UTILITY OF SELECTIVE DIGESTIVE DECONTAMINATION IN MECHANICALLY VENTILATED PATIENTS

Objective: To assess selective digestive decontamination for preventin g nosocomial pneumonia and mortality in mechanically ventilated patien ts. Design: Prospective, randomized, placebo-controlled, double-blind study. Setting: Respiratory intensive care unit of a 1000-bed teaching hospital. Patients: 80 patients receiving mechanical ventilation for more than 72 hours. Interventions: Patients received selective digesti ve decontamination using polymyxin E, tobramycin, and amphotericin B t hrough a nasogastric tube and also topically in the oropharynx; contro l patients received placebo. All patients received intravenous cefotax ime for 4 days or other systemic antibiotics if required. Measurements : Bacteriologic surveillance (three times a week) was done by quantita tively culturing tracheal aspirates, pharyngeal swabs, and gastric jui ce. The diagnosis of pneumonia was based on quantitative cultures of p rotected specimen brush samples (greater than or equal to 10(3) CFU/mL [colony forming units/mL]) or bronchoalveolar lavage fluid (greater t han or equal to 10(4) CFU/mL) and autopsy findings. Results: Bronchial , oropharyngeal, and gastric colonization by gram-negative bacilli and Candida species was lower in the selective digestive decontamination group compared with the placebo group. Nonsignificant differences were found in the incidence of nosocomial infections (28% compared with 37 %; odds ratio, 0.66; 95% Cl, 0.35 to 1.25) and nosocomial pneumonia (1 8% compared with 24%; odds ratio, 0.7; Cl, 0.33 to 1.46) and in the cr ude mortality rate (31% compared with 27%; odds ratio, 1.21; Cl, 0.63 to 2.34) when comparing digestive decontamination with placebo, respec tively. Conclusions: Selective digestive decontamination in our mechan ically ventilated patients significantly decreased the colonization ra te of gram-negative bacilli and of Candida species but not of Staphylo coccus aureus. It did not decrease the incidence of nosocomial pneumon ia, mortality, length of stay, or the duration of mechanical ventilati on.