Mechanisms and risk factors for infection of pulmonary artery catheters and introducer sheaths in cancer patients admitted to an intensive care unit

Pulmonary artery catheters (PACs) are typically inserted for short periods, and the extra-luminal route is assumed to be the overriding source of contamination and/or infection. Our aim was to assess the incidence of PAC and introducer colonization in cancer patients, and to study the mechanisms and risk factors for infection. Patients with a Swan-Ganz catheter admitted to an intensive care unit were prospectively analyzed over 14 months. As soon they were no longer necessary, PACs and introducer sheaths were removed and cultured. We recorded the mean duration of placement, the number of times PACs were handled and the site of insertion. Seventy-nine catheters were inserted in 68 patients. The median (range) duration was three days (0-10) for PACs, and 3.6 days (0-18) for introducers. PAC and/or percutaneous introducer sheath colonization was diagnosed in seven patients (8.9%), but in only one case were both colonized. Colonization rates were 15.5 per 1000 days for PACs and 14.1 per 1000 days for introducers. Introducers were mainly colonized before the 5th day, while PACs were mainly colonized after the 5th day. No PAC or introducer-related local infection or bacteraemia was diagnosed. Colonization was more frequent on catheters inserted into the internal jugular vein. The colonization rate was 5% for PACs and introducers. Our findings suggest that contamination of introducers and PACs may be dissociated and could result from either extraluminal or endoluminal colonization. As three of four PAC colonizations occurred after 5 days, the duration of catheter placement should be considered important. There was little clinical impact of microbial colonization.