Keywords :
Adolescent; Adult; Aged; Bacteremia/diagnosis/epidemiology/etiology; Bacteria/classification/cytology/isolation & purification; Child; Child, Preschool; Female; Fungemia/diagnosis/epidemiology; Hematopoietic Stem Cell Transplantation/adverse effects; Humans; Incidence; Infant; Male; Microbial Sensitivity Tests; Microbiological Techniques; Middle Aged; Predictive Value of Tests; Retrospective Studies; Serotyping; Transplantation, Autologous; Transplantation, Homologous
Abstract :
[en] We studied 622 transplants undertaken between 1982 and 2001 to: (1) determine the incidence, timing and etiology of bacteremias, and (2) examine the ability of routine surveillance cultures to predict bacteremias. A total of 404 episodes (0.65 episode per patient) occurred in 248 patients, due to coagulase-negative staphylococci (n=171, 42%), Gram-negative bacteria (n=129, 32%), streptococci (n=48, 12%), other Gram-positive bacteria (n=33, 8%), anaerobes (n=9, 2%) and fungi (n=14, 3%). Bacteremias were more frequent in allogeneic (0.96 episode/patient) compared to autologous (0.44) transplants (P<0.0001). The overall incidence decreased from 0.92 episode/patient until 1990 to 0.66 in 1991-1996 and 0.55 in 1997-2001 (P<0.0001), but this was only observed in autologous transplants. Among them, 212 (53%) occurred before hospital discharge and 192 (47%) thereafter. This proportion was lower for coagulase-negative staphylococci, other Gram-positive bacteria and Gram-negative bacteria compared to other agents (P=0.001). In 50% of the cases, the agent responsible for the bacteremic episode was present in routine surveillance cultures previously. In conclusion: (1) bacteremias remain a frequent complication, particularly in allogeneic transplantation, even long after hospital discharge; (2) routine surveillance cultures can predict bacteremias in 50% of the cases, but the practical impact of this observation is limited in view of the costs.
Scopus citations®
without self-citations
40