Emergence of Community-Acquired Methicillin-Resistant
Staphylococcus aureus USA 300 Clone as the Predominant Cause of
Skin and Soft-Tissue Infections
Mark D. King, MD, MS; Bianca J. Humphrey, BS; Yun F. Wang, PhD; Ekaterina V. Kourbatova, MD, MPH; Susan M. Ray, MD; and
Henry M. Blumberg, MD
Background: Studies have shown that community-acquired methi-
cillin-resistant Staphylococcus aureus (MRSA) causes S. aureus skin
and soft-tissue infection in selected populations.
Objective: To determine the proportion of infections caused by
community-acquired MRSA, the clinical characteristics associated
with community-acquired MRSA, and the molecular epidemiology
of community-acquired MRSA among persons with community-
onset S. aureus skin and soft-tissue infection.
Design: Active, prospective laboratory surveillance to identify S.
aureus recovered from skin and soft-tissue sources.
Setting: 1000-bed urban hospital and its affiliated outpatient clinics
in Atlanta, Georgia.
Patients: 384 persons with microbiologically confirmed community-
onset S. aureus skin and soft-tissue infection.
Measurements: Proportion of infections caused by and clinical
factors associated with community-acquired MRSA among persons
with community-onset S. aureus skin and soft-tissue infection.
Pulsed-field gel electrophoresis and antimicrobial susceptibility pat-
terns were used to epidemiologically classify community-onset S.
aureus infections. Community-acquired MRSA was defined by
MRSA isolates that either demonstrated a USA 300 or USA 400
pulsed-field type or had a susceptibility pattern showing resistance
only to

-lactams and erythromycin (for isolates not available for
pulsed-field gel electrophoresis).
Results: Community-onset skin and soft-tissue infection due to S.
aureus was identified in 389 episodes, with MRSA accounting for
72% (279 of 389 episodes). Among all S. aureus isolates, 63%
(244 of 389 isolates) were community-acquired MRSA. Among
MRSA isolates, 87% (244 of 279 isolates) were community-ac-
quired MRSA. When analysis was restricted only to MRSA isolates
that were available for pulsed-field gel electrophoresis, 91% (159
of 175 isolates) had a pulsed-field type consistent with community-
acquired MRSA; of these, 99% (157 of 159 isolates) were the
MRSA USA 300 clone. Factors independently associated with com-
munity-acquired MRSA infection were black race (prevalence ratio,
1.53 [95% CI, 1.16 to 2.02]), female sex (prevalence ratio, 1.16
[CI, 1.02 to 1.32]), and hospitalization within the previous 12
months (prevalence ratio, 0.80 [CI, 0.66 to 0.97]). Inadequate
initial antibiotic therapy was statistically significantly more common
among those with community-acquired MRSA (65%) than among
those with methicillin-susceptible S. aureus skin and soft-tissue in-
fection (1%).
Limitations: Some MRSA isolates were not available for molecular
typing.
Conclusions: The community-acquired MRSA USA 300 clone was
the predominant cause of community-onset S. aureus skin and
soft-tissue infection. Empirical use of agents active against commu-
nity-acquired MRSA is warranted for patients presenting with seri-
ous skin and soft-tissue infections.
Ann Intern Med. 2006;144:309-317. www.annals.org
For author affiliations, see end of text.
M
ethicillin-resistant Staphylococcus aureus (MRSA) in-
fections have usually been associated with exposure
to health care settings, but they have recently been recog-
nized in persons without traditional risk factors. These in-
fections have been called community-acquired or commu-
nity-associated MRSA (1). Most cases have been associated
with skin and soft-tissue infection and have been reported
among selected populations, including correctional facility
inmates, homosexual men, and sports teams (2–5). Molec-
ular typing studies in the United States and Australia have
demonstrated that most community-acquired MRSA in-
fections are caused by one of several clones or pulsed-field
types (6, 7). In the United States, 2 clones, designated as
USA 300 and USA 400 by the Centers for Disease Control
and Prevention (CDC), have been identified as the pri-
mary types that cause community-acquired MRSA infec-
tions (6). The outbreaks of MRSA skin and soft-tissue
infections observed in correctional facilities and among
athletes have been associated with the USA 300 pulsed-
field type, while outbreaks associated with severe and fatal
disease in children, as well as skin and soft-tissue infections
in Native American populations, have been associated with
the USA 400 pulsed-field type (6).
The community-acquired MRSA clones have fre-
quently been associated with the Panton–Valentine leuko-
See also:
Print
Editors’ Notes .............................310
Editorial comment..........................368
Related article.............................318
Summary for Patients.......................I-11
Web-Only
CME quiz
Conversion of figures and tables into slides
Annals of Internal Medicine Article
© 2006 American College of Physicians 309
cidin virulence factor and the presence of staphylococcal
chromosome cassette mec (SCCmec) type IV allele (1,
6–9). In contrast, hospital-acquired or health care–associ-
ated MRSA strains usually lack genes for Panton–Valen-
tine leukocidin and are associated with other SCCmec al-
leles (for example, SCCmec type II) (1, 6, 7). In addition to
Panton–Valentine leukocidin and SCCmec type IV, the
community-acquired MRSA USA 300 and USA 400 ge-
notypes have usually demonstrated resistance to

-lactams
and erythromycin while retaining susceptibility to clinda-
mycin, trimethoprim–sulfamethoxazole, and fluoroquino-
lones, whereas health care–associated genotypes are often
multidrug-resistant (1, 6).
Although increasingly reported as a cause of outbreaks
of skin and soft-tissue infection, the proportion of S. aureus
skin and soft-tissue infections caused by community-ac-
quired MRSA in nonoutbreak settings remains poorly de-
fined. We sought to determine the proportion of infections
caused by community-acquired MRSA, the clinical charac-
teristics associated with community-acquired MRSA, and
the molecular epidemiology of community-acquired
MRSA among persons with community-onset S. aureus
skin and soft-tissue infection who were receiving care at a
large hospital and its affiliated clinics in urban Atlanta,
Georgia.
METHODS
Laboratory Surveillance for Community-Onset
S. aureus
Skin and Soft-Tissue Infection
We conducted active, prospective laboratory surveil-
lance from 1 August 2003 to 15 November 2003 to iden-
tify S. aureus isolates of patients with skin and soft-tissue
infections at the Grady Health System. The Grady Health
System includes Grady Memorial Hospital, a 1000-bed
public inner-city hospital, and its affiliated outpatient clin-
ics. We limited surveillance to S. aureus isolates that were
obtained from patients with a community-onset skin and
soft-tissue infection (isolates designated either as an exu-
date or a body fluid culture). We defined community-onset
S. aureus by positive test results from cultures that were
obtained within 72 hours of admission or at an outpatient
visit. An investigator classified community-onset S. aureus
before we performed molecular typing studies or before we
reviewed antimicrobial susceptibility results other than
methicillin susceptibility. We identified S. aureus and
tested antimicrobial susceptibility for all isolates in the
Grady Memorial Hospital Clinical Microbiology Labora-
tory in accordance with the National Committee for Clin-
ical Laboratory Standards (10).
Study Sample and Data Collection
The Emory University Institutional Review Board and
the Grady Research Oversight Committee approved the
study. Grady Memorial Hospital provides medical care for
a primarily medically indigent inner-city population, ap-
proximately 80% of whom are African American, residing
within inner-city Atlanta (Fulton County and DeKalb
County). The hospital serves as a teaching hospital for both
Emory University School of Medicine and Morehouse
School of Medicine. The referral pattern for the hospital
and its affiliated clinics is primarily patient self-referral on
the basis of place of residence (that is, residence within
DeKalb or Fulton Counties) or indigent status, with some
patients referred for tertiary academic medical care. We
retrospectively reviewed computerized medical and labora-
tory records to document patient demographic characteris-
tics, HIV status, presence of end-stage renal disease, hospi-
talization within the previous 12 months, and any history
of MRSA infection or colonization. The reviewer com-
pleted the abstraction process blinded to the molecular typ-
ing results of MRSA isolates. We documented the need for
hospitalization for managing the S. aureus skin and soft-
tissue infection. We obtained S. aureus antimicrobial sus-
ceptibility profiles from the hospital’s microbiology labora-
tory, and we categorized infections as due to either
methicillin-susceptible S. aureus (MSSA) or MRSA. We
reviewed computerized pharmacy records to document the
antimicrobial agent therapy that was used to treat the pa-
tient’s skin and soft-tissue infection. We considered anti-
microbial agent therapy to be adequate if the prescribed
antimicrobial agent had in vitro activity against the isolated
S. aureus strain.
Molecular Typing and Genetic Analyses
We performed pulsed-field gel electrophoresis (PFGE)
using SmaI as a restriction endonuclease on all available
MRSA isolates, as described by Bannerman and colleagues
(11). We digitized gels and saved them as an image for
analysis with BioNumerics software (Applied Maths, Sint-
Context
In community outbreaks of methicillin-resistant Staphylo-
coccus aureus (MRSA), 2 clones predominate in the
United States: USA 300 and USA 400. Little is known
about these infections in the nonoutbreak setting.
Contribution
This study evaluated nonoutbreak community-acquired S.
aureus skin and soft-tissue infections in patients in a large
urban setting. Almost three quarters of the soft-tissue in-
fections were caused by MRSA, and these were predomi-
nantly the USA 300 type. No MRSA USA 300 isolate was
resistant to trimethoprim–sulfamethoxazole or vancomycin,
while most were resistant to erythromycin in addition to

-lactams.
Implications
Antibiotic choice for serious nonoutbreak community-ac-
quired skin and soft-tissue infection should consider high
rates of MRSA in some communities.
—The Editors
Article Community-Acquired MRSA Skin and Soft-Tissue Infections
310 7 March 2006 Annals of Internal Medicine Volume 144 • Number 5 www.annals.org
Martens-Latem, Belgium). We performed cluster analysis
by using the unweighted pair-group method based on Dice
coefficients. We defined pulsed-field type clusters by using
a similarity coefficient of at least 80% and the criteria of
Tenover and colleagues (12). By using the nomenclature
outlined by McDougal and colleagues (6), we categorized
MRSA isolates into 1 of 8 lineages: USA 100 through USA
800.
For some isolates, we determined SCCmec type by
using polymerase chain reaction (PCR)–typing of the mec
gene complex as described by Okuma and colleagues (7).
Similarly, we assessed the presence of Panton–Valentine
leukocidin genes in selected MRSA isolates by using PCR
as previously described by Lina and colleagues (13).
Classification of Community-Onset
S. aureus
Skin and
Soft-Tissue Infections
We classified community-onset S. aureus skin and soft-
tissue infections into 3 groups on the basis of either mo-
lecular typing results (pulsed-field type) of MRSA isolates
or antimicrobial susceptibility pattern in cases in which the
MRSA isolate was not available for PFGE. The 3 groups
were 1) the community-acquired MRSA USA 300/USA
400 group, 2) the “other MRSA” group, and 3) the MSSA
group. The community-acquired MRSA USA 300/USA
400 group was defined by MRSA skin and soft-tissue in-
fections caused by a MRSA isolate that either demon-
strated a USA 300 or USA 400 pulsed-field type (6) or had
an antimicrobial susceptibility profile demonstrating resis-
tance only to

-lactams and erythromycin while retaining
susceptibility to clindamycin, levofloxacin, trimethoprim–
sulfamethoxazole, and vancomycin. The “other MRSA”
group was defined by MRSA skin and soft-tissue infections
caused by a MRSA isolate that either demonstrated a
pulsed-field type other than USA 300 or USA 400 (for
example, USA 100, USA 500, or USA 800 [which have all
been described as health care–associated pulsed-field
types]) or had an antimicrobial susceptibility profile dem-
onstrating resistance to

-lactams, erythromycin, and at
Figure 1. Clinical and molecular epidemiology of
Staphylococcus aureus
skin and soft-tissue infection.
CA-MRSA ⫽ community-acquired methicillin-resistant Staphylococcus aureus; MRSA ⫽ methicillin-resistant Staphylococcus aureus; MSSA ⫽ methicillin-
susceptible Staphylococcus aureus; PFGE ⫽ pulsed-field gel electrophoresis.
ArticleCommunity-Acquired MRSA Skin and Soft-Tissue Infections
www.annals.org 7 March 2006 Annals of Internal Medicine Volume 144 • Number 5 311
least 1 additional antibiotic. We defined the MSSA group
by skin and soft-tissue infections caused by MSSA.
Statistical Analysis
We performed data management and statistical analy-
ses by using Microsoft Excel 2000 software (Microsoft
Corp., Redmond, Washington) and SAS software, version
8.2 (SAS Institute, Cary, North Carolina). We assessed risk
factors for community-acquired MRSA skin and soft-tissue
infection in 2 separate analyses. Our first analysis com-
pared the community-acquired MRSA USA 300/USA 400
group with the MSSA group, and our second analysis com-
pared the community-acquired MRSA USA 300/USA 400
group with the other MRSA group. In addition, we re-
peated each analysis using data that were limited to those
skin and soft-tissue infections in which a MRSA isolate was
available for PFGE to account for any potential misclassi-
fication bias introduced when skin and soft-tissue infec-
tions were classified according to antimicrobial susceptibil-
ity profile.
We initially identified potential risk factors for com-
munity-acquired MRSA skin and soft-tissue infection by
univariate analysis. We calculated prevalence ratios and the
corresponding 95% CIs. Multivariable log-binomial re-
gression models included variables that were statistically
significantly associated with community-acquired MRSA
skin and soft-tissue infection in univariate analysis, as well
as potential confounders and effect modifiers. We tested
interaction among main effect covariates by using 2-way
interaction terms, and we assessed confounding. We chose
the variables included in the final model a priori on the
basis of the biological plausibility of their association with
the outcome of interest, as well as on the basis of statistical
and epidemiologic criteria. We defined a P value of 0.05 or
less as being statistically significant.
Role of the Funding Sources
The Emory Medical Care Foundation, Emory Men-
tored Clinical Research Scholars Program, National Insti-
tutes of Health, National Center for Research Resources,
and National Institute of Allergy and Infectious Diseases
provided funding for the study. The funding sources had
no role in the study design, data collection, data analysis,
and data interpretation or in the writing of the report.
RESULTS
Clinical and Molecular Epidemiology of
Community-Onset
S. aureus
Skin and Soft-Tissue
Infection
We identified microbiologically confirmed communi-
ty-onset S. aureus skin and soft-tissue infection in 389 ep-
isodes of skin and soft-tissue infection among 384 persons.
Of these episodes, 279 (72%) episodes were caused by
MRSA and 110 (28%) episodes were caused by MSSA
(Figure 1). Among the 389 community-onset S. aureus
skin and soft-tissue infections, 244 (63%) were classified
within the community-acquired MRSA USA 300/USA
400 group (159 infections by pulsed-field type and 85 in-
fections by antimicrobial susceptibility profile [for isolates
not available for PFGE]). Among the 279 community-
onset MRSA skin and soft-tissue infections, 244 (87%)
Figure 2. Dendrogram of representative pulsed-field types from methicillin-resistant
Staphylococcus aureus
isolates causing skin
and soft-tissue infections.
The figure shows infections among patients seen at Grady Memorial Hospital (lanes 2 to 7, 9, 11, 13, and 15) and representative methicillin-resistant
Staphylococcus aureus standard-type strains previously published by the Centers for Disease Control and Prevention (CDC) (lanes 1, 8, 10, 12, and 14)
(6). CDC data from McDougal et al. (6).
Article Community-Acquired MRSA Skin and Soft-Tissue Infections
312 7 March 2006 Annals of Internal Medicine Volume 144 • Number 5 www.annals.org
infections were classified within the community-acquired
MRSA USA 300/USA 400 group and 35 (13%) infections
were classified within the other MRSA group (Figure 1).
When we considered only isolates in which PFGE was
performed, 159 of 175 (91%) MRSA isolates had a pulsed-
field type consistent with community-acquired MRSA
USA 300 or USA 400 pulsed-field type and 16 (9%)
MRSA isolates had non– community-associated or health
care–associated pulsed-field types (Figures 1 and 2). The
predominant community-acquired MRSA pulsed-field
type was USA 300, which accounted for 157 of 159 (99%)
community-acquired MRSA isolates studied (Figure 2,
lanes 2 to 7). The other 2 community-acquired MRSA
isolates had a USA 400 pulsed-field type (Figure 2, lane
15). Among the 157 isolates with a USA 300 pulsed-field
type, 121 (77%) isolates were identical to a CDC USA 300
type strain (Figure 2, lanes 2 and 3) and 36 (23%) isolates
were consistent with USA 300 variant strains (Figure 2,
lanes 4 to 7) on the basis of coefficients of similarity of
80% or higher and the criteria of Tenover and colleagues
(12). Figure 2 (lanes 9, 11, and 13) also shows the pulsed-
field types that are representative of non– community-ac-
quired MRSA strains (USA 500, USA 100, and USA 800)
obtained from MRSA isolates of patients with skin and soft-
tissue infections at Grady Memorial Hospital. Two of 2
(100%) community-acquired MRSA USA 300 isolates tested
had Panton–Valentine leukocidin and SCCmec type IV.
Antimicrobial Susceptibility Patterns of MRSA Isolates
according to Pulsed-Field Type
Most USA 300 isolates demonstrated resistance only
to

-lactams and erythromycin (136 of 157 [87%] iso-
lates), with fewer demonstrating resistance to levofloxacin
(15 of 157 [10%] isolates), clindamycin (4 of 157 [3%]
isolates), rifampin (1 of 157 [1%] isolates), or gentamicin
(1 of 157 [1%] isolates) (Table 1). No USA 300 isolate
was resistant to trimethoprim–sulfamethoxazole or vanco-
mycin, and 2 of 157 (1%) isolates demonstrated resistance
only to

-lactams. In contrast, MRSA isolates with non–
community-associated or “health care–associated” pulsed-
field types usually were resistant to other antibiotics in
addition to

-lactams and erythromycin (14 of 16 [88%]
isolates) (Table 1).
Characteristics of Patients with
S. aureus
Skin and
Soft-Tissue Infection
Table 2 shows the characteristics of persons in the
community-acquired MRSA USA 300/USA 400 group,
the other MRSA group, and the MSSA group. When com-
pared with the MSSA group in univariate analysis, persons
in the community-acquired MRSA USA 300/USA 400
group were more likely to be black and female. Persons in
the community-acquired MRSA USA 300/USA 400 group
were less likely to have been hospitalized within the previ-
ous 12 months than those in the MSSA group (Table 2).
Age, HIV status, or the frequency of previous infection or
colonization with MRSA did not differ between the com-
munity-acquired MRSA USA 300/USA 400 group and the
MSSA group. When compared with the other MRSA
group, persons in the community-acquired MRSA USA
300/USA 400 group were more likely to be younger than
50 years of age, to be black, and to be HIV-seronegative.
As observed in the MSSA comparison, persons in the com-
munity-acquired MRSA USA 300/USA 400 group were
also less likely than those in the other MRSA group to have
been hospitalized in the previous 12 months (Table 2). Sex
or the frequency of previous infection or colonization with
MRSA did not statistically significantly differ between the
MRSA groups. Inadequate empirical and definitive antibi-
otic therapy (that is, use of agents without in vitro activity
against the S. aureus isolate recovered from the patient)
were much more common among patients with communi-
ty-acquired MRSA (65% and 43%, respectively) than
among those in the MSSA group (1% and 1%, respec-
tively).
Multivariate Analyses
Black race (prevalence ratio, 1.53 [95% CI, 1.16 to
2.02]), female sex (prevalence ratio, 1.16 [CI, 1.02 to
1.32]), and hospitalization within the previous 12 months
(prevalence ratio, 0.80 [CI, 0.66 to 0.97]) were indepen-
dently associated with the community-acquired MRSA
USA 300/USA 400 group compared with the MSSA group
in multivariate analysis (Table 3). When we repeated the
analysis by using data that were limited to those with skin
and soft-tissue infections in which a MRSA isolate was
available for PFGE, black race (prevalence ratio, 1.87 [CI,
1.25 to 2.81]), female sex (prevalence ratio, 1.25 [CI, 1.04
to 1.50]), and hospitalization within the previous 12
months (prevalence ratio, 0.71 [CI, 0.53 to 0.95]) re-
mained independently associated with the community-ac-
quired MRSA USA 300/USA 400 group (Table 3). The
few cases in the other MRSA group limited construction of
multivariate models comparing the community-acquired
MRSA USA 300/USA 400 group with the other MRSA
group; therefore, we present results of only the univariate
analysis comparing these groups (Table 2).
Table 1. Antimicrobial Susceptibility Patterns of
Methicillin-Resistant
Staphylococcus aureus
Isolates
according to Pulsed-Field Type*
Resistance Pattern Community-Acquired
MRSA USA 300
Group,
n (%)
Other MRSA
Group,
n (%)
†

-Lactam and erythromycin 136 (87) 2 (12)

-Lactam, erythromycin, and
ⱖ1 additional antibiotic
19 (12) 14 (88)

-Lactam only 2 (1) 0 (0)
* Pulsed-field types are defined in accordance with nomenclature outlined by
McDougal et al. (6). MRSA ⫽ methicillin-resistant Staphylococcus aureus.
† The other MRSA group represents MRSA isolates with pulsed-field types other
than USA 300 or USA 400. These included USA 100, USA 500, and USA 800
and are representative of health care–associated genotypes.
ArticleCommunity-Acquired MRSA Skin and Soft-Tissue Infections
www.annals.org 7 March 2006 Annals of Internal Medicine Volume 144 • Number 5 313
