We evaluated the efficacy of linezolid-based salvage therapy compared with glycopeptide-based therapy in patients with persistent (≥7 days) methicillin-resistant Staphylococcus aureus bacteremia (MRSA-B).

All patients with MRSA-B during 2-year period at a tertiary-care hospital were prospectively enrolled. Linezolid-based salvage therapy was classified if patients switched glycopeptides to linezolid with/without carbapenem due to persistent MRSA-B. Covariate adjustment using the propensity score and inverse probability of treatment weighting (IPTW) using the propensity score were performed to control for bias in treatment assignment.

Of 377 patients with MRSA-B, 90 with persistent MRSA-B were included. Of these, 38 (42%) were classified as linezolid-based salvage group and the remaining 52 (58%) as glycopeptide-based therapy group. The duration of persistent bacteremia (median 16 days vs. 10 days; P = 0.008) was longer in linezolid-based salvage group than in the comparator. However, the 30-day mortality (11% vs. 25%; P = 0.08) had a trend toward being lower in linezolid-based salvage group than those in the comparator. Logistic regression models with covariate adjustment and IPTW using propensity scores also revealed that linezolid-based salvage showed a trend toward having better outcome than the comparator, although this did not reach any statistically significance (OR 0.31; 95% CI 0.03–2.95 and OR 0.19; 95% CI 0.01–3.39, respectively).

While having worse prognostic factors compared with glycopeptide-based therapy, linezolid-based salvage therapy revealed a trend toward better outcomes than the comparator. Our data suggest that linezolid-based salvage therapy would be considered in patients with persistent MRSA-B despite the use of glycopeptides therapy.