Abstract
A body of evidence exists that suggests that antimicrobial use in food animals leads to resistance in both pathogenic and commensal bacteria. This study focused on the impact of three different antimicrobial regimes (low-level continuous, pulse, and no antimicrobial) for two antimicrobials (chlortetracycline and tylosin) on the presence of Salmonella spp. and on the prevalence of antimicrobial resistance of both Salmonella spp. and nonspecific Escherichia coli in fecal samples from feeder swine. The prevalence of fecal samples positive for Salmonella spp. significantly decreased between the samples taken at feeder placement compared to samples taken when the animals were close to market weight. Differences in resistance of Salmonella spp. did not appear to be influenced by dosing treatment including the control. Analysis of antimicrobial resistance examining both susceptibility and resistance, as well as MIC outcomes, demonstrated that only resistance to cephalothin increased in E. coli under the pulse chlortetracycline treatment. These results suggest that the dosing regimes examined in this study did not lead to an increase in either the prevalence of Salmonella spp. or the prevalence of antimicrobial resistance in isolates of Salmonella spp. or E. coli.
Microbiological and clinical data indicate that increasing bacterial resistance to antimicrobials is making it more difficult to treat serious bacterial infections in both animals and humans (12, 13, 17, 31). While there may be several contributing causes for the increase in antimicrobial resistance in pathogenic bacteria, most researchers focus on increased use of antimicrobial agents. The “prudent use of antimicrobials” has become a major objective of the human and veterinary medical care establishment (17; Alliance for the Prudent Use of Antimicrobials [http://www.tufts.edu/med/apua/]). Given the possibility of resistant organisms passing from animals to humans (8, 18, 23, 25, 29, 30) that has been recognized by the World Organization of Animal Health (33) and the World Health Organization (31, 32), measures to limit these routes of exposure are necessary.
Cromwell (3) summarizes a number of studies that demonstrate performance benefits, including increases in average daily gain and feed conversion, as well as reductions in morbidity and mortality, of antimicrobial use in multiple phases of swine production. In contrast, Dritz et al. (6) found no benefit related to average daily gain or feed efficiency from administering growth promotion levels of antimicrobials to finishing pigs. More information is needed regarding the effects of reducing the use of antimicrobials in food animals or modifying the way in which antimicrobials are currently delivered.
In 1999, it was estimated that Salmonella spp. (nontyphoidal) was responsible for 30.6% of the estimated 1,809 deaths caused by known food-borne pathogens in the United States (16); it is therefore of public health importance to control the presence of Salmonella in food animals and other segments of the food industry. The use of antimicrobials in food animals has been shown to decrease the likelihood of recovering Salmonella spp. in feces. A recent study by Gebreyes et al. (10) demonstrated a higher animal-level prevalence of Salmonella in antimicrobial-free swine herds (15.2%) compared to conventional herds which used antimicrobials (4.2%). Similar findings were published by Ebner and Mathre (7).
Although both Gebreyes et al. (10) and Ebner et al. (7) found lower prevalences of shedding of Salmonella spp. in feces on farms that used antimicrobials, these researchers did find that the use of antimicrobials selected for resistant populations of the bacteria. There is evidence that antimicrobial use in animals selects for resistance in both pathogenic and commensal organisms (5, 14). A commensal organism of interest, Escherichia coli, may serve as a reservoir of transferable antimicrobial resistance genetic elements (24, 26); laboratory-based studies have shown that E. coli is capable of transferring resistance to other bacterial species, such as Salmonella spp., which are disseminated through the human food chain (1, 16, 27). This mechanism of transfer has been shown to occur within and between many different bacterial genera and has been proposed to be a major cause behind the rapid spread of resistance genes during the last five decades (4).
An evaluation of the risks and benefits in the use of antimicrobial agents in food-animal production is necessary to determine the magnitude of the public health risk; the findings of the present study contribute data to this effort. The objectives here were to determine the impact of three different finisher-pig antimicrobial feeding regimes (low-level continuous, pulse, and no antimicrobial) for chlortetracycline (CTC) and tylosin on the presence of Salmonella and on the prevalence of antimicrobial resistance of both E. coli and Salmonella spp. and to look for evidence of sharing of resistance phenotypes between E. coli and Salmonella spp.
MATERIALS AND METHODS
Treatments and sample collection.
The five treatment protocols, which consisted of the two antimicrobials, CTC and tylosin, each administered by two methods, pulse and low-level continuous, along with an antimicrobial-free control, were randomly assigned to 21 swine barns. Each of the five treatments was randomly assigned to four barns with the exception of the pulse-tylosin treatment, which was assigned to five barns to equalize the number of pens in each treatment. In most barns, three or four pen clusters were further randomly selected for testing but four barns had six pen clusters selected. A pen cluster was three adjacent pens holding a total of 60 pigs (20 per pen), and at every collection six or seven samples were collected (as described below) out of each pen in the cluster for a total of 20 samples/cluster or about one-third of the pigs. Each treatment was assigned to a total of 16 pen clusters, with the exception of the antimicrobial-free treatment, which had only 15 pen clusters.
CTC was fed at 100 g/ton of feed continuously throughout the finishing period until 2 weeks prior to marketing. Continuous tylosin was fed at 40 g/ton for the entire time the pigs were in the barn (17 weeks). Pulse doses of CTC and tylosin, 400 and 100 g/ton, respectively, were fed for 1 week, followed by 3 weeks of no antimicrobial, another week of antimicrobial, and finally concluding with no antimicrobials for the 12 weeks the pigs remained in the barns.
Feeder pigs from a single source were conveniently assigned to pens. One day after arrival for the beginning of the feeding period (period 1), six or seven fecal samples were collected from each pen of a cluster. If possible, when the pigs were observed defecating, fecal samples were obtained prior to it hitting the ground. Otherwise, a portion of the manure was obtained from piles that had presumably just dropped onto the pen floor. Care was taken to avoid touching the floor when collecting the sample. The treatment protocols were initiated at the same time these initial fecal samples were collected. A second set of fecal samples was collected in a similar manner when the pigs were near market weight, approximately 9 weeks after placement (period 2). The fecal samples were placed in plastic containers and packed in ice for overnight shipping to the testing laboratory.
Bacterial isolation and antimicrobial susceptibility testing.
Each fecal sample was processed for isolation of Salmonella spp. as previously described (28). Presumptive Salmonella isolates were serotyped at the National Veterinary Services Laboratories, Ames, IA. For E. coli isolation, approximately 1 g of feces was added to 9 ml of sterile phosphate-buffered saline and vortex mixed, and a loopful was used to streak a CHROMagar ECC plate (NorthEast Medical, Pittsburgh, PA). After incubation for 18 to 24 h at 42°C, presumptive E. coli appeared as blue-green colonies. Presumptive positive colonies were confirmed as E. coli using Vitek (bioMerieux, Durham, NC). A single representative colony was taken from each culture plate. Both Salmonella and E. coli isolates were stored on tryptic soy agar slants at room temperature for short-term storage prior to antimicrobial susceptibility testing.
Antimicrobial MICs for E. coli and Salmonella were determined according to the manufacturer's instructions by using the Sensititre semi-automated broth microdilution antibiotic susceptibility system (Trek Diagnostic Systems, Westlake, OH). MICs were interpreted using Clinical and Laboratory Standards Institute (formerly National Committee for Clinical Laboratory Standards) when available (20, 21). Otherwise, breakpoint interpretations were determined as reported for the National Antimicrobial Resistance Monitoring System (19). Staphylococcus aureus ATCC 29213, E. coli ATCC 25922, Enterococcus faecalis ATCC 29212, and Pseudomonas aeruginosa ATCC 27853 were used as quality control organisms for all antimicrobials except streptomycin, for which official quality control standards have not been set (28, 34). The 16 antimicrobials tested on the custom-made 96-well plate were amikacin, amoxicillin-clavulanic acid, ampicillin, cefoxitin, ceftiofur, ceftriaxone, cephalothin, chloramphenicol, ciprofloxacin, gentamicin, kanamycin, nalidixic acid, streptomycin, sulfamethoxazole, tetracycline, and trimethoprim-sulfamethoxazole.
Statistical analysis. (i) Presence or absence of Salmonella spp.
Multilevel logistic regression analysis, which accounts for the hierarchical nature (barn, pen cluster, and sample within pen cluster) of the data, was implemented to evaluate the presence or absence of Salmonella spp. in the fecal samples using MlWin (version 1.1; Institute of Education, University of London, London, United Kingdom). This analysis also provides an estimate of the proportion of the variance in the presence or absence of Salmonella spp. for each level in the hierarchy. The logistic regression model included independent variables for each treatment (continuous tylosin, continuous CTC, pulse tylosin, and pulse CTC) and for time period (second period). The pigs that did not receive any antimicrobial treatment (control) and the first time period were considered the base or reference levels. The model was constructed by using a backward elimination approach starting with a full model that had all main effects and interactions. The significance level was 0.05.
The presence or absence of E. coli was not analyzed because it was fairly readily isolated from most samples.
(ii) Antimicrobial resistance of Salmonella and E. coli isolates.
Susceptible and intermediate resistance categories were combined for the analysis of susceptibility for both Salmonella and E. coli.
Analysis of changes in antimicrobial resistance of the Salmonella spp. isolates was limited to descriptive statistics, including the calculation of standard errors using SUDAAN software (release 8; Research Triangle Institute, Cary, NC), which accounts for the data hierarchy. In addition to descriptive statistics, the proportion of resistance in E. coli was modeled by using logistic regression in the same manner as for the presence or absence of Salmonella in the analysis above.
(iii) Analysis of MIC data for E. coli.
The MICs for the E. coli isolates were analyzed to look for treatment effects that might not be detectable when analyzing susceptibility and resistance outcome data. For instance, analysis of susceptibility versus resistance could not be implemented when all isolates were susceptible to an antimicrobial. In addition, there could be shifts in MICs within the susceptible (resistant) isolates that might be biologically important. Similarly, there could be changes in MICs, not reflected in changes in the proportion of isolates, that are susceptible or resistant. The MICs were not highly variable in these data, so it was not appropriate to analyze the MIC data as if it were a continuous variable. Instead, the MIC data were considered to be ordinal, which led to the choice of a proportional odds model. The proportional-odds model was constructed by using SUDAAN to account for the clustering of observations within barn and pen. The proportional-odds model evaluates the odds of increasing (or decreasing) one MIC level, while the logistic-regression model estimates the odds of resistance. MICs that were less than or greater than testable limits were assigned values. For example, if a MIC was ≤8 it was assigned a value of 8. If the value was >64 it was assigned a value of 128.
RESULTS
Salmonella prevalence and antimicrobial resistance.
A total of 186 Salmonella isolates were recovered from 3,159 fecal samples (Table 1). The proportion of all fecal samples that were Salmonella positive in the first period (pretreatment) was 7.5% (n = 118) compared to 4.3% in the second period. The percentage of fecal-sample positives in the control group was stable from period 1 to period 2 (2.7 versus 3.3%). The percentage of positive fecal samples numerically decreased in the second period compared to the first period for all four antimicrobial treatments. Fecal samples from pigs in pens treated with pulse CTC had the highest percentage of Salmonella-positive fecal samples in both periods. The three most common Salmonella serotypes—serovar Derby (89.3%), serovar Typhimurium (5.9%), and serovar Mbandaka (3.2%)-occurred in both periods. The percentage of isolates that were serovar Derby decreased from 94.9 to 79.7% from period 1 to period 2, but the decrease was not statistically significant (P = 0.078). Similarly, an increase in the proportion of Salmonella isolates from period 1 to period 2 that were serovar Typhimurium was not significant (P = 0.152). The Salmonella Heidelberg and Meleagradis serotypes only occurred in the second period and accounted for 0.5 and 1.1% of the isolates, respectively. No period or treatment effect was noted for serotype prevalence.
TABLE 1.
Percent of fecal samples positive for Salmonella grouped by treatment and period
| Period and treatment | No. of samples tested | No. of Salmonella-positive samples | % Salmonella positive (SE) |
|---|---|---|---|
| Period 1 | |||
| Continuous tylosin | 320 | 16 | 5.0 (4.9) |
| Continuous CTC | 320 | 31 | 9.7 (1.6) |
| Pulse tylosin | 320 | 19 | 5.9 (4.9) |
| Pulse CTC | 320 | 44 | 13.8 (6.9) |
| No antimicrobials (control) | 300 | 8 | 2.7 (1.5) |
| Total | 1,580 | 118 | 7.5 (2.2) |
| Period 2 | |||
| Continuous tylosin | 319 | 3 | 0.9 (0.6) |
| Continuous CTC | 320 | 11 | 3.4 (0.7) |
| Pulse tylosin | 320 | 14 | 4.4 (2.1) |
| Pulse CTC | 320 | 30 | 9.4 (4.2) |
| No antimicrobials (control) | 300 | 10 | 3.3 (1.7) |
| Total | 1,579 | 68 | 4.3 (1.2) |
The final hierarchical logistic regression model for Salmonella prevalence did not include any interaction or treatment (antimicrobial) terms. The final model was: logit (Psalm-pos) = −3.528 − 0.720 (period 2).
The coefficient can be interpreted as a significant decrease in the probability of a fecal sample being Salmonella spp. positive in the second period compared to the first period (odds ratio [OR] = 0.49, P < 0.001) regardless of treatment. Much of the variation in the presence of Salmonella was at the individual sample level (47%), followed by the pen cluster level (29%) and the barn level (24%).
Analysis of resistance in Salmonella spp. isolates was limited to descriptive statistics because the number of isolates was very small in some treatments and the overall numbers were low in the second period. All Salmonella isolates that were tested for antimicrobial resistance (n = 185, 1 isolate was nonrecoverable after storage) were susceptible to 7 of the 16 antimicrobials tested (amikacin, cefoxitin, ceftriaxone, ciprofloxacin, gentamicin, nalidixic acid, and trimethoprim-sulfamethoxazole). Resistance to amoxicillin-clavulanic acid, ceftiofur, cephalothin, and kanamycin did not occur during all sampling periods (Table 2). Resistance to cephalothin and kanamycin was observed only in the control group and only during the first and second sampling periods, respectively. The prevalence of resistance to streptomycin, sulfamethoxazole, and tetracycline was high for all treatments but varied substantially, largely because of the occurrence of a single susceptible isolate. For example, only three Salmonella isolates were found in the second period of the continuous tylosin treatment, and one of these was susceptible to sulfamethoxazole.
TABLE 2.
Percentage of Salmonella isolates that were resistant to specific antimicrobials grouped by period and treatmenta
| Antimicrobial | Period | % Resistant isolates (SE)b
|
|||||
|---|---|---|---|---|---|---|---|
| Continuous tylosin (n1 = 16, n2 = 3) | Continuous CTC (n1 = 30, n2 = 11) | Pulse tylosin (n1 = 19, n2 = 14) | Pulse CTC (n1 = 44, n2 = 30) | No antimicrobial (n1 = 8, n2 = 10) | All isolates (n1 = 117, n2 = 68) | ||
| Amoxicillin-clavulanic | 1 | 0.0 | 0.0 | 5.3 (6.8) | 0.0 | 0.0 | 0.9 (0.9) |
| acid | 2 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| Ampicillin | 1 | 6.2 (0.0) | 25.8 (24.1) | 73.7 (26.3) | 27.3 (12.7) | 37.5 (13.6) | 32.2 (10.7) |
| 2 | 33.3 (16.1) | 18.2 (18.7) | 50.0 (22.6) | 60.0 (17.2) | 10.0 (7.2) | 42.6 (12.5) | |
| Ceftiofur | 1 | 0.0 | 3.2 (2.0) | 0.0 | 0.0 | 0.0 | 0.8 (0.8) |
| 2 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| Cephalothin | 1 | 0.0 | 0.0 | 0.0 | 0.0 | 12.5 (4.5) | 0.9 (0.9) |
| 2 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| Chloramphenicol | 1 | 0.0 | 0.0 | 15.8 (12.9) | 0.0 | 0.0 | 2.5 (2.0) |
| 2 | 0.0 | 0.0 | 35.7 (30.3) | 10.0 (5.4) | 0.0 | 11.8 (7.4) | |
| Kanamycin | 1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| 2 | 0.0 | 0.0 | 0.0 | 0.0 | 10.0 (7.2) | 1.5 (1.5) | |
| Streptomycin | 1 | 100.0 | 87.1 (12.9) | 100.0 | 100.0 | 100.0 | 96.6 (3.5) |
| 2 | 66.7 (32.2) | 90.9 (10.2) | 85.7 (15.5) | 100.0 | 100.0 | 94.1 (4.0) | |
| Sulfamethoxazole | 1 | 93.8 (0.0) | 80.6 (15.3) | 89.5 (6.0) | 88.6 (2.9) | 100.0 | 88.1 (4.2) |
| 2 | 66.7 (32.2) | 81.8 (14.4) | 85.7 (15.5) | 100.0 | 90.0 (7.2) | 91.2 (4.8) | |
| Tetracycline | 1 | 100.0 | 96.8 (2.0) | 100.0 | 100.0 | 87.5 (13.6) | 98.3 (1.2) |
| 2 | 66.7 (32.2) | 100.0 | 85.7 (15.5) | 96.7 (4.0) | 100.0 | 94.1 (4.0) | |
Pansusceptible antimicrobials (amikacin, cefoxitin, ciprofloxacin, gentamicin, nalidixic acid, and trimethoprim-sulfamethoxazole) were excluded from this table.
n1 and n2 represent the number of isolates in periods 1 and 2, respectively.
E. coli antimicrobial resistance: susceptibility and resistance analysis.
A total of 1,296 E. coli isolates were tested for antimicrobial resistance (Table 3). A high percentage of E. coli isolates were resistant to tetracycline regardless of period, followed by sulfamethoxazole, streptomycin, kanamycin, and ampicillin. Resistance to three antimicrobials—amoxicillin-clavulanic acid, cefoxitin, and ceftiofur—was lower and more sporadic across the treatments and periods. None of the isolates was resistant to amikacin, ceftriaxone, ciprofloxacin, or nalidixic acid.
TABLE 3.
Percentage of E. coli isolates that were resistant to specific antimicrobials grouped by period and treatmenta
| Antimicrobial | Period | % Resistant isolates (SE)b
|
|||||
|---|---|---|---|---|---|---|---|
| Continuous tylosin (n1 = 126, n2 = 150) | Continuous CTC (n1 = 113, n2 = 126) | Pulse tylosin (n1 = 118, n2 = 150) | Pulse CTC (n1 = 121, n2 = 146) | No antimicrobial (n1 = 118, n2 = 128) | All isolates (n1 = 596, n2 = 700)c | ||
| Amoxicillin-clavulanic acid | 1 | 0.8 (0.7) | 3.5 (2.3) | 0.0 | 0.8 (0.5) | 0.0 | 1.0 (0.6) |
| 2 | 2.0 (0.5) | 0.0 | 2.0 (1.2) | 0.0 | 0.0 | 0.9 (0.4) | |
| Ampicillin | 1 | 33.3 (9.5) | 42.5 (4.9) | 39.8 (8.4) | 30.6 (8.3) | 45.8 (8.6) | 38.3 (4.0) |
| 2 | 30.7 (7.1) | 35.7 (6.5) | 32.7 (7.5) | 31.5 (5.6) | 27.3 (1.6) | 31.6 (2.8) | |
| Cefoxitin | 1 | 0.8 (0.7) | 2.6 (1.7) | 0.0 | 0.0 | 0.0 | 0.7 (0.5) |
| 2 | 1.3 (0.6) | 0.0 | 2.0 (1.2) | 0.0 | 0.0 | 0.7 (0.4) | |
| Ceftiofur | 1 | 0.0 | 0.9 (0.6) | 0.0 | 0.0 | 0.0 | 0.2 (0.2) |
| 2 | 0.7 (0.6) | 0.8 (0.8) | 0.0 | 0.0 | 0.8 (0.6) | 0.4 (0.2) | |
| Cephalothin | 1 | 5.6 (4.3) | 3.5 (2.3) | 8.5 (3.1) | 3.3 (0.5) | 7.6 (1.2) | 5.7 (1.3) |
| 2 | 6.0 (1.5) | 6.4 (5.4) | 10.7 (3.8) | 11.0 (4.9) | 3.9 (2.2) | 7.7 (1.9) | |
| Chloramphenicol | 1 | 10.3 (4.4) | 8.0 (1.8) | 9.3 (4.6) | 1.6 (1.6) | 5.1 (1.4) | 6.9 (1.6) |
| 2 | 6.0 (3.4) | 8.7 (5.8) | 12.0 (5.4) | 5.5 (2.9) | 10.2 (4.2) | 8.4 (2.1) | |
| Gentamicin | 1 | 6.4 (4.2) | 3.5 (1.5) | 1.7 (0.9) | 2.5 (0.9) | 2.5 (1.2) | 3.4 (1.0) |
| 2 | 2.0 (1.3) | 9.5 (4.2) | 6.0 (2.6) | 2.7 (1.2) | 6.2 (1.5) | 5.1 (1.3) | |
| Kanamycin | 1 | 64.3 (4.0) | 43.4 (3.7) | 65.2 (6.1) | 45.4 (5.5) | 48.3 (7.0) | 53.5 (3.6) |
| 2 | 28.0 (7.2) | 23.8 (9.0) | 35.3 (9.0) | 26.7 (8.8) | 20.3 (4.6) | 27.1 (3.7) | |
| Streptomycin | 1 | 56.4 (7.0) | 41.6 (5.7) | 56.4 (4.6) | 50.4 (2.5) | 64.1 (3.9) | 53.9 (3.1) |
| 2 | 25.3 (5.2) | 34.9 (5.8) | 34.7 (3.8) | 26.7 (5.4) | 17.2 (4.6) | 27.9 (2.6) | |
| Sulfamethoxazole | 1 | 72.2 (4.5) | 40.7 (8.3) | 62.7 (7.6) | 51.2 (1.9) | 49.2 (7.1) | 55.5 (4.2) |
| 2 | 38.0 (13.1) | 28.6 (11.1) | 40.0 (12.5) | 24.0 (5.8) | 18.8 (2.4) | 30.3 (4.9) | |
| Tetracycline | 1 | 98.4 (1.2) | 90.3 (1.3) | 98.3 (1.6) | 98.4 (1.2) | 95.8 (0.6) | 96.3 (1.0) |
| 2 | 98.7 (0.6) | 98.4 (1.1) | 96.7 (1.5) | 98.0 (2.0) | 94.8 (2.7) | 97.1 (0.8) | |
| Trimethoprim-sulfamethoxazole | 1 | 1.6 (0.6) | 1.8 (0.8) | 2.5 (1.8) | 2.5 (1.9) | 0.8 (0.9) | 1.8 (0.6) |
| 2 | 1.3 (0.8) | 1.6 (1.1) | 0.7 (0.6) | 1.4 (1.0) | 0.8 (0.8) | 1.1 (0.4) | |
Pansusceptible antimicrobials (amikacin, ceftriaxone, ciprofloxacin and nalidixic acid) were excluded from this table.
n1 and n2 represent the number of isolates in periods 1 and 2, respectively. For amikacin, only 592 and 685 isolates were tested in periods 1 and 2, respectively.
Only 594 isolates were tested for resistance to streptomycin in period 1.
Logistic regression models were not constructed for antimicrobials to which all E. coli isolates were susceptible (amikacin, ceftriaxone, ciprofloxacin, and nalidixic acid). Similarly, antimicrobials to which a very high proportion of isolates were either susceptible or resistant, including amoxicillin-clavulanic acid, cefoxitin, ceftiofur, gentamicin, tetracycline, and trimethoprim-sulfamethoxazole were not modeled because sparse data lead to unstable models. The logistic regression analysis was undertaken only for ampicillin, cephalothin, chloramphenicol, kanamycin, streptomycin, and sulfamethoxazole.
The treatment group was not significant in explaining resistance patterns for three of the six antimicrobials (ampicillin, kanamycin, and sulfamethoxazole, Table 4). The prevalence of resistance to kanamycin and sulfamethoxazole significantly decreased from period 1 to period 2 (P < 0.001 for both). Consequently, the ORs comparing period 2 to period 1 were protective for both kanamycin (OR = 0.32; confidence interval [CI] = 0.21 to 0.50) and sulfamethoxazole (OR = 0.35; CI = 0.22 to 0.50). The prevalence of resistance to ampicillin did not change from period 1 to period 2.
TABLE 4.
Logistic regression models for resistance of E. coli isolates to selected antimicrobials and the percentage of variation in resistance due to barn, pen, and samples
| Antimicrobiala | Model (logit)b | % of variation in resistance
|
||
|---|---|---|---|---|
| Barn | Pen | Sample | ||
| Ampicillin | Intercept only | 3.3 | 2.7 | 94.0 |
| Cephalothin* | −2.7 − 0.68(pulse-CTC) + 0.09(period 2) + 1.19(pulse-CTC*period 2) | 13.8 | 2.8 | 83.4 |
| Chloramphenicol* | −2.76 + 0.60(cont-tylosin) + 0.46(period 2) - 1.05(cont-tylosin*period 2) | 31.9 | 2.1 | 66.0 |
| Kanamycin | 0.14 − 1.13(period 2) | 9.5 | 1.8 | 88.7 |
| Streptomycin | 0.26 − 0.60(cont-CTC) − 1.29(period 2) + 1.01(cont-CTC*period 2) | ∼0.0 | 2.7 | 97.3 |
| Sulfamethoxazole | 0.22 − 1.066(period 2) | 14.4 | 2.0 | 83.6 |
*, only the interaction was significant at the 0.05 level. The main effects were kept in the model despite their lack of significance since their interaction was significant.
cont, continuous.
Dosing treatment, period, and their interaction were significant in the remaining 3 models (cephalothin, chloramphenicol, and streptomycin). The evaluation of the effect of the interaction must be made in comparison to the appropriate base level. For example, the interaction for cephalothin should be evaluated by comparing the resistance baseline for pens treated with pulse CTC in period 1 to the results for pens treated with pulse CTC in period 2. In this case, the E. coli from pigs treated with the pulse CTC were 3.6 (95% CI = 1.9 to 6.8) times as likely to be resistant to cephalothin in period 2 than they were in period 1. The interactions were evaluated in a similar manner for chloramphenicol and streptomycin. The OR for chloramphenicol (OR = 0.29, CI = 0.15 to 0.55) indicates that the continuous-tylosin treatment was protective and that chloramphenicol resistance prevalence decreased under continuous-tylosin treatment. For all other treatments (including the control), the resistance to chloramphenicol appeared to increase from period 1 to period 2 (Table 3). The results from the streptomycin model differ from those of the cephalothin and chloramphenicol models. The odds of being resistant to streptomycin in the continuous-CTC treatment was not significantly different from period 1 to period 2 (OR = 0.76; CI = 0.50 to 1.15). However, in comparison, the prevalence of resistance to streptomycin decreased significantly in all other treatments, including the control (Table 3).
E. coli antimicrobial resistance: MIC.
Proportional odds models were not constructed for ceftriaxone, ciprofloxacin, and kanamycin. The data for ceftriaxone and ciprofloxacin were predominantly from a single MIC, yielding unstable models. A proportional odds model was not constructed for kanamycin because the two MIC levels present in the data corresponded to susceptibility and resistance, which makes the analysis equivalent to the logistic model. Models for four antimicrobials—amoxicillin-clavulanic acid, ampicillin, nalidixic acid, and tetracycline—had no significant (P > 0.05) main effects (treatment and period) or interactions. The MIC distributions for the remaining nine antimicrobials, which had models with significant treatment, period, or interactions, are shown in Table 5. Five antimicrobials—amikacin (OR = 2.12, P < 0.0001), cefoxitin (OR = 2.3, P < 0.0001), ceftiofur (OR = 3.0, P < 0.0001), cephalothin (OR = 2.9, P < 0.0001), and chloramphenicol (OR = 0.66, P = 0.002)—exhibited only period effects. The ORs for amikacin, cefoxitin, ceftiofur, and cephalothin were >1.0, indicating that the MICs increased in these antimicrobials regardless of treatment. Two of the remaining antimicrobial models (sulfamethoxazole and trimethoprim-sulfamethoxazole) had both significant period (OR = 0.4, P < 0.0001 and OR = 0.5, P < 0.0001, respectively) and treatment effects (both P < 0.0001). Period effects were protective, which means that the MICs tended to decrease over time. Without an interaction term the treatment effects only represent differences in MICs in pen clusters that were assigned to the treatment and will not be discussed in further detail. The models for both gentamicin and streptomycin indicated that there were significant interactions (P = 0.003 and P = 0.0015, respectively) between periods, which represents differences in MICs between the treatment groups.
TABLE 5.
Percentage of E. coli isolates based on the MICs of nine antimicrobials that had either period, treatment, or period-treatment interactions in the proportional odds model
| Antimicrobial | Treatmenta | Period | % of isolates by MIC (μg/ml):
|
|||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.125 | 0.25 | 0.5 | 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128 | 256 | 512 | 1,024 | |||
| Amikacin | c-tyl | 1 | 7.1 | 46.8 | 37.3 | 7.1 | 1.6 | |||||||||
| 2 | 0.7 | 37.3 | 45.3 | 14.7 | 2.0 | |||||||||||
| c-ctc | 1 | 4.4 | 62.0 | 26.6 | 7.1 | 0 | ||||||||||
| 2 | 0.8 | 50.0 | 34.1 | 13.5 | 1.6 | |||||||||||
| p-tyl | 1 | 5.1 | 51.7 | 37.3 | 5.9 | 0 | ||||||||||
| 2 | 3.3 | 36.0 | 47.3 | 11.3 | 2.0 | |||||||||||
| p-ctc | 1 | 6.6 | 47.1 | 40.5 | 5.8 | 0 | ||||||||||
| 2 | 3.4 | 44.5 | 34.9 | 15.8 | 1.4 | |||||||||||
| None | 1 | 11.0 | 45.8 | 38.1 | 3.4 | 1.7 | ||||||||||
| 2 | 0 | 28.9 | 46.9 | 20.3 | 3.9 | |||||||||||
| Cefoxitin | c-tyl | 1 | 0 | 9.5 | 46.8 | 34.9 | 7.9 | 0 | 0.8 | |||||||
| 2 | 0 | 0 | 28 | 60.7 | 9.3 | 0.7 | 1.3 | |||||||||
| c-ctc | 1 | 0.9 | 6.2 | 48.7 | 32.7 | 8.0 | 0.9 | 2.6 | ||||||||
| 2 | 0 | 0 | 27.8 | 60.3 | 11.9 | 0 | 0 | |||||||||
| p-tyl | 1 | 0.8 | 5.1 | 46.6 | 40.7 | 6.8 | 0 | 0 | ||||||||
| 2 | 0 | 1.3 | 28.7 | 54.7 | 12.7 | 0.7 | 2.0 | |||||||||
| p-ctc | 1 | 0 | 5.0 | 48.8 | 35.5 | 10.7 | 0 | 0 | ||||||||
| 2 | 0 | 2.1 | 34.2 | 49.3 | 13.7 | 0.7 | 0 | |||||||||
| None | 1 | 0.8 | 2.5 | 44.9 | 44.1 | 7.6 | 0 | 0 | ||||||||
| 2 | 0 | 1.6 | 31.2 | 59.4 | 7.8 | 0 | 0 | |||||||||
| Ceftiofur | c-tyl | 1 | 28.6 | 63.5 | 6.4 | 0.8 | 0 | 0 | 0.8 | 0 | ||||||
| 2 | 7.3 | 77.3 | 14.0 | 0.7 | 0 | 0 | 0.7 | 0 | ||||||||
| c-ctc | 1 | 23.9 | 65.5 | 7.1 | 0 | 0.9 | 1.8 | 0.9 | 0 | |||||||
| 2 | 3.2 | 75.4 | 20.6 | 0 | 0 | 0 | 0 | 0.8 | ||||||||
| p-tyl | 1 | 25.4 | 66.1 | 8.5 | 0 | 0 | 0 | 0 | 0 | |||||||
| 2 | 10.7 | 70.0 | 16.0 | 2.7 | 0.7 | 0 | 0 | 0 | ||||||||
| p-ctc | 1 | 23.1 | 66.1 | 9.9 | 0.8 | 0 | 0 | 0 | 0 | |||||||
| 2 | 14.4 | 63.7 | 20.6 | 1.4 | 0 | 0 | 0 | 0 | ||||||||
| None | 1 | 22.9 | 67.8 | 9.3 | 0 | 0 | 0 | 0 | 0 | |||||||
| 2 | 4.7 | 73.4 | 19.5 | 1.6 | 0 | 0 | 0 | 0.8 | ||||||||
| Cephalothin | c-tyl | 1 | 15.1 | 37.3 | 33.3 | 8.7 | 3.2 | 2.4 | ||||||||
| 2 | 1.3 | 18.7 | 53.3 | 20.7 | 3.3 | 2.7 | ||||||||||
| c-ctc | 1 | 6.2 | 37.2 | 45.1 | 8.0 | 0 | 3.5 | |||||||||
| 2 | 1.6 | 19.0 | 57.9 | 15.1 | 3.2 | 3.2 | ||||||||||
| p-tyl | 1 | 11.9 | 36.4 | 35.6 | 7.6 | 4.2 | 4.2 | |||||||||
| 2 | 1.3 | 17.3 | 35.3 | 35.3 | 6.0 | 4.7 | ||||||||||
| p-ctc | 1 | 7.4 | 43.0 | 32.2 | 14.0 | 2.5 | 0.8 | |||||||||
| 2 | 4.1 | 23.3 | 37.7 | 24.0 | 10.3 | 0.7 | ||||||||||
| None | 1 | 3.4 | 34.8 | 44.9 | 9.3 | 5.1 | 2.5 | |||||||||
| 2 | 0.8 | 14.8 | 60.9 | 19.5 | 3.9 | 0 | ||||||||||
| Chloramphenicol | c-tyl | 1 | 7.9 | 54.8 | 22.2 | 4.8 | 7.9 | 2.4 | ||||||||
| 2 | 1.3 | 58.0 | 26.7 | 8.0 | 2.7 | 3.3 | ||||||||||
| c-ctc | 1 | 10.6 | 54.9 | 19.5 | 7.1 | 5.3 | 2.6 | |||||||||
| 2 | 10.3 | 57.9 | 22.2 | 0.8 | 1.6 | 7.1 | ||||||||||
| p-tyl | 1 | 12.7 | 55.9 | 17.8 | 4.2 | 6.8 | 2.5 | |||||||||
| 2 | 6.7 | 56.7 | 24.0 | 0.7 | 6.0 | 6.0 | ||||||||||
| p-ctc | 1 | 19.0 | 57.8 | 16.5 | 5.0 | 0.8 | 0.8 | |||||||||
| 2 | 4.8 | 58.9 | 29.4 | 1.4 | 2.7 | 2.7 | ||||||||||
| None | 1 | 11.0 | 62.7 | 19.5 | 1.7 | 4.2 | 0.8 | |||||||||
| 2 | 3.9 | 56.2 | 28.1 | 1.6 | 2.3 | 7.8 | ||||||||||
| Gentamicin | c-tyl | 1 | 26.2 | 59.5 | 7.1 | 0.8 | 0 | 0 | 4.0 | 2.4 | ||||||
| 2 | 15.3 | 63.3 | 16.0 | 3.3 | 0 | 0 | 2.0 | 0 | ||||||||
| c-ctc | 1 | 27.4 | 57.5 | 8.0 | 0.9 | 0.9 | 1.8 | 3.5 | 0 | |||||||
| 2 | 18.2 | 55.6 | 11.9 | 1.6 | 0.8 | 2.4 | 7.1 | 2.4 | ||||||||
| p-tyl | 1 | 36.4 | 50.0 | 11.9 | 0 | 0 | 0 | 0.8 | 0.8 | |||||||
| 2 | 19.3 | 63.3 | 6.7 | 2.0 | 0.7 | 2.0 | 2.7 | 3.3 | ||||||||
| c-ctc | 1 | 25.6 | 59.5 | 11.6 | 0.8 | 0 | 0 | 1.6 | 0.8 | |||||||
| 2 | 17.8 | 61.6 | 16.4 | 1.4 | 0 | 0 | 1.4 | 1.4 | ||||||||
| None | 1 | 38.1 | 50.8 | 5.9 | 0.8 | 0.8 | 0.8 | 0.8 | 1.7 | |||||||
| 2 | 4.7 | 62.5 | 22.7 | 3.9 | 0 | 0 | 0.8 | 5.5 | ||||||||
| Streptomycin | c-tyl | 1 | 43.7 | 20.6 | 35.7 | |||||||||||
| 2 | 74.7 | 10.7 | 14.7 | |||||||||||||
| c-ctc | 1 | 58.4 | 14.2 | 27.4 | ||||||||||||
| 2 | 65.1 | 7.9 | 27.0 | |||||||||||||
| p-tyl | 1 | 43.5 | 21.4 | 35.0 | ||||||||||||
| 2 | 65.3 | 18.0 | 16.7 | |||||||||||||
| p-ctc | 1 | 49.6 | 15.7 | 34.7 | ||||||||||||
| 2 | 73.3 | 6.2 | 20.6 | |||||||||||||
| None | 1 | 35.9 | 17.1 | 47.0 | ||||||||||||
| 2 | 82.8 | 3.9 | 13.3 | |||||||||||||
| Sulfamethoxazole | c-tyl | 1 | 23.0 | 0 | 0 | 0 | 4.8 | 32.5 | 39.7 | |||||||
| 2 | 56.7 | 0.7 | 0 | 0 | 4.7 | 11.3 | 26.7 | |||||||||
| c-ctc | 1 | 46.9 | 0 | 0 | 0 | 12.4 | 30.1 | 10.6 | ||||||||
| 2 | 69.8 | 0 | 0 | 0 | 1.6 | 11.9 | 16.7 | |||||||||
| p-tyl | 1 | 28.8 | 0 | 0 | 0 | 8.5 | 40.7 | 22.0 | ||||||||
| 2 | 58.0 | 0.7 | 0 | 0 | 1.3 | 18.0 | 22.0 | |||||||||
| p-ctc | 1 | 31.4 | 0.8 | 0 | 0 | 16.5 | 40.5 | 10.7 | ||||||||
| 2 | 74.0 | 0 | 0 | 0 | 2.1 | 12.3 | 11.6 | |||||||||
| None | 1 | 40.7 | 0 | 0 | 0.8 | 9.3 | 23.7 | 25.4 | ||||||||
| 2 | 81.2 | 0 | 0 | 0 | 0 | 3.1 | 15.6 | |||||||||
| Trimethoprim | c-tyl | 1 | 40.5 | 40.5 | 15.9 | 1.6 | 0 | 0 | 1.6 | |||||||
| 2 | 63.3 | 23.3 | 10.7 | 1.3 | 0 | 0 | 1.3 | |||||||||
| c-ctc | 1 | 57.5 | 25.7 | 13.3 | 0.9 | 0.9 | 0 | 1.8 | ||||||||
| 2 | 72.2 | 18.2 | 7.1 | 0.8 | 0 | 0 | 1.6 | |||||||||
| p-tyl | 1 | 55.1 | 33.9 | 8.5 | 0 | 0 | 0 | 2.5 | ||||||||
| 2 | 63.3 | 18.7 | 16.0 | 1.3 | 0 | 0 | 0.7 | |||||||||
| p-ctc | 1 | 54.6 | 28.1 | 10.7 | 4.1 | 0 | 0 | 2.5 | ||||||||
| 2 | 76.0 | 10.3 | 12.3 | 0 | 0 | 0 | 1.4 | |||||||||
| None | 1 | 71.2 | 19.5 | 7.6 | 0.8 | 0 | 0 | 0.8 | ||||||||
| 2 | 83.6 | 11.7 | 3.9 | 0 | 0 | 0 | 0.8 | |||||||||
Treatment coding: c-tyl, continuous tylosin; c-ctc, continuous chlortetracycline; p-tyl, pulse tylosin; p-ctc, pulse chlortetracycline; none, no antimicrobials.
For gentamicin, the MICs significantly increased in period 2 (P < 0.001, Table 6). However, relative to the control, the increase in MICs for the treatments was not as great. For streptomycin, the interpretation is similar except that MICs decreased substantially from the first to the second period. The treatments did not decrease as much as the control in their MICs from period 1 to period 2. This lack of a decrease in the treatments (or tendency toward less change from period 1 to period 2) was most notable for continuous CTC.
TABLE 6.
Proportional odds models for E. coli MIC responses for two antimicrobials
| Antimicrobial | Independent variable | Beta | CI
|
Pa | |
|---|---|---|---|---|---|
| Lower | Upper | ||||
| Gentamicin | Continuous tylosin | 0.56 | −0.01 | 1.13 | 0.0534 |
| Continuous CTC | 0.59 | −0.11 | 1.30 | 0.0989 | |
| Pulse tylosin | 0.18 | −0.59 | 0.95 | 0.6456 | |
| Pulse CTC | 0.56 | −0.11 | 1.22 | 0.0986 | |
| Period 2 | 1.70 | 1.21 | 2.19 | 0.0000* | |
| Continuous tylosin*period 2 | −1.12 | −1.77 | −0.47 | 0.0011* | |
| Continuous CTC*period 2 | −1.09 | −1.97 | −0.20 | 0.0167* | |
| Pulse tylosin*period 2 | −1.00 | −1.70 | −0.31 | 0.0055* | |
| Pulse CTC* period 2 | −1.21 | −2.00 | −0.43 | 0.0030* | |
| Streptomycin | Continuous tylosin | −0.34 | −0.85 | 0.16 | 0.1801 |
| Continuous CTC | −0.88 | −1.45 | −0.31 | 0.0020* | |
| Pulse tylosin | −0.30 | −0.98 | 0.38 | 0.3794 | |
| Pulse CTC | −0.56 | −1.14 | 0.02 | 0.0575 | |
| Period 2 | −2.09 | −2.72 | −1.46 | 0.0000* | |
| Continuous tylosin*period 2 | 0.84 | 0.03 | 1.66 | 0.0431* | |
| Continuous CTC*period 2 | 1.85 | 1.00 | 2.70 | 0.0001* | |
| Pulse tylosin*period 2 | 1.21 | 0.25 | 2.17 | 0.0144* | |
| Pulse CTC*period 2 | 1.15 | 0.24 | 2.06 | 0.0141* | |
*, significant at P = 0.05 level.
DISCUSSION
The potential benefits of any antimicrobial dosing regime must be judged not only in terms of their efficacy in disease prevention or other beneficial outcomes but also in terms of the impact the regimes may have on the prevalence of pathogens and antimicrobial resistance. This study focused on the impact of the regimes on the prevalence of Salmonella and resistance patterns in both Salmonella and E. coli.
The use of a high antimicrobial dose for a short period of time, pulse dosing, is a recent development that has several advantages over continuous low-level antimicrobial feeding (2). Pigs are removed from pens for marketing over a 4- to 5-week period. Under a continuous dosing regime, antimicrobials with a withholding requirement would need to be withheld from the whole pen for the entire last month of the growing period. In contrast, pulse dosing allows periods of time during which groups of animals can be marketed. Pulse dosing, by definition, implies recurrent periods of exposure and nonexposure of bacteria to the antimicrobials. Consequently, during the antimicrobial-free periods the selection pressure in favor of resistant isolates will be absent, which could allow for an environment that favors less-resistant phenotypes. Conversely, a resistant strain could be selected for in the initial exposure period, survive the intermittent period, and potentially have a competitive selection advantage in the second period.
The dosing regimes did not significantly influence the prevalence of Salmonella, but the prevalence of the pathogen significantly decreased from period 1 to period 2. However, numerically, the prevalence of Salmonella declined from period 1 to period 2 in all treatment groups that received antimicrobials, whereas the apparent prevalence in the control group, which was relatively low in both periods, did not decline. Ultimately, the lack of statistical significance may be a reflection of the power of the study. Several studies, including longitudinal studies, have demonstrated the variability of within-herd prevalence estimates for Salmonella (9, 11, 22), which could have affected the power of the present study.
As with the prevalence of Salmonella, the prevalence of resistance of Salmonella isolates did not appear to be influenced by dosing treatment. The second-period sampling was 9 weeks after the first sampling and 4 weeks after the second 2-week pulse. This resulted in a substantial period of time between the last antimicrobial treatment and the sampling date. Mathew et al. (15) noted that apramycin MICs in E. coli declined after removal of the antimicrobial from the feed.
The majority of Salmonella isolates in all treatments and periods was serovar Derby. Potentially, there could have been a shift in resistance patterns if the predominant serotypes changed due to dosing treatment. For example, 10 of 11 of the serovar Typhimurium isolates were pentaresistant (ampicillin, chloramphenicol, streptomycin, tetracycline, and sulfamethoxazole). If serovar Typhimurium had been selected for in any treatment group we might have seen a substantial increase in pentaresistance because none of the other serotypes had that specific phenotype. However, we were not able to detect any pattern that would suggest a serotype selection due to treatment.
Use of both the logistic models and the proportional odds analytical approaches allowed us to examine changes in either resistance or MICs for E. coli isolates that may have been associated with the antimicrobial dosing regime. Another advantage of using both models is that for 10 of the 16 antimicrobials we were unable to create logistic models, but for 8 of those 10 it was possible to construct a proportional odds model. For the five antimicrobials that were evaluated with both models (a proportional odds model was not constructed for kanamycin since the resistance data were equivalent to the MIC data), both analytical procedures gave similar results for ampicillin and streptomycin. Treatment was significant in the logistic model for sulfamethoxazole but not in the proportional odds model. Cephalothin and chloramphenicol both had significant interactions in the logistic model that were not present in the proportional odds models.
Cephalothin was the only antimicrobial against which a significant increase in resistance was identified for the pulse CTC group compared to the other treatment groups and the control group. The biological mechanism to explain this is unknown. Cephalothin is one of the narrow-spectrum cephalosporins which are beta-lactam antimicrobials that inhibit cell wall development, whereas CTC is a tetracycline that is a broad-spectrum antimicrobial that inhibits protein synthesis inside the cell. The distinct difference between these two classes of antimicrobials suggests that this is a coincident observation. Further work at the molecular level is warranted.
There has been concern about commensal E. coli serving as a reservoir of resistance genes for pathogens such as Salmonella. In the present study, we demonstrated that resistance by E. coli to the tested antimicrobials was common. However, this resistance was always mirrored by the resistance patterns seen in Salmonella spp. from the same treatment group. Assessment of the true sharing of resistance genes among the population would best be accomplished through molecular evaluation of isolates.
The present study has demonstrated that the two dosing regimes using two different antimicrobials did not result in increased prevalence of Salmonella or the prevalence of resistance to a number of antimicrobials for Salmonella or E. coli. However, given recent studies demonstrating the small impact of nontherapeutic doses of antimicrobials on swine growth parameters, producers must weigh the costs and benefits for nontherapeutic administration of antimicrobials.
Footnotes
Published ahead of print on 25 January 2008.
REFERENCES
- 1.Blake, D. P., K. Hillman, D. R. Fenlon, and J. C. Low. 2003. Transfer of antimicrobial resistance between commensal and pathogenic members of the Enterobacteriaceae under ileal conditions. J. Appl. Microbiol. 95:428-436. [DOI] [PubMed] [Google Scholar]
- 2.Burch, D. 1990. Pulse dosing versus pneumonia. Pig Int. 1990:20-24. [Google Scholar]
- 3.Cromwell, G. L. 2002. Why and how antimicrobials are used in swine production. Anim. Biotechnol. 13:7-27. [DOI] [PubMed] [Google Scholar]
- 4.Davies, J. 1994. Inactivation of antimicrobials and the dissemination of resistance genes. Science 264:375-382. [DOI] [PubMed] [Google Scholar]
- 5.Docic, M., and G. Bilkei. 2003. Differences in antimicrobial resistance in Escherichia coli, isolates from East-European swine herds with or without prophylactic use of antimicrobials. J. Vet. Med. B Infect. Dis. Vet. Public Health 50:27-30. [DOI] [PubMed] [Google Scholar]
- 6.Dritz, S. S., M. D. Toakch, R. D. Goodband, and J. L. Nessen. 2002. Effects of administration of antimicrobials in feed on growth rate and feed efficiency of pigs in multisite production systems. J. Am. Vet. Med. Assoc. 220:1690-1695. [DOI] [PubMed] [Google Scholar]
- 7.Ebner, P. D., and A. G. Mathre. 2000. Effects of antimicrobial regimens on the fecal shedding patterns of pigs infected with Salmonella Typhimurium. J. Food Prot. 63:709-714. [DOI] [PubMed] [Google Scholar]
- 8.Fey, P. D., T. J. Safranek, M. E. Rupp, E. F. Dunne, E. Ribot, P. C. Iwen, P. A. Bradford, F. J. Angulo, and S. H. Hinrichs. 2000. Ceftriaxone-resistant salmonella infection acquired by a child from cattle. N. Engl. J. Med. 342:1242-1249. [DOI] [PubMed] [Google Scholar]
- 9.Funk, J. A., P. R. Davies, and M. A. Nichols. 2001. Longitudinal study of Salmonella enterica in growing pigs reared in multiple-site swine production systems. Vet. Microbiol. 83:45-60. [DOI] [PubMed] [Google Scholar]
- 10.Gebreyes, W. A., S. Thakur, and W. E. Morrow. 2006. Comparison of prevalence, antimicrobial resistance, and occurrence of multidrug-resistant Salmonella in antimicrobial-free and conventional pig production. J. Food Prot. 69:743-748. [DOI] [PubMed] [Google Scholar]
- 11.Hurd, H. S., J. D. McKean, R. D. Griffith, and M. H. Rostagno. 2004. Estimation of the Salmonella enterica prevalence in finishing swine. Epidemiol. Infect. 132:127-135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kalpana Gupta, M. P. H., T. M. Hooton, and W. E. Stamm. 2001. Increasing antimicrobial resistance and the management of uncomplicated community-acquired urinary tract infections. Ann. Intern. Med. 135:41-50. [DOI] [PubMed] [Google Scholar]
- 13.Kazuki, H., T. Asai, A. Kojima, K. Ishihara, and T. Takahashi. 2006. Role of coresistance in the development of resistance to chloramphenicol in Escherichia coli isolated from sick cattle and pigs. Am. J. Vet. Res. 67:230-235. [DOI] [PubMed] [Google Scholar]
- 14.Mathew, A. G., M. A. Beckmann, and A. M. Saxton. 2001. A comparison of antibiotic resistance in bacteria isolated from swine herds in which antibiotics were used or excluded. J. Swine Health Prod. 9:125-129. [Google Scholar]
- 15.Mathew, A. G., K. N. Garner, P. D. Ebner, A. M. Saxton, R. E. Clift, and S. Liamthong. 2005. Effects of antimicrobial use in sows on resistance of Escherichia coli and Salmonella enterica Typhimurium in their offspring. Foodborne Pathog. Dis. 2:212-220. [DOI] [PubMed] [Google Scholar]
- 16.Mead, P. S., L. Slutsker, V. Dietz, L. F. McCaig, J. S. Bresee, C. Shapiro, P. M. Griffin, and R. V. Tauxe. 1999. Food-related illness and death in the United States. Emerg. Infect. Dis. 5:607-625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Morley, P. S., M. D. Apley, T. E. Besser, D. P. Burney, P. J. Fedorka-Cray, M. G. Papich, J. L. Traub-Dargatz, and J. S. Weese. 2005. Antimicrobial drug use in veterinary medicine. J. Vet. Intern. Med. 19:617-629. [DOI] [PubMed] [Google Scholar]
- 18.Moubareck, C., N. Bourgeois, P. Courvalin, and F. Doucet-Populaire. 2003. Multiple antimicrobial resistance gene transfer from animal to human enterococci in the digestive tract of gnotobiotic mice. Antimicrob. Agents Chemother. 47:2993-2996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.NARMS. 2007. National Antimicrobial Resistance Monitoring System-Enteric Bacteria 2003 Executive Report. [Online.] http://www.fda.gov/cvm/Documents/NARMSExecSum03.pdf.
- 20.NCCLS. 2002. Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria isolated from animals. Approved standard; 2nd edition, M31-A2. National Committee for Clinical Laboratory Standards, Wayne, PA.
- 21.NCCLS. 2002. Performance standards for antimicrobial susceptibility testing; 8th informational supplement, M100-S12. National Committee for Clinical Laboratory Standards, Wayne, PA.
- 22.Nelson, J. M., K. E. Smith, D. J. Vugia, T. Rabatsky-Ehr, S. D. Segler, H. D. Kassenborg, S. M. Zansky, K. Joyce, N. Marano, R. M. Hoekstra, and F. J. Angulo. 2004. Prolonged diarrhea due to ciprofloxacin-resistant campylobacter infection. J. Infect. Dis. 190:1150-1157. [DOI] [PubMed] [Google Scholar]
- 23.Nijsten, R., N. London, A. van den Bogaard, and E. Stobberingh. 1996. Antibiotic resistance among Escherichia coli isolated from faecal samples of pig farmers and pigs. J. Antimicrob. Chemother. 37:1131-1140. [DOI] [PubMed] [Google Scholar]
- 24.Rajic, A., J. Keenliside, M. E. McFall, A. E. Deckert, A. C. Muckle, B. P. O'Connor, K. Manninen, C. E. Dewey, and S. A. McEwen. 2005. Longitudinal study of Salmonella species in 90 Alberta swine finishing farms. Vet. Microbiol. 105:47-56. [DOI] [PubMed] [Google Scholar]
- 25.Schroeder, C. M., C. Zhao, C. DebRoy, J. Torcolini, S. Zhao, D. G. White, D. D. Wagner, P. F. McDermott, R. D. Walker, and J. Meng. 2002. Antimicrobial resistance of Escherichia coli O157 isolated from humans, cattle, swine, and food. Appl. Environ. Microbiol. 68:576-581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Summers, A. O. 2002. Generally overlooked fundamentals of bacterial genetics and ecology. Clin. Infect. Dis. 34(Suppl. 3):S85-S92. [DOI] [PubMed] [Google Scholar]
- 27.Sunde, M., K. Fossum, A. Solberg, and H. Sorum. 1998. Antimicrobial resistance in Escherichia coli of the normal intestinal flora of swine. Microb. Drug Resist. 4:289-299. [DOI] [PubMed] [Google Scholar]
- 28.Wells, S. J., P. J. Fedorka-Cray, D. A. Dargartz, K. Ferris, and A. Green. 2001. Fecal shedding of Salmonella spp. by dairy cows on farm and at cull cow markets. J. Food Prot. 64:3-11. [DOI] [PubMed] [Google Scholar]
- 29.Winokur, P. L., D. L. Vonstein, L. J. Hoffman, E. K. Uhlenhopp, and G. V. Doern. 2001. Evidence for transfer of CMY-2 AmpC beta-lactamase plasmids between Escherichia coli and Salmonella isolates from food animals and humans. Antimicrob. Agents Chemother. 45:2716-2722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Witte, W. 1998. Medical consequences of antimicrobial use in agriculture. Science 279:996-997. [DOI] [PubMed] [Google Scholar]
- 31.World Health Organization. 2000. WHO global principles for the containment of antimicrobial resistance in animals intended for food, p. 5-9. World Health Organization, Geneva, Switzerland.
- 32.World Health Organization. 2002. Antimicrobial resistance. Fact sheet 194. World Health Organization, Geneva, Switzerland. http://www.who.int/mediacentre/factsheets/fs194/en/.
- 33.World Organization for Animal Health (Office International des Epizootes). 2001. Antimicrobial resistance. O. I. E. Rev. Sci. Tech. 20:797-870. [Google Scholar]
- 34.Zhao, S., P. J. Fedorka-Cray, S. Friedman, P. F. McDermott, R. D. Walker, S. Qaiyumi, S. L. Foley, S. K. Hubert, S. Avers, L. English, D. A. Dargatz, B. Salamone, and D. G. White. 2005. Characterization of Salmonella Typhimurium of animal origin obtained from the National Antimicrobial Resistance Monitoring System. Foodborne Pathog. Dis. 2:169-181. [DOI] [PubMed] [Google Scholar]