Prevention of peripheral intravenous catheter-related bloodstream infections: the need for routine replacement

To the Editor: Peripheral intravenous catheters (PIVCs) frequently cause Staphylococcus aureus bacteraemia, and a disproportionate number of episodes involve catheters that have been left in place for ≥ 4 days or have been inserted in emergencies.1 Other studies have shown similar results with catheters left in place for > 48 hours or not routinely replaced after emergency placements.2-4

At Canberra Hospital, we have followed all bloodstream infections (BSIs).5 Since 2002, there have been 52 episodes of BSI associated with PIVCs; of these, 22 have been S. aureus bacteraemia. Where the duration of insertion was known, all but seven episodes involved catheters known to have been in place for > 48 hours (34 episodes) or were inserted in emergency situations or hospital transfers and left in place for > 24 hours (5 episodes) (Box).

In Spain, despite specific recommendations to remove PIVCs within 72 hours, 26% remained in place for longer.6 In participating hospitals where more PIVCs were in place for > 72 hours, PIVC-related BSI rates were threefold higher compared with hospitals that adhered to the recommendations (0.06 v 0.02 per 1000 patient-days).

It is very disconcerting that inappropriate recommendations against the routine replacement of PIVCs have been made,7 based on the occurrence of phlebitis rather than bacteraemia. Phlebitis usually results from non-infective causes (eg, irritation from drugs) and is an inappropriate surrogate marker for infection.

Prevention of severe sepsis is the most important clinical end point. However, its incidence is very low — about one BSI per 3000 catheters.5,6 Thus, any prospective randomised study would need to be extremely large, with tens of thousands of patients in each arm. Studies of this size have never been done and are unlikely.

We believe that we need recommendations based on the best available evidence1-6 — which supports an end point of bacteraemia not phlebitis. We agree that we need good catheter-insertion techniques.7 However, the current evidence strongly suggests that routine replacement of PIVCs at 48–72 hours will result in substantially lower sepsis rates than replacement at later times.