Molecular and clinical epidemiology of vancomycin-resistant Enterococcus faecalis

Journal of Antimicrobial Chemotherapy (2004) 53, 626 630 DOI: 10.1093/jac/dkh138 Advance Access publication 18 February 2004 Molecular and clinical epidemiology of vancomycin-resistant Enterococcus faecalis Simona F. Oprea 1, Najam Zaidi 1, Susan M. Donabedian 1, Mamtha Balasubramaniam 1, Ellie Hershberger 1 and Marcus J. Zervos 1,2 * 1 William Beaumont Hospital, Royal Oak, MI; 2 Wayne State University School of Medicine, Detroit, MI, USA Received 25 November 2003; returned 19 December 2003; revised 6 January 2004; accepted 7 January 2004 Objectives: With the recent emergence of vancomycin-resistant (VR) Staphylococcus aureus, subsequent to the suggested transfer of the vana resistance gene from Enterococcus faecalis, we sought to determine risk factors for acquisition of VR E. faecalis and to evaluate the molecular epidemiology of this less-prevalent and less-studied species of VR enterococcus. Methods: We compared clinical isolates of VR E. faecalis from 71 patients, collected over 12 years in a large community teaching hospital, with isolates from 126 patients with vancomycin-susceptible E. faecalis. Results: Risk factors for VR E. faecalis acquisition by multivariate analysis were nursing home residence (P = 0.0005), haemodialysis (P = 0.009), decubitus ulcers (P = 0.03) and receipt of parenteral vancomycin (P = 0.0002). Twenty-one percent of VR E. faecalis demonstrated vana and 79% vanb resistance. The number of VanA isolates increased over time. Molecular analysis showed vana or vanb in multiple PFGE groups. Conclusions: The results of this study suggest gene dissemination among some isolates and intra-hospital spread of other isolates. The risk factors identified clearly suggest that VR E. faecalis is a nosocomial pathogen and should be considered in infection control practices. Further surveillance of VR E. faecalis is warranted, due to the potential spread of vancomycin resistance among enterococci and staphylococci. Keywords: risk factors, pulse-field gel electrophoresis, E. faecalis Introduction Regarded for a long time as pathogens of low virulence, enterococci have emerged in more recent years as important nosocomial pathogens, causing significant morbidity and mortality. They possess a remarkable capacity to exchange genetic information among themselves and with other genera, which enables them to readily acquire, accumulate and share antibiotic resistance genes. In June 2002, the first clinical isolate of vancomycin-resistant Staphylococcus aureus (VRSA) was identified in a patient from Michigan. 1 Importantly, an Enterococcus faecalis isolate containing a vana gene identical to that of the VRSA strain was isolated, suggesting transfer of the vana gene from E. faecalis to S. aureus. 1,2 Vancomycin resistance is most commonly found in Enterococcus faecium, but it is increasingly seen in the more predominant enterococcal species in humans, E. faecalis. 3 Although the epidemiology of E. faecium is well described, 4 13 little is known about the epidemiology of vancomycin-resistant (VR) isolates of E. faecalis. In this study, we therefore compared the molecular relatedness of VR E. faecalis clinical isolates from a large community teaching hospital in southeastern Michigan, an area where VR E. faecium has been endemic for years 7,9 and from where the patient with VRSA was identified. 1,2 We used PCR analysis and PFGE, in an effort to gain better insight into the epidemiology and mechanisms of vancomycin resistance in E. faecalis. We also conducted a case-control study to identify clinical risk factors associated with the acquisition of VR E. faecalis colonization or infection. Gaining insight into the epidemiology and risk factors associated with VR E. faecalis is essential for designing more effective infection control measures. With the recently recognized transfer of vancomycin resistance genes from E. faecalis to S. aureus, understanding of the epidemiology of VR E. faecalis is urgently needed.... *Correspondence address. Wayne State University School of Medicine, William Beaumont Hospital, Research Institute, 3811 W 13 Mile Road, Royal Oak, MI 48073, USA. Tel: +1-248-551-0419; Fax: +1-248-551-5069; E-mail: mzervos@beaumont.edu... 626 JAC vol.53 no.4 The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.

Epidemiology of vancomycin-resistant Enterococcus faecalis Materials and methods The study was carried out at a 999 bed tertiary care community teaching hospital located in suburban Detroit. Approval from the institutional review board was obtained before initiation of the study. All adult patients infected or colonized with VR E. faecalis, irrespective of site, were included as cases. If the patients were infected or colonized with the same strain isolated from the same and/or separate site at different points in time, the first event was considered for risk-factor analysis. Duplicate isolates (from the same patient and as determined by PFGE type) were excluded from the risk factors analysis. Controls were randomly selected from a database of patients infected or colonized with vancomycin-susceptible (VS) E. faecalis, and matched for length of hospital stay with cases. Controls were excluded from analysis if evidence of colonization or infection with VR E. faecalis or VR E. faecium prior to isolation of VS E. faecalis was found. Through retrospective chart review, we collected data on patient demographic characteristics, underlying diseases, source of VR E. faecalis or VS E. faecalis, Acute Physiology and Chronic Health Evaluation (APACHE) II scores, antimicrobial use, receipt of immunosuppressive drugs, presence of invasive devices at the time of VR E. faecalis or VS E. faecalis isolation, nursing home residence, location in the hospital, including intensive care unit (ICU) stay, co-infection with methicillinresistant S. aureus (MRSA), multidrug-resistant Gram-negative bacilli (MDR GNB) or C. difficile, and recent surgeries or procedures. The presence of potential risk factors was assessed over the 3 month period preceding isolation of enterococci, with the exception of type and duration of specific antimicrobial therapy, which was recorded during the preceding 2 weeks. 14 Data regarding the following antibiotics were collected: parenteral and oral vancomycin, penicillins (including penicillin/ β-lactamase inhibitor combinations), cephalosporins, fluoroquinolones, aminoglycosides, clindamycin, metronidazole, aztreonam and carbapenems. Immunosuppressive drugs included cyclosporin, cyclophosphamide, azathioprine, tacrolimus, methotrexate, steroids (the equivalent of more than 10 mg of prednisolone a day) and cytotoxic chemotherapy. Bacterial strains Clinical isolates of VR E. faecalis identified at a single hospital during 1991 October 2002 were analysed. Isolates were identified from the Microbiology Laboratory database and confirmed as E. faecalis by using standard biochemical reactions. 15 Susceptibility of isolates to vancomycin (Eli Lilly and Co., Indianapolis, IN, USA) and ampicillin (Sigma, St. Louis, MO, USA) was determined by a microdilution assay in accordance with NCCLS recommendations. 16 The presence of β-lactamase was detected by nitrocefin (Glaxo, Middlesex, UK). PFGE and PCR analysis Genomic DNA was prepared in agarose plugs, digested with SmaI (New England BioLabs, Beverly, MA, USA) and electrophoresed on a CHEF- DRIII (Bio-Rad Laboratories, Richmond, CA, USA), as described previously. 4 Isolates with up to six band differences were placed in strain groups using published consensus criteria for strain delineation. 17 PCR was performed using published methods. 18 Statistical analysis All categorical variables were analysed by contingency tables using either the Two-Tailed Fisher s Exact tests or the Cochran Mantel Haenzel tests wherever appropriate. All continuous variables were analysed using the Two-Tailed Student s t-tests. All P values less than α of 0.05 (probability of type I error) were considered to be significant throughout this study. Risk factors identified in the univariate analysis were included in multivariate stepwise (forward) analysis to determine independent risk factors for acquiring VR E. faecalis as opposed to VS E. faecalis. Statistical analyses were performed using the SAS System for Windows version 8.2. Results Results of susceptibility testing Eighty-three clinical isolates of VR E. faecalis from 74 patients were identified during February 1991 October 2002. Isolates were from blood (21), urine (39), wounds (10), intra-abdominal sources (6), catheter tip (4), respiratory secretions (1), joints (1) and stool (1). Vancomycin MICs were uniformly 256 mg/l for VanA isolates and 16 512 mg/l for VanB isolates. Eight VanB isolates had vancomycin MICs of 16 mg/l. Among VR E. faecalis isolates, 96% were susceptible to ampicillin (one resistant isolate in 1995, one in 1999; susceptibility for one isolate was unavailable), whereas 99% of VS E. faecalis were susceptible to ampicillin (one resistant isolate in 2001). None produced β-lactamase. PCR and PFGE results Seventeen isolates (20%) demonstrated vana and 66 (80%) vanb resistance. The 17 VanA isolates were placed in seven PFGE groups (A1-7) with one group (A1) containing 53% of the isolates. The 66 VanB isolates were placed in 16 PFGE groups (B1-B16) with one group (B1) containing 58% of the isolates (Table 1). One isolate positive for the vanb gene had the same PFGE pattern as the isolates in the largest VanA PFGE group and was negative for the vana gene. This isolate also had a much lower MIC (16 mg/l) compared with the MIC of the vana containing isolates ( 256 mg/l) in this group. There was an increasing number of VR E. faecalis isolates over time, as well as an increase in recovery of VanA isolates (Table 1). All isolates recovered before 1999 carried the vanb resistance gene. After 1999, VanB isolates continued to predominate; however, in 2002 the number of VanA isolates (45%) approached that of VanB isolates (55%). Four of the VanA isolates belonged to unique PFGE groups (A2, 4, 5, 6). Strain types A3 and A7 were isolated from two different patients each, one 2 days and one 8 days apart. Although these isolates were clonally related, the band pattern of the two strains in each of these groups was not identical (A3a and A3b, and A7a, A7b, respectively). The remaining nine VanA isolates belonged to PFGE type A1. They were all isolated during October 2001 May 2002. Among the 66 VanB isolates, eight were unique (B2, 4, 9 14). Nineteen VanB isolates were divided between six PFGE groups. The remaining 38 VanB isolates all belonged to one group (B1), and were recovered in our hospital during 1995 2002. Of note, 14 of the 16 VanB isolates from 2001 belonged to the same PFGE group (B1). Three patients were each colonized or infected with more than one strain of VR E. faecalis, some were also associated with different MICs, sometimes as close as 2 weeks apart (range 14 113 days). Four patients were each colonized or infected with more than one strain of VR E. faecalis with the same MICs and PFGE type, although 1 2 band variations existed between strains. Results of case-control study A total of 71 cases (patients with VR E. faecalis) and 126 controls (patients with VS E. faecalis) were included in this part of the analysis. Cases and controls were comparable in age, gender distribution, source of specimens, severity of illness as assessed by APACHE II 627

S. F. Oprea et al. Table 1. Summary of PFGE and PCR analysis of vancomycin-resistant E. faecalis strains PFGE type No. isolates Year(s) Resistance Dissemination A1 9 2001 2002 vana clonal, intra-hospital A2 1 1999 vana unique A3 2 2002 vana intra-hospital A4 1 2002 vana unique A5 1 2002 vana unique A6 1 2002 vana unique A7 2 2002 vana intra-hospital Seven groups 17 (20%) VanA B1 38 1995, 1996, 1998, vanb clonal, intra-hospital 2000 2002 B2 1 1996 vanb unique B3 4 1995 vanb intra-hospital B4 1 1997 vanb unique B5 2 1991 vanb intra-hospital B6 3 1996, 2002 vanb sporadic, intra-hospital B7 4 1995, 1996, 1998 vanb sporadic B8 2 1996 vanb intra-hospital B9 1 1996 vanb unique B10 1 2000 vanb unique B11 1 2002 vanb unique B12 1 2001 vanb unique B13 1 2001 vanb unique B14 1 1999 vanb unique B15 4 2002 vanb intra-hospital B16 1 2000 vanb unique 16 groups 66 (80%) VanB scores, time from admission to positive culture for VR E. faecalis or VS E. faecalis, ICU stay, receipt of immunosuppressants, number of devices, gastrointestinal and non-gastrointestinal surgery or procedure in the preceding 3 months. Most enterococcal specimens (both VR E. faecalis and VS E. faecalis) were isolated from urine samples, followed by blood and wounds. Risk factors identified by univariate analysis were nursing home residence (P < 0.0001, odds ratio 5.50, 95% confidence interval 2.83 10.48), diabetes (P = 0.0439, odds ratio 1.84, 95% confidence interval 1.01 3.33), haemodialysis (P = 0.0007, odds ratio 6.20, 95% confidence interval 1.92 20.04), decubitus ulcers (P < 0.0001, odds ratio 4.85, 95% confidence interval 2.24 10.49), presence of gastrostomy tube (P = 0.0093, odds ratio 2.72, 95% confidence interval 1.26 5.88), co-infection/colonization with MRSA (P < 0.0001, odds ratio 6.74, 95% confidence interval 3.25 13.95), MDR GNB (P < 0.0001, odds ratio 6.33, 95% confidence interval 2.51 16), Clostridium difficile (P < 0.0001, odds ratio 7.17, 95% confidence interval 2.50 20.58), prior use of antimicrobials (P = 0.006, odds ratio 3.48, 95% confidence interval 1.37 8.82), use of vancomycin (P < 0.0001, odds ratio 6.05, 95% confidence interval 2.81 13.04) and aztreonam (P = 0.018, odds ratio 9.23, 95% confidence interval 1.93 44.05). Absence of co-infection/colonization with other nosocomial pathogens (P < 0.0001, odds ratio 0.13, 95% confidence interval 0.07 0.26) and use of penicillins (P = 0.0254, odds ratio 0.33, 95% confidence interval 0.12 0.91) emerged as protective factors. The independent risk factors that remained significant in a multivariate logistic analysis are shown in Table 2. Table 2. Risk factors independently associated with VR E. faecalis as determined by multivariate analysis Risk/protective factor P value Odds ratio Discussion 95% confidence interval Risk factor ESRD on HD 0.0097 7.63 1.636 35.608 parenteral vancomycin 0.0002 7.60 2.658 21.730 ECF residence 0.0005 4.51 1.933 10.525 decubitus ulcers 0.0285 3.07 1.125 8.375 Protective factor absence of co-infection 0.0010 0.25 0.112 0.572 penicillins 0.0081 0.12 0026 0.578 ESRD, end-stage renal disease; HD, haemodialysis; ECF, extended care facility. This is the first study that focuses on the molecular and clinical epidemiology of VR E. faecalis. Molecular analysis suggested vana and vanb gene dissemination among isolates and showed intra-hospital spread of other strains. A shift from vanb- to vana-containing isolates occurred over the period of the study. Receipt of chronic haemodialysis, nursing home residence, presence of decubitus ulcers and prior exposure to vancomycin were found to be independent risk factors for VR E. faecalis acquisition. 628

Epidemiology of vancomycin-resistant Enterococcus faecalis Historically, among enterococci, E. faecalis has been the predominant nosocomial pathogen. 19 Why E. faecium has become by far the most prevalent VR Enterococcus (VRE), whereas E. faecalis has retained its susceptibility to vancomycin, is unknown. Since its first report in 1988, the prevalence of VRE has increased from 0.3% of isolates during 1989 1993 to 12% during 1998 2002. 20 Recent data indicate that up to 50% of E. faecium and 3% of E. faecalis are resistant to vancomycin, 3 posing significant therapeutic and epidemiological challenges. Among enterococcal isolates, E. faecium has attracted almost exclusive attention, due to its resistance to multiple antimicrobials. In vitro transfer of resistance from E. faecalis to S. aureus had been shown years ago. 21 It is unclear why, despite a rapid and constant increase in VRE since the late 1980s, in vivo transfer of resistance to S. aureus occurred only recently and involved an E. faecalis strain, the less prevalent VRE. The recent report of the first isolate of VRSA carrying a vana resistance determinant probably acquired from E. faecalis, 1,2 makes our understanding of VR E. faecalis epidemiology essential. Earlier studies of the molecular epidemiology of VR E. faecium, revealed vana resistance in the majority of isolates (92%) tested, 4,7,20 whereas in one study of E. faecalis, isolates demonstrated predominantly vanb resistance (89.4%). 22 This predominance of vanb resistance in VR E. faecalis was also observed throughout our study, and no VanA isolates were seen until 1999. After their appearance, however, the number of VanA isolates increased rapidly and by the end of 2002 their number (13/29 or 45%) approached that of VanB E. faecalis. This shift to vana is of concern since this gene confers high-level resistance to glycopeptides and is the type of gene implicated in the transfer of vancomycin resistance from E. faecalis to S. aureus. The vana resistance determinant is known to have increased transferability, 23 as compared with vanb, which may explain the rapid increase in the number of VanA isolates, once they appeared, observed in our study. Our results are in agreement with those of earlier studies of hospitalacquired VR E. faecium and VS E. faecalis, which indicated clonal dissemination as a major mechanism for the spread of isolates. 6,9 Clonal intra-hospital spread was observed for both VanA and VanB isolates in our study. Although there was a diversity of PFGE strain types of VR E. faecalis (seven for VanA and 16 for VanB), both VanA and VanB types had one PFGE group that contained over half of the isolates. In previous studies, a considerable amount of heterogeneity among VR E. faecium strains was observed, even in epidemiologically related isolates. 9 We made the same observation in our strains of VR E. faecalis, suggesting that there is not only clonal dissemination of epidemic strains, but for some isolates, also probable transmission of resistance genes, either by way of a plasmid or transposon, from organism to organism. Importantly, eight isolates, all containing the vanb gene, had vancomycin MICs of 16 mg/l, and were thus reported as intermediate by the Clinical Microbiology Laboratory. This observation was made earlier and we agree with Quintiliani et al. 23 that the cutoff point for vancomycin resistance needs further evaluation, and furthermore that the role of vanb in the clinical setting needs to be determined. Earlier studies have associated several factors with an increased risk of developing VRE colonization and infection. Among these, one often cited is prior antimicrobial use, especially use of vancomycin, cephalosporins and agents active against anaerobic bacteria. 10,24 26 Our study supports a link between prior exposure to parenteral vancomycin and increased risk of acquisition of VR E. faecalis. In this study, we chose patients with susceptible strains as controls, as in a variety of earlier studies. 10,14,19,25 Choosing patients with susceptible strains as controls may overestimate the odds ratio of vancomycin use. An association, although weaker, between vancomycin use and emergence of VRE was also found in studies that used controls without enterococcal infection or colonization, 27 and is consistent with the concept that antibiotic treatment exerts selection pressure that promotes resistance. Time at risk, defined as time from admission to the hospital to positive culture, and co-morbidities were similar in cases and controls, eliminating these factors as potential confounding variables. Although prolonged use of cephalosporins was found to be significantly associated with VR E. faecalis by univariate analysis, it was eliminated as a risk factor after multivariate analysis. The protective effect we observed with prior use of penicillins has been reported before for enterococcus. 28 Susceptibility to ampicillin was high both in VR and VS E. faecalis isolates (96% and 99%, respectively) and did not increase over time. Anti-anaerobic agents were used more in patients who later developed VR E. faecalis infection or colonization, but this association did not reach statistical significance. Some of many other exposures associated in prior studies with infection or colonization with VRE in general, or VR E. faecium in particular, include length of hospital stay, transfer from another hospital, transfer between hospital floors, proximity to a known colonized or infected patient, receipt of enteral feedings, severe underlying illness, haematological malignancy, prior bone marrow transplant, renal insufficiency, instrumentation, prolonged ventilation and prolonged total parenteral nutrition. 10,25,26,29 None of these studies has specifically evaluated VR E. faecalis. All the host risk factors that were found in this study to be independently associated with VR E. faecalis (nursing home residence, receipt of haemodialysis, presence of decubitus ulcers) increase the likelihood of spread of organisms and thus, are not surprising. They are important since they define a population at risk that can be targeted by more intensive infection control measures. We found that absence of co-infection with other nosocomial pathogens was a protective factor in terms of acquisition of VR E. faecalis. This is in concordance with prior observations of increased frequency of infection or colonization with other nosocomial pathogens in patients with VRE. 29 The risk factors identified clearly suggest that VR E. faecalis is a nosocomial pathogen, with an epidemiology largely similar to that previously described of VR E. faecium. 10,12,25 The clinical relevance of the observations described in this study should be considered in infection control practices. Rapid spread of VR E. faecalis from a clonal origin, as seen with both VanA and VanB isolates, and the rapid rise in the number of VanA isolates, calls for implementation of careful isolation and infection control measures. Judicious use of antibiotics is of prime importance. Further surveillance of VR E. faecalis (with targeting of patient populations at increased risk nursing home residents, patients on haemodialysis, patients with decubitus ulcers) is warranted, due to the potential spread of vancomycin resistance genes not only among enterococci, but also from enterococci to staphylococci, both major causes of serious infections in hospitalized patients. Acknowledgements The authors thank Dora Vager from the Microbiology Research Laboratory, and John Koerber and Anita Gorial from the Pharmacy Department for their invaluable contribution to this study. 629

S. F. Oprea et al. References 1. Chang, S., Sievert, D. M., Hageman, J. C. et al. (2003). Infection with vancomycin-resistant Staphylococcus aureus containing the vana resistance gene. New England Journal of Medicine 348, 342 7. 2. Flannagan, S. E., Chow, J. W., Donabedian, S. M. et al. (2003). Characterization of the plasmid content of an Enterococcus faecalis VRE isolated from a patient also colonized by a Staphylococcus aureus with a VanA phenotype. Antimicrobial Agents and Chemotherapy 47, 3954 9. 3. Edmond, M. B., Wallace, S. E., McKlish, D. K. et al. (1999). Nosocomial bloodstream infections in United States hospitals: a three-year analysis. Clinical Infectious Diseases 29, 239 44. 4. Donabedian, S., Chow J. W., Shlaes, D. M. et al. (1995). DNA probes and contour-clamped homogenous field electrophoresis for identification of enterococci to the species level. Journal of Clinical Microbiology 33, 141 5. 5. Morris, J. G. Jr, Shay, D. K., Hebden, J. N. et al. (1995) Enterococci resistant to multiple antimicrobial agents, including vancomycin: establishment of endemicity in a university medical center. Annals of Internal Medicine 123, 250 9. 6. Hayden, M. K. (2000). Insights into the epidemiology and control of infection with vancomycin-resistant enterococci. Clinical Infectious Diseases 31, 1058 65. 7. Donabedian, S., Hershberger, E., Thal, L. A. et al. (2000). PCR fragment length polymorphism analysis of vancomycin-resistant Enterococcus faecium. Journal of Clinical Microbiology 38, 2885 8. 8. Manjunath, P. P., Rodvold, K. A., Schreckenberger, P. C. et al. (2002). Risk factors associated with the development of infection with linezolid- and vancomycin-resistant Enterococcus faecium. Clinical Infectious Diseases 35, 1269 72. 9. Thal, L. A., Donabedian, S., Robinson-Dunn, B. et al. (1998). Molecular analysis of glycopeptide-resistant Enterococcus faecium isolates collected from Michigan hospitals over a 6-year period. Journal of Clinical Microbiology 36, 3303 8. 10. Tornieporth, N. G., Roberts R. B., John, J. et al. (1996). Risk factors associated with vancomycin-resistant Enterococcus faecium infection or colonization in 145 matched case patients and control patients. Clinical Infectious Diseases 23, 767 72. 11. Bopp, L. H., Schoonmaker, D. J., Baltch, A. L. et al. (1999). Molecular epidemiology of vancomycin-resistant enterococci from 6 hospitals in New York State. American Journal of Infection Control 27, 411 7. 12. Papanicolau, G. A., Meyers, B. R., Meyers, J. et al. (1996). Nosocomial infections with vancomycin-resistant Enterococcus faecium in liver transplant recipients: risk factors for acquisition and mortality. Clinical Infectious Diseases 2, 760 6. 13. Chow, J. W., Kuritza, A., Shlaes, D. M. et al. (1993). Clonal spread of vancomycin-resistant Enterococcus faecium between patients in three hospitals in two states. Journal of Clinical Microbiology 6, 1609 11. 14. Vergis, E. N., Hayden, M. K., Chow, J. W. et al. (2001). Determinants of vancomycin resistance and mortality rates in enterococcal bacteremia: a prospective multicenter study. Annals of Internal Medicine 135, 484 92. 15. Facklam, R. R., Sahm, D. F. & Teixeira, L. M. (1999). Enterococcus. In Manual of Clinical Microbiology, 7th edn (Murray, P. R., Ed.), pp. 297 305. ASM Press, Washington, DC, USA. 16. National Committee for Clinical Laboratory Standards. (1997). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically Fourth Edition: Approved Standard M7-A4. NCCLS, Wayne, PA, USA. 17. Tenover, F. C., Arbeit, R. D., Goering, R. V. et al. (1995). Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. Journal of Clinical Microbiology 33, 2233 9. 18. Clark, N. C., Cooksey, R. C., Hill, B. C., et al. (1993). Characterization of glycopeptide-resistant enterococci from US hospitals. Antimicrobial Agents and Chemotherapy 37, 2311 7. 19. Zervos, M. J., Kaufman, C. A., Therasse, P. M. et al. (1987). Nosocomial infection by gentamicin-resistant Streptococcus faecalis. An epidemiologic study. Annals of Internal Medicine 106, 687 91. 20. National Nosocomial Infections Surveillance (NNIS) system report; data summary from January 1992 June 2002. (2002). American Journal of Infection Control 30, 458 75. 21. Noble, W. C., Virani, Z., Cree, R. G. (1992). Co-transfer of vancomycin and other resistance genes from Enterococcus faecalis NCTC 12201 to Staphylococcus aureus. FEMS Microbiology Letters 72, 195 8. 22. Malani, P. N., Thal, L., Donabedian, S. M. et al. (2002). Molecular analysis of vancomycin resistant Enterococcus faecalis from Michigan hospitals during a 10 year period. Journal of Antimicrobial Chemotherapy 49, 841 3. 23. Quintiliani, R., Evers, S. & Courvalin, P. (1993). The vanb gene confers various levels of self-transferable resistance to vancomycin in enterococci. Journal of Infectious Diseases 167, 1220 3. 24. Fridkin, S. K., Edwards, J. R., Courval J. M. et al. (2001). For the Intensive Care Antimicrobial Resistance Epidemiology (ICARE) Project and the National Nosocomial Infections Surveillance (NNIS) System Hospitals. The effect of vancomycin and third-generation cephalosporins on prevalence of vancomycin-resistant enterococci in 126 U. S. Adult Intensive Care Units. Annals of Internal Medicine 135, 175 83. 25. Bhavnani, S. M., Drake, J. A., Forrest, A. et al. (2000). A nationwide, multicenter, case-control study comparing risk factors, treatment, and outcome for vancomycin-resistant and -susceptible enterococcal bacteremia. Diagnostic Microbiology and Infectious Diseases 36, 145 58. 26. Safdar, N. & Maki, D. G. (2002). The commonality of risk factors for nosocomial colonization and infection with antimicrobial-resistant Staphylococcus aureus, enterococcus, gram-negative bacilli, Clostridium difficile, and Candida. Annals of Internal Medicine 136, 834 44. 27. Carmeli, Y., Samore, M. H. & Huskins, W. C. (2002). The association between antecedent vancomycin treatment and hospital-aquired vancomycin-resistant enterococci. Clinical Infectious Diseases 34, 1558 63. 28. Donskey, C. J., Chowdhry, T. K., Hecker, M. T. et al. (2000). Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. New England Journal of Medicine 343, 1925 32. 29. Donskey, C. J., Ray, A. J., Hayen, C. K. et al. (2003). Colonization and infection with multiple nosocomial pathogens among patients colonized with vancomycin-resistant enterococcus. Infection Control and Hospital Epidemiology 24, 242 5. 630