ObjectiveTo determine whether selective digestive decontamination (SDD) had some negative impact on the bacterial resistance observed in strains isolated from samples from patients receiving nonabsorbable antibiotics and cefazolin.DesignCase-control study.SettingIntensive care unit of a university tertiary-care hospital.PatientsOver a 6-yr period, 360 multiple trauma patients (case patients) submitted to SDD were compared with 360 patients not receiving SDD (controls).InterventionsSDD consisted of polymyxin E, gentamicin, and amphotericin B and was applied on the buccal mucosa and provided in the nares and the stomach. For the first 3 days, systemic cefazolin (1 g three times a day) was provided. Resistance analysis was performed in case patients and controls on samples collected at predetermined intervals.Measurements and Main ResultsSDD was used in a small subset of patients admitted to the intensive care unit (360 of 5987 over the 6-yr study period). A relative overgrowth of gram-positive cocci was observed. Methicillin resistance ofStaphylococcus epidermidiswas increased (SDD 76%, controls 63%,p< .05) but not that ofStaphylococcus aureus(SDD 20%, controls 18%). Resistance of Enterobacteriaceae,Pseudomonas aeruginosa, andAcinetobacterto &bgr;-lactamines and aminoglycosides was the same in SDD patients and controls.ConclusionsWhen used in a small subset of patients who have been shown to derive benefit from it (patients who have experienced multiple trauma), SDD has a moderate impact on microbial ecology. However, surveillance cultures are indispensable because the absence of resistance to SDD antibiotics determines the long-term safety of the SDD prophylaxis.