Objective: To test the hypothesis that the outcome of patients with ventilator-associated pneumonia (VAP) depends on both their baseline severity at VAP onset and the adequacy of empirical antibiotic therapy.
Design and setting: Prospective clinical study in six intensive care units in Paris, France.
Patients: One hundred and forty-two patients with VAP after >/= 48 h of mechanical ventilation.
Measurements and results: Patients were compared according to whether adequate antibiotics were started when VAP was first suspected (D0). At day 0, the rate of adequate antibiotic therapy was 44.4% and rose to 92% at day 2. Outcomes were recorded at the ICU and hospital discharge. Overall, no significant mortality difference was found with and without adequate early antibiotics. When patients were also classified based on the initial Logistic Organ Dysfunction score (LOD), mortality was significantly higher with inadequate early antibiotic therapy in the groups with LOD </= 4 (ICU mortality: 37% vs 7%, P=0.006; hospital mortality: 44% vs 15%, P=0.01). A multivariate logistic regression confirmed that inadequate antibiotic therapy increased mortality in patients with LOD </= 4 after adjustment on other prognostic factors.
Conclusions: Inadequate empirical treatment seemed to be associated with a poor prognosis only in patients with LOD </= 4. These results need to be confirmed by further studies before any reappraisal of current guidelines for empirical antibiotic therapy of VAP can be envisaged.