[Polyurethane versus polyethylene: in vivo randomized study of infectious complications of central catheterization].

Polyurethane (PU) and polyethylene (PE) catheters were distributed by randomization among adult ICU patients to evaluate the impact of the catheter polymer on the rate of catheter-related sepsis (CRS). The two catheters were otherwise strictly identical. Three hundred central venous catheters were randomized and inserted in the subclavian or internal jugular vein, at the discretion of the clinician. Mean duration of insertion was 9.2 (+/- 3) days for both catheter types and mean number of line openings was 159 (+/- 60). A bacteriologic culture using a variant of Maki's technique was performed on the 205 catheters removed before patient discharge. No significant differences were found between the two catheters. Total number of tip infections was 33 and polyurethane and polyethylene prevalence ratio was 0.7. A clinical evaluation was performed for the 183 catheters inserted for a least 48 hours (in 142 patients). Patients were divided into four clinical categories according to previously reported definitions (Brun-Buisson et al., 1987). Clinical tip-infection rate was 4.4% and polyurethane and polyethylene ratio was 0.7. For three additional catheters (1.6%), only the Luer-lock was infected. These three catheters were made of polyethylene and were inserted into the internal jugular vein. The Luer-lock was made of polypropylene.