Methicillin-Resistant Staphylococcus aureus Colonization, Behavioral Risk Factors, and Skin and Soft-Tissue Infection at an Ambulatory Clinic Serving a Large Population of HIV-Infected Men Who Have Sex with Men

Community-associated (CA) methicillin-resistant Staphylococcus aureus (MRSA) is increasingly recognized as an important cause of skin and soft-tissue infection (SSTI) [1]. However, the pathogenesis and risk factors for CA-MRSA SSTI remain incompletely understood [2].

The impact of CA-MRSA has been considerable in clinical settings that serve large urban populations of men who have sex with men [3, 4]. Lee et al. [3] found that risky behaviors, such as anonymous sex, illicit drug use, and contact with persons who have skin infection, were associated with MRSA SSTI. In another cohort of human immunodeficiency virus (HIV)-infected patients, Mathews et al. [4] observed an increased risk of SSTI associated with an elevated HIV load and CD4 cell counts <50 cells/µL, whereas trimethoprim-sulfamethoxazole prophylaxis appeared protective.

Along with behavioral risk factors and comorbid conditions, S. aureus colonization appears to increase the risk of SSTI [5]. Although nasal colonization is the best understood, up to 20% of individuals may have gastrointestinal or perineal colonization [6]. Indeed, the frequency of groin and buttock infections [7] suggests that perineal colonization may be relevant to CA-MRSA SSTI [2].

To better understand SSTI risk among persons with nares or perianal MRSA colonization and to evaluate potential behavioral and demographic risk factors, we conducted a prospective cohort study involving nearly 800 outpatients.

Methods. Fenway Community Health is located in Boston, Massachusetts, and provides health care to a large gay, lesbian, bisexual, and transgender population and to >1200 HIV-infected patients. Beth Israel Deaconess Medical Center is a 675-bed tertiary care hospital in Boston and is the primary referral center for patients from Fenway Community Health. This study was approved by the institutional review boards of Beth Israel Deaconess Medical Center and Fenway Community Health.

A total of 798 patients (age, ⩾18 years) were sequentially enrolled from the Fenway Community Health medicine clinics during the period from October 2005 through January 2007. After written informed consent was obtained, participants completed the study questionnaire. Participants' bilateral anterior nares, perianal region, and any open wounds were sampled with rayon-tipped swabs (BBL Cultureswabs; Becton Dickinson). Perianal samples were self-collected. Twenty-eight persons refused to provide perianal swabs but were permitted to enroll in the study.

Samples were plated on blood agar and mannitol salt agar and placed in enrichment broth at 35°C for 48 h. Morphologically distinct colonies were identified as S. aureus using Staphaurex latex agglutination (Remel). If discordant S. aureus culture results were obtained from nares and perianal samples, the broth samples were subcultured. MRSA isolates were further characterized by pulsed-field gel electrophoresis.

Participants were observed for 1 year for the development of SSTI. One investigator (J.D.S.), who was blinded to participants' baseline colonization status, performed data extraction from clinic notes, emergency department records, and microbiology logs. Only SSTIs that were diagnosed at outpatient clinics or in an emergency department were included. SSTIs that had been present at the time of enrollment were not counted as incident infections. Patients were classified as having multiple SSTIs if new signs of infection developed at an anatomically distinct site and if prior infection had been classified as resolved by the clinician.

Wound specimens were obtained for culture at the clinicians' discretion. Nares isolates were not counted unless the patient had evidence of a nares SSTI. Susceptibility testing was performed at the Beth Israel Deaconess Medical Center laboratory or at Quest Diagnostics (Madison, NJ).

Analyses were conducted using Stata, version 9.2 (Stata Corp). P values <.05 were considered to be statistically significant. All P values are 2-sided, and exact 95% confidence intervals (CIs) were calculated for odds ratios (ORs). Fisher's exact test was used for categorical variables. Multivariate analysis was performed using a stepwise forward-selection logistic regression model with likelihood ratio tests. To assess model stability, we also constructed logistic regression models by sequentially adding variables that had been identified as significant in the univariate analysis, after eliminating collinearvariables.

Results. A total of 795 participants had data available for follow-up. The median age of participants was 41 years, and the majority (610 of 795) were white men. Nearly 70% of participants (547 of 795) self-identified as men who have sex with men, and nearly one-third (243 of 795) had HIV infection.

Overall, 30 (3.8%) of 795 participants had MRSA identified from any site at enrollment (table 1). MRSA was isolated from nares for 26 (3.3%) of 794 participants, from perianal swabs for 12 (1.6%) of 767 participants, and from wound swabs for 5 (83%) of 6 participants. Most isolates (22 of 30) had pulsed-field gel electrophoresis patterns consistent with USA300.

Twenty-nine participants (3.6%) had an SSTI at enrollment. Seventy-three persons (9.2%) developed 93 SSTIs, for an incidence of 11.7 cases per 100 person-years. Patients with SSTI presented a median of 125 days after enrollment. More than 20% of persons with any SSTI (16 of 73) developed recurrent SSTI, with mean (± standard deviation) of 1.27±0.56 infections per year for persons with any SSTI. Culture specimens were obtained from 41 (56%) of 73 patients with an SSTI, of which 22 (53.7%) of 41 yielded MRSA. Most cases presented as cellulitis (31 cases [42.5%]) or furuncles (26 cases [35.6%]). The lower extremities were most frequently involved (22 cases [30.1%]).

Isolation of MRSA at baseline from the nares (OR, 4.81; 95% CI, 1.73–12.13), the perianal region (OR, 14.86; 95% CI, 3.90–60.64), or wound swabs (OR, 15.4; 95% CI, 1.72–186.1) were all strongly associated with SSTI. Of the 30 patients who were colonized with MRSA at enrollment, 11 (36.7%) subsequently developed SSTI, compared with 62 (8.1%) of 765 participants without MRSA ( P<.001). Recurrent SSTI was more common among persons with perianal colonization (OR, 3.31; 95% CI, 0.42–22.09) or nares colonization (OR, 2.4; 95% CI, 0.33–14.30), but the increase was not statistically significant.

HIV-infected individuals were more likely to develop SSTI, but neither CD4 cell count nor receipt of antiretroviral therapy was found to be significantly protective against SSTI (table 2). Prophylaxis with trimethoprim-sulfamethoxazole was associated with a non-statistically significant protective effect. Prior SSTIs that required antibiotic treatment or surgical drainage were strongly associated with recurrent infection. SSTI risk was also increased among users of illicit drugs, especially crystal methamphetamine, and among persons who engaged in anonymous sex, had greater numbers of sexual partners, or reported a history of sexually transmitted diseases. Being a man who have sex with men and anal intercourse were also risk factors for SSTI.

In the final multivariate model, MRSA perianal colonization—but not nares or wound colonization—remained statistically significantly associated with SSTI (OR, 10.34; 95% CI, 2.84–37.60). Prior skin infection (OR, 4.08; 95% CI, 2.31–7.22) and crystal methamphetamine use (OR, 4.98; 95% CI, 2.60–9.55) were also significant risk factors. Perianal colonization, previous SSTI, and crystal methamphetamine use remained independent predictors in a subanalysis restricted to culture-confirmed MRSA SSTI.

Discussion. In this study, we prospectively investigated the relationship between nares and perianal MRSA colonization and SSTI. We also assessed SSTI risk associated with a range of demographic and behavioral characteristics. As hypothesized, MRSA colonization was a strong risk factor for SSTI; more than one-third of the MRSA-colonized persons developed SSTI, similar to prior estimates in a population of soldiers [8].

The association between perianal MRSA colonization and SSTI was perhaps our most intriguing finding. To our knowledge, this is the first prospective study to have demonstrated such a relationship in the era of CA-MRSA, although prior work from different settings had suggested perianal colonization might be important. Nearly 50 years ago, a study involving patients with recurrent furunculosis found that 56% had perineal colonization [9]. More recently, Squier et al. [10] found that rectal S. aureus colonization was associated with an increased risk of subsequent infection among surgical inpatients.

In addition to MRSA colonization, higher-risk sexual behaviors and illicit drug use—in particular, use of crystal methamphetamine—were also strongly associated with SSTI. Crystal methamphetamine users are more likely to have unprotected sex and multiple partners [11], which could promote the spread of CA-MRSA. These findings lend strength to the hypothesis that skin-skin contacts are important for CA-MRSA transmission [2].

Given our clinics' large population of HIV-infected patients, we were also interested in whether SSTI risk was related to the extent of immunosuppression. Two studies suggested that SSTI risk is increased at lower CD4 cell counts [4, 7]. However, like Lee et al. [3], we did not find a statistically significant association between CD4 cell count and SSTI.

This investigation's strengths include its prospective design, relatively large size, and broad assessment of possible SSTI risk factors. Although our health centers serve a large population of HIV-infected men who have sex with men, the spectrum of SSTIs reported here are similar to that reported more widely [1].

This study also had several limitations. Our ability to compare baseline and clinical isolates was limited, because the latter were not accessible for additional molecular testing. Also, perianal and nares colonization status were assessed at one time point and may have changed prior to infection; however, this would have been expected to bias our results toward the null and would not be expected to invalidate the observed association between baseline colonization and SSTI. Because we used passive case-finding, it is possible that not all SSTIs were identified. However, we believe that the majority were captured; >93% of participants had subsequent follow-up.

In conclusion, we found that prior SSTI and behaviors possibly linked to more frequent skin-skin contact, such as greater numbers of sexual partners and illicit drug use, were significantly associated with SSTI. We also demonstrated that perianal MRSA colonization is associated with increased SSTI risk, as recognized for nares colonization [5]. Most (10 of 12), but not all, patients with MRSA perianal colonization had nares colonization. Routine nares screening therefore likely fails toidentify some MRSA-colonized persons, providing support to proposals for collecting perianal swabs as part of MRSA surveillance programs [12]. Perianal colonization may also be relevant to SSTI recurrence, but better understanding of MRSA colonization dynamics is needed.

• Received December 2, 2008.
• Accepted February 18, 2009.

• © 2009 by the Infectious Diseases Society of America