Epidemiology of Methicillin‐Resistant Staphylococcus aureus and Vancomycin‐Resistant Enterococcus in a Rural State
Background. Most data on methicillin‐resistant Staphylococcus aureus (MRSA) and vancomycin‐resistant Enterococcus (VRE) isolates come from large tertiary care centers. Infection control personnel need to understand the epidemiology of MRSA and VRE across the continuum of care, including small rural hospitals, to develop effective control strategies.
Objective. To describe the epidemiology of MRSA and VRE in Iowa.
Setting. Fifteen hospitals in Iowa.
Methods Between July 1998 and June 2001, a total of 1,968 S. aureus isolates and 1,845 Enterococcus isolates from patients infected with these pathogens were examined. Multivariate models were developed to evaluate patient and institutional risk factors for MRSA infection and VRE infection.
Results. The proportion of S. aureus isolates resistant to methicillin was 31%, and the proportion of Enterococcus isolates resistant to vancomycin was 6%. Independent risk factors for MRSA infection included residence in a long‐term care facility, age of more than 60 years, hospitalization in a hospital with less than 200 short‐term care beds, and acquiring the infection in the hospital. Independent risk factors for VRE infection included use of a central venous catheter, residence in a long‐term care facility, acquisition of infection in the hospital, and hospitalization in a hospital with more than 200 short‐term care beds.
Conclusions. In Iowa, the epidemiology of MRSA differ from those of VRE. MRSA has become established in small rural hospitals. Effective MRSA control strategies may require inclusion of all hospitals in a state or region.
Received February 3, 2005; accepted June 14, 2005; electronically published February 28, 2006.
Methicillin‐resistant Staphylococcus aureus (MRSA) and vancomycin‐resistant enterococci (VRE) are important antimicrobial‐resistant pathogens. Both MRSA and VRE have traditionally been considered as hospital‐acquired organisms, and they share many risk factors for acquisition.1 In addition, the incidence of infection is increasing for both pathogens in US hospitals.2‐4 MRSA now accounts for approximately 60% of S. aureus bloodstream infections in US intensive care units (ICUs), whereas more than a quarter of hospital‐acquired enterococcal infections are due to VRE.2
MRSA and VRE infection rates continue to increase in the United States despite widespread adherence to current guidelines to prevent and control spread of antimicrobial‐resistant pathogens. Thus, novel infection control approaches are needed if we are to reverse these trends.5 Some healthcare epidemiologists have argued that the aggressive measures, including active surveillance or “search and destroy” approaches,6 used by some countries in northern Europe and Scandinavia should be adopted in the United States to control the spread of MRSA and VRE.7,8 Currently, the most aggressive approaches to control of MRSA and VRE in the United States focus almost exclusively on the hospital environment, specifically large academic medical centers, many of which are in urban areas. Such control efforts are likely to fail if reservoirs of antimicrobial resistance exist outside these settings and if nonurban settings are ignored.
To develop effective control strategies for MRSA and VRE, therefore, infection control personnel need to understand the epidemiology of these organisms across the continuum of healthcare settings. Because one quarter of the US population resides in rural locations, which are often served by small hospitals, infection control personnel need more information about antibiotic resistance patterns in rural environments. We initiated a statewide longitudinal sentinel surveillance system in July 1998 to describe the epidemiology of antimicrobial resistance in Iowa.9,10 We present 3 years of data from the Emerging Infections and the Epidemiology of Iowa Organisms (EIEIO) study that address the epidemiology of MRSA and VRE infection in a rural midwestern state.
Methods
A statewide hospital‐based surveillance network was established. The 15 participating centers were chosen on the basis of geographic location, population distribution, and number of hospital beds. The number of beds in each hospital ranged from approximately 86 to more than 858, and centers were geographically distributed in locations throughout Iowa, including Sioux City and Council Bluffs in western Iowa, Clinton in eastern Iowa, Mason City and Spencer in northern Iowa, and Ottumwa and Burlington in southern Iowa. The annual number of hospital admissions ranged from 2,500 to more than 40,000, and 6 participating centers were classified as rural or rural‐referral hospitals. Together, the participating centers accounted for more than half of the short‐term care beds in Iowa during the study period. Each hospital had a microbiology laboratory that received specimens from both inpatients and outpatients. During each quarter, the microbiology laboratories at each participating medical center submitted the first 10 consecutive S. aureus and Enterococcus isolates from unique patients considered to be infected (not colonized) at any body site with one of these organisms. The organisms were submitted to the University of Iowa (Iowa City) for identification and susceptibility testing with reference methods.
The rate of methicillin resistance was defined as the proportion of all clinical S. aureus isolates submitted that were resistant to oxacillin, and the rate of vancomycin resistance was defined as the proportion of all clinical Enterococcus isolates submitted that were resistant to vancomycin. The isolates submitted were not from screening cultures of nares or rectal swab specimens.
Susceptibility Testing
Antimicrobial susceptibility was determined by the broth microdilution method described by the Clinical and Laboratory Standards Institute.11,12 Isolates of S. aureus were tested for susceptibility to oxacillin, erythromycin, clindamycin, gentamicin, tetracycline, chloramphenicol, trimethoprim‐sulfamethoxazole, rifampin, and ciprofloxacin. Enterococcus isolates were tested for susceptibility to ampicillin, chloramphenicol, gentamicin, streptomycin, teicoplanin, and vancomycin.
Epidemiologic Data Collection
Each participating hospital had an infection control program staffed by at least 1 infection control professional (ICP) who provided the epidemiologic and patient‐related data necessary for the study. The laboratory informed the ICP when isolates were submitted and provided the ICP with the medical record number for each patient and the specimen identification number for each isolate. The ICPs recorded the history of antimicrobial use and surgical procedures for each patient. The ICPs also noted whether the patient had resided in a long‐term care facility (LTCF) during the previous 6 months and whether the infection was nosocomial. Infections were determined to be nosocomial according to each ICP's interpretation of Centers for Disease Control and Prevention criteria; in most cases, this meant that signs and symptoms began >48 hours after hospital admission. The data collection form was revised in July 2000 to include questions about the anatomical site of infection and whether, in the past 30 days, the patient was in an intensive care unit, was immunosuppressed, had a central venous catheter, or was treated with renal dialysis.
Demographic data for centers in the study were obtained from the Association of Iowa Hospitals and Health Systems. Hospital size was measured on the basis of the number of short‐term care beds, and hospitals were classified as urban, rural referral, or rural, on the basis of Medicare definitions.
Statistical Methods
The Pearson χ2 test was used to assess the correlation between the outcome variable (MRSA or VRE infection) and other categorical variables. In bivariate analyses, the Cochran‐Mantel‐Haenszel test was used to test for trends, when appropriate. To identify variables that were highly associated with MRSA or VRE infection, all variables associated with outcome variables that had a P value of less than .2 and all first‐order interactions were included in multivariate logistic regression models constructed with a backward stepwise selection method. The Akiake information criterion was used for selecting models that best explained the data. The Hosmer‐Lemeshow goodness‐of‐fit statistic was used to assess the models' fit. Pearson and deviance residuals were checked. Spearman correlation analysis was performed to identify the correlation between MRSA and VRE infection rates and hospital size. All tests were performed with an α level of .05. We used SAS, version 9.0 (SAS Institute), for all analyses.
Results
The 15 centers submitted 1,968 S. aureus isolates (range, 114‐155 isolates per center), of which 619 (31%) were methicillin resistant (range, 9%‐51% per center). Fourteen centers submitted epidemiologic data for 1,047 (53%) of these isolates (median, 58% of isolates per center [range, 9%‐92% of isolates per center]), 337 (32%) of which were MRSA. The basic demographic and clinical data, including hospital ward, age, and sex and specimen type, were similar for all S. aureus isolates and for the isolates for which epidemiologic data were available.
Nine of the medical centers were in an urban location (ie, a metropolitan statistical area, as defined by the US Census Bureau), 5 were rural‐referral hospitals (ie, hospitals that had operating characteristics similar to those of urban hospitals but were located in rural communities), and 1 was a rural medical center. The single rural center did not provide epidemiologic data. All 6 hospitals in rural areas were grouped together for purposes of statistical analyses.
Bivariate analyses identified a number of factors associated with MRSA infection (Table 1). Stepwise logistic regression analysis identified the following independent risk factors for MRSA infection: residence in a LTCF (
), age greater than 60 years (
), hospitalization in an institution with <200 beds (
), and acquisition from a source other than the community (ie, designated as “not community acquired”) (
) (Table 2, Figure 1).
Figure 1. Methicillin resistance rates among clinical Staphylococcus aureus isolates (MRSA rate) versus hospital size, according to number of beds. See Methods for the definition of the rate of methicillin resistance.
The 15 centers submitted 1845 enterococcal isolates (range, 106‐139 isolates per center), of which 118 (6%) were vancomycin resistant (range, 1%‐22% per center). Of the 118 VRE isolates, 76% had a vanA phenotype (teicoplanin resistant), and 24% had a vanB phenotype (teicoplanin susceptible). Fourteen centers submitted epidemiological information for 985 (53%) of these isolates (median, 57% of isolates per center; range, 9%‐90% of isolates per center), 71 (6%) of which were resistant to vancomycin. The basic demographic data, including patient location, age, sex, and specimen type, were similar for all enterococcal isolates and the isolates for which epidemiologic data were available. We excluded from the analyses all the enterococcal isolates submitted by one of the centers, which was an outlier with a VRE rate well in excess of all the other hospitals, regardless of size (Figure 2). We suspect that an ongoing outbreak may explain this outlier. Bivariate analyses identified a number of factors associated with infection with a vancomycin‐resistant isolate (Table 3). Stepwise logistic regression analysis identified the following independent risk factors for VRE infection: residence in a LTCF (
), admission to a hospital with >200 short‐term care beds (
), acquisition from a source other than the community (ie, designated as “not community acquired”) (
), and use of a central venous catheter (
) (Table 4). The correlation between VRE infection and a hospital size of more than 200 short‐term care beds was confirmed by a Spearman correlation coefficient of 0.72308 (
; Figure 2).
Figure 2. Vancomycin resistance rates among clinical enterococcal isolates (VRE rate) versus hospital size, according to number of beds. See Methods for the definition of the rate of vancomycin resistance.
Discussion
Most studies of MRSA and VRE infection and colonization have been performed at single academic teaching hospitals, many of which are located in large metropolitan areas. Using data from these studies, investigators have described risk factors for infection and carriage of MRSA and VRE.1,5 Some healthcare epidemiologists have emphasized the commonality of factors associated with infection due to resistant organisms, including advanced age, underlying disease, severity of illness, use of central venous catheters, use of antimicrobial agents, proximity to an infected or colonized patient, contact with the healthcare system, and length of hospital stay.1 However, a recent large survey of more than 400 US hospitals found that antimicrobial resistance rates were strongly correlated with certain hospital characteristics, including number of beds, affiliation with an academic institution, and geographical region.3 Thus, data from single academic institutions may not be generalizable to rural hospitals. To develop a more accurate description of the epidemiology of resistant organisms, healthcare epidemiologists must conduct multicenter studies that include small hospitals and rural hospitals.
We examined the epidemiology of MRSA and VRE infections across a spectrum of hospitals participating in a unique statewide surveillance program in Iowa. These data suggested that, although some risk factors are common to both MRSA and VRE infection, the epidemiology of these 2 organisms differ in Iowa hospitals. For example, the majority (59%) of the MRSA specimens in this study came from hospitals with <200 beds. Thus, our results indicate that MRSA infection is endemic in small Iowa hospitals, most of which are located in rural areas. Bivariate analysis showed that both hospitalization in a rural location and hospitalization in a facility with <200 beds were risk factors for MRSA infection. In the multivariate model, small hospital size and older age remained statistically significant risk factors, but rural location did not, indicating that rural location may be a confounder of the other 2 variables, because small hospitals are frequently located in rural areas and because a substantial percentage of Iowa’s rural population is older than 60 years.
In contrast, our statewide data suggest that VRE infection is mostly confined to larger urban hospitals. Hospitalization in a smaller hospital appeared to be protective against VRE infection, in contrast to the situation for MRSA infection. The risk factors for VRE infection (eg, presence of a central venous catheter and previous intensive care unit stay) seemed to be associated with the presence of underlying disease and the severity of illness, which may be associated with hospitalization in a larger hospital or with early transfer to a larger referral hospital.
These findings have implications for MRSA and VRE control in Iowa. For example, the most resource‐intensive MRSA or VRE control measures, such as active surveillance or “search and destroy”5,6 approaches, are more likely to be used in large tertiary care centers. For organisms such as VRE, for which there may not be a large reservoir of resistance outside of large tertiary care hospitals,13 control measures used in selected larger facilities may reduce rates of infection over time. However, because MRSA is common and well established in smaller rural Iowa hospitals and among elderly Iowans residing in LTCFs, use of aggressive approaches in only a few hospitals is not likely to cause sustained decreases in the rates of MRSA infection in hospitals statewide. A uniform approach that is practiced across the continuum of care is likely necessary to successfully control MRSA in Iowa. Moreover, we have found that community‐acquired MRSA carrying staphylococcal chromosomal cassette type IV and genes encoding Panton‐Valentine leukocidin are geographically widespread in Iowa.14 The emergence of community‐acquired MRSA outside of healthcare settings expands the reservoir of MRSA and adds additional challenges and complexity to control efforts.
Of note, one common risk factor for both MRSA infection and VRE infection was residence in a LTCF, suggesting that both organisms are endemic in these settings in Iowa. This suggestion was confirmed by a recent survey of LTCFs in Iowa.15 Any active surveillance program would therefore need to include recent stay in a LTCF as a risk factor that prompts the performance of surveillance cultures for detection of MRSA and/or VRE. The challenges of controlling the spread of resistant organisms within LTCFs must be taken into account in any strategy to decrease the spread of MRSA and VRE. Iowa has established statewide guidelines for control of antibiotic resistance that specifically address LTCFs.16
Our study has several limitations. Only approximately half of the isolates submitted were accompanied by data forms detailing risk factors, a relatively small number of isolates per center were examined (40 per year), and the determination of infection was subjectively made at each individual center (leading to a risk of misclassification of colonization as infection).
Nonetheless, the data we gathered in this surveillance program have helped guide our efforts to control antimicrobial resistance in Iowa. For example, when we presented the data on the epidemiology of MRSA and VRE in Iowa during dozens of educational sessions across the state, we found that some personnel in small rural hospitals still believed that antimicrobial resistance is primarily a problem restricted to large tertiary care centers. These data helped us to emphasize the scope of the problem in Iowa and to motivate healthcare workers in small rural hospitals to improve infection control practices. However, given the resources required for aggressive, comprehensive, and consistent strategies to control MRSA and VRE, such programs are not likely to be implemented until antimicrobial resistance is viewed less as a patient safety issue in specific hospitals and more as a widespread public health crisis. The increasing incidence of severe community‐acquired MRSA infection17,18 and the emergence of vancomycin‐resistant S. aureus in the United States19‐22 certainly warrant such a paradigm shift.
Acknowledgments
We are grateful for financial support provided by the Centers for Disease Control and Prevention’s Prevention Epicenter Program (CCU715091‐07), Abbott Laboratories, Glaxo Wellcome, Ortho‐McNeil Pharmaceuticals, Pfizer, SmithKline Beecham Pharmaceuticals, and Wyeth‐Ayerst Laboratories. We are also indebted to the 15 medical centers and numerous individuals involved in EIEIO. We thank the following infection control professionals who made this study possible: Dee Pederson, Yvonne Hunt, Shelley Zarling, Barb Livingston, Paula Simplot, Kim Horn, Bill Farmer, Barb Dean, Brenda DePue‐Weber, Kay Detweiler, Jackie Roth, Janet Hefel, Kelly Sterk, Mary Fer and Angie Ross, Nancy Coulter, and LaVonne Beneke.
References
- 1. Safdar N, Maki DG. The commonality of risk factors for nosocomial colonization and infection with antimicrobial‐resistant Staphylococcus aureus, enterococcus, gram‐negative bacilli, Clostridium difficile, and Candida. Ann Intern Med 2002;136:834‐844.
- 2. Centers for Disease Control and Prevention. National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2003, issued August 2003. Am J Infect Control 2003;31:481‐498.
- 3. Diekema DJ, BootsMiller T, Vaughn T, et al. Antimicrobial resistance trends and outbreak frequency in United States hospitals. Clin Infect Dis 2004;38:78‐85.
- 4. Edmond MB, Wallace SE, McClish DK, Pfaller MA, Jones RN, Wenzel RP. Nosocomial bloodstream infections in US hospitals: a three‐year analysis. Clin Infect Dis 1999;29:239‐244.
- 5. Muto CA, Jernigan JA, Ostrowsky BE, et al. SHEA guideline for preventing nosocomial transmission of multidrug‐resistant strains of Staphylococcus aureus and enterococcus. Infect Control Hosp Epidemiol 2003;24:362‐386.
- 6. Verhoef J, Beaujean D, Blok H, et al. A Dutch approach to methicillin‐resistant Staphylococcus aureus. Eur J Clin Microbiol Infect Dis 1999;18:461‐466.
- 7. Jarvis WR. Controlling antimicrobial‐resistant pathogens. Infect Control Hosp Epidemiol 2004;25:369‐372.
- 8. Farr BM, Jarvis WR. Would active surveillance cultures help control healthcare‐related methicillin‐resistant Staphylococcus aureus infections? Infect Control Hosp Epidemiol 2002;23:65‐68.
- 9. Beekmann SE, Diekema DJ, Brueggemann AB, et al. Epidemiology of methicillin‐resistant Staphylococcus aureus and prevalence of true community‐acquired MRSA in an Iowa statewide surveillance network. In: Program and abstracts of the Fourth Decennial International Conference on Nosocomial and Healthcare‐Associated Infections; March 5‐9, 2000; Atlanta, GA. Abstract P‐T2‐42.
- 10. Diekema DJ, Messer SA, Coffman SA, et al. The epidemiology of candidemia: three year results from the Emerging Infections and the Epidemiology of Iowa Organisms study. J Clin Microbiol 2002;40:1298‐1302.
- 11. National Committee for Clinical Laboratory Standards (NCCLS). Methods for Dilution Antimicrobial Tests for Bacteria That Grow Aerobically. Approved standard M7‐A4. Wayne, PA: NCCLS; 2000.
- 12. National Committee for Clinical Laboratory Standards (NCCLS). Performance Standards for Antimicrobial Susceptibility Testing. Supplemental tables M100‐S10. Wayne, PA: NCCLS; 2000.
- 13. Stevenson KB, Searle K, Stoddard GJ, Samore MH. Methicillin‐resistant Staphylococcus aureus and vancomycin‐resistant enterococci in rural communities, Western United States. Emerg Infect Dis 2005;11:895‐903.
- 14. Ince D, Winokur PL, Beekmann S, et al. Prevalence of Panton‐Valentine leukocidin genes among methicillin‐resistant Staphylococcus aureus in Iowa. In: Program and abstracts of the Society for Healthcare Epidemiology of America 15th Annual Scientific Meeting; April 9‐12, 2005; Los Angeles, CA. Abstract 312.
- 15. Kreman T, Hu J, Pottinger J, Herwaldt LA. Survey of long‐term care facilities in Iowa for policies and practices regarding residents with methicillin‐resistant Staphylococcus aureus or vancomycin‐resistant enterococci. Infect Control Hosp Epidemiol 2005;26:811‐815.
- 16. Iowa Department of Public Health. Report of the Iowa Antibiotic Resistance Task Force: A Public Health Guide. 2004. Available at: http://www.idph.state.ia.us/adper/common/pdf/cade/antibioticreport.pdf. Accessed February 20, 2006.
- 17. Centers for Disease Control and Prevention. Four pediatric deaths from community‐acquired methicillin‐resistant Staphylococcus aureus—Minnesota and North Dakota, 1997‐1999. JAMA 1999;282:1123‐1125.
- 18. Herold BC, Immergluck LC, Maranan MC, et al. Community‐acquired MRSA in children with no identified predisposing risk. JAMA 1998;279:593‐598.
- 19. Centers for Disease Control and Prevention. Staphylococcus aureus resistant to vancomycin—United States, 2002. MMWR Morb Mortal Wkly Rep 2002;51:565‐567.
- 20. Centers for Disease Control and Prevention. Vancomycin‐resistant Staphylococcus aureus—Pennsylvania, 2002 [published correction appears in MMWR Morb Mortal Wkly Rep 2002; 51:931]. MMWR Morb Mortal Wkly Rep 2002;51:902.
- 21. Centers for Disease Control and Prevention. Vancomycin‐resistant Staphylococcus aureus—New York, 2004. MMWR Morb Mortal Wkly Rep 2004;53:322‐323.
- 22. Michigan Department of Community Health. Bureau of laboratory broadcast fax. Available at: http://www.michigan.gov/documents/VRSA_Feb05_HAN_118391_7.pdf. Accessed May 31, 2005.