A Randomized Cross-Over Clinical Trial to Compare 3.15% Chlorhexidine/70% isopropyl alcohol (CHG) vs 70% isopropyl alcohol alone (alcohol) and 5s vs 15s scrub for routine disinfection of needleless connectors (NCs) on central venous catheters (CVCs) in an adult medical intensive care unit (ICU)
Background: Disinfecting NCs before access prevents introduction of microbes and reduces the risk of CVC- associated bloodstream infection. The best disinfection method is unknown.
Methods: We conducted a prospective, randomized, crossover clinical trial in a 24-bed ICU to compare 2 disinfectants (CHG vs alcohol) and 2 scrub times (5s vs 15s) for routine disinfection of CVC NCs (negative displacement). Interventions were assigned randomly to each of 2 clinically identical, geographically separate ICU regions. Disinfectant was crossed over twice; scrub times remained the same. Blank stickers were affixed to fronts of prep pads to blind providers to disinfectant. Fidelity to the interventions was monitored by surreptitious observation.
NCs from centrally or peripherally inserted CVCs that had been in place ≥24 hours were eligible for study. Tunneled, antibiotic-impregnated, and dialysis CVCs and CVC introducers were excluded. NCs were changed routinely every 4 days. Biofilm contamination of NC interiors was assessed by sonication and culture. The primary outcome measure was proportion of contaminated NCs.
Results: From March 26, 2012 – June 14, 2013, 1323 NCs were evaluated for eligibility. 509 NCs from 159 catheters (141 patients) were collected and processed. No deviation from assigned disinfectant was observed (101 observations). Mean scrub time was 7s (95% CI, 6-9s) in 5s arms and 9s (95% CI 8-11s) in 15s arms (p=0.071). Proportion of NC contamination was lower in CHG vs alcohol arms, but the difference was significant only for 5s scrub time (Table).
In exploratory analysis, subclavian vein CVC site vs other anatomic sites was associated with lower risk of NC contamination (RR=0.28, 95% CI 0.07-1.07, p=0.042.) In a multilevel mixed effects model in which NCs were nested within lines, the scrub-time + disinfectant interaction term was significant (p<0.001) in predicting NC contamination. Distributions of contaminating microbes were similar across arms (51% coagulase-negative staphylococci, 3% Staphylococcus aureus, 26% other Gram positive spp., 9% Gram negative rods, 10% Candida spp.)
Conclusion: CHG scrub resulted in less NC contamination than alcohol scrub but the difference was significant only for the shorter scrub time.
M. K. Hayden,
PDI Healthcare: Product was contributed to the hospital participating in this clinical trial at no cost., Chlorhexidine and alcohol prep pads
L. M. Portillo, None
B. Xu, None
S. Weiner, None
L. Fogg, None
H. Zhang, None
V. Didwania, None
K. Lolans, None
R. Donlan, None
R. A. Weinstein, None
W. Trick, None