Microbial Colonization of Needleless Intravenous Connectors and the Male Luer End of IV Administration Sets: Does the Partner Matter?

Background:   Use of positive-displacement, needleless intravenous connectors (NC) has been associated with increased rates of bloodstream infection (BSI).  The U.S. FDA has recently instructed makers of these NCs to determine if their devices increase the risk for BSIs.  Whether neutral displacement NCs or the male luer (ML) end of IV administration sets play a role in central line associated BSIs (CLABSI) is unclear. 

Objective:   To determine a) the microbial colonization rates of neutral displacement NCs and ML of IV administration set ends, and b) if cross-contamination of NC, ML and/or the bloodstream occurs.

Methods:   NCs and MLs from patients admitted to 5 different intensive care units at Loyola University Medical Center (LUMC) were cultured using direct plating and broth enhancement in the LUMC microbiology lab.  Isolates from NC:ML pairs and device:blood pairs were typed using pulse field gel electrophoresis if collected sets yielded concordant growth.  CLABSI was defined according to CDC/NHSN definitions.

Results:   We cultured 279 devices (212 NC, 67 ML) from 78 patients.  52 (25%) NCs and 25 (37%) MLs cultured positive. Use of NC for 7 or more days was associated with increased microbial colonization of NCs (Odds Ratio [95% confidence interval]) 2.3 [1.2, 4.3] and MLs 3.1 [1.1, 8.7].  38 (49%) patients had simultaneous cultures of both NCs and MLs.  Of these, 11 pairs from 7 patients (18%) yielded concordant microbial growth (table), 15 (39%) were discordant and 16 (42%) cultured negative.  2 NC:ML pairs (1 P. aeruginosa, 1 coagulase negative staphylococcus) were indistinguishable by molecular typing.  30 blood cultures from 18 patients were positive during hospitalization in the study units and 1 patient was admitted from a long-term care facility with Klebsiella pneumoniae septicemia. Of these, 8 patients fulfilled criteria for CLABSI, 5 had clinically insignificant positive blood cultures and 6 had bacteremia from another source.  2 patients with CLABSI (Enterobacter aerogenes and K. pneumoniae) had central venous catheters and NCs changed upon diagnosis.  NC cultures obtained more than 1 week later were positive for similar organisms.  One NC:blood culture pair (K. pneumoniae) was closely related by molecular typing (2 band difference).

Conclusions:   Both NC and the ML of IV administration sets are colonized at similar rates by similar organisms and serve as potential reservoirs for CLABSI or clinically insignificant positive blood cultures.  Molecular data confirm that cross-contamination of NC, ML and the bloodstream occurs.  Colonization of the ML may have greater significance due to its potential to introduce microorganisms into the flow tract, which cannot be disinfected by scrubbing the hub.