Objective: To evaluate the secular trends in MRSA BSIs after the introduction of a nosocomial MRSA control intervention.
Design: Before-after study.
Setting: An 850-bed community hospital with an ICU and vascular surgery, neurosurgery, bone marrow transplantation, and AIDS units. MRSA is endemic at this hospital; the prevalence of methicillin resistance among patients with S. aureus infection is greater than 50%.
Patients: Among all inpatients, MRSA BSI was identified, its origin defined, and incidence rates calculated by ward and origin.
Intervention: A MRSA control program was implemented based on active surveillance cultures to identify MRSA-colonized patients, followed by isolation using contact precautions. Incidence rates of MRSA BSI during the intervention (i.e., July 1, 1997, to December 31, 2001) and preintervention (i.e., January 1, 1996, to June 30, 1997) periods were compared.
Results: Sixty-nine MRSA BSIs were identified. When compared with the preintervention period, the incidence rate of MRSA BSI was reduced from 0.64 to 0.30 per 1000 admissions (RR, 0.46; CI95, 0.25-0.87; P = .02) during the intervention period. The impact was greater in the ICU, with an 89% reduction (RR, 0.11; CI95, 0.01-0.98; P = .03), and for CVC-associated MRSA BSIs, with an 82% decrease (RR, 0.17; CI95, 0.05-0.55; P = .002). Methicillin resistance among S. aureus blood isolates decreased from 46% to 17% (RR, 0.36; CI95, 0.22-0.62; P = .0002).
Conclusion: A reduction in MRSA bacteremia is achievable through use of the MRSA "search and isolate" intervention even in a hospital with high rates of endemic MRSA.