Selective decontamination regimens in French ICUs: association with reduced infection and resistance emergence.

Background: Despite randomized controlled trials with favorable results, few intensive care units (ICUs) implemented selective decontamination (SD) for ICU-acquired infection prevention. We aimed to evaluate, SD implementation and associated-effects in a large network of French ICUs.

Methods: This study was conducted using the healthcare-associated infection surveillance cohort "REA-REZO" involving 193 participating ICUs. All patients receiving invasive mechanical ventilation for > 24 h were included. In addition to standard of care (SOC), six ICUs applied a SD strategy during the study period. The primary endpoint was the rate of ICU-acquired infection. A propensity-score matched analysis was conducted using non-parsimonious regression model. The secondary endpoint was the rate of colonization by multidrug resistant organisms (MDRO) during the ICU stay.

Results: Among 81,661 patients with invasive mechanical ventilation for longer than 24 h, 2727 patients receiving SD were matched with 2 727 receiving SOC. The ICU-acquired infection incidence was lower in the SD group as compared with the SOC group (Incidence Rate Ratio = 0.66 [0.60-0.73]; p < 0.001) although the ICU mortality was similar (31.9% vs 32.5%, respectively p = 0.689). Acquisition of MDRO was lower in the SD group than in the SOC group (40 (1.5%) patients vs. 139 (5.1%) patients p < 0.001).

Conclusions: These results showed that a strategy of SD was associated with reduced ICU-acquired infection incidence and decreased emergence of MDRO, while the mortality was not affected.