OUP user menu

Persistence of vancomycin-resistant enterococci (VRE) on Norwegian broiler farms

Katrine Borgen, Marit Sørum, Hilde Kruse, Yngvild Wasteson
DOI: http://dx.doi.org/10.1111/j.1574-6968.2000.tb09348.x 255-258 First published online: 1 October 2000


Five Norwegian broiler farms previously identified as housing broilers carrying vancomycin-resistant enterococci (VRE) were examined for the presence of VRE 4 years after avoparcin was banned. Environmental samples were obtained from empty, cleaned broiler houses. Faecal samples were collected weekly from the flock housed after the environmental sampling. The hatchery from where the chicks originated was also sampled. VRE were found to be present in the farm environment after depopulation and cleanup of the broiler houses. Within 3 weeks after introduction to the farm, all broiler flocks tested positive for VRE. VRE were not isolated from the hatchery.

  • Vancomycin-resistant enterococcus
  • Broiler farm
  • Avoparcin
  • Persistence

1 Introduction

Vancomycin-resistant enterococci (VRE) were first detected outside hospital environments in 1993 [1]. Soon after, an association between the use of avoparcin as a feed additive and occurrence of VRE in the animal husbandry environment was documented [24]. Because of possible health hazards related to VRE outside hospitals, avoparcin has been prohibited in many European countries, and was finally withdrawn from the market by its manufacturer, Roche Vitamins, in early 2000. In Norway, avoparcin was used as a feed additive for broilers and turkeys from 1986 until it was banned in 1995. The poultry production is based on an ‘all in–all out’ practice to prevent transmission of microbiological agents between flocks. All the birds in a flock are slaughtered at the same time. Subsequently, the house is cleaned, disinfected and left empty for at least 2–3 weeks before a new flock is introduced. This farm management procedure combined with the ban of avoparcin was expected to gradually reduce the prevalence of VRE in poultry production. Studies from Denmark [5], Germany [6], and Italy [7] have shown a decrease in the VRE prevalence in animal production after avoparcin was banned in the respective countries. In Norway, however, the prevalence of VRE in poultry has remained unchanged from 1995 to 1998 [8]. The aim of the present work was to study if the VRE reservoir previously identified originated from the hatchery, the feed or the farms themselves.

2 Materials and methods

2.1 Collection of samples

During the spring and summer of 1999, environmental sampling was performed on five broiler farms previously identified as housing VRE-carrying broilers and in the hatchery from which all the five farms receive their day old chicks. From the farms, samples of clean shavings, foodstuff, and various materials from inside the empty broiler house, as well as manure and soil from the surrounding environment were obtained. The material collected at the hatchery included eggshell, litter, dust, and a dead chick. In addition, the floors, walls, fixtures, and equipments in both the broiler houses and the hatchery were sampled by the use of swabs and selective contact agar plates. From each of the five farms, 57–66 samples were obtained whereas 41 samples were obtained from the hatchery. All environmental samples were brought immediately to the laboratory for analysis. From the broiler flock housed after the environmental sampling, faecal samples were collected weekly throughout the birds’ 5 weeks life span. The farmers were instructed on how to fill sterile tubes with fresh faecal material from the floor in the broiler house, and the samples were sent by mail without delay to the laboratory. All faecal samples were kept frozen at −80°C prior to analysis.

2.2 Questionnaire

At the time of the environmental sampling, the farmers were interviewed using a questionnaire, in which the questions were grouped into four main categories: farm management, feed and water supply, hygienic measures, and building constructions.

2.3 Isolation of VRE

2.3.1 Environmental samples

Collected material and swab samples obtained from the five farms and the hatchery were added to Azide dextrose broth (Oxoid, Bastingstoke, Hampshire, England) and incubated at 37°C overnight for enrichment. Subsequently, 0.1 ml material was spread onto Slanetz and Bartley enterococcus agar plates (SB) (Oxoid) with vancomycin (Sigma, St. Louis, MO, USA). For the samples from the farms, plates containing 32 mg l−1 vancomycin were used (SBv2), whereas for the hatchery samples, plates containing 6 mg l−1 vancomycin were used (SBv1). SBv2 contact agar plates were used for direct isolation of VRE from the farms, whereas SB contact agar plates without vancomycin were used for sampling the hatchery. Typical enterococcal colonies on SB were transferred to SBv1. All SB, SBv1, and SBv2 plates were incubated at 37°C for 48 h.

2.3.2 Faecal samples

The faecal samples were analysed without enrichment. Five grams from each sample were dissolved in 45 ml peptone water (Difco, Detroit, MI, USA). From the suspended material, 0.1 ml was plated onto SBv2 and incubated as described above.

From both the environmental and faecal samples, typical colonies from SBv1 and SBv2 were subcultured onto blood agar plates (Difco), and one colony from each positive sample was selected for species identification and testing of susceptibility to vancomycin.

2.4 Species identification

Presumptive identification of Enterococcus spp. was performed on the basis of typical colony morphology, Gram stain, and absence of catalase production. Identification to species level was performed by the RAPID ID32 Strep test kit (bioMérieux, Marcy-l'Etoile, France) and further confirmed by species specific PCRs targeting the ddl genes of Enterococcus faecium and Enterococcus faecalis and the vanC1 gene of Enterococcus gallinarum[9].

2.5 Antimicrobial susceptibility testing and vanA PCR

The minimal inhibitory concentration (MIC) of vancomycin was determined using Etest (AB Biodisk, Solna, Sweden). The breakpoints applied were according to the National Committee of Clinical Laboratory Standards’ (NCCLS’) guidelines (M100-S9, January 1999). Vancomycin-resistant isolates were examined for the presence of the vanA gene by PCR [10]. A selection of vanA-carrying isolates was tested for antimicrobial susceptibility to nine antimicrobials by disk diffusion using Neo Sensitabs™ according to the instructions of the manufacturer (Rosco, Taastrup, Denmark). The following antimicrobials were included: ampicillin, bacitracin, chloramphenicol, erythromycin, gentamicin, spiramycin, streptomycin, tetracycline, and virginiamycin. The isolates were categorised as sensitive, intermediate, or resistant (S, I or R), and those interpreted as either I or R were selected for further MIC determination using Etest (AB Biodisk). Positive and negative control strains were included in all experiments as previously described [8].

3 Results

3.1 Isolation and identification of VRE

VRE were not isolated from any of the 41 samples from the hatchery. As for the five broiler farms, VRE were isolated from 0/29, 6/30, 15/28, 24/30, and 29/30 of the contact agar plate samples, and from 10/32, 19/32, 17/29, 28/33, and 30/36 of the environmental samples and swabs. For two broiler flocks, VRE were not isolated from the first, and the first and second faecal samples, respectively. Otherwise, VRE were isolated from all faecal samples of the five flocks. The results are summarised in Table 1. As only one VRE isolate from each positive sample was selected, a total of 178 isolates from the environmental and 22 isolates from the faecal samples were identified and analysed for the presence of vanA. Of the 178 environmental isolates, 135 were identified as E. faecium and 43 as Enterococcus hirae. Of the 22 faecal isolates, 18 were identified as E. faecium, three as E. hirae, and one as Enterococcus durans. For each farm, ten environmental isolates and one faecal isolate were selected for further antimicrobial susceptibility testing (55 isolates in total).

View this table:
Table 1

Proportion of environmental samples from five broiler farms positive for vancomycin-resistant enterococci (VRE), isolation of VRE from weekly obtained poultry faecal samples, and information on some hygienic precautions on these farms

No. of VRE positive samples/no. of environmental samplesVRE in 5 g faecal samplesHygienic precautions
FarmDirect isolationaIsolation after enrichmentbWeekHigh pressure cleaningDisinfectionHygiene barrier in use
  • aContact agar plates on walls, floors, and fixtures inside the broiler house.

  • bSwabs from walls, floors, and fixtures inside the broiler house, and samples mainly from feed, shavings, manure, and soil.

3.2 Antimicrobial susceptibility

The VRE isolates harboured the vanA gene as shown by PCR. No resistance to ampicillin or aminoglycosides was observed. Twenty-nine isolates were categorised as I or R to erythromycin and/or tetracycline by disk diffusion. The MIC values determined by Etest are presented in Fig. 1.

Figure 1

Distribution of minimal inhibitory concentration (MIC) values for erythromycin (EM) and tetracycline (TC) in 29 vancomycin-resistant enterococci (VRE) isolates. The MIC values were determined using Etest (AB Biodisk) according to the manufacturers recommendations.

3.3 Questionnaire

The five farmers practised the ‘all in–all out’ management to various degrees and the differences were mainly in disinfection and use of a hygiene barrier at the entrance of the broiler house. Some hygienic precautions applied are summarised in Table 1. The broiler houses varied from old restored wooden houses to modern concrete buildings made for broiler production. The flock size ranged from 4500 to 18 000 birds. The farms had poultry production only, except for one farm, which also had about 30 sheep.

4 Discussion

This study shows that VRE persist in the poultry farm environment in the absence of avoparcin as a selective force, even after depopulation and cleanup of the broiler houses, and for some farms, use of thorough hygienic measures. As enterococci comprise a part of the normal intestinal microflora both in poultry and other animals and are naturally tolerant to heat treatment, elimination of enterococci from the empty broiler houses is unlikely.

In the present study, VRE were isolated from the feeding machinery and feed troughs inside the broiler houses, whereas no VRE could be isolated from the feed obtained from feed mill samples. The bulk feed storage bin and feeding machinery inside the broiler houses might serve as an important source for VRE, since avoparcin initially was introduced to the farms through the feed.

To our knowledge, no other studies on survival of VRE in animal husbandry environments have so far been published. Survival of Salmonella enteritidis in poultry environments has, however, been studied by Davies and Wray. They showed that S. enteritidis persisted for at least one year inside poultry houses and that this persistence was associated with dust particles in food troughs, as well as with incomplete disinfection and cleansing procedures, and presence of rodents and wild bird populations on the farms [11,12]. The present study revealed incomplete cleansing and disinfection procedures on three of the five farms, and the walls and floors inside the broiler houses often had a rough surface where microorganisms easily hide. The survival of S. enteritidis and VRE in poultry environments is not directly comparable. Although both having an enteric reservoir there are major differences in cell wall composition and resistance to external environmental factors between the two bacterial species.

A recent Norwegian study of competitive fitness of isogenic vancomycin-resistant and vancomycin-sensitive E. faecium strains showed that carriage of the vanA genes did not reduce the fitness of the vancomycin-resistant strains enough for the vancomycin-sensitive strains to have a clear competitive advantage [13]. An environmental adaptation of VRE strains on Norwegian broiler farms could partially explain the observed persistence despite removal of avoparcin. Co-selection due to other antimicrobial agents is unlikely, as most of the VRE strains isolated were susceptible to all antimicrobial agents tested. Other selective genes carried on plasmids may influence on the persistence of vanA-carrying enterococci.

No VRE was isolated from the environmental samples from the hatchery delivering day old chicks to the farms, and a vertical transmission of VRE is unlikely. Even though a limited number of samples were taken, analysis for VRE in the hatchery was performed with 6 mg l−1 vancomycin, a minimal selective force. These results confirm the association between previous avoparcin usage and occurrence of VRE [24], as the hatchery has never been exposed to avoparcin.

The persistence of VRE in the broiler farm environment shown in this study is the most likely explanation for the continuing high prevalence of VRE seen in Norwegian poultry production after removal of avoparcin from the feed [8]. According to our results, the broiler chicks are colonised by VRE soon after arrival on the farm, although the study is limited and general conclusions concerning the epidemiology of VRE should not be drawn. From a public health point of view, the reservoir of vanA-VRE present on broiler farms is a reason for concern, and strategies should be developed to diminish this population of VRE.


The authors would like to thank the participating farmers for collaboration in collection of samples. This research work was financially supported by The Research Council of Norway, grant no. 117130/112.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
View Abstract