Share this article on:

A Multivariate Model to Classify Methicillin-Resistant Staphylococcus aureus Infections as Health Care–Associated or Community-Associated Pathogen

Joshi, Manjari MBBS

Infectious Diseases in Clinical Practice: January 2012 - Volume 20 - Issue 1 - p 1–2
doi: 10.1097/IPC.0b013e3182440dba
Editorial Comment

From the R Adams Cowley Shock Trauma Center, University of Maryland School of Medicine, Baltimore, MD.

Correspondence to: Manjari Joshi, MBBS, R Adams Cowley Shock Trauma Center, University of Maryland School of Medicine, Room T3R51, 22 S Greene St, Baltimore, MD 21201. E-mail:

The author discloses that she has received recent research funding from Pfizer, Synthes, and Atox-Bio. She has served on the speaker bureau of Pfizer and also served as a consultant for Pfizer and Synthes.

Infections caused by methicillin-resistant Staphylococcus aureus (MRSA) are associated with considerable morbidity and mortality. Methicillin-resistant Staphylococcus aureus has been predominantly a health care–associated (HA-MRSA) pathogen occurring in patients who are chronically debilitated with multiple underlying comorbid conditions. In the 1990s, community-acquired MRSA (CA-MRSA) emerged and differed from HA-MRSA in that it occurred in young healthy individuals. It was also associated with predominantly skin and soft tissue infections, as well as with more severe infections.1,2 It has been postulated that most isolates of CA-MRSA contain genes encoding for a toxin called Panton-Valentine leukocidin (PVL). This toxin is associated with more severe infections such as necrotizing pneumonia and invasive skin and soft tissue infections.3

Earlier on, these 2 strains of MRSA had distinct characteristics. However, as the reservoir of the CA-MRSA increased in the community, it started spreading to the health care environment, displacing HA-MRSA.4,5 In fact, a mathematical model has been developed, which suggests that eventually, CA-MRSA will become the dominant MRSA strain in hospital and health care facilities.6 Despite this spread, the rates of HA-MRSA are increasing in the community, as it is often associated with health care exposure. Therefore, the boundaries between HA-MRSA and CA-MRSA are progressively blurring.

However, for now, CA-MRSA is still distinct from HA-MRSA based on genotypic, phenotypic, and epidemiologic characteristics. Several sets of criteria have been proposed, including one by the Centers for Disease Control and Prevention,7 to distinguish the 2 strains of MRSA:

  • - No medical history of MRSA colonization or infection;
  • - No permanent medical devices such as indwelling catheters;
  • - No hospitalization, admission to long term care facility, dialysis or surgery in the past year;
  • - Diagnosis of MRSA made in either the outpatient setting or within 48 hours of hospital admission.

Several other factors such as age, susceptibility patterns, type of infection, and preexisting illness have also been considered in the classification scheme. To date, no clinical schemes have been completely reliable. Unfortunately, laboratory methods such as pulsed-field gel electrophoresis (PFGE), which are not routine tests, are the criterion standard for differentiating MRSA strains.

In the current issue, Sievert et al8 use the laboratory-based molecular typing PFGE as a standard to develop a clinical model that can predict whether an infection was caused by HA-MRSA or CA-MRSA. The investigation used a subset of the Michigan surveillance database to improve the clinical prediction of MRSA infections. Previously, in their setting, 3 data points—health care risk factors, infection type, or susceptibility pattern—were used to classify MRSA infections. Their study demonstrates that 4 variables (infection type, susceptibility pattern, age, and hospitalization), when considered together, were more likely to improve accuracy of categorizing infections as CA-MRSA or HA-MRSA.

Several questions arise from this study. These are:

  1. In a setting of a hospitalized patient, is it important to routinely distinguish HA-MRSA from CA-MRSA?

If one answers “yes,” one can justify that CA-MRSA can produce more severe infections because of PVL toxin production. Theoretical rationale for detecting PVL-producing strains might be to direct physicians to add a protein (toxin) synthesis inhibitor such as clindamycin or linezolid to the therapeutic regimen. Furthermore, if such an infection occurs in a debilitated host, it can have more dire consequences. Therefore, it is clinically relevant to know whether the infection is caused by CA-MRSA or HA-MRSA.

If one answers “no,” one can deduce that if an infection with MRSA is anticipated, the treatment for HA-MRSA will automatically cover for CA-MRSA and no different antimicrobial therapy is recommended. However, such infections have infection control (IC) implications (shown below).

  • 2. Would it be important from an epidemiological point of view to differentiate HA-MRSA from CA-MRSA?

As the community reservoir of CA-MRSA expands, more patients with CA-MRSA will be hospitalized. As the severity of infections increases, the length of stay will increase. This not only has financial implications but also means that with increasing length of stay, more in-hospital transmission can occur. Therefore, in addition to maximal use of IC strategies such as hand hygiene, screening, and decolonization, other novel strategies might also be investigated to reduce transmission of MRSA.9

However, in a community setting, IC measures cannot be used effectively. Therefore, the community reservoir CA-MRSA may continue to grow and feed into the hospital setting.

  • 3. Are there data from in vitro studies that can have clinical implications?

Community-acquired MRSA carries the gene cassette staphylococcal cassette chromosome mec type IV, which is much smaller than the staphylococcal cassette chromosome mec types I, II, and III carried by HA-MRSA. Therefore, the smaller size may lead to more efficient transmission of resistance.10 Additionally, in vitro studies have demonstrated that the growth rate of CA-MRSA is 1.33 times faster than the growth rate of HA-MRSA. Potentially, this could lead to enhanced colonization and transmission of CA-MRSA.11 Lastly, CA-MRSA strains can carry the PVL gene, which produces cytotoxins associated with tissue necrosis and destruction and therefore results in more severe infections.

Despite all these questions, the discrimination of the type of MRSA infection in a health care setting is important. Molecular methods such as PFGE typing are expensive, require special training, and are not universally available. A simple clinical model that uses easily obtainable, accurate, and reliable parameters can be a useful modality for predicting CA-MRSA infection from HA-MRSA. However, because the epidemiology of MRSA is constantly changing, the clinical models that distinguish CA-MRSA from HA-MRSA will need to be dynamic and constantly changing. In the future, it will be useful to develop models that should be tested prospectively with a larger number of patients and in different clinical settings.

Furthermore, this model has several limitations. It was based on retrospective data, had a disproportionate number of HA-MRSA, and was limited by the number of PFGE results available.

The study also demonstrated that by using this model, the ability to identify HA-MRSA was much greater than the ability to identify CA-MRSA. This observation was linked to the increasing presence of CA-MRSA in the health care setting and in the hospital environment. The concept of this study is important and proves the point that accurate clinical models need to be developed to distinguish CA-MRSA from HA-MRSA. However, whether such a model will be valid in the constantly changing world of MRSA is questionable.

Back to Top | Article Outline


1. Daum RS. Community-acquired methicillin resistant Staphylococcus aureus infections. Pediatr Infect Dis J. 1998; 17: 745–746.
2. King MD, Humphrey BJ, Wang YF, et al.. Emergence of community-acquired methicillin-resistant Staphylococcus aureus USA 300 clone as the predominant cause of skin and soft tissue infections. Ann Intern Med. 2006; 6: 223–229.
3. Miller LG, Perdreau-Remington F, Rieg G, et al.. Necrotizing fasciitis caused by community-associated methicillin-resistant Staphylococcus aureus in Los Angeles. N Engl J Med. 2005; 352: 1445–1453.
4. Seybold U, Kourbatova EV, Johnson JG, et al.. Emergence of CA-MRSA USA 300 genotype as a major cause of health-care associated bloodstream infections. Clin Infect Dis. 2006; 42: 647–666.
5. Popvich KJ, Weistein RA, Hota B. Are community associated MRSA strains replacing traditional nosocomial MRSA strains? Clin Infect Dis. 2008; 46: 795–798.
6. D’Agata EMC, Webb GF, Horm MA, et al.. Modeling the invasion of community-acquired methicillin-resistant Staphylococcus aureus into hospitals. Clin Infect Dis. 2009; 48: 274–284.
7. Centers for Disease Control and Prevention. Community-associated MRSA information for clinicians. Available at: Accessed November, 2011.
8. Sievert DM, Boulton ML, Wilson ML, et al.. A multivariate model to classify methicillin-resistant Staphylococcus aureus infections as healthcare or community-associated. Clin Infect Dis. 2012; 20: 42–48.
9. File TM. Impact of community-acquired methicillin-resistant Staphylococcus aureus in the hospital setting. Cleveland Clinic Journal of Medicine. 2007; 74 (suppl 4): S6–S11.
10. Weber JT. Community-associated MRSA. Clin Infect Dis. 2005; 41 (suppl 4): S269–S272.
11. Okuma K, Iwakawa K, Turnridge JD, et al.. Dissemination of new MRSA clones in the community. J Clin Microbiol. 2002; 40: 4289–4294.
© 2012 Lippincott Williams & Wilkins, Inc.