Systemic inflammatory response and increased risk for ventilator-associated pneumonia: A preliminary study

Objective: Inflammatory markers have been assessed for the diagnosis and follow-up of ventilator-associated pneumonia (VAP), but their potential role in predicting the risk for VAP is unknown. We prospectively assessed the evolution of cytokines in mechanically ventilated patients and their predictive and diagnostic role for VAP. Design: Prospective observational study. Setting: Medical intensive care unit. Patients: Mechanically ventilated patients. Exclusion criteria were active infection at admission and subsequent extrapulmonary infection. Interventions: None. Measurements and Main Results: Sequential measurements of interleukin (IL)-1, IL-6, IL-8, IL-10, and tumor necrosis factor-alpha were done in 44 ventilated patients. VAP was suspected in 20 cases and microbiologically confirmed in nine. At admission, demographics, severity scores, and clinical and standard laboratory values did not discriminate patients with and without VAP, but the median (interquartile range) serum levels of IL-6 were higher in patients who subsequently developed VAP, compared with those without VAP (235 [141-803] vs. 113 [60-170] pg/mL, p = 0.015). The sensitivity and specificity of IL-6 to predict VAP was 71% and 78%, respectively, using 198 pg/mL as optimal cutoff, with relative risk (95% confidence interval) 8.9 (1.4-56.3). When VAP was suspected, serum levels of IL-6 were higher in patients with confirmed compared with nonconfirmed VAP (1131 [496-1987] vs. 236 [115-357] pg/mL, p = 0.016). The sensitivity and specificity to discriminate between confirmed and nonconfirmed VAP was 71% and 89%, respectively, using 620 pg/mL as optimal cutoff, with relative risk (95% confidence interval) 15.0 (1.2-185.2). Conclusions: IL-6 at admission is an early and accurate indicator of patients at increased risk for VAP. IL-6 is also accurate in discriminating patients with VAP from other causes of pulmonary infiltrates. Extrapolation of these results to the overall population of critically ill patients is limited by the small number of patients. (Crit Care Med 2009; 37:1691-1695)