Objective: Despite the lack of evidence to support routine scheduled replacement of dialysis catheters (DCs) this practice continues to be widely used in many intensive care units (ICUs). This study evaluated whether additional risks of catheter-related infection (CRI) are incurred with a conservative attitude in critically ill cancer patients.
Design and setting: Prospective, observational study over a 14-month period in a 15-bed medicosurgical unit in a comprehensive cancer center.
Patients: Seventy-nine double-lumen DCs were evaluated in 47 patients. Incidence rates of infection per 1000 days of catheter use were examined over 7-day periods.
Measurements and results: The mean indwelling time was 6.9+/-5.5 days. Twelve DCs (15.2%) were removed for suspected CRI. Catheter-tip cultures remained negative in 74 cases (93.7%). Overall, one bacteremic CRI, two colonization episodes, and two contaminations were diagnosed, leading to DC colonization and DC-related bacteremia incidence rates of, respectively, 5.4 and 1.8 per 1000 days. When the catheter colonization rate was examined at 7-day intervals, the incidence rate was similar whatever the indwelling time: 5.8, 4.8, and 6.0 per 1000 days, respectively, for the 49 catheters left in place for 7 days or less, 8-14 days (21 DCs), and more than 14 days (9 DCs). The DC colonization incidence rate was similar to that of the 42 short-term catheters inserted during the same period in the same patients (5.9 per 1000 days).
Conclusions: The stable low risk for DC-related infections over time does not support the rationale for scheduled replacement, even in immunocompromised cancer patients.