Groshong or implanted catheter infections in ambulatory haematological patients.

Incidence rates of bacteraemia and catheter-related infections were measured prospectively amongst haematological patients having long-term catheters and hospitalised in the ambulatory care unit between November 2005 and October 2006. The following risk factors were collected: age, sex, catheter type, follow-up duration, level of personal hygiene, pathology, number of lines of treatment, autograft and erythropoietin treatment. 340 patients were included, having 353 catheters (100 of the Groshong-type, followed during 17,621 days, and 253 of the type with implantable ports, followed during 51,049 days). 0.13 catheter-related infections and 0.07 bacteraemia per 100 catheter days were observed with the Groshong-type catheter, whereas 0.05 (P<10(-5)) catheter-related infections and 0.05 (P=0.048) bacteraemia were observed amongst patients with implantable ports. A multivariate analysis (Cox method taking into account the length of follow-up) on risk factors highlighted a significant effect of the type of catheter on catheter-related infections (Groshong versus implantable port OR=5.74, P<10(-3)), and of several factors on bacteraemia (lymphoma versus other pathologies OR=3.19, P=0.041; erythropoietin treatment OR=2.88, P=0.009; autograft OR=3.35, P=0.011; number of lines of treatment OR=0.68, P=0.047). It was not possible to determine if poor levels of personal hygiene had a significant impact, due to large numbers of missing data. These results, consistent with other studies, are not only useful in validating prevention policy but also in demonstrating lack of catheter traceability.