Widespread emergence of methicillin resistance in community-acquired Staphylococcus aureus infections in Denver.

Background: Increasing rates of methicillin resistance among outpatient Staphylococcus aureus infections led us to assess the epidemiology and outcome of a local outbreak.

Methods: A retrospective cohort study of outpatient skin and soft tissue infections due to S aureus in 2003.

Results: From 2002 to mid-2004, the percentage of outpatient S aureus isolates resistant to methicillin increased from 6 to 45%. In multivariate analysis, only male sex and age greater than 18 years were associated with methicillin resistance. Methicillin resistance was common (>15%) among isolates from patients in nearly all subgroups evaluated. Pulsed field gel electrophoresis showed isolates related to USA 300, but methicillin-resistant strains had unusually high rates of quinolone resistance.

Conclusions: A single strain of methicillin-resistant S aureus is responsible for the increase in skin infections in outpatients without traditional risk factors for infection with an antibiotic-resistant strain. In areas with high rates of methicillin-resistant S aureus outpatient infections, we recommend non-[beta]-lactam antibiotics for initial treatment of skin and soft tissue infections.

Key Words: antibiotic resistance, epidemiology, methicillin resistance, outpatient, Staphylococcus aureus

**********

Resistance to [beta]-lactam antibiotics, often termed methicillin resistance, is an increasing problem in the treatment of community-acquired Staphylococcus aureus infections. Early reports showed that most patients with community-acquired methicillin-resistant S aureus had easily identifiable risk factors for infection with a drug-resistant isolate, such as previous hospitalization, prolonged intravenous drug use, or previous antibiotic use. (1,2) However, the patients in a number of recent reports of community-acquired methicillin-resistant infection do not have these classic risk factors. Outbreaks of methicillin-resistant infections have been described in a number of communal settings (jails, athletic teams) (3,4) as well as in patients without identifiable risk factors in the broader community. (5,6)

The emergence of community-acquired methicillin-resistant S aureus raises a number of important clinical and public health questions. Identification of those groups with an increased risk of methicillin-resistant infection could allow for targeting with alternate antibiotic therapy. In addition, efforts to control and prevent methicillin-resistant infections in the community would benefit from more detailed information on the epidemiology of antibiotic-resistant strains. We recently noted a marked increase in community-acquired methicillin-resistant S aureus in Denver, Colorado. We performed a retrospective study to evaluate the epidemiology of community-acquired methicillin-resistant infections, their clinical manifestations, and response to treatment.

Methods

Setting

Denver Health is an integrated health care system consisting of 9 community and 11 school-based clinics, the public health department, and a 350-bed acute care hospital with associated specialty-care clinics. Notably, there are no chronic care facilities in the system. Denver Health provides care to approximately 160,000 persons per year, predominantly low-income and medically indigent patients of Denver County, the center of metropolitan Denver.

Microbiology and pulsed field gel electrophoresis

All microbiologic testing was done in a central laboratory. Catalase-positive, Gram-positive cocci were confirmed as S aureus, using a commercial latex agglutination kit (Remel Inc, Lenexa, KS). Antibiotic susceptibility testing was performed with the use of an automated system (Microscan Walk-Away system; Dade-Behring, Inc, West Sacramento, CA). Isolates with a mean inhibitory concentration of greater than 2 [micro]g/mL to oxacillin in the automated system had confirmatory testing for [beta]-lactam resistance using Mueller Hinton agar with 4% NaCl and 6 [micro]g/mL oxacillin (Remel Inc, Lenexa, KS). (7) During the time period of this study, the D-test was not used to assess clindamycin susceptibility.

Staphylococcal genomic DNA was extracted from 20 randomly selected community-acquired methicillin-resistant isolates using the GenePath System Group 1 Reagent Kit (Bio-Rad Laboratories, Richmond, CA). Samples were digested with SmaI (5 U/[micro]L), using the rapid procedure protocol (Bio-Rad) and DNA fragments were resolved in 1% agarose using the GenePath System protocol No. 5 (Bio-Rad). The similarity between isolates was determined by visual comparison of photographs of isolate-banding patterns stained with ethidium bromide (Bio-Rad). Using the Tenover method, (8) we compared the strain relatedness of the community samples with each other, with a strain of community-acquired MRSA common in the Midwest United States (MW2 strain obtained through the Network on Antimicrobial Resistance in Staphylococcus aureus Program), and with strains previously characterized as dominant inpatient strains at our hospital. Isolates that differed by greater than 6 bands were distinct. Representatives of the common PFGE types were tested for Panton-Valentine leukocidin (PVL), using previously described techniques. (9)

Temporal trend in community-acquired MRSA

We used the laboratory database to identify all patients from January 2002 through June 2004 who had S aureus isolated from blood, wound, or a soft tissue site. During this time period, bacterial cultures were obtained at the discretion of the responsible health care provider. The use of culture to identify resistant strains was encouraged after April 2004. Using patients from 2003 only, we defined as community-acquired infections those cases whose culture was obtained in an outpatient clinic, emergency room or urgent care center, or within 24 hours of hospital admission.

Cohort study

Selection of subjects. We conducted a retrospective cohort study of all patients with community-associated S aureus infection in 2003. Patients whose medical records suggested a skin or soft tissue infection (erythema, warmth, pain, swelling, or abscess) were included in this review. Patients were excluded from analysis if they had any history of methicillin-resistant S aureus infection or colonization. The local institutional review board approved this study.

Risk factor and outcome assessment. We used computerized medical records and databases to assess the presence of potential risk factors for community-acquired methicillin-resistant infection. The medical records review was performed by an investigator who was aware of the resistance pattern of the S aureus isolate. We defined recent hospitalization or residence in a long-term care facility as any documented institutionalization in the 12 months before presentation. Use of an outpatient central venous catheter was defined as present if used in the past 3 months. Injection drug use was defined as present if use in the past 3 months was documented in the chart. Patients with HIV infection were identified from the Expanded Surveillance Initiative-Adult and Adolescent Spectrum of HIV Disease database, a prospective study of all persons with HIV infection in the Denver Health system. (10) For purposes of analysis, we combined persons with documented negative HIV testing with those of unknown HIV serostatus. We used medical records from hospitalizations and the Denver Health outpatient pharmacy database to assess for antibiotic exposure within the 6 months before presentation. This database contains records of prescriptions filled in all outpatient pharmacies in the Denver Health system.

Outcome assessments included the number of days from presentation until the patient was given an antibiotic having in vitro activity against the S aureus strain isolated, the number of health care encounters in the 30 days before and after the date the first culture was obtained, the percentage of patients requiring hospital admission or surgical debridement in the operating room, death within 30 days after the first culture date, and the number of patients with a recurrent or persistent skin infection (defined as a recurrence of symptoms within 90 days after the first culture date or persistence of symptoms after 7 days of appropriate therapy).

Data analysis. We assessed temporal changes in methicillin resistance among community-acquired isolates using [chi square] for trend. The association between demographic and clinical characteristics and the presence of a methicillin-resistant isolate were evaluated by using the [chi square] test with Yates correction. Because children have been a risk group for methicillin-resistant infections in some reports, (11,12) we compared the rate of resistance of those younger than 18 years of age with those 18 and over. Factors having an associated P-value less than 0.20 were included in an adjusted logistic regression model. Analyses were performed with the EpiInfo, version 6.0 (Centers for Disease Control and Prevention, Atlanta, GA), SAS, version 6.2 (SAS Institute Inc, Cary, NC), and Microsoft Excel (Renton, WA) software packages.

Results

Temporal trend in community-acquired MRSA

There was no apparent trend in the number of S aureus isolates from community-acquired infections (Fig. 1). However, there was a dramatic increase in the proportion of isolates that were methicillin-resistant, from 6% in the first quarter of 2002 to 45% by mid 2004 (P < 0.001).

Retrospective cohort study

During 2003, 249 outpatients had a culture growing S aureus, 70 with methicillin resistance and 179 with methicillin susceptibility. From that group, 13 patients with methicillin-resistant and 43 with methicillin-susceptible isolates were excluded for the following reasons: no documentation of skin or soft tissue infection at time of culture (47); history of previous methicillin-resistant infection or colonization (7); no medical record available (2). After exclusions, 193 patients were included in the analysis, 57 (30%) with methicillin-resistant isolates and 136 (70%) with methicillin-susceptible isolates.

Factors associated with methicillin-resistant infection

In an unadjusted analysis, patients infected with methicillin-resistant isolates were more often male (RR, 2.2; 95% CI, 1.2-3.8), HIV infected (RR, 1.9; 95% CI, 1.2-2.9), homeless (RR, 1.5; 95% CI, 0.97-2.4) and greater than 18 years of age (RR, 2.5; 95% CI, 0.99-6.5) (Table 1). Among patients with diagnosed HIV infection, there was no apparent association between infection with methicillin-resistant isolate and CD4 lymphocyte count or HIV RNA levels (data not shown). An adjusted logistic regression showed that being older than 18 years of age and male sex were associated with methicillin-resistant infection (Table 1).

[FIGURE 1 OMITTED]

There was no apparent association between community-acquired methicillin-resistant infection and previous hospitalization, diabetes, outpatient central venous catheter, injection drug use in the past 3 months, incarceration, or presence of chronic wounds. In addition, recent prior antibiotic exposure was not associated with an increased risk of methicillin resistance. The rate of methicillin resistance was 21% (15 of 71) among those who had received an antibiotic prescription during a hospitalization or from an outpatient pharmacy in the Denver Health system in the preceding 6 months, compared with 34% (42 of 122) among those who had no documented antibiotic use. Evidence for use of the Denver Health outpatient pharmacy system for the patients in this cohort was as follows: 154 (80%) had received a prescription drug from an outpatient pharmacy since July 2002, and 95% (183 of 193) filled the prescription used to treat the S aureus infection evaluated in this study at a Denver Health pharmacy.

In an effort to define demographic or clinical groups that might be at low risk of methicillin-resistant infections, we examined groups without putative risk factors (Table 1). Having a methicillin-resistant isolate was common in all subgroups evaluated ([greater than or equal to]15%), with the exception of patients younger than 18 years of age (13%, 4 of 31).

Clinical manifestations of S aureus infections

Methicillin-resistant infections were not more severe than methicillin-susceptible infections; similar proportions required hospital admission (60% versus 65%) and/or surgery (43% versus 41%). Though blood cultures were not performed on the entire cohort, among patients sampled, the proportion with bacteremia was somewhat lower in the group with methicillin-resistant isolates (32% [18 of 57] versus 55% [34 of 136], P value, 0.07).

Treatment and outcomes

Compared with patients with susceptible isolates, patients with methicillin-resistant infections were less likely to receive initial antibiotic therapy to which the infecting strain was susceptible (Table 2). Median time to receipt of an antibiotic with in vitro activity was 3 days for patients with resistant isolates, compared with 0 days for patients with susceptible isolates. Despite available susceptibility data, 25% of all patients with resistant isolates had no record of an effective antibiotic being prescribed.

Outcomes of infection were similar in the two groups (Table 2). Patients with resistant and susceptible isolates had similar numbers of health care visits in the 30 days before and 30 days after the date of positive culture. Similar percentages of patients from both groups had recurrence of infection within 90 days. Only two deaths occurred within 30 days of culture, both among patients with susceptible isolates (P = 0.5).

Microbiology

Methicillin-resistant isolates were more often resistant to non-[beta]-lactam antibiotics than methicillin-susceptible isolates, including erythromycin (50% versus 34%), clindamycin (25% versus 4%), and the fluoroquinolone antibiotics (51% versus 3%). There was no significant difference in resistance to tetracycline (9% versus 3%) or trimethoprim-sulfamethoxazole (0% versus 2%). All isolates were susceptible to vancomycin, linezolid, and rifampin.

Molecular characterization of MRSA isolates

Pulsed field gel electrophoresis was performed on 20 methicillin-resistant isolates to compare relatedness between community-acquired strains as well as to our endemic hospital strains (Fig. 2). Of the study isolates, 10 were identical (E strain) and 4 additional isolates were related (E1-E4). These were distinct from our endemic hospital strain (A) and two sporadic hospital strains (B and C). Two additional community isolates (F and G) were distinct from both our hospital strains and the predominant community strain. Of the remaining community isolates, one was identical with our endemic hospital strain, one was identical with a sporadic hospital strain (B), and the final two isolates did not produce interpretable patterns. Therefore, 16 of 20 strains tested (80%) had not been previously identified as hospital strains, and only 2 strains (10%) had been previously identified as strains circulating in our hospital before the end of 2003.

[FIGURE 2 OMITTED]

The E strain and related strains (E1-E4) had never been identified among a hospitalized patient before December 2003. In December 2003, two hospitalized patients were diagnosed with infections with strain E. One had signs of infection within the first day after admission but did not have cultures obtained until day 4; the other patient was known to have been colonized with methicillin-resistant S aureus before his hospitalization. The E strain was not related to the MW2 strain, a strain of community-acquired methicillin-resistant S aureus common in the midwest United States. The E strain was positive for the PVL toxin, whereas strains E1, E2, and F were negative for this toxin.

Discussion

Like other major cities in the United States, central Denver has recently had a dramatic increase in community-acquired methicillin-resistant S aureus infections. In less than 3 years, the proportion of resistant isolates increased from 6 to 45%. Methicillin resistance was more common among adult men in our study, but rates of resistance were substantial (>15%) in nearly all of the demographic and clinical groups we evaluated. Most community-acquired methicillin-resistant strains had pulsed field gel electrophoresis patterns distinct from the endemic strain transmitted in our hospital. Patients with methicillin-resistant infections did not have worse outcomes than patients with susceptible infections, despite significant delays in the initiation of appropriate antimicrobial therapy.

Risk factors for methicillin-resistance have long been thought to be markers of chronic illness and exposure to the health care system: age older than 60 years, previous hospitalization (especially in an intensive care unit), and chronic open wounds. (13) Early reports on community-acquired methicillin-resistant infections often found these associations with the health care system, along with injection drug use and previous antibiotic use. (1,2) Studies of community carriage of methicillin-resistant S aureus found that prevalence was generally low (0.26 to 2.8%) and associated with previous hospitalization. (14,15) Therefore, community-acquired methicillin-resistant S aureus was thought to represent the spillover of resistant strains from the hospital to high-risk persons in the community. Two independent lines of evidence from our study demonstrate that the epidemiology of community-acquired methicillin-resistant infections has changed dramatically since these early reports. First, hospitalization and other indicators of frequent exposure to health care settings (presence of diabetes mellitus, chronic central venous catheter use, recent antibiotic use) were not associated with methicillin-resistant infection, suggesting that methicillin-resistant S aureus is being transmitted in settings other than health care facilities. Moreover, methicillin resistance was common in nearly all demographic and clinical subgroups, suggesting widespread community-based transmission of antibiotic-resistant strains. Second, pulsed field gel electrophoresis showed that community-acquired methicillin-resistant infection was usually due to strains not previously identified in our hospital. Our findings reinforce other recent studies demonstrating the lack of traditional risk factors among patients with community-acquired methicillin-resistant infections. (5,6,11,16,17)

The initial report of four pediatric deaths from community methicillin-resistant infections, (11) coupled with the discovery of this strain's strong association with the PVL toxin, (18,19) led to concern that a bacterial strain may have emerged having both antibiotic resistance and increased virulence. Subsequent studies have had inconsistent results, with some studies suggesting more severe infections and worse outcomes among community-acquired methicillin-resistant S aureus, (20,21) whereas others, including our own, have not found an association between [beta]-lactam resistance and clinical outcomes. (6,17,22) Furthermore, our study found no difference in health care utilization, hospital admission, surgery, death, or recurrence of infection between patients with resistant and susceptible infections.

Community-acquired methicillin-resistant S aureus is not generally associated with the multiclass antibiotic resistance commonly seen with hospital strains. The lack of multidrug resistance in community-acquired strains has been associated with a novel chromosomal antibiotic resistance cassette (type IV). (18) This locus carries only the mecA resistance gene (conveying resistance to [beta]-lactam antibiotics) and is smaller and presumably more mobile than types I-III, which are typically found in hospital-acquired methicillin-resistant S aureus. This probably conveys a fitness advantage outside of the antibiotic pressures of the hospital (23) and may explain its rapid dissemination in the community. In our outbreak, resistance to the fluoroquinolone class was surprisingly high, at 51%. The higher rates of resistance to antibiotic classes other than [beta]-lactams seen in our study raises concerns that methicillin-resistant strains may be able to incorporate resistance mechanisms without sacrificing fitness. This will require further investigation.

Our study has several important limitations. First, we might have overestimated the rate of methicillin resistance among community-acquired S aureus infections if clinicians more often obtained bacterial culture and susceptibility testing among patients with unresponsive infections or from patients having perceived risk factors for antibiotic resistance that were not recorded in the medical record. However, the small number of health care visits in the month before S aureus diagnosis suggests that bacterial cultures were not reserved for patients with inadequate responses to initial empiric therapy, and the similarity in health care usage by patients with resistant and susceptible isolates suggests that differential use of bacterial culture did not confound our results. Second, the presence of risk factors for methicillin resistance was not uniformly recorded in the medical record. However, we only collected the information recorded before the receipt of the susceptibility test results, so it is doubtful that the information was biased, even if incomplete. Finally, we used an objective measure of prior antibiotic use--receipt of an antibiotic during a prior hospitalization or from one of the outpatient pharmacies in our system--but some patients may have received antibiotics from other sources.

The widespread emergence of community-acquired methicillin-resistant S aureus has important clinical implications. Based on the results of this study, we recommend that all patients admitted to our hospital with serious infections suspected to be due to S aureus be treated with an antibiotic active against resistant strains, generally vancomycin. Whether high rates of methicillin resistance should change initial antibiotic use for less serious S aureus infections remains controversial. Patients with methicillin-resistant isolates in our study commonly had delays before receiving an appropriate antibiotic but did not have a higher rate of adverse treatment outcomes. We suspect that the lack of association between timing of appropriate antibiotic use and clinical outcomes is that many S aureus infections can be effectively treated with appropriate incision and drainage, and without adjunctive antibiotics. However, when antibiotic therapy is thought to be clinically indicated, we now recommend treatment with antibiotics that have predictable activity against the common methicillin-resistant strains in our community, generally trimethoprim/sulfamethoxazole or doxycycline.

Acknowledgments

We thank Dr. Lance Peterson and Suzanne Paule (Evanston Northwestern Healthcare) for their assistance with strain interpretation and for performing PVL testing.

References

1. Crane LR, Levine DP, Zervos MJ, et al. Bacteremia in narcotic addicts at the Detroit Medical Center, I: microbiology, epidemiology, risk factors, and empiric therapy. Rev Infect Dis 1986;8:364-373.

2. Lewis E, Saravolatz LD. Comparison of methicillin-resistant and methicillin-sensitive Staphylococcus aureus bacteremia. Am J Infect Control 1985;13:109-114.

3. Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus infections among competitive sports participants: Colorado, Indiana, Pennsylvania, and Los Angels County, 2000-2003. MMWR Morb Mortal Wkly Rep 2003;52:793-795.

4. Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus skin and soft tissue infections in a state prison: Mississippi, 2000. MMWR Morb Mortal Wkly Rep 2001;50:919-922.

5. Baggett HC, Hennessy TW, Leman R, et al. An outbreak of community-onset methicillin-resistant Staphylococcus aureus skin infections in southwestern Alaska. Infect Control Hosp Epidemiol 2003;24:397-402.

6. Sattler CA Jr, Mason EO, 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;21:910-917.

7. National Committee for Clinical Laboratory Standards. Performance Standards for Antimicrobial Disk Susceptibility Testing-Sixth Edition: Approved Standard M2-A6. 6th ed. Wayne, PA; 1998.

8. Tenover FC, Arbeit RD, Goering RV, et al. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J Clin Microbiol 1995;33:2233-2239.

9. 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;29:1128-1132.

10. Farizo K, Buehler J, Chamberland M, et al. Spectrum of disease in persons with human immunodeficiency virus infection in the United States. JAMA 1992;267:1798-1805.

11. Centers for Disease Control and Prevention. Four pediatric deaths from community-acquired methicillin-resistant Staphylococcus aureus: Minnesota and North Dakota, 1997-1999. MMWR Morb Mortal Wkly Rep 1999;48:707-710.

12. Mongkolrattanothai K, Boyle S, Kahana MD, et al. Severe Staphylococcus aureus infections caused by clonally related community-acquired methicillin-susceptible and methicillin-resistant isolates. Clin Infect Dis 2003;37:1050-1058.

13. Lucet JC, Chevret S, Durand-Zaleski I, et al. Prevalence and risk factors for carriage of methicillin-resistant Staphylococcus aureus at admission to the intensive care unit: results of a multicenter study. Arch Intern Med 2003;163:181-188.

14. Charlebois ED, Bangsberg DR, Moss NJ, et al. Population-based community prevalence of methicillin-resistant Staphylococcus aureus in the urban poor of San Francisco. Clin Infect Dis 2002;34:425-433.

15. Shopsin B, Mathema B, Martinez J, et al. Prevalence of methicillin-resistant and methicillin-susceptible Staphylococcus aureus in the community. J Infect Dis 2000;182:359-362.

16. Baggett HC, Hennessy TW, Rudolph K, et al. Community-onset methicillin-resistant Staphylococcus aureus associated with antibiotic use and the cytotoxin Panton-Valentine leukocidin during a furunculosis outbreak in rural Alaska. J Infect Dis 2004;189:1565-1573.

17. Groom AV, Wolsey DH, Naimi TS, et al. Community-acquired methicillin-resistant Staphylococcus aureus in a rural American Indian community. JAMA 2001;286:1201-1205.

18. Baba T, Takeuchi F, Kuroda M, et al. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 2002;359:1819-1827.

19. 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;9:978-984.

20. Gillet Y, Issartel B, Vanhems P, et al. Association between Staphylococcus aureus strains carrying gene for Panton-Valentine leukocidin and highly lethal necrotizing pneumonia in young immunocompetent patients. Lancet 2002;359:753-759.

21. Ellis MW, Hospenthal DR, Dooley DP, et al. Natural history of community-acquired methicillin-resistant Staphylococcus aureus colonization and infection in soldiers. Clin Infect Dis 2004;39:971-979.

22. Herold BC, Immergluck LC, Maranan MC, et al. Community-acquired methicillin-resistant Staphylococcus aureus in children with no identified predisposing risk. JAMA 1998;279:593-598.

23. Okuma K, Iwakawa K, Turnidge JD, et al. Dissemination of new methicillin-resistant Staphylococcus aureus clones in the community. J Clin Microbiol 2002;40:4289-4294.

We share the same biology, regardless of ideology.
--Sting

Megan J. Clancy, MD, Amy Graepler, MT, Peter E. Breese, MSPH, Connie S. Price, MD, and William J. Burman, MD

From the Departments of Internal Medicine and Public Health, Denver Health and the Department of Medicine (Division of Infections Diseases), University of Colorado Health Sciences Center, Denver, CO.

Reprint requests to Dr. William Burman, 605 Bannock MC 2600, Denver, CO 80204. Email: [e-mail]

Accepted June 30, 2005.

RELATED ARTICLE: Key Points

* There is a rapid increase in community-acquired methicillin-resistant Staphylococcus aureus skin and soft tissue infections.

* These infections are widespread in the community, and we were unable to identify an at-risk group.

* These infections did not have worse outcomes than infections from methicillin-susceptible isolates.

* These infections were caused by a single pulsed field type that is responsible for outbreaks in other parts of the country.

Table 1. Demographic and clinical factors associated with a methicillin-
resistant isolate among patients with community-acquired S aureus skin
and soft tissue infections

                                    Unadjusted analysis
                                    MRSA infected/
                                    total n of this  Relative risk
Risk factor                         risk group (%)   (95% CI)

Gender
  Men                               45/122 (37)      2.2 (1.2-3.8)
  Women                             12/71 (17)
Age
  [greater than or equal to] 18 yr  53/162 (32)      2.5 (0.99-6.5)
  < 18 yr                            4/31 (13)
Race
  White                             27/83 (32)       0.77 (0.42-1.45)
  Hispanic/Latino                   22/85 (26)       1.37 (0.73-2.58)
  Black                              7/18 (39)       0.62 (0.23-1.71)
  Other                              1/7 (14)        2.58 (0.30-21.97)
HIV serostatus
  HIV positive                      14/30 (47)       1.9 (1.2-2.9)
  HIV negative or unknown           43/163 (26)
Homelessness
  Homeless in the past year         18/45 (45)       1.5 (0.97-2.4)
  Not homeless                      39/148 (26)
Hospitalization
  Hospitalized within the past      16/73 (22)       0.6 (0.4-1.1)
    year
  Not hospitalized                  41/120 (34)
Injection drug use
  Injection drug use in past 3 mo    7/24 (29)       1.0 (0.5-2.0)
  No injection drug use             50/169 (30)
Diabetes mellitus
  Present                            3/15 (20)       0.4 (0.1-1.1)
  Absent                            54/178 (30)
Antibiotic use within 6 mo (a)
  Antibiotic use                    15/71 (21)       0.51 (0.26-1.0)
  No antibiotic use                 42/122 (34)
Chronic wound
  Present                           10/30 (33)       0.53 (0.2-1.2)
  Absent                            47/163 (29)
Outpatient central venous catheter
  in past 3 mo
  Outpatient catheter                1/6 (17)        0.6 (0.1-3.4)
  No outpatient catheter            56/187 (30)
Incarceration at time of infection
  Incarcerated                       2/7 (29)        1.0 (0.3-3.2)
  Not incarcerated                  55/186 (30)

                                    Adjusted analysis
Risk factor                         Odds ratio (95% CI)

Gender
  Men                               2.06 (0.92-4.6)
  Women
Age
  [greater than or equal to] 18 yr  3.7 (1.2-12.9)
  < 18 yr
Race
  White                             0
  Hispanic/Latino                   1.68 (0.53-5.4)
  Black                             1.05 (0.48-2.3)
  Other                             0.28 (0.03-2.6)
HIV serostatus
  HIV positive                      1.7 (0.6-4.5)
  HIV negative or unknown
Homelessness
  Homeless in the past year         1.3 (0.6-2.9)
  Not homeless
Hospitalization
  Hospitalized within the past      0.9 (0.4-2.2)
    year
  Not hospitalized
Injection drug use
  Injection drug use in past 3 mo   --
  No injection drug use
Diabetes mellitus
  Present                           0.4 (0.1-1.6)
  Absent
Antibiotic use within 6 mo (a)
  Antibiotic use                    0.47 (0.2-1.2)
  No antibiotic use
Chronic wound
  Present                           0.84 (0.3-2.3)
  Absent
Outpatient central venous catheter
  in past 3 mo
  Outpatient catheter               --
  No outpatient catheter
Incarceration at time of infection
  Incarcerated                      --
  Not incarcerated

(a) Antibiotic used during a hospitalization or dispensed from a Denver
Health outpatient pharmacy within the past 6 months.

Table 2. Outcomes of community-acquired S aureus skin and soft tissue
infections by methicillin susceptibility of the isolate

                                     Methicillin    Methicillin
                                     resistant      susceptible
Treatment                            (n = 57)       (n = 136)    P value

Median delay (d) to antibiotic        3.0 (0, 4.5)   0           N/A
  therapy with in vitro activity
  against the S aureus isolate
  (IQR) (a)
No record of antibiotic treatment    14 (25)         0           < 0.001
  with an agent having in vitro
  activity against the S aureus
  isolate, n (%)
Required operative debridement,      25 (43)        77 (41)        0.73
  n (%)
Outcome
Median health care visits within      1 (0, 1)       1 (0, 2)      0.48
  30 days before the positive
  culture (IQR)
Median health care visits within      1 (0, 3)       1 (0, 2)      0.76
  30 days after the positive
  culture (IQR)
Hospitalized, n (%)                  32 (60)        87 (65)        0.50
Recurrence or persistent infection,   7 (12)         9 (7)         0.27
  n (%)
Death within 30 days, n (%)           0              2 (1)         1.0

(a) Interquartile range (25th to 75th percentile).