Management and epidemiologic analyses of an outbreak due to methicillin-resistant Staphylococcus aureus

Following implementation of special measures to control a nosocomial outbreak of methicillin-resistant Staphylococcus aureus (MRSA), we used immunoblot typing in conjunction with antimicrobial susceptibility testing to investigate the epidemiology of this event and to determine whether this outbreak represented the failure of infection control measures to limit the spread of previously endemic MRSA strains or the introduction of a new strain of MRSA.

Isolates of MRSA recovered from hospitalized patients were initially categorized on the basis of antimicrobial susceptibility results. Organisms susceptible to ciprofloxacin and/or trimethoprim/sulfamethoxazole were recovered from patients at a relatively constant rate prior to December 1988 and were categorized as endemic isolates. Subsequently, there was an outbreak due to organisms resistant to both of these antibiotics; these were therefore categorized as outbreak isolates. Isolates were later characterized by immunoblot typing. Prior to this analysis, isolates were given code numbers so that clinical and epidemiologic data as well as resistance patterns were not known until this testing was complete.

Between January 1986 and November 1988, an average of 3.9 patients per month acquired nosocomial MRSA in the Sepulveda Veterans Administration Medical Center. In contrast, from December 1988 to October 1989, 369 MRSA isolates were collected from 125 patients (an average of 11.4 patients per month). Prior to December 1988, all tested nosocomial isolates of MRSA were susceptible to ciprofloxacin and/or to trimethoprim/sulfamethoxazole. In contrast, the outbreak was due to spread of MRSA isolates resistant to these antibiotics. Immunoblot typing of 204 isolates from 98 individuals identified five distinct immunoblot types of which types B and C were by far the most common. Type B was highly associated with outbreak isolates, whereas type C was associated with endemic isolates (p less than 0.001). All sequential isolates from single patients that belonged to different susceptibility categories demonstrated discordant immunoblot types. In contrast, concordant immunoblot types were observed for 25 of 27 sequential isolates that displayed minor variations in antimicrobial resistance. The institution of more stringent infection control measures was followed by the return of nosocomial MRSA acquisition rates to pre-outbreak levels. Although novobiocin and trimethoprim/sulfamethoxazole were extensively used to treat patients harboring outbreak and endemic isolates, respectively, in no instance was the initial MRSA isolate from any patient resistant to novobiocin and only 6% of initial endemic isolates displayed trimethoprim/sulfamethoxazole resistance. A modest, significant increase in the resistance of endemic isolates to various other antimicrobial agents was noted however.

Immunoblot analyses provided strong, corroborative evidence that at least two separate strains of MRSA were present during the outbreak and that a newly introduced strain with a distinctive antimicrobial resistance pattern was primarily responsible for the rapid spread of MRSA during the outbreak. The observation that previously effective infection control measures failed to prevent the nosocomial spread of a newly introduced community-acquired MRSA strain suggests that a single set of control measures may not be equally efficacious against all strains of MRSA. In this regard, previously reported variations in resistance to topical antimicrobials and/or antiseptics, and differences in virulence factors such as colonization potential, invasiveness, and survival on fomites, may warrant further study. Control of the outbreak strain of MRSA in our institution did occur after the implementation of more strenuous isolation procedures.(ABSTRACT TRUNCATED)