Infection prevention strategies for carbapenem-resistant and ESBL-producing gram-negative bacteria in intensive care units: A systematic review with exploratory component network meta-analysis

Background Healthcare-associated infections caused by multidrug-resistant Gram-negative bacteria remain a major challenge in intensive care units (ICUs). Infection prevention and control strategies are often implemented as multimodal bundles beyond standard precautions, but the effectiveness of individual components remains uncertain. Methods We conducted a systematic review of interventional studies evaluating infection prevention strategies in adult ICUs, complemented by an exploratory component network meta-analysis to examine the comparative effects of multimodal interventions. The primary outcome was acquisition of carbapenem-resistant Gram-negative bacilli or extended spectrum β-lactamase–producing Enterobacterales during ICU stay. Effects were compared with standard precautions, and certainty of evidence was assessed using the GRADE approach for network meta-analysis. Results Seventeen studies met the inclusion criteria, and 12 contributed to the component network meta-analysis (36,264 patients; 2,542 acquisition events). Two multimodal strategies were associated with lower estimated acquisition compared with standard precautions: active surveillance combined with chlorhexidine bathing, contact precautions, and enhanced environmental cleaning (RR 0.34; 95% CI 0.14–0.82), and chlorhexidine bathing combined with decolonization (RR 0.39; 95% CI 0.18–0.84). No individual IPC component demonstrated a significant independent effect. Overall certainty of evidence was low because of heterogeneity, indirectness, and imprecision. Conclusions The available evidence suggests that multimodal infection prevention strategies may reduce acquisition of multidrug-resistant Gram-negative organisms in ICUs, whereas isolated interventions beyond standard precautions show uncertain benefit. Findings from the component network meta-analysis should be interpreted cautiously given the substantial clinical heterogeneity and low certainty of the available evidence.