Infectious risk of replacing venous catheters by the guide-wire technique

During the perioperative period cardiac surgical patients are often monitored by pulmonary artery (PA) catheters. This catheter, which is floated through the right heart into the pulmonary artery, enables the intensivist to measure and calculate indices of myocardial performance. After a variable period of time this invasive monitoring can often be abandoned following cardiovascular stabilization in these patients, but patients usually still require a central venous (CV) access for diagnostic and therapeutic purposes. To place this CV catheter either a de novo puncture at a new site or a guide-wire change at the existing exit site through the PA catheter in place can be performed. Each de novo puncture is associated with a risk of traumatising internal vessels or organs. In contrast, guide-wire change avoids this risk but inherits a potential risk of transferring bacteria by manipulation of contaminated lines. Our study included 159 consecutive cardiac surgical patients in whom PA monitoring was established preoperatively and terminated within a period of up to 72 hours postoperatively. At random the PA catheter was replaced by a CV line either by de novo puncture or by guide-wire change. All CV lines were left in place for 7 days according to standard practice in our intensive care unit (ICU). After removal of CV catheters all catheter tips were cultured semiquantitatively by rollplate technique according to Maki (26) and subsequent immersion broth culture. A positive culture was defined as growth of one up to 15 colonies on the agar-plate or any microbial growth in the broth. A significant colonization was assumed in catheters yielding more than 15 colony forming units (cfu) on the blood agar plate (26). Our results show a significant risk of colonization and catheter-related infection associated with the guide-wire technique as opposed to the de novo puncture. The figures for relevant colonization were 33.3% in the guide-wire group as opposed to 10.5% in the de novo group. However, this difference was noted only in the subgroup in which replacement of PA catheters by CV catheters was performed beyond 48 h after initial insertion of PA catheters. Within the time intervals of 24 and 25 to 48 h, respectively, we could not detect any significant difference between groups.