Catheter-related bloodstream infection in adult haematology patients: catheter removal practice and outcome.

The aim of the present study was to describe the practice of central venous catheter (CVC) removal and outcomes of catheter-related bloodstream infection (CR-BSI) in adult haematology patients. Patients were identified retrospectively according to diagnosis coding of inpatient episodes and evaluated when, on examination of medical records, there had been evidence of sepsis with strong clinical suspicion that the source was the CVC. Demographic and bacteriological data, as well as therapeutic measures and clinical outcomes, were recorded. One hundred and three patient episodes were evaluated. The most frequent type of CVC was the Hickman catheter and the most frequently isolated pathogen was coagulase-negative staphylococci. Twenty-five percent of episodes were managed with catheter removal. Treatment failure, defined as recurrence of infection within 90 days or mortality attributed to sepsis within 30 days, occurred significantly more frequently in the group managed without catheter removal (52.5% versus 4%, P < 0.05). Specifically, 90-day recurrence was more common when the catheter was retained (46% versus 0%). However the difference in 30-day attributable mortality (7% versus 4%) was not significantly different. Notably, no significant difference between the two groups emerged in respect of other measured characteristics that had been considered as potential determinants of outcome. More frequent CVC removal for CR-BSI, in this population, should be considered. Management of CR-BSI without catheter removal is associated with treatment failure, morbidity and carries significant resource implications.