ECCMID edition:18 location:Barcelona, Spain date:19-22 April 2008 AVA Conference edition:22 location:Savannah, US date:11-14 September 2008
Needleless positive-pressure mechanical valve connectors: are they safe?
V. Cossey, M.J. Thelissen, G.A. Goossens, M. Stas, A. Schuermans. (Leuven, BE)
Objectives: To determine whether the introduction of a needleless positive-pressure mechanical valve connector in adult and paediatric hematology-oncology units influenced the rate of catheter-related bloodstream infections (BSI) and the type of causative micro-organisms.
Methods: A mechanical valve (MV) connector system (CLC2000®, ICU Medical) was introduced in November 2005 and replaced miscellaneous conventional open systems (COS) with a standard Luer-Lock removable cap. The main objective of this change was to reduce the use of heparin flushes in long-term tunnelled catheters (Hickman®, CR Bard).
The retrospective analysis included all patients with new Hickman catheters, inserted during two different study periods. The catheter-related BSI rate was observed during 6 months for each system and reported as number of BSI per 1000 catheter days. The results with COS (January through June 2005) were compared with a similar period with MV-use (September 2006 through February 2007). Microbiological characteristics of bacteraemias occurring during each period were also compared. The hospital policy for care of Hickman catheters did not change over both study periods.
Results: During the COS-period, 39 Hickman catheters were inserted in 15 children and 24 adults with a total dwell time of 1544 and 1899 catheter days respectively. During the MV-period 60 Hickman catheters were placed in 13 children and 47 adults with a total dwell time of 1181 and 3223 catheter days respectively.
Both in the paediatric and adult population, the rate of catheter-related BSI during the MV period was significantly higher than during the COS-period: resp. 11 vs 1.29 infections per 1000 catheter days (p =.009) and 12.4 vs 6.32 infections per 1000 catheter days (p = 0.019).
During the MV-period, the percentage of polymicrobial BSIs increased from 0 to 84.6% in children and from 8.3% to 21% in adult patients.
Conclusion: A significant increase in the rate of BSI was found among hematology and oncology patients with newly inserted Hickman catheters concomitantly with a change from a conventional open system to a mechanical valve connector, despite additional educational sessions regarding proper use of the device. The risk and type of BSI are clearly associated with the catheter cap design which promotes microbial contamination. The observed effects are more pronounced in children