Catheter-related sepsis in patients on intravenous nutrition: a prospective study of quantitative catheter cultures and guidewire changes for suspected sepsis.

One hundred and ninety-five central venous catheters used for intravenous nutrition in 113 patients were studied prospectively. Catheter-related sepsis (CRS), defined by recovery of the same organism from the catheter tip and peripheral blood cultures, occurred with only 3.3 per cent of catheters or 2.3 per 1000 days of therapy. In contrast, CRS was suspected with 30 per cent of catheters and catheter contamination occurred in 37 per cent. Contamination was defined by a positive catheter tip culture without recovery of the same organism from the blood. CRS was present in 4 of 12 cases (33 per cent) with greater than 1000 colony forming units on the catheter tip but in only 2 of 54 (4 per cent) with fewer organisms. Thirty-eight cases suspected of having CRS were randomized to have catheter removal and later replacement, or replacement over a guidewire. There were no significant differences in the catheter contamination rate but there were significantly fewer problems of insertion in the guidewire group. However, transfer of Klebsiella sp., during the guidewire procedure, resulted in subsequent sepsis in one case. It is concluded that replacement of catheters over a guidewire is a safe and convenient way of establishing whether sepsis is catheter-related. Because organisms may be transferred, the procedure is not an appropriate treatment for catheter-related sepsis.