Antimicrobial prophylaxis in acute leukaemia: prospective randomized study comparing two methods of selective decontamination.

In a prospective study the efficacy of two regimens for selective decontamination of the digestive tract was studied in patients with acute leukaemia during remission induction therapy. Seventy-eight patients were randomized to receive either a combination of cotrimoxazole, polymyxin B and nystatin (group A) or a combination of nalidixic acid, polymyxin B, neomycin and nystatin. With both regimens the gastrointestinal tract could be decontaminated equally effectively from potential pathogens. In the oropharyngeal region the decontamination from Enterobacteriaceae was significantly better in group A (P less than 0.01). In both groups less than 10% of the acquired infections were caused by gram-negative bacilli and no gram-negative septicaemia occurred in either group. The median time interval until the first acquired infection was 17 days in group A and 36 days in group B, respectively (P less than 0.05). It is concluded that regimen A might be more effective than regimen B though both regimens prevent reliably severe gram-negative infections.