K-3399. Recurrent Bacteremia (RB) - When and Which?
Session: Poster Session: Blood Stream Infections
Monday, October 27, 2008: 12:00 AM
Room: Hall C
Background: RB is a well known clinical problem that has been studied to some extent. However, there is a lack of population-based (PB) data regarding frequency and time frame of RB as well as the overall distribution of microorganisms (MO) and foci of infection (FI). Methods: All clinically significant episodes of bacteremia have been recorded prospectively in a Danish PB research database since 1992. Data included MO, patient demographics and FI as determined in collaboration with physicians. The date of the draw of the first positive culture was the baseline date. We used the following criteria to define RB: an interval of 1) >48 hours was required if a different MO was recovered in a subsequent blood culture and 2) >30 days if the MO were identical, unless there was unequivocal evidence for a new FI. We included all patients with a first episode (FE) of bacteremia from 1996 through 2005 in whom RB occurred within 12 months. Results: Among 8,672 patients with bacteremia, 1003 (12%) had RB, of which 74% occurred within 3 months. 58% of the RB were nosocomial as compared to 48% in the FE. The proportion of fungemia and poly-microbial bacteremia was higher for RB than for FE (22% vs. 17%). In 737 patients (74%) both episodes were mono-microbial and the same MO was found in 274 patients (37%). Among these the most frequent MO were Escherichia coli (48%) and Staphylococcus aureus (20%). Identical FI in both episodes were found in 386 (52%) of the RB patients; the urinary tract (15%) being the most frequent identical FI. Compared to the FE, an unknown FI was more frequent in RB (29% vs. 23%). This was also true for i.v. devices (15% vs. 13%).
Conclusions: In this PB study, 12% of the bacteremic patients had RB within one year, of which 74% occurred within 3 months. More often RB was nosocomial, polymicrobial or caused by yeasts, compared to the FE. RB should be studied in a broader perspective in order to identify risk factors and preventive strategies.
Christian Østergaard1, Henrik Schønheyder2, Jenny Knudsen3, Kim Gradel2, Niels Frimodt-Møller, MD4, Ulrich Jensen5 and  U. S. Jensen, None., (1)Copenhagen University Hospital Herlev, Herlev, Denmark, (2)Aalborg Hosp., (3)Hvidovre Hosp., (4)Statens Serum Institut, (5)Statens Serum Inst., Copenhagen, Denmark

See more of: Blood Stream Infections
See more of: Poster Session