Diagnosing catheter-related bloodstream infections is important but not alw
ays easy and a failure to make the diagnosis may have serious consequences.
A high rate of unnecessary catheter removal is noted. We retrospectively c
ompared the clinical and usual methods of microbiological diagnoses of cath
eter-related sepsis to the speed of detection of the catheter versus periph
eral blood cultures using the Bact-Alert system. We analyzed 50 files of pa
tients with central indwelling devices: 16 single lumen catheters and 34 im
planted ports. Twenty-one catheters were classified as infected, and we obs
erved an earlier positivity of catheter versus peripheral blood in all case
s, but significant for 19 patients. According 50 standard diagnosis methods
, 29 catheters were estimated non-infected, a more rapid detection of perip
heral culture was reported for 17 specimens and, for another eight patients
, the time of detection was equal to blood culture drawn from the catheter:
In this group, four discrepancies were recorded with a differential time i
n favor of sepsis related to catheters ranging from 0.5 to 2 hours. Because
of its simplicity and low cost, we believed that this method could be the
first step of a diagnosis of catheter-related sepsis and could, therefore,
avoid unjustified removal, in particular for the implanted ports for which
the diagnostic methods are less codified than for catheters. A prospective
study is ongoing; the design of the study focuses only on implanted ports.
(C) 2000 Editions scientifiques et medicales Elsevier SAS.