Eight-Year Surveillance, Genotyping and Strain-Relatedness of Nosocomial Methicillin-Resistant Staphylococcus aureus (MRSA) at Intermountain Healthcare Hospitals in Salt Lake City, Utah

Background: Outbreaks of nosocomial MRSA infections caused by SCCmec type IV-carrying community-associated MRSA strains (CA-MRSA) are being reported more frequently.  These strains are of concern in hospital settings due to their propensity to cause rapidly invasive infections.  We investigated if CA-MRSA strains have caused nosocomial MRSA infections at four Intermountain Healthcare hospitals.

Objective:

To determine the contribution of community acquired MRSA strains to the burden of healthcare-associated MRSA infections in an urban hospital system and to detect changes in the types of infections associated with community strains.

Methods: Infection preventionists at our facilities use a locally developed database program to detect and record nosocomial infections using standard CDC definitions. Since 1993 we have stored MRSA isolates from the Clinical Microbiology Laboratory. We genotyped nosocomial MRSA isolates for staphylococcal chromosomal cassette mec (SCCmec) using the Applied Biosystems 7300 Real Time PCR.  We used DiversiLab rep-PCR analysis to determine strain relatedness to a reference library of pulse field gel electrophoresis (PFGE) determined lineages.  DiversiLab strain typing was completed on all SCCmec type IV isolates and a representative number of SCCmec type II isolates.  We used the S-statistic for trend analysis.

Results: We identified 621 patients with nosocomial MRSA infection between January 1, 2001 and December 31, 2008.  We recovered and performed genotyping on one representative MRSA isolate from each of 341 (55%) of the 621 patients.  Of the isolates that were available for genotyping and strain relatedness studies, SCCmec type IV MRSA strains represented 24 (7%) of the 341 isolates. Overall these strains were highly similar (similarity index > 95%) to USA 300/USA 500 PFGE strains; however for the latter 2 years of the study 56% of SCCmec type IV isolates were not similar to any MRSA reference strain in the DiversiLab library.   SCCmec II MRSA strains were all highly similar to the USA 100 PFGE type.  For the period 2001-06 SCCmec IV strains caused 2.2% of infections, while from 2007-08 these strains caused 28% of nosocomial MRSA infections, a significant increase over time. The types of infection caused by the two strains were similar, with the exception of increased skin and soft tissue infections for the CA-MRSA strains (18% vs. 4%).

Conclusions: Nosocomial MRSA infections due to strains carrying SCCmec IV occurred uncommonly but increased significantly over the study period in our four urban hospitals.  In recent years the CA-MRSA strains causing nosocomial infections in our institutions have been increasing in diversity and are frequently unrelated to common reference strains.