Risk Factors Associated with Methicillin Resistance among Staphylococcus aureus Infections in Veterans

Transcription

1 infection control and hospital epidemiology january 2010, vol. 31, no. 1 original article Risk Factors Associated with Methicillin Resistance among Staphylococcus aureus s in Veterans Natalie L. McCarthy, MPH; Patrick S. Sullivan, PhD, DVM; Robert Gaynes, MD; David Rimland, MD background. Methicillin-resistant Staphylococcus aureus (MRSA) is an emerging concern in infectious disease practice. Although MRSA infections occur in a wide variety of anatomic sites, the majority of studies considering the risk factors for methicillin resistance among S. aureus infections have focused on MRSA bacteremia. objective. To describe risk factors associated with methicillin resistance among S. aureus infections at different anatomic sites. methods. We collected information on the demographic and clinical characteristics of patients examined at the Atlanta Veterans Affairs Medical Center with S. aureus infections during the period from June 2007 through May We used multivariate logistic regression to describe factors significantly associated with methicillin resistance. results. There were 568 cases of S. aureus infection among 528 patients. We identified 352 cases (62%) of MRSA infection and 216 cases (38%) of methicillin-sensitive S. aureus infection. The adjusted odds of methicillin resistance were higher among infections that occurred among patients who had a prior history of MRSA infection (odds ratio [OR], 3.9 [95% confidence interval {CI}, ]) or resided in a long-term care facility during the past 12 months (OR, 2.0 [95% CI, ]) but were lower for infections that occurred among patients who had undergone a biopsy procedure during the past 12 months (OR, 0.7 [95% CI, ]). Most cases of infection were community-onset infections (523 [92%] of 568 cases), and about one-half (278 [49%]) were not healthcare associated. conclusions. Compared with previous studies of methicillin resistance among patients with S. aureus bacteremia, we found similar factors to be associated with methicillin resistance among S. aureus isolates recovered from more diverse anatomic sites of infection. Of note, nearly one-half of our cases of MRSA infection were not healthcare associated. Infect Control Hosp Epidemiol 2010; 31:36-41 Antibiotic resistance among bacterial pathogens is widely recognized as an emerging issue in infectious disease practice. The National Nosocomial s Surveillance system documented a steady increase in the rate of resistance to antibiotics of Staphylococcus aureus isolates recovered from samples obtained from intensive care unit patients with S. aureus infection during the period from 1998 through Of particular concern is methicillin-resistant S. aureus (MRSA), which has emerged as a major contributor to morbidity and mortality for hospitalized patients. When compared with infections due to methicillin-sensitive S. aureus (MSSA), infections due to MRSA are associated with a significantly higher mortality rate. 2-4 Therefore, it is important to examine the factors that may be associated with MRSA infection and to determine whether associated factors have changed in recent years. Most studies of factors associated with methicillin resistance have focused on bacteremia due to S. aureus. 2-7 Limited data are available for a more heterogeneous group of infections due to S. aureus. 8-9 Our study examined differences in demographic and clinical characteristics between patients with MRSA infection and patients with MSSA infection among all anatomic sites of infection found in a recent cohort. methods We conducted a retrospective exploratory analysis to evaluate the risk factors for methicillin resistance among infections due to S. aureus, for all anatomic sites of infection. The study population consisted of inpatients and outpatients treated at the Atlanta Veterans Affairs Medical Center. Information was collected on all patients with either an MRSA infection or an MSSA infection, or both, diagnosed during the period June 1, 2007, through May 31, MRSA infections and MSSA infections were identified by means of analysis of the database of the microbiology laboratory at the Atlanta Veterans Affairs Medical Center. A line listing of all clinical cultures that grew MRSA and/or MSSA was generated from the computerized patient record system (Vista) at the medical center by using From the Rollins School of Public Health (N.L.M., P.S.S.) and the School of Medicine (R.G., D.R.), Emory University, and the Veterans Affairs Medical Center (R.G., D.R.), Atlanta, Georgia. Received April 2, 2009; accepted July 27, 2009; electronically published November 24, by The Society for Healthcare Epidemiology of America. All rights reserved X/2010/ $ DOI: /649017

2 mrsa and mssa risk factors 37 the microbiology option for specific organisms. Surveillance cultures for nasal colonization, which used samples obtained at admission to the medical center, were excluded from the analysis. Duplicate cultures from the same clinical infection were also excluded. To classify the types of infections, an experienced infectious disease physician (D.R.) reviewed the electronic medical record and used all available information, including anatomic site of culture sampling, radiographic study results, and clinical information, to define a true infection caused by MRSA or MSSA. The physician determined the standards for classifying infections a priori, by use of the Centers for Disease Control and Prevention (CDC) criteria for defining and classifying infections. 10 Cultures not associated with a true infection were not included. A minimum of 30 days between episodes was required to define a separate episode of infection. The infections were then categorized according to anatomic site, in the following categories: skin or soft tissue, bone, bloodstream, urinary tract, respiratory tract, surgical site, ear, sinuses, and other. A x 2 test was used to determine whether there were statistically significant differences in anatomic site of infection between MRSA infections and MSSA infections. We also created a classification for serious infections, which comprised bacteremia and any other infection necessitating hospitalization. We used the CDC definition to classify the infections as hospital-acquired infections, healthcare-associated infections, and community-acquired infections. Hospital-acquired infections were defined as those in which a culture sample obtained 48 hours or more after admission to the hospital yielded a result positive for S. aureus; otherwise, they were considered to be community-onset infections. Healthcareassociated infections, either hospital acquired or community onset, were defined according to the CDC definition as those in which patients had been hospitalized, had undergone surgery or dialysis, resided in a long-term care facility, or used an indwelling catheter during the past 12 months. 11 Information pertinent to the investigation of risk factors was obtained by means of medical chart review. We collected demographic data, including age, race, and sex, and data on risk factors for S. aureus infections, including invasive procedures during the 12 months before the diagnosis of MRSA infection or MSSA infection. Data regarding prior diagnoses of chronic obstructive pulmonary disease (COPD), diabetes mellitus, human immunodeficiency virus (HIV) infection, renal disease, and compromised skin integrity (such as psoriasis, eczema, ulcers, and unspecified cutaneous lesions) were collected. Renal disease was defined as abnormal renal function, documented by a glomerular filtration rate of less than 60, which was calculated using the Cockcroft-Gault formula. 12 Other putative risk factors included the presence of a central venous catheter, residence in a long-term care facility, hospitalization, or a stay in the intensive care unit, all during the 12 months before the index infection. We collected data on any antibiotic use during the 3 months before infection, and any past MRSA infections were also documented. Finally, we collected data on invasive medical procedures undergone during the past 12 months: any biopsy, cardiac catheterization, endoscopy, dialysis, or surgery. We used SAS software, version 9.2 (SAS Institute), to perform the analyses. Descriptive statistics were used to summarize the characteristics of all patients included in the study according to the type of infection (MRSA or MSSA). Since patient age was nonnormally distributed, for each type of infection, we report a median age and interquartile range. We compared serious infections according to type of infection by use of the x 2 test. The Wilcoxon rank sum test was used to compare distributions of age for patients with MRSA infection and patients with MSSA infection. For categorical data, proportions were compared by use of the x 2 test. To build our model, we started with a list of demographic and clinical factors known or suspected to be associated with methicillin resistance. Because all of the variables in the study were potential predictors for methicillin resistance, we considered for the multivariable model only those factors associated with methicillin resistance with a P value of less than.20 in bivariate analysis. To account for patients with multiple infections, we used the GENMOD procedure in SAS (SAS Institute) with an exchangeable covariance matrix to calculate the adjusted odds ratios (ORs) while controlling for repeated observations of patients. We assessed collinearity of all pairs of possible risk factors by use of condition indices and variance decomposition proportions. 13 The backward selection procedure was used to build the final model. At each step, the variable with the highest insignificant P value was removed, and the model was adjusted with the remaining variables. This process continued until each of the variables in the model was significant, with a P value of less than.05, while controlling for the other variables in the model. Once all of the first-order terms for the models were entered, we tested for statistical significance of all 2-way interactions of main effects. We used the logistic regression model to calculate adjusted ORs and 95% confidence intervals (CIs). We conducted a subset analysis that included only the bacteremia infections, to compare our results with those found in other studies. For this analysis, we calculated adjusted ORs by use of a logistic regression model, using similar methods to those described above. However, as a result of the low number of bacteremia infections ( n p 72), we used an a of.10 for bivariate results as a criterion for entry into the multivariate model, and we calculated 90% CIs around the ORs. results During the 1-year study period, there were a total of 785 culture results positive for S. aureus, including duplicates and contaminants. There were 568 cases of S. aureus infection. We identified 352 cases (62%) of MRSA infection and 216 cases (38%) of MSSA infection. There were a total of 528 infected patients, 33 of whom had more than 1 infection

3 38 infection control and hospital epidemiology january 2010, vol. 31, no. 1 table 1. Bivariate Analysis of Selected Risk Factors Associated with Methicillin Resistance in Staphylococcus aureus s among Patients at the Atlanta Veterans Affairs Medical Center, June 2007 May 2008 Patient characteristic due to MRSA (n p 352) due to MSSA (n p 216) P Sex.09 Male 335 (95) 198 (92) Female 17 (5) 18 (8) History of MRSA infection!.001 Yes 113 (32) 14 (6) No 239 (68) 202 (94) Antibiotic use a.04 Yes 151 (43) 74 (34) No 201 (57) 142 (66) Residence in a long-term care facility b.001 Yes 41 (12) 8 (4) No 311 (88) 208 (96) Hospitalized in the past 12 months.03 Yes 130 (37) 61 (28) No 222 (63) 155 (72) Biopsy in the past 12 months.02 Yes 24 (7) 27 (12) No 328 (93) 189 (88) Dialysis in the past 12 months.04 Yes 26(7) 7(3) No 326 (93) 209 (97) COPD c.04 Yes 58(16) 22(10) No 294 (84) 194 (90) Hospital onset d.43 Yes 31 (9) 15 (7) No 321 (91) 201 (93) Healthcare associated e.08 Yes 190 (54) 100 (46) No 162 (46) 116 (54) note. Data are no. (%) of cases. COPD, chronic obstructive pulmonary disease; MRSA, methicillin-resistant S. aureus; MSSA, methicillin-susceptible S. aureus. a During past 3 months. b During past 12 months. c Concurrent with MRSA or MSSA infection. d Culture sample obtained at least 48 hours after hospitalization. e During past 12 months: hospitalization, surgery, receipt of dialysis, residence in a longterm care facility, or use of indwelling catheters. during the study period. Of the 33 patients with more than 1 infection, 24 had multiple MRSA infections, 6 had multiple MSSA infections, and 3 had 1 MSSA infection and also 1 MRSA infection. The large majority of cases of MRSA infection were cases of community-onset infection (ie, 321 [91%] of 352 cases). For cases of MRSA infection and cases of MSSA infection (total, 568 cases), the median age of patients was 60 years (range, years). In 533 (94%) of these 568 cases of infection, the patients were male. In 299 cases (53%), the patients were non-hispanic white; in 253 cases (45%), the patients were non-hispanic black; in 4 cases (1%), the patients were Hispanic; and in 12 cases (2%), the race of the patient was unknown. Table 1 gives the results of the bivariate analysis for the factors found to be associated with methicillin resistance with a P value of less than.20. We did not include race, intensive care unit stay, cardiac catheterization, presence of central venous catheter, use of endoscopy, surgery, diabetes, renal disease, HIV infection, and compromised skin in the model, because they were not associated with methicillin resistance with a P of less than.20. Table 1 also presents the distribution of hospital-onset and healthcare-associated infections among the cases of MRSA infection and the cases of MSSA infection. Table 2 gives the results for the multivariate analysis. Previous MRSA infection (OR, 3.9 [95% CI, ]) and stay in a long-term care facility in the past year (OR, 2.0 [95%

4 mrsa and mssa risk factors 39 table 2. Crude Odds Ratios (ORs) and Adjusted ORs from Multivariate Analysis of Risk Factors Associated with Methicillin Resistance in s Due to Staphylococcus aureus among Patients at the Atlanta Veterans Affairs Medical Center, June 2007 May 2008 Patient characteristic Crude OR (95% CI) Adjusted OR (95% CI) Sex 1.8 ( ) History of MRSA infection 6.8 ( ) 3.9 ( ) Use of antibiotics a 1.4 ( ) Residence in long-term care facility b 3.4 ( ) 2.0 ( ) Past hospitalization b 1.5 ( ) Biopsy b 0.5 ( ) 0.7 ( ) Receipt of dialysis b 2.4 ( ) COPD c 1.7 ( ) note. CI, confidence interval; COPD, chronic obstructive pulmonary disease; MRSA, methicillin-resistant S. aureus. a During past 3 months. b During past 12 months. c Concurrent with study infection. CI, ]) were associated with higher odds of having an MRSA infection, compared with having an MSSA infection. A biopsy in the previous 12 months (OR, 0.7 [95% CI, ]) was associated with lower odds of having an MRSA infection, compared with having an MSSA infection. Approximately one-quarter (13 [25%] of 51) of the biopsies were skin biopsies, and the rest varied among type: liver, foot, rectum, colon, stomach, bladder, dental (extraction of dental fragment), lung, neck node, bone marrow, duodenum, larynx, anus, prostate, pancreas (computed tomography guided biopsy), tongue, neck mass, jaw, nasal septum, kidney, ileum, and nose. No 2-way interaction between main effects was statistically significant in the final model, and therefore none were retained. Table 3 describes the distribution of the anatomic sites of infection. Approximately one-half of the cases of MRSA infection and the cases of MSSA infection were cases of skin or soft-tissue infection, and the rest of the cases of infection varied among the other anatomic sites. Cases of MSSA infection were more likely than cases of MRSA infection to be surgical site infections ( P p.03) or ear infections ( P p.02). There was no significant difference in the proportion of serious infections (defined as cases of bacteremia or infection requiring hospitalization) between cases of MRSA infection and cases of MSSA infection ( P p.87). When we examined only cases of bacteremia due to S. aureus, we found that there were 72 cases of this type of infection, 44 (61%) of which were due to MRSA. Cases of bacteremia were approximately twice as likely to occur among non-hispanic white patients (46 [66%] of 70) as among non- Hispanic black patients (24 [34%] of 70), and none of the cases of bacteremia occurred among Hispanic patients. The age range for patients with bacteremia was years. In the final logistic regression subset analysis of cases of bacteremia, the odds of methicillin resistance among bloodstream infections due to S. aureus were significantly higher for patients with a history of central venous catheter use (OR, 1.9 [90% CI, ]), compromised skin (OR, 2.4 [90% CI, ]), or COPD (OR, 6.4 [90% CI, ]). discussion The results of our study confirmed previously identified risk factors associated with methicillin resistance in S. aureus infections, as well as risk factors associated with MRSA bacteremia specifically. We also noted a lower proportion of healthcare-associated disease than had been expected on the basis of the history of MRSA in healthcare settings We believe that the recent high frequency of community-acquired MRSA infections contributes to the lack of differences between risk factors associated with MRSA infection and risk factors associated with MSSA infection. table 3. Anatomic Sites of Due to Staphylococcus aureus among Patients at the Atlanta Veterans Affairs Medical Center, June 2007 May 2008 Anatomic site due to MRSA (n p 352) due to MSSA (n p 216) P a Skin or soft tissue 187 (53) 99 (46).09 Bone 18 (5) 13 (6).64 Bloodstream 44 (12) 28 (13).87 Urinary tract 49 (14) 22 (10).19 Respiratory tract 16 (5) 11 (5).77 Surgical site 31 (9) 32 (15).03 Ear 1 (0) 5 (2).03 Sinuses 3 (1) 5 (2).16 Other 3 (1) 1 (0).36 note. Data are no. (%) of cases. MRSA, methicillin-resistant S. aureus; MSSA, methicillin-susceptible S. aureus. a P values for the classifications of ear, sinuses, and other were calculated by use of the Fisher exact test; P values for the remaining classifications were calculated by use of the x 2 test.

5 40 infection control and hospital epidemiology january 2010, vol. 31, no. 1 Cases of bacteremia accounted for a limited number of infections in our study. Of the 568 cases of infection due to S. aureus, 496 (87%) were cases of infection other than bacteremia, and 286 (50%) were cases of skin or soft-tissue infection. Thus, we believe that our approach of considering all S. aureus infections gives a broader picture of the risk factors associated with methicillin resistance. Although there was no significant difference between the number of cases of serious infection due to MRSA and the number of cases of serious infection due to MSSA, 187 (33%) of the 568 cases were considered serious (72 cases [13%] were bacteremia, and 115 cases [20%] were cases of other types of infection requiring hospitalization). This evidence indicates that attention should be given to the risk factors for bacteremia due to S. aureus, as well as other types of S. aureus infection. There was a strong association between previous MRSA infection and methicillin resistance for the index infection in our analysis. Our observed OR of 3.9 is consistent with those of other studies, which have found that the odds of having an MRSA infection are approximately 3 4 times greater for patients with a history of MRSA infection than for patients with no history of MRSA infection Residency in a longterm care facility during the past 12 months as a risk factor for MRSA infection is also consistent with the results of other studies. 14,21 The protective association between biopsy and methicillin resistance is an unexplained finding. However, our observed association was weak and was based on a small number of observations and should be interpreted with caution unless others confirm this finding. In the subset analysis of cases of bacteremia, the odds of methicillin resistance were great for infections that occurred in patients with COPD, in patients with a central venous catheter, or in patients with compromised skin. These findings were consistent with those of reports published elsewhere Overall, there were few significant differences between cases of MRSA infection and cases of MSSA infection. This may be due to the fact that we observed more cases of communityacquired MRSA infection than expected, consistent with recent reports In our study, of the 352 cases of MRSA infection, 321 (91%) were cases of community-onset infection and 162 (46%) were cases of community-acquired infection (ie, not healthcare associated according to the CDC definition). MSSA infection is thought to be mostly a communityassociated disease; however, in our study, nearly one-half of both types of S. aureus infection (methicillin resistant and methicillin susceptible) were community acquired. The appearance of both types of infection in the community may explain the limited distinction in healthcare-related risk factors between MRSA infection and MSSA infection, as others have previously reported elsewhere. 27,29,30 In addition, relatively few cases of MRSA infection (31 [9%] of 352) actually occurred in the hospital. These findings may have implications for the success of the MRSA prevention and control efforts recently advocated for hospitals. 31 There are some limitations to our study. Because we collected our data through medical chart review, we were limited in the types of data we could collect. Data on exposures that may have occurred in the community, such as contact with people with infections due to S. aureus, were not systematically collected or recorded in the medical records and therefore could not be considered in our analysis. In addition, because this was a retrospective study, there is a possibility of information bias in the collection of historical information from medical records. There is also a chance of selection bias, in that some patients with MRSA infection or MSSA infection may not have been tested by culture but were treated empirically without a diagnostic test. These infections would not be included in the study, so we may have underestimated the true burden of disease. If such bias in culturing was differential, in which certain types of infections were more likely to be submitted for culture by clinicians, then our data on the characteristics of infections might also be affected. Because these data were collected in a specialized clinical population (veterans), the findings cannot be generalized to the US population. Finally, the absence of culture testing for analysis of molecular subtype limits the ability to distinguish community-acquired infections from healthcare-associated infections according to molecular criteria. This last issue has become less relevant, as mixtures of both resistance patterns are now seen in the community and in healthcare settings. In conclusion, previous MRSA infection and residency in a long-term care facility were significantly associated with a greater risk of MRSA infection than MSSA infection. A biopsy procedure in the past year had a weak association with MSSA infection and should be considered for further study with a larger population. The risk factors for MRSA infection were not as distinct from those of MSSA infection, as described elsewhere. We believe that the striking increase in community-acquired MRSA infections explains the similarities in clinical history between MRSA infections and MSSA infections in our cohort. acknowledgments Financial support. Centers for Disease Control and Prevention (grant 07FED706507). Potential conflicts of interest. All authors report no conflicts of interest relevant to this article. Address reprint requests to David Rimland, MD, Medical Specialty Service Line (111-RIM), Veterans Affairs Medical Center, 1670 Clairmont Road, Decatur, GA (david.rimland@va.gov). references 1. National Nosocomial s Surveillance (NNIS) system report, data summary from January 1992 through June 2004, issued October Am J Infect Control 2004; 32: Cosgrove SE, Sakoulas G, Perencevich EN, Schwaber MJ, Karchmer AW, Carmeli Y. Comparison of mortality associated with methicillin-resistant and methicillin-suceptible Staphylococcus aureus bacteremia: a meta-analysis. Clin Infect Dis 2003; 36: Blot SI, Vandewoude KH, Hoste EA, Colardyn FA. Outcome and at-

6 mrsa and mssa risk factors 41 tributable mortality in critically ill patients with bacteremia involving methicillin-susceptible and methicillin-resistant Staphylococcus aureus. Arch Intern Med 2002; 162: Shurland S, Zhan M, Bradham DD, Roghmnn M. Comparison of mortality risk associated with bacteremia due to methicillin-resistant and methicillin-susceptible Staphylococcus aureus. Infect Control Hosp Epidemiol 2007; 28: Wang J, Chen S, Wang J, et al. Comparison of both clinical features and mortality risk associated with bacteremia due to community-acquired methicillin-resistant Staphylococcus aureus and methicillin-susceptible S. aureus. Clin Infect Dis 2008; 46: Gonzalez C, Rubio M, Romero-Vivas J, Gonzalez M, Picazo JJ. Bacteremic pneumonia due to Staphylococcus aureus: a comparison of disease caused by methicillin-resistant and methicillin-susceptible organisms. Clin Infect Dis 1999; 29: Laupland KB, Ross T, Gregson DB. Staphylococcus aureus bloodstream infections: risk factors, outcomes, and the influence of methicillin resistance in Calgary, Canada, J Infect Dis 2008; 198: Sadoyama G, Gontijo Filho PP. Risk factors for methicillin resistant and sensitive Staphylococcus aureus infection in a Brazilian university hospital. Braz J Infect Dis 2000; 4: Munckhof WJ, Nimmo GR, Carney J, et al. Methicillin-susceptible, nonmultiresistant methicillin-resistant and multiresistant methicillin-resistant Staphylococcus aureus infections: a clinical, epidemiological and microbiological comparative study. Eur J Clin Microbiol Infect Dis 2008; 27: Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008; 36: Klevens RM, Morrison MA, Fridkin SK, et al. Community-associated methicillin-resistant Staphylococcus aureus and healthcare risk factors. Emerg Infect Dis 2006; 12: Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16: Kleinbaum DG, Klein M. Logistic Regression: A Self-Learning Text. 2nd ed. New York, NY: Springer; Jorgensen JH. Laboratory and epidemiologic experience with methicillin-resistant Staphylococcus aureus in the United States. Eur J Clin Microbiol 1986; 5: Wenzel RP, Nettleman MD, Jones RN, Pfaller MA. Methicillin-resistant Staphylococcus aureus: implications for the 1990s and effective control measures. AmJMed1991; 91(Suppl 3B):221S 227S. 16. Mulligan ME, Murray-Leisure KA, Ribner BS, et al. Methicillin-resistant Staphylococcus aureus: a consensus review of the microbiology, pathogenesis, and epidemiology with implications for prevention and management. AmJMed1995; 98: Haley CC, Mittal D, LaViolette A, Jannapureddy S, Parvez N, Haley RW. Methicillin-resistant Staphylococcus aureus infection or colonization present at hospital admission: multivariable risk factor screening to increase efficiency of surveillance culturing. J Clin Microbiol 2007; 45: Moran GJ, Krishnadasan A, Gorwitz RJ, et al. Methicillin-resistant S. aureus infections among patients in the emergency department. N Engl J Med 2006; 355: Huang SS, Platt R. Risk of methicillin-resistant Staphylococcus aureus infection after previous infection or colonization. Clin Infect Dis 2003; 36: Beam JW, Buckley B. Community-acquired methicillin-resistant Staphylococcus aureus: prevalence and risk factors. J Athl Train 2006; 41: Bradley S, Terpenning MS, Ramsey MA, et al. Methicillin-resistant Staphylococcus aureus: colonization and infection in a long-term care facility. Ann Intern Med 1991; 115: Bakowski E, Wey SB, Medeiros EAS. Risk factors for bacteremia and predictors of mortality of patients with bloodstream infection with methicillin-resistant Staphylococcus aureus. Am J Infect Dis 2008; 4: Wang F, Chen Y, Chen T, Liu C. Risk factors and mortality in patients with nosocomial Staphylococcus aureus bacteremia. Am J Infect Control 2008; 36: Chi C, Wong W, Fung C, Yu K, Liu C. Epidemiology of community-acquired Staphylococcus aureus bacteremia. J Microbiol Immunol Infect 2004; 37: Lowy FD, Aiello AE, Bha M, et al. Staphylococcus aureus colonization and infection in New York state prisons. J Infect Dis 2007; 196: Vanden F, Naimi T, Enright MC, et al. Community-acquired methicillinresistant Staphylococcus aureus carrying Panton-Valentine leukocidin genes: worldwide emergence. Emerg Infect Dis 2003; 9: Miller LG, Perdreau-Remington F, Bayer AS, et al. Clinical and epidemiologic characteristics cannot distinguish community-associated methicillin-resistant Staphylococcus aureus infection from methicillin-susceptible S. aureus infection: a prospective investigation. Clin Infect Dis 2007; 44: Crum NF, Lee RU, Thornton SA, et al. Fifteen-year study of the changing epidemiology of methicillin-resistant Staphylococcus aureus. Am J Med 2006; 119: Diederen BMW, Kluytmans JAJW. The emergence of infections with community-associated methicillin resistant Staphylococcus aureus. J Infect 2006; 52: Rollason J, Bastin L, Hilton AC, et al. Epidemiology of community-acquired methicillin-resistant Staphylococcus aureus obtained from the UK West Midlands region. J Hosp Infect 2008; 70: Muto C, Jerigan J, Ostrowsky B, et al. SHEA guideline for preventing nosocomial transmission of multidrug-resistant strains of Staphylococcus aureus and Enterococcus. Infect Control Hosp Epidemiol 2003; 24: