RESUMEN
BACKGROUND: Linezolid is frequently used in critically ill patients with ventilator-associated pneumonia. Its potent activity against Gram-positive microorganisms and its high tissue penetration may favour Gram-negative colonization and infection. The aim of our study was to evaluate the risk for Gram-negative infections in critically ill patients treated with linezolid or vancomycin. METHODS: The cases of all patients admitted over an 18-month period to a hepatic intensive care unit for ≥ 1 week, and treated with linezolid or vancomycin, were retrospectively reviewed. The main clinical characteristics and infections due to Gram-negative bacteria in the month after starting linezolid or vancomycin were obtained. RESULTS: Seventy-one patients treated with linezolid and 68 treated with vancomycin fulfilled the inclusion criteria. Co-morbidities were similar in both groups. Patients on linezolid treatment had a longer stay in the ICU (mean ± standard deviation 41 ± 38 days vs 18.4 ± 13 days), received this treatment later (14.3 ± 15.1 days vs 6.3 ± 6.5 days), had a higher mean serum creatinine concentration (1.71 ± 1.18 mg/dl vs 1.04 ± 1.04 mg/dl), more often required haemodiafiltration (29.6% vs 13.2%), and 30 day-mortality was higher (42.3% vs 20.6%) than in patients receiving vancomycin. More than 95% in both groups received a broad-spectrum beta-lactam in addition to linezolid or vancomycin. The rate of Gram-negative infection during the following month was 28.2% in the linezolid group and 26.5% in the vancomycin group (p > 0.5). CONCLUSIONS: Linezolid was more frequently used in critically ill patients with longer ICU stay and renal failure. The rate of infection due to Gram-negative microorganisms was similar in patients who received linezolid or vancomycin.