Excess ICU mortality attributable to ventilator-associated pneumonia:: The role of early vs late onset

Objective: To determine the impact of ventilator-associated pneumonia (VAP) on ICU mortality, and whether it is related to time of onset of pneumonia. Design: Prospective cohort study. Setting: 16-bed medical-surgical ICU at a university-affiliated hospital. Patients and measurements: From 2002 to 2003, we recorded patients receiving mechanical ventilation for > 72 h. Patients developing an infection other than VAP were excluded. Patients definitively diagnosed with VAP (n = 40) were cases and patients free of any infection acquired during ICU stay (n = 61) were controls. The VAP-attributed mortality was defined as the difference between observed mortality and predicted mortality (SAPS II) on admission. Results: Mechanical ventilation was longer in VAP patients (25 +/-20 vs 11 +/- 9 days; p < 0.001), as was ICU stay (33 +/-23 vs 14 +/-12 days; p < 0.001). In the non-VAP group, no difference was found between observed and predicted mortality (27.9 vs 27.4%; p > 0.2). In the VAP group, observed mortality was 45% and predicted mortality 26.5% (p < 0.001), with attributable mortality 18.5%, and relative risk (RR) 1.7 (95% CI 1.12-23.17). No difference was observed between observed and predicted mortality in early-onset VAP (27.3 vs 25.8%; p > 0.20); in late-onset VAP, observed mortality was higher (51.7 vs 26.7%; p < 0.01) with attributable mortality of 25% and an RR 1.9 (95% CI 1.26-2.63). Empiric antibiotic treatment was appropriate in 77.5% of episodes. No differences in mortality were related to treatment appropriateness. Conclusions: In mechanically ventilated patients, VAP is associated with excess mortality, mostly restricted to late-onset VAP and despite appropriate antibiotic treatment.