[Pneumonia prevention in long-term mechanically ventilated patients: selective skin decontamination according to Stoutenbeek or prevention of colonization according to Unertl? A prospective randomized comparison of both treatments].

In a prospective randomised study, the effects of two different colonisation prophylaxis techniques on colonisation and pulmonary infection were investigated in 40 critically ill patients with long-term ventilatory support (greater than or equal to 4 days). 20 patients were selectively decontaminated with 4 x 100 g polymyxin E, 4 x 80 mg tobramycin and 4 x 500 mg amphotericin B, administered through the gastric tube and with an antimicrobial sticky paste in the oropharynx (group I). 20 patients received 50 mg of polymyxin B and 80 mg of gentamicin dissolved in 10 ml of 0.9% saline at 6 h intervals into nose, oropharynx and stomach as well as 300 mg of amphotericin B in the oropharynx only (group II). All patients received cefotaxime systemically in the first 3 days. In group I gram-negative aerobic bacteria in the pharynx decreased from 35% to 0%, in group II from 40% to 10% and in the rectum from 80% to 61% (10% in the second week) in Group I and from 100% to 73% (33% in the second week) in group II. The decrease in gram-negative microorganisms was accompanied by an increase in the frequency of Staphylococcus epidermidis. In group I, two patients developed pneumonia and two patients urinary tract infections, in group II two patients suffered from pneumonia and 3 patients urinary tract infections. Both regimes are effective methods of prophylaxis for lowering colonisation with gram-negative aerobic bacteria and the frequency of pneumonia in patients requiring long-term mechanical ventilation. A possible selection of gram-positive bacteria must be appropriately monitored.

RCT Entities:   Population Interventions Outcomes