In situ endoluminal brushing: a safe technique for the diagnosis of catheter-related bloodstream infection

Abstract

We assessed the safety of the endoluminal brush technique for the in situ diagnosis of central-venous-catheter (CVC)-related bloodstream infection (CRBSI). The endoluminal brush was used to investigate patients with suspected CRBSI by sampling the CVC lumen to within 3-5 cm of the catheter tip (the brush was kinked to mark the length required). Quantitative peripheral blood cultures were taken 1 min pre-brushing and 1 min and 1h post-brushing. In addition, a 1 mL sample of through-catheter blood was aspirated immediately pre- and post-brushing, and aliquots were directly plated on to blood agar to determine the load of CVC-associated bacteria. After CVC removal, catheter tips were sent for culture using the Maki roll and modified Cleri flush techniques. Of 139 suspected cases, 61 (45%) had confirmed episodes of CRBSI. No patients experienced fever or rigors, or had evidence of embolic phenomena after CVC endoluminal brushing. There was no difference overall in pre-brushing and 1-min post-brushing peripheral blood bacterial counts (P < 0.702). Peripheral and CVC blood bacterial counts were significantly reduced 1h post-brushing (P < 0.03 and P < 0.001, respectively), possibly due to removal of endoluminal biomass. The endoluminal brush technique can be safely used to diagnose CRBSI without the need for CVC removal, provided that the brush is not allowed to protrude beyond the tip of the catheter.