Central venous catheter-related infection: does insertion site still matter? A French multicentric cohort study.

Abstract

We aim to evaluate the association between central venous catheter (CVC) insertion site and microbiological CVC complications in a nationwide cohort. This study was conducted using the healthcare-associated infection surveillance cohort "REA-REZO" involving 193 intensive care units (ICUs). All CVC inserted and removed during the same ICU stay between January 1st 2018 and December 31st 2022 were eligible but only those whose tips were sent for microbiological analysis were included. Primary objective was to describe CVC insertion sites and subsequent catheter-related bloodstream infection (CRBSI). Out of 126,997 CVCs, 71,314 were not sent for tip culture, and only 55,663 CVCs were included, (30,548 in internal jugular [IJ], 14,423 in femoral and 10,692 in subclavian [SC] sites). The incidence of CRBSI was 0.7 [0.6-0.8] in the IJ site, 0.7 [0.6-0.9] in the femoral site, and 0.6 [0.4-0.7] CRBSI per 1000 CVC days in the SC site (p = 0.248). The multivariable Poisson regression model showed no differences of CRBSI incidence rates between the three insertion sites. Microorganisms observed in CRBSI were coagulase-negative Staphylococci (27.9%), Enterobacterales (27.5%), non-fermenting Gram-negative Bacilli (10.4%), Candida sp. (16.9%), and Staphylococcus aureus (16.9%). Low CRBSI incidence rates were reported. CRBSI incidences rates were similar in the three insertion sites. Uncertainty remains due to potential selection bias since many CVCs had to be excluded.