Fifteen-Year Study of the Changing Epidemiology of Methicillin-Resistant Staphylococcus aureus

Transcription

1 The American Journal of Medicine (2006) 119, CLINICAL RESEARCH STUDY AJM Theme Issue: Infectious Disease Fifteen-Year Study of the Changing Epidemiology of Methicillin-Resistant Staphylococcus aureus Nancy F. Crum, MD, MPH, a Rachel U. Lee, MD, a Scott A. Thornton, MS, b O. Colin Stine, PhD, c Mark R. Wallace, MD, a Chris Barrozo, d Ananda Keefer-Norris, c Sharon Judd, RN, a Kevin L. Russell, MD, MTM&H d a Infectious Diseases Division, Naval Medical Center San Diego, San Diego, Calif; b Navy Environmental and Preventive Medicine Unit 6 (NEPMU6), Pearl Harbor, Hawaii; c University of Maryland School of Medicine, Baltimore, Md; d Naval Health Research Center, San Diego, Calif. ABSTRACT PURPOSE: The study s purpose was to elucidate the evolutionary, microbiologic, and clinical characteristics of methicillin-resistant Staphylococcus aureus (MRSA) infections. METHODS: MRSA cases from military medical facilities in San Diego, from 1990 to 2004, were evaluated and categorized as community-acquired or nosocomial. Sequence type, staphylococcal chromosomal cassette gene type, and Panton-Valentine leukocidin gene status were determined for a subset of isolates. RESULTS: Over the 15-year period, 1888 cases of MRSA were identified; 65% were community acquired. The incidence (155 infections/ person-year in 2004) and household-associated cases rapidly increased since Among persons with community-acquired MRSA, 16% were hospitalized and only 17% were initially given an effective antibiotic. Community-acquired MRSA cases compared with nosocomial MRSA cases were more often soft-tissue and less often urinary, lung, or bloodstream infections (P.001). Patients with community-acquired MRSA were younger (22 vs 64 years, P.001) and less likely to have concurrent medical conditions (9% vs 98%, P.001). Clindamycin resistance increased among community-acquired MRSA isolates during 2003 and 2004 compared with previous years (79% vs 13%, P.001). Genetically, nosocomial MRSA isolates were significantly different than those acquired in the community. Although community-acquired MRSA isolates were initially diverse by 2004, one strain (staphylococcal chromosomal cassette type IV, sequence type 8, Panton-Valentine leukocidin gene positive) became the predominant isolate. CONCLUSIONS: Community-acquired and intrafamilial MRSA infections have increased rapidly since Our 15 years of surveillance revealed the emergence of distinct community-acquired MRSA strains that were genetically unrelated to nosocomial MRSA isolates from the same community Elsevier Inc. All rights reserved. KEYWORDS: Methicillin-resistant Staphylococcus aureus; MRSA; Community-acquired; Epidemiology The data in this article was presented in part at the 43rd Annual Meeting of the Infectious Diseases Society of America, San Francisco, California, October 6 to 9, 2005; Abstract This work was supported by the Department of Defense. The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Navy, Department of Defense, or United States Government. Requests for reprints should be addressed to Nancy F. Crum, MD, MPH, c/o Clinical Investigation Department (KCA), Naval Medical Center San Diego, Bob Wilson Drive, Ste. 5, San Diego, CA address: NFCrum@nmcsd.med.navy.mil. Staphylococcus aureus infections have always been among the most common causes of morbidity and mortality worldwide and have been described as the persistent pathogen. 1 Methicillin-resistant Staphylococcus aureus (MRSA) was first described in 1961 in the United Kingdom and has become a common worldwide pathogen. MRSA is thought to have appeared because of the selection pressure of antibiotic use within hospitals. 2,3 The first community-acquired MRSA infections occurred in 1980 and were associated with spread from hospitals into the community. 4 Over the past decade, community-acquired MRSA infections without apparent connection to health care facilities have occurred /$ -see front matter 2006 Elsevier Inc. All rights reserved. doi: /j.amjmed

2 944 The American Journal of Medicine, Vol 119, No 11, November 2006 with increasing frequency. 5-7 Deaths among children in the late 1990s led to heightened concern and indicated that community-acquired MRSA strains may have an increased virulence. 8 Since 2000, community-acquired MRSA cases have increasingly been reported, causing outbreaks among prison inmates, athletes, and military trainees The evolutionary history of community-acquired MRSA continues to be debated; recent evidence suggests that although some community-acquired MRSA isolates are derived from strains traceable to health care facilities, others developed de novo in the community Most community-acquired MRSA isolates seem to have distinct microbiologic and epidemiologic characteristics from nosocomial MRSA. Community-acquired MRSA isolates typically carry the meca gene on the type IV staphylococcal chromosomal cassette in contrast with type II in nosocomial MRSA cases. 14,18 Community-acquired MRSA also frequently carries the genes encoding Panton-Valentine leukocidin, a cytotoxin virulence factor associated with soft tissue infections and necrotizing pneumonias. 19,20 Clinically, community-acquired MRSA, compared with nosocomial MRSA, usually affects younger people with fewer comorbidities and has CLINICAL SIGNIFICANCE a propensity for skin and soft tissue infections. 21 Finally, community-acquired MRSA is typically sensitive to more antibiotics than nosocomial MRSA, likely because of a smaller staphylococcal chromosomal cassette with less ability to carry antimicrobial resistance genes ,22 We report our 15-year experience with MRSA infections among U.S. military beneficiaries in San Diego, Calif. This large, heterogenic, and definable population offer a unique opportunity to add to our understanding of the changing epidemiology of MRSA, from a nosocomial-acquired infection to community-acquired infection, and the subsequent health care implications. This is the first single-site study comparing community-acquired and nosocomial MRSA isolates over a 15-year time period. The incidence of community-acquired MRSA infections has dramatically increased since 2002, and an increasing proportion of cases are the result of intrafamilial spread. Despite that most community-acquired MRSA cases occur in relatively young persons, we found that infections were often severe with a hospitalization rate of 16%. Community-acquired and nosocomial MRSA isolates differ from one another epidemiologically and genetically, suggesting that most community-acquired MRSA isolates did not emerge from local nosocomial MRSA strains. METHODS We evaluated all positive cultures for MRSA at U.S. military clinics and the tertiary hospital in San Diego from 1990 to Our patient population consisted of 247,000 active duty military personnel, dependents (including spouses and children), and retirees of the armed services. All cultures obtained at U.S. naval medical clinics in the San Diego area are referred to the Naval Medical Center San Diego microbiology laboratory. MRSA was defined by the Clinical Laboratory Standards Institute 23 criteria by disk diffusion using an oxacillin screening plate (minimal inhibitory concentration 4 mg/l); E-test confirmation was made in cases with borderline results. All MRSA isolates were stored at 70 C. As part of an active surveillance program, the microbiology staff reports all positive MRSA cultures to the infectious disease division, and an infection control officer records and tracts these data. Nosocomial MRSA was defined as more than 48 hours of hospitalization, recent ( 1 year) hospital stay, multiple medical visits for chronic conditions such as dialysis, or family member with a recent history of hospital-acquired MRSA. All positive cultures were evaluated by review of computer data, and when necessary, written medical charts, to ascertain whether the isolate represented a true infection or a colonizer. An infection was defined as MRSA from a sterile body fluid (eg, blood, synovial, or cerebrospinal fluid); purulent drainage; urine culture with associated pyuria; or sputum culture representative of the lower respiratory tract ( 25 polymorphonuclear leukocytes and 10 squamous epithelial cells per field) with an otherwise unexplained pneumonia (infiltrate/opacity verified by a staff radiologist) on chest radiograph. A colonizer was defined as an isolate taken from the nares for surveillance purposes or from a site without evidence that MRSA was the cause of an infection. A subset of patients (n 310) underwent an extensive review of their inpatient and outpatient records. This group consisted of all MRSA cases in 2002 as well as patients from 1990 to 2001 and 2003 to 2004; the latter cases were selected to ensure a distribution from various months of each year and were chosen without knowledge of clinical information. Data collected included demographics (age, sex, race), risk factors for MRSA, medical history, recent surgical history, hospitalizations, antibiotic therapy, and outcomes. Antibiograms were reviewed for antibiotic sensitivity testing performed by the diskdiffusion method (Kirby-Bauer). A sample of this group s MRSA isolates (134 community-acquired and 85 nosocomial) underwent genetic evaluation by multilocus sequence typing using seven loci (arcc, aroe, glpf, gmk, pta, tpi, and yqil) to determine sequence type as defined

3 Crum et al 15-Year Study of MRSA 945 Figure 1 Number of methicillin-resistant Staphylococcus aureus (MRSA) isolates from 1990 to 2004, Naval Medical Center San Diego, San Diego, California. by the Web database ( Standard primers and protocols were used. 24 Trace files from an ABI 3730xl were read, and a contig was established from bidirectional sequencing using Phred and Phrap. 25,26 Polymerase chain reaction was used to detect the staphylococcal chromosomal cassette of methicillin-resistance gene type and the presence of Panton-Valentine leukocidin genes. 27,28 This testing was performed at the Naval Health Research Center, San Diego, Calif, and the University of Maryland School of Medicine, Baltimore, Md. Descriptive statistics were performed, and comparisons were made using the Fisher exact test for dichotomous variables and rank-sum tests for continuous variables. A P value less than.05 was considered to be statistically significant (STATA 8.2 software; StataCorp, College Station, Tex). RESULTS Methicillin-Resistant Staphylococcus aureus Cases from 1990 to 2004 A total of 1888 MRSA isolates were found over a 15-year period at our institution; data to establish the case as community-acquired or nosocomial MRSA were available on 1846 isolates (97.8%). Of these, 1208 (65.4%) were community-acquired MRSA cases and 638 (34.6%) were nosocomial MRSA cases (Figure 1). Of the 1888 isolates, 1533 (81.2%) represented an MRSA infection, 348 (18.4%) represented a colonization, and 7 (0.4%) could not be classified. Community-acquired MRSA was the cause of the majority of the infections resulting from MRSA (n 1044, 68.1%) in our cohort, and 16.4% (171/1044) of communityacquired MRSA infections required hospitalization. Total MRSA isolates increased from 10 in 1990 to 632 in Community-acquired MRSA increased most rapidly from 2002 to 2004, with an annual incidence in 2004 of infections per 100,000 population. There was a seasonal trend with a peak in the number of community-acquired MRSA infections that occurred during the third quarter of each year during 2002 to 2004 (autocorrelation technique with a lag of 4 quarters, P.05) (Figure 2). The sites of infection for community-acquired and nosocomial MRSA cases are shown in Table 1. Communityacquired MRSA infections were most commonly soft tissue (95% of cases) and rarely involved other body sites; only 0.3% were bacteremic. Comparing community-acquired with nosocomial MRSA infections, the former were statistically more likely to be skin or soft tissue infections and less likely to be isolated from the urine, sputum, blood, or body fluid, or associated with a foreign device. The majority (99%) of community-acquired MRSA cases involved a single focus of infection, whereas nosocomial MRSA cases had multiple locations of infection in 10% of cases (P.001). A repeat positive MRSA culture from the same patient was defined as more than 1 month between isolates, which did not occur during the same hospitalization. Fifteen percent (281/1888) of isolates occurred in patients who had previously tested positive for MRSA; of these, 194 were new infections and 87 represented colonization. On examination of individual patients, 206 of 1607 (12.8%) had repeated positive MRSA cultures (mean 1.37, range 1-7). Of these patients, 143 (69.4%) experienced recurrent MRSA infections, whereas the remainder (30.6%) had repeated positive cultures that represented colonization. We also examined all 1888 isolates for cases that occurred within the same household; results are shown in Figure 3. The number of affected household members

4 946 The American Journal of Medicine, Vol 119, No 11, November 2006 Figure 2 Seasonal trends in community-acquired MRSA isolates. ranged from two to four persons (mean 2.3), and secondary cases typically developed within 1 to 3 months of the first family positive MRSA culture. Intrafamilial MRSA isolates increased sharply after 2002, and most were the result of community-acquired MRSA. Of the 632 MRSA isolates in 2004, 63 (10%) were among family members; this figure includes both the index and secondary cases. Forty-one (65%) of the 63 isolates from households in 2004 represented acute infections, and all were the result of community-acquired MRSA. Subsets of Methicillin-Resistant Staphylococcus aureus Cases We evaluated a subset of MRSA cases in depth for demographic, clinical, and microbiologic features (Table 2). Patients with community-acquired MRSA were younger (median of 22 vs 64 years, P.001) and less likely to have concurrent medical conditions (9% vs 98%, P.001). Community-acquired MRSA cases were more common among males, possibly reflecting the underlying distribution Table 1 Site of Infections in Community-acquired and Nosocomial Methicillin-resistant Staphylococcus aureus Cases Site of Infection* Community-acquired MRSA N 1044 Nosocomial MRSA N 457 P Value Soft tissue/abscess 989 (94.7%) 218 (47.7%).001 Urine 11 (1.1%) 56 (12.3%).001 Sputum 4 (0.4%) 87 (19.0%).001 Blood 3 (0.3%) 71 (15.5%).001 Catheter/foreign device 0 (0.0%) 52 (11.4%).001 Ear/nose/sinus/throat 12 (1.2%) 12 (2.7%).04 Cerebrospinal fluid 0 (0.0%) 3 (0.7%).001 Other 23 (2.2%) 19 (4.2%).04 Not documented 8 (0.8%) 1 (0.2%).29 Multiple simultaneous sites of infection 6 (0.6%) 45 (9.8%).001 Site of soft tissue Total 484 Total 178 Extremity 338 (69.8%) 95 (53.4%).001 Buttocks/genital 68 (14.0%) 18 (10.1%).20 Face/head/neck 35 (7.3%) 9 (5%).32 Trunk 43 (8.9%) 56 (31.5%).001 MRSA methicillin-resistant Staphylococcus aureus. *Percentages do not equal 100% because some patients had MRSA from multiple sites. Other includes isolates from eye, body fluids, vagina, bone, and heart valves. Some cases did not have the exact location of the infection documented.

5 Crum et al 15-Year Study of MRSA 947 Figure 3 Household spread of MRSA. of our young, active-duty population. Patients with nosocomial MRSA infections were more likely to have cancer, a wound, or a foreign device than those with communityacquired MRSA (P.001). In addition, more patients with nosocomial MRSA had a history of recent antibiotic use (75 vs 41%, P.001). Community-acquired versus nosocomial MRSA isolates were less often resistant to clindamycin (19% vs 81%, P.001), erythromycin (80% vs 98%, P.001), and trimethoprim/sulfamethoxazole (2% vs 8%, P.01). All MRSA isolates were sensitive to minocycline and vancomycin. During the years 1993 to 2002, community-acquired MRSA isolates maintained a low rate (12.8%) of clindamycin resistance; however, during 2003 and 2004, the rate increased to 78.9% (P.001). Nosocomial MRSA isolates have always demonstrated high rates of clindamycin resistance. An antibiotic effective against MRSA was used initially in 25% of infections; more patients with nosocomial MRSA were begun on an antibiotic, which effectively treated MRSA compared with community-acquired MRSA (43% vs 17%, P.001). The mortality rate among patients with nosocomial MRSA was higher than those with community-acquired MRSA (19% vs 0%); this finding was likely related to the patients other medical conditions. A subgroup of isolates underwent genetic analysis (Tables 2 and 3). Genetically, 95% of community-acquired MRSA isolates had the type IV staphylococcal chromosomal cassette, 70% were sequence type 8, and 69% were positive for the Panton-Valentine leukocidin genes. Nosocomial MRSA isolates were significantly different; 94% had the type II staphylococcal chromosomal cassette, 77% were sequence type 5, and 0% were positive for the Panton- Valentine leukocidin genes (P.001). Some sequence types and staphylococcal chromosomal cassette types were present in both community-acquired and nosocomial MRSA isolates, for example, sequence types 5, 8, 59, 83, and staphylococcal chromosomal cassette types II and IV. Before 2002, community-acquired MRSA was caused by a variety of sequence types including 1, 5, 8, 30, 59, 83, and 231. In 2002, there was an MRSA outbreak at a San Diego training facility; 29 a community-acquired MRSA strain with the type IV staphylococcal chromosomal cassette and a sequence type of 8 was the predominant strain. In our study cohort since 2002, community-acquired MRSA cases have largely been the result of this strain. Likewise, nosocomial MRSA isolates initially appeared diverse in terms of the sequence type, but over time were predominantly staphylococcal chromosomal cassette type II and sequence type 5. We evaluated the relationship between the sequence type and the presence of the Panton-Valentine leukocidin genes with various outcome measures among community-acquired MRSA cases. Sequence type 8 required hospitalization in 19% of cases, versus 5% among patients with other sequence types (P.06); days of hospitalization among those with sequence type 8 were longer than those without sequence type 8 (8.4 vs 2.5 days, P.14). The mean white blood count was higher (12 300/mm 3 vs 8600/mm 3, P.02) among those with sequence type 8. Community-acquired MRSA isolates that were positive for the Panton-Valentine leukocidin genes were compared with those that were negative; there were no significant differences in hospitalization rates or white blood count. No patient with community-acquired MRSA died.

6 948 The American Journal of Medicine, Vol 119, No 11, November 2006 Table 2 Comparisons Between Community-acquired and Nosocomial Methicillin-resistant Staphylococcus aureus Cases Factor Community-acquired MRSA N 191 Nosocomial MRSA N 119 P Value Age, median (range), y 22 (0.5 87) 64 (0.5-99).001 Sex, male 145/191 (75.9%) 69/119 (58%).001 Race White 66/95 (69.5%) 82/118 (69.4%).54 African American 8/95 (8.4%) 14/118 (11.9%) Hispanic 7/95 (7.4%) 4/118 (3.4%) Other 14/95 (14.7%) 18/118 (15.3%) Medical history 17/191 (8.9%) 116/119 (97.5%).001 Median number of preexisting medical 0 (0-5) 3 (0-7).001 conditions (range) Cancer 0/191 (0%) 35/119 (29.4%).001 Current wound or foreign device in place 6/191 (3.1%) 64/105 (61%).001 Surgical history 5/191 (2.6%) 82/118 (69.5%).001 Prior antibiotic use 34/84 (40.5%) 63/84 (75%).001 First antibiotic effective against MRSA isolate 23/132 (17.4%) 23/53 (43.4%).001 Mortality 0/191 (0%) 23/119 (19.3%).001 White blood count 10.8 ( ) 12.1 (2.4-35).08 Antibiotic susceptibility (sensitive) Minocycline 166/166 (100%) 84/84 (100%) Clindamycin 154/191 (80.6%) 21/112 (18.8%).001 Erythromycin 37/185 (20%) 2/112 (1.8%).001 Trimethoprim-sulfamethoxazole 183/186 (98.3%) 107/116 (92.2%).01 Panton-Valentine leukocidin gene positive 93/134 (69.4%) 0/85 (0%).001 Staphylococcal chromosomal cassette type II 7 (5.2%) 80 (94.1%).001 III 0 (0%) 3 (3.5%) IV 127 (94.8%) 2 (2.4%) Sequence type 1 15 (11.2%) 0 (0%) (5.2%) 65 (76.5%) 8 94 (70.1%) 1 (1.8%) (8.1%) 0 (0%) 36 0 (0%) 9 (10.6%) 59 5 (3.7%) 2 (2.4%) MRSA methicillin-resistant Staphylococcus aureus. DISCUSSION To our knowledge, this is the first single-site study that compares community-acquired MRSA and nosocomial MRSA isolates over a 15-year time period. Our report demonstrates that the incidence of community-acquired MRSA infections has dramatically increased since 2002 and that an increasing proportion of community-acquired MRSA cases are the result of intrafamilial spread. The community-acquired and nosocomial MRSA isolates from this study, which examines a single geographic location, are both epidemiologically and genetically different, suggesting that most community-acquired MRSA isolates did not emerge from local nosocomial MRSA strains. Finally, MRSA strains appear to have become less diversified over time; staphylococcal chromosomal cassette type IV, sequence type 8, Panton-Valentine leukocidin gene positive isolate is now the predominant community-acquired MRSA strain. From 1990 to 2001, there was an increase in the incidence of community-acquired MRSA infections in the San Diego military community; 30 however, in 2002, a major increase occurred. By 2004, our rate of community-acquired MRSA infections was at least 155 cases per 100,000, a rate significantly higher than recently reported by Fridkin et al, 31 who examined three communities in the United States and found a rate of 18 to 25.7 cases per 100,000 population. 31 The specific reasons for the higher incidence among our cohort is unknown but may be related to close living conditions and recurrent skin trauma among our military personnel; further data on these potential risk factors are needed. Household transmission increased in 2003 and 2004 among our cohort. Ten percent of the MRSA isolates in 2004 involved household-related cases; most of these were soft tissue infections. Both adults and children were involved. These data confirm that intrafamilial spread is increasing among patients with community-acquired MRSA. Studies previously showed that MRSA may spread to close contacts within the same household and community, and a recent report suggests that family pets may serve as

7 Crum et al 15-Year Study of MRSA 949 Table 3 Genetic Analysis Data for Community-acquired and Nosocomial Methicillin-resistant Staphylococcus aureus Isolates by Year Year (No.) Community-acquired MRSA Staphylococcal Chromosomal Cassette Type/ Panton-Valentine Leukocidin Gene Positive 1/Sequence Type (Number of Isolates) N (10) IV-1-1 (3) II-0-5 (1) IV-0-8 (1) II-0-36 (3) II (1) II-0-8 (1) 1994 (9) IVa-0-83 (1) II-0-5 (2) IV-1-1 (1) IIIa (1) IV-1-30 (1) II (1) II-0-83 (1) II-0-36 (1) 1995 (9) IV-1-8 (1) II-0-36 (2) IV-1-5 (1) II-0-5 (2) IV-1-30 (2) IV-1-1 (1) 1996 (8) II-0-30 (2) II-0-5 (2) IV-0-8 (1) II-0-36 (1) IV-1-59 (2) 1997 (13) IV-1-30 (1) II-0-5 (7) IV-1-1 (4) II (1) 1998 (13) IV-1-30 (1) II-0-5 (9) II-0-5 (1) IV-0-5 (1) II-0-36 (1) 1999 (8) IV-1-30 (1) II-0-5 (3) IV-1-1 (2) IIIa-0-59 (1) IVa-0-5 (1) 2000 (15) IV-1-30 (1) II-0-5 (6) IV-1-8 (1) II (1) IV-0-8 (3) IV-1-1 (1) IV-0-5 (1) IV-0-59 (1) 2001 (13) IV-0-30 (1) II-0-5 (8) II (1) II (1) IV-1-8 (2) 2002 (92)* IV-1-8 (48) II-0-5 (13) IV-0-8 (19) IIIa-0-59 (1) IV-0-1 (2) IV-0-45 (1) IV-0-59 (2) II-0-36 (1) IV-0-5 (1) II (1) IV-1-30 (1) II-0-5 (2) 2003 (11) IV-1-8 (7) II-0-5 (3) IV-1-1 (1) 2004 (18) IV-1-8 (10) II-0-5 (7) II-0-8 (1) Nosocomial MRSA Staphylococcal Chromosomal Cassette Type/ Panton-Valentine Leukocidin Gene Positive 1/Sequence Type (Number of Isolates) N 85 MRSA methicillin-resistant Staphylococcus aureus. *In 2002, an outbreak occurred at one of the Navy training sites (reference 29); 35 of the isolates reported in our article were from this outbreak: 26 isolates were IV-1-8, 5 isolates were IV-0-8, and 4 isolates had no genetic analysis performed. MRSA reservoirs. 35 We also noted that 13% of all patients in our study had multiple positive MRSA cultures, of which 69% were recurrent infections. These data suggest that eradication may be an important strategy in the prevention of MRSA infections. Community-acquired MRSA causes mainly skin and soft tissue infections. In our study, 95% of these cases represented soft tissue abscesses, cellulites, or furunculosis; this percentage is higher than previously reported. 21 Military members may be at particular risk of soft tissue infections because of frequent skin abrasions associated with training exercises; this is a known risk factor among athletes. 12 Despite the fact that community-acquired MRSA infections were largely confined to the soft tissues and occurred in otherwise healthy young persons, 16% required hospitalization.

8 950 The American Journal of Medicine, Vol 119, No 11, November 2006 An effective first antibiotic active against the specific MRSA isolate was more often used in nosocomial MRSA than community-acquired MRSA cases (43% vs 17%, P.001). This may reflect a higher index of suspicion for MRSA infections in nosocomial cases. Other studies also have found a high rate of ineffective initial antibiotic use. 21,31 Practitioners should maintain a high index of suspicion for MRSA as the cause of infections, especially among patients failing -lactam antibiotic therapy as well as high-risk groups (minorities, low socioeconomic status, and groups in closed settings, such as inmates and military personnel). 14 All community-acquired MRSA isolates in this series were sensitive to minocycline and vancomycin; 98% were susceptible to trimethoprim-sulfamethoxazole. Clindamycin resistance increased to 79% during 2003 and 2004, making this an unreliable treatment option; others have also noted increasing clindamycin resistance among community-acquired MRSA isolates. 36,37 Community-acquired MRSA isolates became less genetically diverse over the 15-year study period; most 2002 to 2004 isolates were staphylococcal chromosomal cassette type IV, Panton-Valentine leukocidin gene positive, sequence type 8. This may be related to an outbreak at a military training facility in 2002 involving the same strain and subsequent dissemination through the local community. 29 Nosocomial MRSA isolates also became more confined to a specific strain type suggesting that certain community-acquired and nosocomial MRSA strains may become endemic in a given geographic region replacing other strains over time because of selective genetic advantages. Studies from different geographic locations have reported other predominant MRSA sequence types: Taiwan with type 59, Europe with type 80, and the United States with types 1, 8, and 30; these findings suggest coevolution of various community-acquired MRSA clones in different geographic locations. 42 Our study has advantages over other cohorts, including a well-defined population who had close medical follow-up and available MRSA isolates spanning over a decade. In addition, the population from which our community-acquired MRSA cases occurred had little or no contact with health care settings, an issue that confounded other investigations. Study limitations included that many of our cases were active duty members who live in close proximity to one another on ships or in military housing; thus our findings may not be generalizable to other population groups. Another potential limitation is the possibility of patients obtaining medical care outside of our health care system; we believe that this is an uncommon occurrence given the free access to medical care provided to our beneficiaries. There is the possibility that more wound cultures were obtained over time; however, this cannot be the sole explanation for the dramatic increase in community-acquired MRSA cases. In addition, many infections are still treated without a culture being performed, which would underestimate the true rate of MRSA infections. Finally, genetic analysis of MRSA isolates were performed on a subset of patients, and many (75/134 community-acquired MRSA isolates) of the isolates were from 2002, the year that an outbreak occurred; this may have led to overestimating the clonal nature of our isolates. CONCLUSION Community-acquired MRSA infections and household-related cases have dramatically increased since Most community-acquired MRSA involves the soft tissues; infections may be severe. Although a prime opportunity existed for nosocomial MRSA to spread and become established as community-acquired MRSA strains in this defined community, our 15 years of surveillance revealed the emergence of distinct MRSA strains uniquely adapted to the community. Further research on other specific virulence factors of community-acquired MRSA is ongoing, 16,43,44 but one must question: Will the attributes that allowed the selective advantage of community-acquired MRSA isolates in the community foster spread and colonization back to our hospitals, resulting in more aggressive nosocomial MRSA infections in the future? References 1. Sheagren JN. Staphylococcus aureus. The persistent pathogen. N Engl J Med. 1984;310: Jevons MP. Celbenin -resistant staphylococci. BMJ. 1998;339: Hryniewicz W. Epidemiology of MRSA. Infection. 1999;27:S13-S Centers for Disease Control. Community-acquired methicillin-resistant Staphylococcus aureus infections Michigan. MMWR Morb Mortal Wkly Rep. 1981;30: Salgado CS, Farr BM, Calfee DP. Community-acquired methicillinresistant Staphylococcus aureus: a meta-analysis of prevalence and risk factors. Clin Infect Dis. 2002;36: 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: Herold BC, Immergluck LC, Maranan MC, et al. Community-acquired methicillin-resistant Staphylococcus aureus in children with no identified predisposing risk. JAMA. 1998;279(8): Centers for Disease Control and Prevention. Four pediatric deaths from community-acquired methicillin-resistant Staphylococcus aureus, Minnesota and North Dakota, MMWR Morb Mortal Wkly Rep. 1999;48: Zinderman CE, Conner B, Malakooti MA, et al. Community-acquired methicillin-resistant Staphylococcus aureus among military recruits. Emerg Infect Dis. 2004;10: Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus in correctional facilities Georgia, California, and Texas, MMWR Morb Mortal Wkly. Rep 2003;52: Nguyen DM, Mascola L, Bancroft E. Recurring infections in a football team. Emerg Infect Dis. 2005;11: Kazakova SV, Hageman JC, Matava M, et al. A clone of methicillinresistant Staphylococcus aureus among professional football players. N Engl J Med. 2005;352: Charlebois ED, Perdreau-Remingon F, Kreiswirth B, et al. Origins of community strains of methicillin-resistant Staphylococcus aureus. Clin Infect Dis. 2004;15: Eady EA, Cove JH. Staphylococcal resistance revisited: communityacquired methicillin-resistant Staphylococcus aureus an emerging problem for the management of skin and soft tissue infections. Curr Opin Infect Dis. 2003;16: Fey PD, Said-Salim B, Rupp ME, et al. Comparative molecular analysis of community- or hospital-acquired methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother. 2003;47:

9 Crum et al 15-Year Study of MRSA Deresinski S. Methicillin-resistant Staphylococcus aureus: an evolutionary, epidemiologic, and therapeutic odyssey. Clin Infect Dis. 2005; 40: Chambers HF. The changing epidemiology of Staphylococcus aureus? Emerg Infect Dis. 2001;7(2): Daum RS, Ito T, Hiramatsu K, et al. A novel methicillin-resistance cassette in community-acquired methicillin-resistant Staphylococcus aureus isolates of diverse genetic backgrounds. J Infect Dis. 2002;186: Diep BA, Sensabaugh GR, Somboona NS, et al. Widespread skin and soft-tissue infections due to two methicillin-resistant Staphylococcus aureus strains harboring the genes for Panton-Valentine leucocidin. J Clin Microbiol. 2004;42: Francis JS, Doherty MC, Lopatin U, et al. Severe community-onset pneumonia in healthy adults caused by methicillin-resistant Staphylococcus aureus carrying the Panton-Valentine leukocidin genes. Clin Infect Dis. 2005;40: Naimi TS, LeDell KH, Como-Sabetti K, et al. Comparison of community- and health care-associated methicillin-resistant Staphylococcus aureus infection. JAMA. 2003;290(22): Almer LS, Shortridge VD, Nilius AM, et al. Antimicrobial susceptibility and molecular characterization of community-acquired methicillin-resistant Staphylococcus aureus. Diagn Microbiol Infect Dis. 2002; 43: Clinical and Laboratory Standards Institute [formerly National Committee for Clinical Laboratory Standards]. Performance standards for antimicrobial disk susceptibility tests. Approved standard M2-A8. Wayne, PA: Clinical and Laboratory Standards Institute; Enright MC, Day NP, Davies CE, et al. Multilocus sequence typing for the characterization of methicillin-resistant (MRSA) and methicillinsusceptible (MSSA) clones of Staphylococcus aureus. J Clin Microbiol. 2000;38: Ewing B, Hillier L, Wendl MC, Green P. Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res. 1998;8: Ewing B, Green P. Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res. 1998;8: Oliveira DC, Tomasz A, De Lencastre H. Secrets of success of a human pathogen: molecular evolution of pandemic clones of methicillin-resistant Staphylococcus aureus. Lancet. 2002;2: Lina G, Piemont Y, Godail-Gamot F, et al. Involvement of Panton- Valentine leukocidin-producing Staphylococcus aureus in primary skin infections and pneumonia. Clin Infect Dis. 1999;29: Campbell KM, Vaughn AF, Russell KL, et al. Risk factors for community-acquired methicillin-resistant Staphylococcus aureus (MRSA) infections in military trainees: review of an outbreak in San Diego, California, J Clin Microbiol. 2004;42: Kallen AJ, Driscoll TJ, Thornton S, et al. Increase in communityacquired methicillin-resistant Staphylococcus aureus at a Naval Medical Center. Infect Control Hosp Epidemiol. 2000;21: Fridkin SK, Hageman JC, Morrison M, et al. Methicillin-resistant Staphylococcus aureus disease in three communities. N Engl J Med. 2005;352: Hollis RJ, Barr JL, Boebbeling BN, et al. Familial carriage of methicillin-resistant Staphylococcus aureus and subsequent infection in a premature neonate. Clin Infect Dis. 1995;21: L Heritueau F, Lucet JC, Scanvic A, Bouvet E. Community-acquired methicillin-resistant Staphylococcus aureus and familial transmission. JAMA. 1999;282: Calfee DP, Durbin LJ, Germanson TP, et al. Spread of methicillinresistant Staphylococcus aureus (MRSA) among household contacts of individuals with nosocomially acquired MRSA. Infect Control Hosp Epidemiol. 2003;24: Manian FA. Asymptomatic nasal carriage of mupirocin-resistant, methicillin-resistant Staphylococcus aureus (MRSA) in a pet dog associated with MRSA infection in household contacts. Clin Infect Dis. 2003;36:e Braun L, Craft D, Williams R, et al. Increasing clindamycin resistance among methicillin-resistant Staphylococcus aureus in 57 Northeast United States military treatment facilities. Pediatr Infect Dis J. 2005; 24: Lewis JS II, Jorgensen JH. Inducible clindamycin resistance in Staphylococci: should clinicians and microbiologists be concerned? Clin Infect Dis. 2005;40: Wang C-C, Lo W-T, Chu M-L, Siu LK. Epidemiological typing of community-acquired methicillin-resistant Staphylococcus aureus isolates from children in Taiwan. Clin Infect Dis. 2004;39: Holmes A, Ganner M, McGuane S, et al. Staphylococcus aureus isolates carrying Panton-Valentine leucocidin genes in England and Wales: frequency, characterization, and association with clinical disease. J Clin Microbiol. 2005;43: Faria NA, Oliveira DC, Westh H, et al. Epidemiology of emerging methicillin-resistant Staphylococcus aureus (MRSA) in Denmark: a nationwide study in a country with low prevalence of MRSA infection. J Clin Microbiol. 2005;43: Mishaan AMA, Mason EO Jr, Martinez-Aguilar G, et al. Emergence of a predominant clone of community-acquired Staphylococcus aureus among children in Houston, Texas. Pediatr Infect Dis J. 2005;24: Vandenesch F, Naimi T, Enright MC, et al. Community-acquired methicillin-resistant Staphylococcus aureus carrying Panton-Valentine leukocidin genes: worldwide emergence. Emerg Infect Dis. 2003;9: Baba T, Takeuchi F, Kuroda M, et al. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet. 2002; 359: Said-Salim B, Mathema B, Braughton K, et al. Differential distribution and expression of Panton-Valentine leucocidin among communityacquired methicillin-resistant Staphylococcus aureus strains. J Clin Microbiol. 2005;43: