Selective decontamination of subglottic area in mechanically ventilated patients with multiple trauma.

OBJECTIVE: To determine whether selective decontamination locally in the subglottic area (SDSA) reduces tracheal colonization and prevents ventilator-associated pneumonia (VAP) in patients with multiple trauma. DESIGN AND SETTING: A prospective randomized, controlled, clinical study in a 14-bed general intensive care unit of a university hospital. PATIENTS: 79 consecutive multiple trauma patients admitted to the ICU who were expected to be mechanically ventilated for more than 5 days; 61 patients completed the protocol. INTERVENTION: Patients were randomly assigned to receive SDSA using a continuous infusion of a suspension containing three nonabsorbable antibiotics (polymyxin, tombramycin, and amphotericin B; n=30) or placebo ( n=31). MEASUREMENTS: The incidence of bronchial and gastric colonization and the number of cases of VAP were recorded. Gastric fluid and tracheal secretion cultures were obtained soon after intubation and thereafter every 4 days. Etiological diagnosis of VAP was based on samples taken by a specific protected double catheter set. RESULTS: VAP developed in 5 of 30 (16.6%) patients receiving SDSA and 16 of 31 (51.6%) patients receiving placebo. Negative bronchial secretion cultures were found in 14 of 30 (46.6%) patients in the SDSA group and in only 3 of 31 (9.6%) patients in the control group. No patient with negative bronchial secretion culture developed VAP. No significant differences in outcome were found. CONCLUSIONS: The SDSA is an effective and safe type of chemoprophylaxis against tracheal colonization and can significantly reduce the incidence of VAP in mechanically ventilated patients with multiple trauma.