Arterial Line Insertions in the ICU: To Gown or Not To Gown?

Abstract & Commentary

By Andrew M. Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle, is Associate Editor for Critical Care Alert.

Dr. Luks reports no financial relationship to this field of study.

Synopsis: Although the incidence of catheter-related bloodstream infections from arterial catheters is low, arterial catheters have similar colonization rates and catheter-related bloodstream infections as concurrently sited and identically managed central venous catheters.

Source: Koh DBC, et al. Crit Care Med. 2008 [epub ahead of print] PMID: 18091549

Most of the attention on catheter-related bloodstream infections (CR-BSI) in the ICU focuses on central venous catheters (CVC), a bias that likely derives, in part, from the 2002 Centers for Disease Control guidelines which stipulate that arterial catheters (AC) have "low infection rates—rarely associated with bloodstream infections."1 Koh and colleagues conducted a single-center, prospective cohort study to determine whether this approach is valid or whether more attention should be focused on the infective potential of ACs.

Over a 24-month period, the authors examined all ACs and CVCs inserted in ICU patients by attending and resident-level physicians using full barrier precautions (except in rare cases of extreme patient instability) and chlorhexidine skin preparation. Catheters were inserted in the ICU, operating room or emergency room and the authors excluded any catheters inserted at an outside hospital. Multiple AC brands were used and all sites were covered with a transparent sterile dressing and accessed only by ICU nursing staff. No antibiotic-coated or tunneled CVCs were included in the study. At the time of catheter removal (typically on discharge from the ICU to the wards), the distal 3 to 5 cm of the catheter was collected and sent for semi-quantitative culture. Results of patient blood cultures were also tracked during the study period. The authors defined catheter colonization as > 15 colony forming units on the catheter tip culture and CR-BSI as > 15 colony forming units plus a positive blood culture within 48 hours of catheter removal with the same microorganism and antibiotic sensitivity.

During the study period, 321 ACs were sited for an average of 3.4 + 3.5 days and observed for a total of 1082 catheter days, while 618 CVCs were sited for an average of 6.5 + 4.9 days and observed for 4040 catheter days. 5.3% of ACs and 11% o CVC were colonized with bacteria. The colonization and CR-BSI rates per 1,000 catheter days were 15.71 (95% CI 9.5-25.9) and 0.92 (95% CI 0.13 - 6.44), respectively for ACs and 16.83 (95% CI 13.3 - 21.3) and 2.23 (95% CI 1.12 - 4.44), respectively for CVCs. There was one instance of CR-BSI due to an AC and there were no statistically significant differences in colonization rates between ACs and CVCs. AC Colonization rates were higher when the catheters were inserted in the operating and emergency rooms, while there was a non-statistically significant trend toward higher AC colonization rates when catheters were inserted by residents rather than attending-level physicians and when catheters were inserted in the femoral sites. The majority (79%) of colonization events were due to coagulase-negative staphylococci while Staphylococcus aureus, Corynebacterium species and enterococcus accounted for the remainder of cases.


While full barrier precautions have become the standard of care and a standard practice for CVC insertions at our institution, it is not an infrequent occurrence to see resident or attending physicians inserting arterial catheters without being fully gowned or applying an appropriate sterile field to their work site. The patient's arm is sometimes positioned poorly on the edge of the bed with crumpled linen, wires, and the patient gown in close proximity to the insertion site. When asked why they used this approach on a particular patient, the majority of physicians typically respond with a comment that closely resembles that from the 2002 Centers for Disease Control guidelines regarding the low incidence of arterial catheter related infections. The data from Koh and colleagues should lead to reevaluation of such practices, particularly when one considers that this is not the first study to demonstrate similar rates of AC and CVC colonization in the ICU.2 What is particularly noteworthy is that the ACs had a high rate of colonization equal to that of CVCs even though full barrier precautions were employed during their insertion. This would lead one to suspect that the colonization rates and perhaps even the CR-BSI rates, might, in fact, be higher when poor sterile techniques are used with AC insertion.

It should be noted that full barrier precautions have not been shown to decrease the incidence of AC-related blood stream infections3 and, as noted earlier, the increased colonization rates have not been associated with an increased rate of bloodstream infections. However, in an era when Medicare and other insurance providers will be cutting or eliminating reimbursement for preventable complications, such as catheter-related urinary tract infections, it is worth considering whether we need to improve the sterility of our arterial line insertion practices, particularly in patients we anticipate will require their catheters for long duration ICU stays.


  1. O'Grady NP, et al. MMWR Recomm Rep. 2002: 51:1-29.
  2. Traore O, et al. Crit Care Med. 2005:33:1276-1280.
  3. Rijnders BJ, et al. Clin Infect Dis. 2003:36:743-748.