|Home | About | Journals | Submit | Contact Us | Français|
There is growing consensus that community-associated methicillin-resistant Staphylococcus aureus (MRSA) plays an important role in infections both in the community and in the healthcare setting.1 Risk factors for healthcare-associated MRSA have been well studied, but less is known about the risk for colonization with S. aureus or MRSA in the community.
Because they share crowded, often unsanitary living conditions, typically manifest poor underlying health, and have limited access to hygiene facilities and healthcare services, individuals who are homeless may have an increased risk for MRSA colonization. The aim of this study was to investigate the prevalence of S. aureus and MRSA colonization among a population of homeless individuals and to determine risk factors for S. aureus and MRSA colonization.
A prevalence study was conducted at 3 shelter locations in Columbus, Ohio (a men's emergency shelter, a women's emergency shelter, and a men's long-term shelter), and at a “Homeless Stand-Down,” a multiday outreach event for homeless individuals in Cleveland, Ohio. At each site, participants were informed of the study by posters and flyers throughout the facility and self-referred to a private area where the study was conducted.
After providing written informed consent, participants were asked to complete a brief questionnaire on risk factors for S. aureus and MRSA colonization, including demographic characteristics, current health status, substance use, recent living arrangements, and hygiene practices. Nasal swab samples were collected from each participant with a dry, cotton-tipped applicator before and after instillation of normal saline (BBL CultureSwab; Becton Dickinson). Specimens were preincubated in a nonselective liquid media (BBL Trypticase Soy Broth, Becton Dickinson) at 37°C for 12 hours, and S. aureus was identified using routine microbiologic procedures. S. aureus isolates were subcultured on oxacillin resistance screening agar (ORSAB, Oxoid) for the identification of MRSA.
Overall prevalence rates were calculated using the SPSS statistical package, version 14 (SPSS). Assuming a baseline prevalence of 30%, a sample size of 215 was determined to provide a prevalence estimate with a margin of error of 6%. Differences between individuals who tested positive and individuals who tested negative for S. aureus and MRSA were compared using χ2 tests and Student t tests. Logistic regression was used to build a model of predictor variables for MRSA colonization with variables that were significant at the P < .10 level. The effects of predictor variables were quantified by estimating prevalence odds ratios (ORs) and 95% confidence intervals (CIs).
Seventy-five (34.9%) of 215 participants had positive results for S. aureus. Of 71 tested S. aureus specimens, 55 (77.5%) were MRSA, for an overall estimate of MRSA colonization prevalence of 25.6%.
Individuals who tested positive for MRSA reported higher frequencies of recent antibiotic use (23.6% vs 12.3%; P < .04), history of alcoholism (10.9% vs 4.5%; P < .08), and current smoking (85.5% vs 73.1%; P < .06) and a lower frequency of having stayed with a friend for at least 1 night in the previous 30 days (5.5% vs 16.0%; P < .03), compared with individuals who tested negative for MRSA. Similar results were observed for individuals who tested positive for S. aureus, compared with individuals who tested negative for S. aureus. Differences in other characteristics did not approach statistical significance (Table).
Results of multivariable logistic regression analysis revealed 3 variables associated with MRSA colonization. In the final multivariable regression model, recent antibiotic use (OR, 2.73 [95% CI, 1.10–6.77]) and history of alcoholism (OR, 5.14 [95% CI, 1.24–21.4]) increased the risk of MRSA infection, whereas living with a friend was protective (OR, 0.18 [95% CI, 0.04–0.72]).
This study documents the prevalence of S. aureus and MRSA colonization among a population of homeless individuals. Although the S. aureus colonization prevalence in this sample is similar to prior population estimates,2 the prevalence of MRSA colonization (25.6%) was much higher than has been previously reported for individuals in other community settings (1%–2%).2-5 On the basis of these findings, the rate of MRSA colonization among the homeless may be 10–20 times higher than the rate among the general population. This finding is different from the 2.8% prevalence rate found in a study of the urban poor conducted in San Fransisco in 1999–2000.6
The risk factors that were identified for S. aureus and MRSA colonization were consistent with those found in other studies. Living with a friend was associated with lower odds for both S. aureus and MRSA colonization, and these lower odds may represent less exposure to congregate living situations and relatively better social support and access to hygiene facilities. We were not able to identify specific hygiene-related practices or health conditions that were associated with an increased risk of S. aureus or MRSA colonization, such as the frequency of handwashing, showering, or doing laundry.
There are several limitations of the current study. Because we did not determine the genetic subtypes of the isolates, we were unable to control for the effects of shelter-level or community-level MRSA prevalence. Similar to what has been demonstrated in family members of MRSA-colonized individuals,7,8 it is likely that our sample contained clusters of individuals highly colonized by the same bacterial strains.
Future studies should address S. aureus and MRSA colonization in the homeless and other vulnerable populations.9,10 Identification of specific risk factors and broad intervention to improve the living conditions of high-risk populations may be an important aspect of MRSA prevention control activities. Interventions might include improved individual hygiene, decontamination of colonized individuals, cleaning and improved sanitation of shelters, and coordinated community efforts to address homelessness as a public health problem.
Potential conflicts of interest. All authors report no conflicts of interest relevant to this article.