We compared, in three intensive care units, colonization of hubs with hub p
rotection boxes or hubs with needleless closed connectors; 137 central veno
us catheters and 451 hubs were randomized in two groups with similar charac
teristics. Catheter and hub colonization were not different between the two
groups. Among 30 colonized catheters, the same isolate was found in only t
wo hubs; hub contamination rarely is responsible for catheter colonization
in short-term catheters. Further studies are required to evaluate the benef
it of protected hubs compared with unprotected hubs (Infect Control Hosp Ep
idemiol 2000;21: 10-42).