Geographic diversity of Methicillin Resistant Staphylococcus aureus (MRSA) infections in the community setting

Background: MRSA is recognized as a community infection and epidemiological studies suggest geographic variations occurs within regions.  We hypothesize that skin and soft tissue infections (SSTI) are primarily transmitted in the community and will cluster geographically, where as healthcare-associated infections (HAIs), such as blood stream infections, are transmitted in the healthcare setting and will be more randomly distributed upon community geographic analysis.
Objective:  To determine geographic distribution of  MRSA infections from a large academic tertiary medical center (MC) and 7 referring smaller community hospitals (CH) to help understand MRSA distribution and transmission patterns in and between communities.     
Methods: We compared the geographic, demographic, and clinical information for 937 patients with MRSA infections from January 2007 to October 2009:  189 retrospective archived MRSA isolates from MC; 306 consecutive prospective isolates from CH; and 443 randomized prospective isolates from MC.   ArcGIS 9.3 was utilized to geocode home addresses of patients and United States Census Bureau 2000 census data was used for comparison between zip code (ZC) areas.  

Results: The isolates were distributed into 241 different zip codes, the number of isolates per ZC varied from 0 to 37.  The distances between CH and MC ranged from 27 to 121 miles.  Race, community versus HAIs, types of infection, and antibiotic resistance varied geographically (antibiotic susceptibility not shown).  SSTI clustered in discrete geographic areas significantly more often than other types of infections.  Distribution of MRSA isolates by specimen type, number and percent of isolates, number of ZC areas, range of the number of isolates in each ZC area, and the average number of patients per zip code are listed below.  Six zip code areas had the highest number of isolates (range 23 to 47).  Areas surrounding the CH had higher incidence of SSTI whereas MC had higher prevalence of blood stream and other HAIs.  Comparison of US census and zip code data showed most high prevalence areas to be lower socio-economic as compared to other areas.  

Specimen type	SSTI	OTHER	BLOOD	URINE	SPUTUM
# of isolates (%)	271 (29%)	286 (30%)	188 (20%)	50 (0.05%)	142 (15%)
# of zip codes	92	116	102	37	80
Range	0 - 37	0 - 10	0 - 8	0 - 5	0 - 12
# of patients/ zip code	2.9	2.46	1.8	1.4	1.8

                                         

   

                                     

                                       

                                                                          

                                                                                                                                         

                                        

                               

Conclusions: Diverse geographic distribution exists amongst MRSA isolates from MC and CH.   CH tended to have clusters of SSTI cases strongly suggesting a predominant community origin whereas MC had more HAIs consistent with non-community source of cases.  Utilizing geographic analysis can broaden our understanding of MRSA origins, mode of transmission, and predicted antimicrobial susceptibilities to guide appropriate management and control.  Combining geographic data with molecular typing will likely further enhance our understanding of MRSA transmission and acquisition.