Objective: The use of topical polymyxin and tobramycin to prevent intensive care infections is controversial. Moreover, these antibiotics are ineffective against methicillin-resistant Staphylococcus aureus. A decontamination regimen using mupirocin and chlorhexidine could prevent acquired infections, including those involving S. aureus. Because these two regimens could have a complementary role, we evaluated their effects when given both alone and combined.
Design: The authors conducted a multiple-center, placebo-controlled, randomized, double-blind study performed according to a 2 x 2 factorial design.
Setting: The study was conducted at three polyvalent medical intensive care units at university-affiliated hospitals in France.
Patients: Adult patients (age, > or =18 yrs) intubated for <48 hrs who were likely to be ventilated for >48 hrs.
Intervention: Two regimens were used: topical administration of polymyxin/tobramycin (or placebo) and nasal mupirocin with chlorhexidine body washing (or nasal placebo with liquid soap). The patients (n = 515) received polymyxin/tobramycin alone (n = 130), mupirocin/chlorhexidine alone (n = 130), both regimens (n = 129), or all placebos (n = 126) for the period of mechanical ventilation plus 24 hrs.
Measurements and main results: The incidence of total infections acquired from the date of randomization until the termination date of study treatments plus 48 hrs was assessed. There were fewer acquired infections with both regimens than with polymyxin/tobramycin alone (odds ratio, 0.44; 95% confidence interval, 0.26-0.75; p = .003), mupirocin/chlorhexidine alone (0.43; 0.25-0.73; p = .002), or all placebos (0.42; 0.25-0.72; p = .001). There were no differences between polymyxin/tobramycin alone (0.95; 0.59-1.54; p = .84) and mupirocin/chlorhexidine alone (0.98; 0.60-1.58; p = .92) vs. all placebos. The probability of freedom from infection was higher with both regimens than with polymyxin/tobramycin alone (p = .002), mupirocin/chlorhexidine alone (p < .001), or all placebos (p < .001). Infection rates were also significantly lower with both regimens than with polymyxin/tobramycin alone (p = .017), mupirocin/chlorhexidine alone (p < .001), or all placebos (p < .001).
Conclusion: Acquired infections were substantially reduced by mupirocin/chlorhexidine plus polymyxin/tobramycin, whereas each regimen given alone was ineffective. Whether both regimens could increase Candida infections deserves further investigation.