antimicrobial with both agar dilution and automated microbroth dilution. Exact agreement as determined by the same MIC for each isolate by agar and microbroth dilution was 59.9%, and agreement as defined within the 1 dilution error for test interpretation was 97.1% for ciprofloxacin (Table 3). The measures of dispersement of isolate MIC distributions as defined by MIC 50 and MIC 75 were identical. However, the MIC 90 was determined to be one dilution higher for agar dilution than for microbroth dilution. The overall level of agreement between the two tests was determined to be excellent with a k of The overall percentage of isolates determined to be resistant was similar for microbroth dilution (2.9%) as compared with agar dilution (2.3%). This difference was not significant (P = 0.74). Erythromycin Erythromycin did not perform as consistently across both techniques. Exact agreement as determined by the same MIC for each isolate by agar and microbroth dilution was 27.9%, and agreement as defined within the 1 dilution error for test interpretation was 79.7% for this antimicrobial (Table 3). SensiTitre Ò tended to define isolate MICs at one dilution lower when there was disagreement between the two tests. However, this was only evident in the MIC 50. The other summary measures agreed exactly between the two tests including percentage of resistant isolates, MIC 75 and MIC 90. The tendency for the microbroth dilution to classify isolates with a lower MIC was reflected in a k of 0.494, which would be considered to be fair to good agreement. Nalidixic acid Similar to the fluoroquinolone ciprofloxacin, the nalidixic acid results were similar by both techniques. Exact agreement as determined by the same MIC for each isolate by agar and microbroth dilution was 51.7%, and agreement as defined within the 1 dilution error for test interpretation was 90.1% for nalidixic acid (Table 3). The measures of isolate distributions were the same for agar dilution and microbroth dilution for MIC 50, MIC 75 and MIC 90. Individual isolate classification resulted in a good agreement as determined by k of While the percentage of resistant isolates, in Table 3, was higher for microbroth dilution (4.0 versus 2.3%). This was not significantly different (P = 0.35). Tetracycline Our dairy isolates demonstrated a broad range of MICs to this antimicrobial. However, in order to obtain isolates with which to test the other drugs, isolates co-resistant to tetracycline were overrepresented in this comparison. Exact agreement for tetracycline as determined by the same MIC for each isolate by agar and microbroth dilution was 41.9%, and agreement as defined within the 1 dilution error for test interpretation was 84.9% for tetracycline (Table 3). The measures of isolate distributions were the same for agar dilution and microbroth dilution for MIC 50, MIC 75 and MIC 90. Individual isolate classification resulted good agreement as determined by k of While the percentage of resistant isolates, in Table 3, was higher for microbroth dilution (90.7 versus 88.4%) this was not significantly different (P = 0.48). Discussion As noted previously, the aim of this study was not to describe the antimicrobial susceptibility of C. jejuni from dairy sources, so the results should not be considered as representative of the resistance patterns from this source. Isolates were chosen to represent the broadest range of MIC values, and consequently there were more resistant dairy isolates selected for this study. 31 Page 4 of 6

5 Comparison of microbroth dilution to agar dilution Table 3. Agreement between agar dilution and SensiTitre Ò microbroth dilution Antimicrobial agents ampicillin ciprofloxacin erythromycin nalidixic acid tetracycline no. % no. % no. % no. % no. % Exact agreement Within 1 dilution Within 2 dilutions Throughout the literature and across researchers there has been much difficulty in standardizing the laboratory techniques for C. jejuni. This disparity is pervasive at the isolation and identification level, 20 during the determination of appropriate quality control organisms, 32,33 as well as standardizing acceptable antimicrobial testing methods. 26,29,34,35 The lack of standardization for C. jejuni in testing antimicrobial resistance makes it very difficult to discuss trends in susceptibility across research groups and consequently to identify risk factors which may be contributing to those trends globally. Since NCCLS has validated that C jejuni ATCC should be used for quality control evaluation of the performance of antimicrobial techniques, it is encouraging that this strain performed with similar consistency for both agar dilution and microbroth dilution. We also demonstrated that C. jejuni performed within the quality control test ranges with both testing techniques. Overall, our level of agreement between the two techniques of agar dilution and microbroth dilution was good to excellent and varied by the antimicrobial being tested as demonstrated by the kappa values for the descriptive classification of isolates as resistant or susceptible and varied from to The percentage of concordant results (both tests determined an isolate to be susceptible or resistant) was 96.5% and is reflected in the good to excellent agreement of the kappa values for each antimicrobial. We also demonstrated that no significant differences were obtained in the classification of descriptive statistics for the isolates being tested. In summarizing the measures of dispersion and central tendency of our isolate collection, we found that agar dilution and microbroth dilution both detected isolates in a similar manner according to percentage of resistant isolates, MIC 50, MIC 75 and MIC 90. The most disagreement between agar dilution and microbroth dilution across these measures was by no more than 1 log 2 dilution, which is essentially within the agreement that can be anticipated by the inherent error in reading MICs manually. 30 Agreement defined through MIC outcome values demonstrated that microbroth dilution performed in an acceptable manner to agar dilution testing of our C. jejuni isolates. Exact agreement for MIC outcome ranged from 18.9% for ampicillin to 59.9% for ciprofloxacin. This finding is extremely encouraging for the fluoroquinolone family of antimicrobials, since valid monitoring for the development of decreased susceptibility of C. jejuni to this drug is a concern worldwide. 15,19 As the definition of agreement within the inherent error of manually reading MIC values has been determined to be 1 log 2 dilution, 30 we demonstrated that microbroth dilution using automated inoculation with a SensiTitre Ò system produced acceptable results. Essential agreement of 1 log 2 dilution ranged from 79.7% for ampicillin to 97.1% agreement for ciprofloxacin. For all of our isolates, the essential agreement was 86.7% (note: weighted for sample size re: ampicillin). Other authors have reported levels of agreement of 78.7% in comparing microbroth dilution with agar dilution, % in comparing microbroth dilution with Etest, % in comparing Etest with agar dilution 36 and 61.9% in comparing Etest with agar dilution. 29 However, these authors tested different antimicrobials, fewer overall isolates and also included Campylobacter coli in their comparison. 26,29,36 It should be noted that due to the limited use of antimicrobials in dairy cattle, much lower levels of resistance were observed in our isolates compared with authors studying poultry, swine or human isolates. 13,16,24,25 Therefore, studies conducted using isolates with more exposure to antimicrobial selective pressure and more diverse range of MIC may yield different levels of agreement than we have demonstrated. In summary, we have demonstrated that microbroth dilution using a SensiTitre Ò autoinoculation system and commercially prepared antimicrobial panels (Trek Diagnostics, Westlake, OH, USA) may be both cost and time efficient while yielding results comparable to agar dilution. For researchers processing a high volume of samples, microbroth dilution may provide an acceptable alternative for testing C. jejuni isolates. Validation of performance of the antimicrobials included for a customized microbroth panel with agar dilution will help ensure comparable results. The opportunity to more efficiently follow trends in C. jejuni may help elucidate trends and risk factors that may be contributing to the reduced susceptibility of this organism to antimicrobials of interest to both human and veterinary medicine. Acknowledgements We thank RoseAnn Miller, Morgan Wallace, Katherine May and Robert Hanson for their technical assistance. This work was supported by USDA NRI (Epidemiological Aspects of Food Safety), the Michigan Agricultural Experiment Station and the Population Medicine Center. References 1. Altekruse SF, Swerdlow DL, Stern NJ. Campylobacter jejuni. In: Tollefson L, ed. The Veterinary Clinics of North America, Food Animal Practice: Microbial Food Borne Pathogens. Philadelphia, PA, USA: W. B. Saunders Company, 1998; Allos MB. Campylobacter jejuni infections: update on emerging issues and trends. Clin Infect Dis 2001; 32: Page 5 of 6

6 Halbert et al. 3. Blaser MJ. Epidemiologic and clinical features of Campylobacter jejuni infections. J Infect Dis 1997; 176: S Rees JH, Soudain SE, Gregson NA et al. Campylobacter jejuni infection and Guillain-Barre syndrome. N Engl J Med 1995; 333: Nachamkin I, Allos BM, Ho T. Campylobacter species and Guillain-Barre syndrome. Clin Microbiol Rev 1998; 11: Manser PA, Dalziel RW. A survey of Campylobacter in animals. J Hyg 1985; 95: Petersen L, Nielsen EM, Engberg J et al. Comparison of genotypes and serotypes of Campylobacter jejuni isolated from Danish wild mammals and birds and from broiler flocks and humans. Appl Environ Microbiol 2001; 76: Jacob-Reitsma W. Campylobacter in the food supply. In: Nachamkin I, Blaser M, eds. Campylobacter, 2nd edn. Washington, DC: ASM Press, 2000; Kramer JM, Frost JA, Bolton FJ et al. Campylobacter contamination of raw meat and poultry at retail sale: identification of multiple types and comparison with isolates from human infection. J Food Prot 2000; 63: Luo N, Sahin O, Lin J et al. In vivo selection of Campylobacter isolates with high levels of fluoroquinolone resistance associated with gyra mutations and the function of the CmeABC efflux pump. Antimicrob Agents Chemother 2003; 47: McDermott PF, Bodeis SM, English LL et al. Ciprofloxacin resistance in Campylobacter jejuni evolves rapidly in chickens treated with fluoroquinolones. J Infect Dis 2002; 185: Moore JE, McLernon P, Wareing D et al. Characterization of fluoroquinolone-resistance Campylobacter species isolated from human being and chickens. Vet Rec 2002; 150: Aarestrup FM, Nielsen EM, Madsen M et al. Antimicrobial susceptibility patterns of thermophilic Campylobacter spp. from humans, pigs, cattle, and broilers in Denmark. Antimicrob Agents Chemother 1997; 41: Neu HC. The crisis in antibiotic resistance. Science 1992; 257: Padungtod P, Kaneene JB. Campylobacter spp. in human, chickens, pigs and their antimicrobial resistance. J Vet Med Sci 2003; 65: Saenz Y, Zarazaga M, Lantero M et al. Antibiotic resistance in Campylobacter strains isolated from animals, foods, and humans in Spain Antimicrob Agents Chemother 2000; 44: Threlfall EJ, Ward LR, Frost JA et al. The emergence and spread of antibiotic resistance in food-borne bacteria. Int J Food Microbiol 2000; 62: VanDenBogaard AE. Antimicrobial resistance-relation to human and animal exposure to antibiotics. J Antimicrob Chemother 1997; 40: Piddock LJ, Ricci V, Pumbwe L et al. Fluoroquinolone resistance in Campylobacter species from man and animals: detection of topoisomerase genes. J Antimicrob Chemother 2003; 51: Correy JE, Post DE, Colin P et al. Culture media for the isolation of Campylobacters. Int J Food Microbiol 1995; 26: National Committee for Clinical Laboratory Standards. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically: Approved Standard 23. NCCLS, Wayne, PA, USA, National Committee for Clinical Laboratory Standards. MIC Testing Supplemental Tables: Approved Standard M100. NCCLS, Wayne, PA, USA, Cabrita J, Rodrigues J, Braganca F et al. Prevalence, biotypes, plasmid profile and antimicrobial resistance of Campylobacter isolated from wild and domestic animals from northeast Portugal. J Appl Bacteriol 1992; 73: Aquino MH, Filgueiras AL, Ferreira MC et al. Antimicrobial resistance and plasmid profiles of Campylobacter jejuni and Campylobacter coli from human and animal sources. Lett Appl Microbiol 2002; 34: NARMS. National Antimicrobial Resistance Monitoring System: Enteric Bacteria 2000 Annual Report. Atlanta, GA: Centers for Disease Control and Prevention, Luber P, Bartelt E, Genschow E et al. Comparison of broth microdilution, e test, and agar dilution methods for antibiotic susceptibility testing of Campylobacter jejuni and Campylobacter coli. J Clin Microbiol 2003; 41: Green AM. Patterns of occurrence of Campylobacter in organic and conventional dairy farms in Midwestern and Northeastern United States (MS thesis). East Lansing, MI, USA: Michigan State University, Zwald A, Ruegg P, Kaneene J et al. Management practices and reported antimicrobial usage on conventional and organic dairy farms. J Dairy Sci 2004; 87: Ge B, Bodeis S, Walker RD et al. Comparison of the E-test and agar dilution for in vitro antimicrobial susceptibility testing of Campylobacter. J Antimicrob Chemother 2002; 50: Barry AL, Braun LE. Reader error in determining minimal inhibitory concentrations with microdilution susceptibility test panels. J Clin Microbiol 1981; 13: Halbert LW. Patterns of reduced antimicrobial susceptibility of Campylobacter isolated from organic and conventional dairy farms in the Midwestern and northeastern United States (PhD dissertation). East Lansing, MI, USA: Michigan State University, Denton JH. Performance standards for Campylobacter are not warranted. Food Tech 2002; 56: Hakanen A, Kotilainen P, Siitonen A et al. Quality control strains used in susceptibility testing of Campylobacter spp. J Clin Microbiol 2002; 40: Hanson RA. The development of standardized in vitro antimicrobial susceptibility testing methods for Campylobacter spp. (MS thesis). East Lansing, MI, USA: Michigan State University, Wallace FM, Fedorka-Cray PJ. Evaluation of three different methods for antimicrobial susceptibility screening of Campylobacter. In: General Meeting of the American Society for Microbiology, Los Angeles, CA, American Society for Microbiology, Washington, DC, USA. 36. Oncul O, Zarakolu P, Oncul O et al. Antimicrobial susceptibility testing of Campylobacter jejuni: a comparison between E-test and agar dilution method. Diagn Microbiol Infect Dis 2003; 45: Page 6 of 6