Objective: To evaluate the efficacy of two regimens, selective digestive decontamination (SDD) and sucralfate, in the prophylaxis of nosocomial infection in critically ill patients requiring prolonged mechanical ventilation. Design: Prospective, randomised, multi-centre trial. Setting: Medical and surgical intensive care units in 10 hospitals in Catalonia, Spain. Subjects: A total of 129 patients requiring mechanical ventilation for at least four days participated in the study. The absence of infection and use of systemic or topical antibiotic therapy prior to the diagnosis of nosocomial pneumonia was also required. Interventions: Patients were randomly assigned to the control group (n=42) in which histamine H2-receptor antagonists were given; the SDD group (n=41) using topical polymyxin E, tobramycin, and amphotericin B in the pharynx and stomach; and the sucralfate group (n=46). Patients in both regimens received intravenous cefotaxime for four days. Endpoints: The follow-up period was limited to ICU stay and included from the patient's entry into the study until extubation with a cutoff limit of 40 days, or until exclusion from the study due to development of nosocomial infection and/or administration of antibiotics for any reason. Mortality attributable to nosocomial pneumonia or another infection was defined as death with progressive respiratory failure and/or septic shock during the course of infection. Measurements and main results: SDD patients showed a lower incidence of nosocomial infection (36.6%) than sucralfate-treated patients (54.3%) and controls (71.4%) (p=0.006). Nosocomial pneumonia was more frequent among controls (50%) than in SDD (17%) and sucralfate (30%) groups (p=0.005). Early-onset pneumonia (within the four days of ICU stay) was less common in either SDD-or sucralfate-treated patients (28.6% and 14.3%) than in controls (61.9%) (p=0.015). A statistically significant difference (p=0.037) in late-onset pneumonia was found between SDD and sucralfate using Kaplan-Meier estimates. The development of infection other than nosocomial pneumonia was also significantly different between the SDD group and controls and between SDD and sucralfate. Conclusions: Both SDD and sucralfate regimens reduced the occurrence of nosocomial infection in mechanically ventilated patients, particularly in the occurrence of nosocomial pneumonia. Selective decontamination was more effective than sucralfate in preventing 'late-onset' pneumonia and nosocomial infection other than pneumonia.
Read full abstract