Cluster-Randomized Controlled Trial of Enhanced Carbapenem-Resistant Enterobacterales Prevention Program in General Medicine Wards, Siriraj Hospital

Abstract

Background: Carbapenem-resistant Enterobacterales (CRE) colonization is a major risk factor for infection. Most infection prevention and control (IPC) strategies rely on private-room isolation, but evidence of their effectiveness in resource-limited settings is scarce. Methods: From February to October 2021, we conducted a cluster-randomized controlled trial in 6 general medical wards at Siriraj Hospital, enrolling adults with ≥1 CRE risk factor. Wards were randomized to standard infection control care (sIC) or an enhanced CRE prevention program (eIC) comprising sIC plus monthly staff education, real-time notifications of CRE acquisition, and contact-precaution reminders. Active stool/rectal CRE surveillance was performed at enrollment and weekly. Primary outcomes were the CRE acquisition incidence and CRE acquisition–free time. Results: A total of 363 patients were included: 174 in the intervention group (1684 patient-days) and 189 in the control group (1517 patient-days). The cumulative incidence of CRE acquisition was slightly lower in the intervention group (36.8% vs 46.6%; P = .06), with a significantly lower incidence rate per patient-day (0.038 vs 0.058; P = .007). In a post hoc analysis excluding acquisitions within 24 hours, the cumulative incidence was similar between groups (25.7% vs 33.6%; P = .16). The probability of remaining CRE-free showed an unadjusted hazard ratio (HR) of 0.72 [95% CI, 0.52–1.00; P = .05]. After adjusting for prior antibiotic use, the adjusted HR was 0.75 [95% CI, 0.54–1.05; P = .09]. There were no differences in all-cause mortality or length of hospital stay. Conclusions: Carbapenem-resistant Enterobacterales acquisition incidence was high in this setting. The enhanced CRE prevention program tended to reduce CRE acquisition and prolong CRE-free survival. Larger studies are needed to explore benefits on morbidity and mortality.