Bacteriology of selective decontamination: efficacy and rebound colonization.

In earlier studies concerning the use of selective decontamination (SD) we recognized some patients to have an incomplete or failed decontamination which had negative consequences for their outcome. We also experienced patients to have a rebound colonization with potentially pathogenic microorganisms (PPM) after the withdrawal of SD medication, thereby endangering the hospital environment as sources of possibly resistant microorganisms. In a prospective observational cohort-study, we have studied 135 patients on a surgical ICU in a University hospital. On admission 51% of all patients harboured PPM, of which 64% and 79% were eliminated form the oropharynx and gut, respectively. A total of 49 episodes of SD (32%) had to be defined as ineffective, partly due to late or inadequate administration of medication. Rebound colonization with nosocomial aerobic PPM was seen in 20 of 90 successfully decontaminated episodes, but there were no infections. Eighty-four of 123 surveillance cultures (68%) revealed a complete recolonization with flora in a composition not different from the pre-admission flora. Withdrawal of SD medication 5 days after extubation proved satisfactory because no infections ensued. Surveillance cultures are obligatory to determine whether SD is effective as one third of patients receiving SD experienced bacterial colonization with PPM and therefore had no benefit from this prophylaxis. In addition, these results indicate that an intact and functional colonization resistance develops in patients after cessation of effective SD medication, but only when physiological defence mechanisms are presumably restored. In the seven years of experience with SD, increasing antibiotic resistance has not been in the study hospital.