A comparison of long-term outcomes after meticillin-resistant and meticillin-sensitive Staphylococcus aureus bacteraemia: an observational cohort study.

BACKGROUND: Findings from previous studies have suggested that outcomes after meticillin-resistant Staphylococcus aureus (MRSA) bacteraemia are worse than after meticillin-sensitive S. aureus (MSSA) bacteraemia. We assessed whether patients who had MRSA bacteraemia had a higher risk of death and recurrent infections than those who had MSSA bacteraemia.

METHODS: For this observational cohort study, we assessed data from the microbiology laboratory database at the Royal Perth Hospital (WA, Australia). Data were for all patients who had an episode of MRSA bacteraemia between July 1, 1997, and June 30, 2007, and, by use of a computer-generated randomisation sequence, a randomly selected subgroup of patients who had an episode of MSSA bacteraemia (patients with one or more set of blood cultures positive for S. aureus). The primary outcomes were survival time and subsequent infection-related hospital readmissions, analysed by Cox proportional hazards regression with adjustment for important prognostic factors.

FINDINGS: Of the 583 patients who had an episode of MRSA or MSSA bacteraemia, we used data for the 582 who had complete data linkage: 185 patients who had MRSA bacteraemia and 397 patients who had MSSA bacteraemia. The crude survival time of patients after MRSA bacteraemia was shorter than it was for patients with MSSA bacteraemia (14 months [IQR 1-86] vs 54 months [3-105]; hazard ratio 1·46, 95% CI 1·18-1·79; p=0·01). The adverse association between MRSA and all-cause mortality (0·98, 0·77-1·30; p=0·87) or infection-related mortality (1·22, 0·89-1·69; p=0·22) were not statistically significant after adjustment for important prognostic factors including age, comorbidities, severity of acute illness, metastatic infections, and long-term care facility resident status. After adjustment for these confounding factors, we saw no difference in infection-related hospital readmissions between patients who had MRSA bacteraemia and those who had MSSA bacteraemia (odds ratio 0·95, 95% CI 0·59-1·53; p=0·83).

INTERPRETATION: Long-term outcomes after MRSA bacteraemia were worse than those after MSSA bacteraemia through its confounding associations with other prognostic factors. Our findings might have implications for management strategies to control MRSA colonisation.

FUNDING: The Medical Research Foundation of Royal Perth Hospital.