Jf. Abete et al., Risk factors associated with colonization and bacteremia in non-tunnelled central venous catheters, REV CLIN ES, 200(3), 2000, pp. 126-132
Objective. To identify risk factors for colonization and bacteremia among p
atients with non-tunnelled central venous catheters.
Materials and methods. A prospective study was conducted of a cohort of pat
ients carrying non-tunnelled central venous catheters. Different parameters
were obtained and the degree of its association with colonization of the d
istal portion of the catheter or with bacteremia associated with colonizati
on was estimated. The CDC (centers for Disease Control) diagnostic criteria
of colonization and catheter-related bacteremia were used.
Results. A total of 118 catheters were eventually analyzed, corresponding t
o 114 patients, with a catheterization mean time of 14 +/- 8 days (mean +/-
SD); out of these 114 patients, 51 were colonized and in 22 the presence o
f associated bacteremia was confirmed. The parameters associated with a hig
her risk for catheter colonization included length of colonization, femoral
location, number of lumina and a vital prognosis lower than one month. All
these factors, with the exception of the increase in the number of lumina,
showed an independent association with colonization on the multivariate an
alysis [catheterization length tin weeks): OR 1.46; 95%CI: 1.0-2.11; femora
l location: OR 3.73; 95%CI: 1.16-11.9; vital prognosis lower than one month
: OR 12.7; 95%CI: 1.4-112.7]. As for risk for catheter-related bacteremia,
the univariate analysis showed an association with catheterization length a
nd a vital prognosis loner than one month; the latter was the only factor t
hat maintained an independent association in the multivariate analysis (OR
5.75; 95%CI: 1.17-28.27).
Conclusion. The present study documents the relevance of prolonged catheter
ization as a consistent risk for colonization of non-tunnelled central veno
us catheters. This risk increases independently in canalization at femoral
site and particularly among severely ill patients. The presence of these fa
ctors allows the identification of a high risk population for the developme
nt of catheter-related bacteremia.