• D B C Koh, I K Robertson, M Watts, and A N Davies.
• University of Tasmania, Intensive and Coronary Care Unit, Launceston General Hospital and Clifford Craig Medical Research Trust, Launceston, Tasmania. dkoh@utas.edu.au
• Anaesth Intensive Care. 2010 Jul 1;38(4):678-84.

AbstractPeripheral arterial catheters are used for the continuous monitoring of blood pressure and repeated blood sampling in critically ill patients, but can be a source of catheter-related bloodstream infection. A common assumption is that the more frequently an arterial catheter is accessed, the greater the likelihood of contamination and colonisation to occur We sought to determine whether the accessing frequency has an influence on the rate of colonisation in a peripheral arterial catheter A retrospective, unmatched, nested case control study was conducted in our intensive care unit. The intensive care unit charts of 96 arterial catheters from 83 patients were examined to measure the number of times each respective arterial catheter was accessed. Multivariate Cox proportional hazards regression was used to compare the rate of accessing of arterial catheters and account for varying arterial catheter in situ duration. Arterial catheters which had a high access rate of 8.1 or more times/day (five colonised of 32 patients: hazards ratio 1.69, 95% confidence interval 0.52 to 5.49; P = 0.77), or a medium access rate of 6.7 to 8.0 times/day (six colonised of 32 patients: hazards ratio, 1.35, 95% confidence interval, 0.37 to 4.92: P = 0.65) were not significantly more colonised when compared to arterial catheters which had a low access rate of O to 6.6 times/day (six colonised of 32 patients), adjusted for arterial catheter insertion site and place in hospital where the arterial catheter insertion was performed. We were unable to demonstrate that the accessing frequency of an arterial catheter was a major predisposing factor for the likelihood of colonisation. Other mechanisms other than hub colonisation should be investigated further.

19 keywords...

hide…