Prevention of nosocomial infection in a pediatric intensive care unit (PICU) through the use of selective digestive decontamination

Objective: To assess the effectiveness of selective digestive decontaminati on (SDD) on the control of nosocomial infection (NI) in critically ill pedi atric patients. Design: A prospective, randomized, non-blinded and controll ed clinical microbiology study. Setting: The pediatric intensive care unit (PICU) of a tertiary level pediatric university hospital. Criteria for incl usion: Patients 1 month to 14 years old, who underwent some kind of manipul ation or instrumentation (mechanical ventilation, vascular cannulation, mon itoring of intracranial pressure, thoracic or abdominal drainage, bladder c atheterization, peritoneal dialysis, etc.) and/or presented a neurological coma requiring a stay in the PICU of 3 or more days. Patients: Over a perio d of 2 years, 244 patients met the inclusion criteria; 18 patients were wit hdrawn because of protocol violation. The treatment group comprised 116 pat ients and the control group, 110 patients. Intervention: The treatment grou p received a triple therapy of colimycin, tobramycin and nystatin administe red orally or via nasogastric tube every 6 hours. All patients with mechani cal ventilation or immune-depression received decontamination treatment of the oropharyngeal cavity with hexitidine (Oraldine(R) 0.5 mg/ml) every 6-8 hours in accordance with the PICU's conventional protocol. Method: Up to 10 types of nosocomial infection were diagnosed following criteria of the Cen ters for Disease Control (CDC). The severity and manipulation of the patien ts on admission was assessed using the therapeutic intervention scoring sys tem (TISS) and multi-organ system failure scores (MOSF). Measurements and m ain results. Univariant analysis: SDD did not significantly reduce the inci dence of NI, antibiotic use, the length of stay, or :mortality; although a small percentage of respiratory and urinary tract infections was detected, catheter-related bacteremia was the most common infection. Multivariant ana lysis: Controlling the risk factors for each child through log regression s howed that SDD acted as a protective factor for more than 90% of the sample with respect to the appearance of respiratory and urinary tract infections , reducing the risk of such infections to 1/5 and 1/3, respectively. Conclu sions: SDD was effective in controlling respiratory and urinary tract infec tions in children admitted to the PICU, but it did not reduce the incidence of other types of nosocomial infection.