Community-Associated Methicillin-Resistant Staphylococcus aureus, Minnesota, 2000-2003

Jessica M. Buck; Kathryn Como-Sabetti; Kathleen H. Harriman; Richard N. Danila; David J. Boxrud; Anita Glennen; Ruth Lynfield


Emerging Infectious Diseases. 2005;11(10) 

In This Article


In 2000, 12 sentinel hospitals in Minnesota (6 in the 7-county Minneapolis–St. Paul metropolitan area and 6 in greater Minnesota) began reporting all cases of MRSA isolated in their respective microbiology laboratories to the Minnesota Department of Health (MDH). Characteristics of these sentinel sites have been described elsewhere.[7]

Infection control practitioners from each hospital completed a case report form for patients with a positive MRSA culture obtained during 2000–2003. Patient medical records were reviewed to determine the type of infection, history of underlying illness (injection drug use, diabetes, malignancy, chronic heart or lung conditions, chronic skin conditions), or immunosuppressive therapy (defined as long-term systemic steroid use, excluding topical creams, steroids used only for short-course treatment, and inhaled steroids used for asthma) and any history of patient healthcare exposures as defined in the CA-MRSA case definition. The hospital laboratories submitted CA-MRSA isolates to MDH.

All patients with cultures obtained during 2000–2002 who met the CA-MRSA case definition based on medical record review were interviewed to confirm their classification (patient culture dates 2000–2002) and to assess possible CA-MRSA risk factors (patient culture dates 2001–2002). Patients identified at 4 of the 12 sentinel sites during 2003 who had no exclusionary healthcare exposures noted on medical record review were contacted to confirm CA-MRSA classification and conduct risk factor interviews. In addition, a random sample of 2003 patients from the remaining 8 sentinel sites were interviewed to confirm CA-MRSA classification. Informed consent was obtained from all patients before telephone interview.

US Census data from 2000 were used to provide median income by zip code[25] as a proxy for case household income. The University of Minnesota and MDH Institutional Review Boards reviewed and approved the study.

The Centers for Disease Control and Prevention (CDC) Active Bacterial Core Surveillance Program defined a CA-MRSA case as a patient with an MRSA infection and no history of the following: surgery, hospitalization, or residence in a long-term care facility within the year before infection, presence of a percutaneous device or indwelling catheter, dialysis within the previous year, hospitalization >48 h before MRSA culture, or previous MRSA infection or colonization.

Patients were classified as confirmed CA-MRSA case-patients if the medical record review and interview did not show any of the above healthcare risk factors. Patients were classified as probable CA-MRSA case-patients if the medical record review did not show any healthcare risk factors, but the interview was not completed (because of patient refusal, inability to locate, or language barriers).

CA-MRSA patients identified from prospective sentinel surveillance with culture dates in 2000 and 2003 were included in this subanalysis if they had an SSTI (e.g., abscess, cellulitis, folliculitis, wound infection [nonsurgical]) or infection in a normally sterile site caused by CA-MRSA. CDC's Active Bacterial Core Surveillance Program definition of sterile site infections was used to define cases of invasive CA-MRSA disease. This definition defines a normally sterile site as a portion of the body in a healthy state in which no microorganisms are found and includes the following: blood, cerebrospinal fluid, pleural fluid, peritoneal fluid, pericardial fluid, bone, joint fluid, internal body site (lymph node, brain, heart, liver, spleen, vitreous fluid, kidney, pancreas, or ovary), or other normally sterile site. Although cases of necrotizing pneumonia caused by CA-MRSA have been reported,[26] CA-MRSA specimens isolated only from sputum were not included in our subanalysis because sputum was not defined as a sterile site.

All MRSA isolates submitted to MDH were tested to confirm Staphylococcus aureus identification by using a tube coagulase test[27] (Difco Laboratories, Detroit, MI, USA). Testing for antimicrobial susceptibility was performed by using a broth microdilution panel (PML Microbiologicals, Wilsonville, OR, USA) containing the following 11 antimicrobial agents: ciprofloxacin, gentamicin, trimethoprim/sulfamethoxazole, clindamycin, tetracycline, erythromycin, rifampin, linezolid, mupirocin, vancomycin, and oxacillin. Clinical and Laboratory Standards Institute (CLSI, formerly National Committee for Clinical Laboratory Standards) breakpoints were used to determine levels of resistance for all antimicrobial agents except mupirocin, for which no CLSI breakpoints exist.[28] A standard of <4 µg/mL was used as a breakpoint for susceptibility to mupirocin.[29]

Molecular subtyping of MRSA isolates was performed by PFGE and digestion with the restriction endonuclease SmaI.[30] Patterns were evaluated both visually and with BioNumerics software (Applied Maths, Kortrijk, Belgium) by using the dice coefficient. Indistinguishable patterns must visually appear identical, and the DNA patterns must differ by <1.5% with respect to molecular weight. MRSA isolates were considered part of a CA-MRSA pulsed-field type (PFT) if they were ≥80% similar to the USA300 or USA400 reference strains based on Dice coefficients. MRSA isolates were considered part of an HA-MRSA PFT if they were ≥80% similar to USA100, USA200, or USA500–800 reference strains.[30]

The Yates continuity corrected chi-square test was used to test for trends with EpiInfo version 6.2 (CDC, Atlanta, GA, USA), and univariate analysis of the data was performed with EpiInfo 2000 (CDC). Multivariate logistic regression was used to evaluate the association of the type of MRSA infection (SSTI versus invasive disease) with microbiologic and molecular features of the MRSA isolates. Demographic characteristics associated with the type of infection in the univariate analysis were controlled for in the multivariate analysis model. An α≤0.05 significance level was required for predictors to remain in the model. Multivariate analysis was accomplished by using SAS version 8.0 for Windows (SAS Institute, Cary, NC, USA).