National prevalence of methicillin-resistant Staphylococcus aureus in inpatients at US health care facilities, 2006

National prevalence of methicillinresistant Staphylococcus aureus in inpatients at US health care facilities, 2006 William R. Jarvis, MD,a JoAnn Schlosser, MA,b Raymond Y. Chinn, MD,c Samantha Tweeten, MPH, PhD,d and Marguerite Jackson, PhD, RNe Hilton Head Island, South Carolina; Washington, DC; and San Diego, California

Background: Despite methicillin-resistant Staphylococcus aureus (MRSA) being endemic in virtually all US health care facilities, there are no data on the prevalence of MRSA in US health care facilities. Methods: We conducted a national prevalence survey of MRSA in inpatients at US health care facilities. The survey was developed, received institutional review board approval, and then distributed to all members of the Association for Professionals in Infection Control and Epidemiology, Inc. (APIC). Members were asked to complete the survey on one day during the period October 1 to November 16, 2006, reporting the number of inpatients with MRSA infection or colonization and facility-specific information. Results: Personnel at 1237 hospitals completed the survey. Complete facility data were provided for 1187 (96%) of these health care facilities. All states were represented (mean, 23 facilities per state; range, 1-99). Respondents reported 8654 MRSA-colonized/infected patients in 187,058 inpatients; the overall MRSA prevalence rate was 46.3 per 1000 inpatients (34 infections and 12 colonizations per 1000 inpatients). Active MRSA surveillance testing was conducted by 29% of respondents: 54% used routine media, 38% used selective media, and 8% used polymerase chain reaction. Detailed data were provided on 7994 (92.4%) MRSAcolonized/infected patients. Our data suggest that approximately 70% of isolates were more consistent with health care-associated MRSA (HA-MRSA) than community-associated MRSA. Conclusion: Our survey documents a much higher MRSA prevalence rate than previous studies using different methodologies. The majority of MRSA in inpatients appears to be HA-MRSA. Given that most facilities did not perform active surveillance testing, these are minimum estimates of the national burden of MRSA in US health care facilities. (Am J Infect Control 2007;35:631-7.)

Since methicillin-resistant Staphylococcus aureus (MRSA) was first reported as a health care-associated infection (HAI) pathogen in US hospitals, it has rapidly become endemic in virtually all US health care facilities.1 From Jason and Jarvis Associates,a Hilton Head Island, SC; Association for Professionals in Infection Control and Epidemiology, Inc.,b Washington, DC; Sharp Memorial Hospital,c San Diego, CA; County of San Diego Health and Human Services Agency,d San Diego, CA; and University of California San Diego School of Medicine,e San Diego, CA. Address correspondence to William R. Jarvis, MD, 135 Dune Lane, Hilton Head Island, SC 29928. E-mail: [email protected]. Reprints are not available. Supported by the APIC Scientific Research Council (previously the Research Foundation) and the APIC. This study was presented at the 2007 Annual APIC meeting, San Jose, CA. 0196-6553/$32.00 Copyright ª 2007 by the Association for Professionals in Infection Control and Epidemiology, Inc. doi:10.1016/j.ajic.2007.10.009

Despite the fact that it is the most common multidrugresistant organism causing HAIs, no comprehensive national health care facility prevalence or incidence study data were available before our survey. Several studies have estimated the burden of MRSA in US health care facilities. In 1982, Boyce and Causey2 reported the results of a 1980 survey of 341 US hospitals and found that 145 (56%) had more than 1 MRSAinfected patient. They also found that these patients were more likely to be at large university teaching hospitals. From 1975 to 1980, the number of hospitals reporting MRSA increased from 24 facilities in 17 states to 112 facilities in 36 states. In 2005, Kuehnert et al3 estimated MRSA incidence using data from the National Hospital Discharge Survey (NHDS). Using International Classification of Disease (ICD)-9 codes for 1999-2000, they estimated that, at 475 hospitals, discharge diagnoses for 125,969 hospitalizations likely were due to MRSA, providing a calculated overall MRSA incidence of 3.95 per 1000 discharges. Most recently, Elixhauser and Steiner of the Agency for 631

632

Vol. 35 No. 10

Healthcare Research and Quality updated this study and found that, in 2005, the number of MRSA hospital stays increased to 368,600; this was a 10-fold increase from 1995, a 3-fold increase from 2000, and a 30% increase from 2004 (http://www.hcup-us.ahrq.gov/ reports/statbriefs/sb35.pdf). In 2006 Klevens et al4 analyzed data from the intensive care unit (ICU) component of the Centers for Disease Control and Prevention’s (CDC) National Nosocomial Infections Surveillance (NNIS) system, which included 1268 ICUs in 337 hospitals. Between 1992 and 2003, they found that the number of MRSA infections increased from 660 to 2184 and that the proportion of all S aureus infections caused by MRSA increased from 35.9% to 64.4%. Each of the above studies has many limitations. Some are 25 years old and included a small number of US facilities. Those using the CDC’s NNIS ICU component data include only a small number of US hospitals (;211 of ;5500 US hospitals or a 4% sample) in the ICU component. Furthermore, CDC NNIS hospitals have variable regional representation, are a nonrandom (convenience) sample, are biased toward large teaching hospitals, do not include hospital-wide data, do not include nonacute care facilities or acute care facilities #100 beds, and only include HAIs (not MRSAcolonization or HA-MRSA with community-onset [CO] or community-associated [CA] MRSA) data. Analyses using the NHDS database include only 475 acute care hospitals (no Veteran’s Administration [VA], military, or nonacute care). The NHDS database does not include antimicrobial susceptibility data (requiring data from another source for this information), does not use standardization of HAI definitions, does not include data on MRSA colonization, nor does it differentiate HA- versus CO- or CA-MRSA. Because of the limitations of existing data, the increasing importance of MRSA, the national debate about whether attempts to control MRSA (as has been done in Northern Europe and Western Australia) should be attempted in the United States, and, if so, what control measures should be recommended, led us to attempt to determine the national prevalence of MRSA in US health care facilities.

METHODS Our survey was developed, approved by the Association for Professionals in Infection Control and Epidemiology, Inc. (APIC) Research Foundation (APIC-RF; now the Scientific Research Council) and the APIC Board of Directors, beta tested at several APIC-RF member hospitals, and then approved by an institutional review board. To reach all APIC members, APIC announced the survey on August 16 and 30, September

Jarvis

15 and 27, October 1 and 25, and November 8, 2006, through various e-mail or/and via print communications during that period. The electronic survey was placed on the APIC secure Web site and was available in hard copy, for those without Internet access. All APIC members were invited to participate and to encourage non-APIC infection control professionals (ICPs) to do so. Each ICP determined whether additional institutional review board review was needed at his or her facility. ICPs were asked to identify one day between October 1 and November 10, 2006, on which to conduct a MRSA prevalence survey. On the day selected, participants were asked to identify all patients known to be infected or colonized with MRSA. No patient cultures or patient interviews were requested or required. The survey questionnaire had 2 parts. Part I included facility data such as location (state), type (public, private; adult, children’s, women’s; acute, long-term care), number of licensed beds, number of inpatients on the survey day; and types of ICUs, specialty services, HAI surveillance performed (site and patient populations), and HAI surveillance definitions used. In addition, data were collected on the number of known MRSA inpatients on the survey date, what isolation measures are taken for MRSA-culture positive patients (and if different by ward vs ICU), whether active surveillance testing is done routinely to detect MRSA-colonized patients, and, if yes, what populations are tested, and what microbiologic methods are used. Part II of the survey included detailed data on each patient identified as MRSA colonized or infected, including age, gender, service, duration from admission to MRSA-positive test; whether the patient was identified by clinical cultures, active surveillance testing or both, colonization, or infection status (if infected, the site); underlying conditions; and susceptibility to 6 antimicrobials (ie, levofloxacin, clindamycin, tetracycline, gentamicin, trimethoprim-sulfamethoxazole, and rifampin). If respondents had Internet access, all survey responses were entered directly by hospital personnel onto the data collection form on the APIC secure Web site. If respondents did not have Internet access, they completed paper copies of the survey that then were faxed to APIC headquarters, at which the data were entered into the database. The distribution of hospitals responding to the survey was compared with the distribution of all US acute care facilities from the American Hospital Association (AHA) hospital statistics tables for 2005.6 Comparisons were made using SPSS Base 8.0 for Windows (SPSS, Chicago, IL). All other data analyses were conducted using Excel (Microsoft Corp, Redman, WA) or Epi Info version 3.3 (Centers for Disease Control and Prevention, October 2004).

Jarvis

RESULTS There were 1237 health care facility respondents who completed the survey questionnaire. Complete facility data were provided for 1187 (96%) hospitals. When we compared the distribution of these respondent health care facilities by size and geographic distribution to that of the AHA, our survey respondents were significantly more likely than AHA hospitals to be $200 beds (49.6% vs 28.1%, respectively), equally likely to be 100 to 199 beds (23.0% vs 22.9%, respectively), and significantly less likely than AHA hospitals to be ,100 beds (26.0% vs 48.9%, respectively) (Table 1). When compared by census division, our survey facilities were significantly more likely than AHA hospitals to be in the New England, Middle Atlantic, and East North Central regions and significantly less likely to be in the West South Central, West North Central, and Mountain regions (Table 2). There were no significant differences between survey respondents and AHA hospitals for the South Atlantic, East South Central, or Pacific regions. There were health care facility respondents from every US state (mean, 23; range, 1-99 health care facilities) (Fig 1). Respondent health care facilities had an average of 1.6 ICPs, 45% performed hospital-wide HAI surveillance, and 55% performed targeted HAI surveillance. Most (.72%) used contact isolation for identified MRSA-infected or -colonized patients. The health care facility respondents reported 8654 MRSA colonized or infected patients in 187,058 inpatients, giving an overall MRSA prevalence of 46.3 per 1000 inpatients. There were 34 MRSA infections per 1000 inpatients and 12 MRSA colonizations per 1000 inpatients. The MRSA prevalence rate varied from 0 in South Dakota (only 1 health care facility reporting) to 91 per 1000 inpatients in Hawaii (Fig 2). Detailed data were provided on 7926 (91.6%) of the MRSA colonized/infected patients reported: 54% were male, 80% were detected by clinical culture, and 69% were detected ,48 hours after admission. Infection sites included (data not mutually exclusive) 28% skin and soft tissue, 17% pulmonary, 13% bloodstream, 10% urinary tract, and 8% surgical site. The majority (69%) of MRSA patients were on medical services at the time MRSA testing was obtained. Of MRSA isolates tested, 33% were susceptible to clindamycin, and ,18% were susceptible to levofloxacin. Using the latter 3 criteria (ie, time from admission to diagnosis, site of infection, and susceptibility to clindamycin and levofloxacin) to differentiate HA-MRSA from CAMRSA, approximately 70% of the isolates were more consistent with HA-MRSA. Of the responding health care facilities, 29% performed active MRSA surveillance testing. Populations in which such testing was performed included (not

December 2007

633

Table 1. Comparison of bed sizes of facilities represented by respondents to APIC survey, 2006, and US registered hospitals,5 2005 Number AHA AHA beds (range) frequency percent

APIC survey

APIC Difference percent (P value)

6-24 25-49 50-99 100-199

419 1138 1257 1319

7.3 19.8 21.8 22.9

18 134 157 273

1.5 11.3 13.2 23.0

200-299 300-399 400-499 5001 Missing Total

709 416 196 302

12.3 7.2 3.4 5.2

198 162 90 139 16 1187

16.7 13.6 7.3 11.7 1.3

5756

,.0001 ,.0001 ,.0001 .9501 NS ,.0001 ,.0001 ,.0001 ,.0001

NS, difference is not statistically significant.

mutually exclusive) long-term care facility transfers (42%), other health care facility transfers (33%), patients with repeated admissions (20%), patients on selected wards (18%), ICU patients (16%), or dialysis patients (14%). Of those performing active MRSA surveillance testing, 54% used routine media, 35% used selective media, and 9% used polymerase chain reaction.

DISCUSSION Since the introduction of MRSA into US health care facilities, there has been an interest in the infection control community to determine periodically the incidence or prevalence of MRSA. Unfortunately, none of the current or past existing databases provide adequate information to make direct assessments of the current prevalence or incidence of MRSA in US health care facilities. The CDC’s NNIS system no longer collects hospital-wide component data, the data reported from the ICU component represent fewer than 350 US hospitals, and NNIS does not include any facilities with 100 or fewer beds (the median size of a US hospital is less than 100 beds).4 Furthermore, the NNIS system only collects data on infections, but no data on MRSA colonization or CO- or CA-MRSA are included. Similarly, the NHDS only includes data from 475 hospitals and does not include any antimicrobial susceptibility testing results,3 requiring that indirect estimates of the rate of MRSA discharge diagnoses be made using another microbiology database. In addition, a number of studies have shown the inadequacy of administrative databases compared with prospective collection of data by ICPs.6-8 Nevertheless, the NNIS data show that the rates of MRSA infections have doubled in the ICU component between 1992 and 2003, and the proportion of S aureus HAIs caused by MRSA has tripled

634

Jarvis

Vol. 35 No. 10

Table 2. Comparison of American Hospital Association and Association for Professionals in Infection Control and Epidemiology, Inc, survey respondent acute care hospital geographic distributions Census division* Division 1: New England (CT, ME, MA, NH, RI, VT) Division 2: Middle Atlantic (NJ, NY, PA) Division 3: South Atlantic (DE, DC, FL, GA, MD, NC, SC, VA, WV) Division 4: East North Central (IL, IN, MI, OH, WI) Division 5: East South Central (AL, KY, MS, TN) Division 6: West North Central (IA, KS, MN, MO, NE, ND, SD) Division 7: West South Central (AR, LA, OK, TX) Division 8: Mountain (AZ, CO, ID, MT, NV, NM, UT, WY) Division 9: Pacific (AK, CA, HI, OR, WA) Missing data 1 Guam (n 5 1) Total

AHA frequency

AHA percent

206

4.2

474

APIC percent

Difference (P value)

86

7.2

,.0001

9.6

174

14.7

,.0001

743

15.1

189

15.9

744

15.1

230

19.4

.9303 NS .0003

438

8.9

86

7.2

678

13.7

110

9.3

.0717 NS ,.0001

738

15.0

112

9.4

,.0001

367

7.4

52

4.4

.0002

548

11.1

139

11.7

.5512 NS

9 1187

0.8

4936

APIC survey

AHA, American Hospital Association; APIC, Association for Professionals in Infection Control and Epidemiology, Inc; NS, difference is not statistically significant. *US census divisions for community hospitals only (excludes federal and some specialty facilities)

during this period.4 The NHDS analysis found that the estimated rate of MRSA discharge diagnoses was 3.95 per 1000 discharges.3 More recent analyses of the NHDS found that, by 2005, the number of MRSA-associated hospital stays had increased from 125,969 to 368,600, a 30% increase in the number of discharges with such a diagnosis over the 2004 level (http:// www.hcup-us.ahrq.gov/reports/statbriefs/sb35.pdf). Because of the limitations in the databases used for previous estimates of the magnitude of MRSA in US health care facilities, we sought to determine the prevalence by asking ICPs in the health care facilities to identify all patients with MRSA at their facilities on one day in the fall of 2006. We believe this was a more direct and accurate method to determine the ‘‘true’’ magnitude of the burden of MRSA. The 1237 health care facilities responding represent at least 24% of all US acute care facilities (based on AHA 2004 data). This is the largest sample of US health care facilities ever reporting their MRSA prevalence. In addition, our respondent health care facilities were more closely reflective of the size and geographic distributions of all US acute care facilities than those included in any previous estimate of the magnitude of MRSA in US health care facilities. Our respondents reported 8654 MRSA-colonized or -infected patients in 187,058 inpatients. Thus, the overall MRSA prevalence in US health care facilities was 46.3 per 1000 inpatients. This is 8 to 11 times higher than any previous incidence estimates (using

different methodologies). Our MRSA rate should be considered a minimum estimate because some facilities, eg, larger health care facilities or those with higher numbers of MRSA patients, may not have completed the survey; and fewer than 30% of health care facilities reported performing any active surveillance testing for MRSA (thus failing to detect the MRSA-colonized patient); and, even those who did perform MRSA active surveillance testing, usually used less sensitive methods, such as routine nonselective media. As a part of the survey, we wanted to differentiate HA- from CA-MRSA. This is particularly important because of the recent emergence of CA-MRSA. The study design did not include obtaining and testing the MRSA isolates or conducting interviews with the identified MRSA patients to determine recent health care exposure. However, we did collect 3 variables, each of which often have been used either alone or in combination in previous studies to categorize HA-, CA-, or CO-MRSA.9-22 These include the site of infection (skin/soft tissue or not), the duration from admission to MRSA detection (48 hours or less), and antibiotic susceptibility for selected antimicrobials. Using all 3 of these criteria, we estimate that approximately 70% of the isolates reported were likely to be HAMRSA, and approximately 30% were more consistent with CA-MRSA. Our data suggest that, although CAMRSA has received enormous recent media attention and researcher interest, HA-MRSA continues to account for the majority of the MRSA burden in US

Jarvis

December 2007

635

Fig 1. The APIC National MRSA Inpatient Survey: results-number of responding facilities by state.

Fig 2. The APIC National MRSA Inpatient Survey: MRSA prevalence rates by state. health care facility inpatients. Currently, the majority of all MRSA infections in inpatients is detected by clinical culture, within 48 hours of admission, and involves sites of infection other than skin and soft tissue. The majority of the patients in our survey was located on medical services at the time MRSA was detected, contrary to the general perception that MRSA is most common in ICU patients.

Our survey had several limitations. First, all US health care facilities did not respond, and we do not know whether the nonrespondent facilities were similar to the respondent facilities. However, given the large number of health care facilities responding and the geographic and size distribution of these facilities, we believe that our results strongly suggest trends that could be generalized to all US health care facilities.

636

Jarvis

Vol. 35 No. 10

Second, the small number of respondents in some states makes generalizations difficult, and caution should be used in calculating or interpreting a MRSA prevalence rate in states with small numbers (eg, ,5 respondents and 20% of the health care facilities in the state) of respondents. However, many states had a large proportion (.25%) of the health care facilities in that state reporting; in such states, there can be greater confidence about the accuracy of the MRSA prevalence rate point estimate. Third, only 29% of respondent facilities performed any MRSA active surveillance testing. Previous studies have indicated that only 16% to 36% of patients with MRSA would be detected with clinical cultures alone.23,24 Of those performing MRSA active surveillance testing, most used routine nonselective media, which has been shown to be much less sensitive than either selective media or polymerase chain reaction. Thus, it is likely that the true MRSA prevalence in US health care facilities is higher than we detected. Fourth, without confirmatory genotyping and staphylococcal cassette mec-A testing of each isolate and interviewing each MRSA-positive patient for recent antimicrobial and health care exposure, it is not possible to absolutely differentiate HA- and CA- or CO-MRSA. Even with these data, there is considerable debate about the best definitions to use to differentiate HA-, CO-, and CA-MRSA.25-29 Many previous authors have used a single criterion of detection of MRSA positivity at greater than 48 hours after health care facility admission to differentiate HA- from COMRSA or CA-MRSA.12,16,18-21,29 Others have shown that the majority of CA-MRSA infections involve the skin and soft tissue and that the ‘‘classical’’ CA-MRSA is much less resistant than the ‘‘classical’’ HA-MRSA strain.10,11,14,16-22,26 To attempt an ‘‘accurate’’ estimate, we evaluated our data in terms of all 3 of these criteria and believe that approximately 70% of the MRSA patients reported are likely to be HA-MRSA and should continue to be the focus of our prevention efforts. However, the continued emergence of CA-MRSA, which also can be introduced and transmitted within health care facilities, cannot be neglected. We hope our data will alert health care facility administrative and infection control personnel to the importance of implementation of comprehensive evidence-based MRSA prevention measures.24,30-32 These include conducting a risk assessment to identify the patients at greatest risk of MRSA infection or colonization; use of active surveillance testing to identify this reservoir upon admission; placing identified MRSA-colonized or -infected patients in contact isolation; reinforcing the importance of hand hygiene for the care of all patients, including after contact with identified MRSA patients or their potentially contaminated environment; and reinforcing the importance of good environmental cleaning

throughout the facility. Through such comprehensive control measures, we believe that health care facilities in the United States can be as successful in controlling MRSA as health care facilities in Northern Europe or Western Australia33 and that the US MRSA prevalence will begin to decrease measurably in the near future. The authors thank the APIC Scientific Research Council, the APIC Board, Kathy Wayre, Denise Graham, and Jill Midgett for their assistance and support in conducting this survey.

References 1. Haley RW, Hightower AW, Khabbaz RF, Thornsberry C, Martone WJ, Allen JR, et al. The emergence of methicillin-resistant Staphylococcus aureus infections in United States hospitals: possible role of the house staff-patient transfer circuit. Ann Intern Med 1982;97:297-308. 2. Boyce JM, Causey WA. Increasing occurrence of methicillin-resistant Staphylococcus aureus in the United States. Infect Control 1982;3:377-83. 3. Kuehnert MJ, Hill HA, Kupronis BA, Tokars JI, Solomon SL, Jernigan DB. Methicillin-resistant-Staphylococcus aureus hospitalizations, United States. Emerg Infect Dis 2005;11:868-72. 4. Klevens RM, Edwards JR, Tenover FC, McDonald LC, Horan T, Gaynes R. National Nosocomial Infections Surveillance System. Changes in the epidemiology of methicillin-resistant Staphylococcus aureus in intensive care units in US hospitals, 1992-2003. Clin Infect Dis 2006;43:387-8. 5. American Hospital Association. AHA Hospital Statistics 2005. Chicago, IL: AHA; 2007. 6. Simonds DN, Jarvis WR, Horan TC, Kelly R, Jarvis WR. Detecting pediatric nosocomial infections: how do infection control and quality assurance personnel compare? Am J Infect Control 1997;25:202-8. 7. Sherman ER, Heydon KH, St. John KH, Texner E, Rettig SL, Alexander SK, et al. Administrative data fail to accurately identify cases of healthcare-associated infection. Infect Control Hosp Epidemiol 2006;27: 332-7. 8. Wright SB, Huskins WC, Dokholyan RS, Goldmann DA, Platt R. Administrative databases provide inaccurate data for surveillance of longterm central venous catheter-associated infections. Infect Control Hosp Epidemiol 2003;24:946-9. 9. Moran CA, Hadler JL. Population-based incidence and characteristics of community-onset Staphylococcus aureus infections with bacteremia in 4 metropolitan Connecticut areas, 1998. J Infect Dis 2001;184: 1029-34. 10. Rosario-Rosado RV, Rene AA, Jones B. Descriptive analysis of patients with community-onset and hospital-onset methicillin-resistant Staphylococcus aureus infections. Infect Control Hosp Epidemiol 2004;25:171-3. 11. King MD, Humphrey BJ, Wang YF, Kourbarova EV, Ray SM, Blumberg HM. Emergence of community-acquired methicillin-resistant Staphylococcus aureus USA 300 clone as the predominant cause of skin and soft-tissue infections. Ann Intern Med 2006;144:309-17. 12. File TM Jr. Impact of community-acquired methicillin-resistant Staphylococcus aureus in the hospital setting. Cleve Clin J Med 2007;74(Suppl 4): S6-11. 13. Furuya EY, Cook HA, Lee MH, Miller M, Larson E, Hyman S, et al. Community-associated methicillin-resistant Staphylococcus aureus prevalence: how common is it? A methodological comparison of prevalence ascertainment. Am J Infect Control 2007;35:359-66. 14. Popovich K, Hota B, Rice T, Aroutcheva A, Weinstein RA. Phenotypic prediction rule for community-associated methicillin-resistant Staphylococcus aureus. J Clin Microbiol 2007;45:2293-5. 15. Tsuji BT, Rybak MJ, Cheung CM, Amjad M, Kaatz GW. Communityand health care-associated methicillin-resistant Staphylococcus aureus: a comparison of molecular epidemiology and antimicrobial activities of various agents. Diagn Microbiol Infect Dis 2007;58:41-7.

Jarvis 16. Skiest DJ, Brown K, Cooper TW, Hoffman-Roberts H, Mussa HR, Elliott AC. Prospective comparison of methicillin-susceptible and methicillin-resistant community-associated Staphylococcus aureus infections in hospitalized patients. J Infect 2007;54:427-34. 17. Huang H, Flynn NM, King JH, Monchaud C, Morita M, Cohen SH. Comparisons of community-associated methicillin-resistant Staphylococcus aureus (MRSA) and hospital-associated MSRA infections in Sacramento, California. J Clin Microbiol 2006;44:2423-7. 18. Frank AL, Marcinak JF, Mangat PD, Schreckenberger PC. Communityacquired and clindamycin-susceptible methicillin-resistant Staphylococcus aureus in children. Pediatr Infect Dis J 1999;18:993-1000. 19. Herold BC, Immergluck LC, Maranan MC, Lauderdale DS, Gaskin RE, Boyle-Vavra S, et al. Community-acquired methicillin-resistant Staphylococcus aureus in children with no identified predisposing risk. JAMA 1998;279:593-8. 20. Gorak EJ, Yamada SM, Brown JD. Community-acquired methicillinresistant Staphylococcus aureus in hospitalized adults and children without known risk factors. Clin Infect Dis 1999;29:797-800. 21. Sattler CA, Mason EO Jr, Kaplan SL. Prospective comparison of risk factors and demographic and clinical characteristics of community-acquired, methicillin-resistant versus methicillin-susceptible Staphylococcus aureus infection in children. Pediatr Infect Dis J 2002;21:910-7. 22. Martinez-Aquilar G, Hammerman WA, Mason EO Jr, Kaplan SL. Clindamycin treatment of invasive infections caused by communityacquired, methicillin-resistant and methicillin-susceptible Staphylococcus aureus in children. Pediatr Infect Dis J 2003;22:593-8. 23. Salgado CD, Farr BM. What proportion of hospital patients colonized with methicillin-resistant Staphylococcus aureus are identified by clinical microbiological cultures? Infect Control Hosp Epidemiol 2006;27: 116-21. 24. Muto CA, Jernigan JA, Ostrowsky BE, Richet HM, Jarvis WR, Boyce JM, et al. SHEA guideline for preventing nosocomial transmission of multidrug-resistant strains of Staphylococcus aureus and enterococcus. Infect Control Hosp Epidemiol 2003;24:362-86. 25. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128-40.

December 2007

637

26. Salgado CD, Farr BM, Calfee DP. Community-acquired methicillinresistant Staphylococcus aureus: a meta-analysis of prevalence and risk factors. Clin Infect Dis 2003;36:131-9. 27. Friedman ND, Kaye KS, Stout JE, McGarry SA, Trivette SL, Briggs JP, et al. Health-care-associated bloodstream infections in adults: a reason to change the accepted definition of community-acquired infections. Ann Intern Med 2002;137:791-7. 28. Lesens O, Hansmann Y, Brannigan E, Hopkins S, Meyer P, O’Connel B, et al. Healthcare-associated Staphylococcus aureus bacteremia and the risk for methicillin resistance: is the Centers for Disease Control and Prevention definition for community-acquired bacteremia still appropriate? Infect Control Hosp Epidemiol 2005;26:204-9. 29. Layton MC, Hierholzer WJ Jr, Patterson JE. The evolving epidemiology of methicillin-resistant Staphylococcus aureus at a university hospital. Infect Control Hosp Epidemiol 1995;16:12-7. 30. Seigel JD, Rhinehart E, Jackson M, Chiarello L, and the Healthcare Infection Control Practices Advisory Committee. Management of multidrug-resistant organisms in healthcare settings, 2006. Available at: http://www.cdc.gov/ncidod/dhqp/index.html. Accessed November 2007. 31. Seigel JD, Rhinehart E, Jackson M, Chiarello L, and the Healthcare Infection Control Practices Advisory Committee. 2007 Guideline for isolation precautions: preventing transmission of infectious agents in healthcare settings. Available at: http://www.cdc.gov/ncidod/dhqp/ gl.isolation.html. Accessed November 2007. 32. Association for Professionals in Infection Control and Epidemiology, Inc. Guide to the elimination of methicillin-resistant Staphylococcus aureus (MRSA) transmission in hospital settings. APIC, Washington, DC. MRSA implementation guide. Available at: http://www.apic.org/ Content/NavigationMenu/GovernmentAdvocacy/MethicillinResistant StaphylococcusAureusMRSA/Resources/MRSAguide.pdf. Accessed November 2007. 33. Salgado CD, Farr BM. The importance of infection control in controlling antimicrobial resistant pathogens. In: Jarvis WR, editor. Hospital infections. Philadephia: Wolters Kluwer/Lippincott Williams and Wilkins; 2007. p. 673-88.