R. Ivanova

018 NEW FEATURES OF ENDOCARDITIS WITH THE CHANGE OF MILLENNIUM

The aim of this study was to describe the characteristics of HAIE and to establish a comparison between health care and community-acquired episodes. HAIE was defined as either IE manifesting >48 hours after admission to the hospital or IE acquired in association with a significant invasive procedure performed during a stay and/or manipulation in a hospital setting within 6 months before diagnosis. Results: HAIE accounted for 16% of 793 cases. Compared with community-acquired infection, patients with HAIE tended to be older (60.1 vs. 53.6; P= 0.0001) and had more co-morbidities (Charlson index, 3.3 vs. 1.8; P= 0.0001) and staphylococcal infections (58.3% vs. 24.8%). Intra-hospital mortality (44.9 vs. 24.2%) was higher in the HAIE group. Vascular manipulation constituted the main cause of bacteremia responsible for HAIE (63%). Septic shock (OR, 10.13; 95% CI, 3.18 32.12; P= 0.0001) and severe heart failure (OR, 2.79; 95% CI, 1.09 7.13; P= 0.03) were independent predictors of intrahospital mortality in the HAIE cohort. Conclusions: The present study demonstrates that HAIE attacks a fragile population and it is principally caused by microorganisms strongly related to vascular manipulations. Extremely careful management of vascular accesses is needed in order to minimize the risk of secondary bacteraemias.

037 STAPHYLOCOCCUS AUREUS PROSTHETIC VALVE ENDOCARDITIS

Background: The incidence of enterococcal bacteraemia (EB) is increasing. We ought to appraise the incidence and clinical characteristics of EB in the current decade. Methods: From 2001 2008, all adults with EB were included. Data on demographics, co-morbidities, severity (Charlson and Pitt scores), and clinical characteristics were obtained. Two periods were compared: 2001 2004 (P1) and 2005 2008 (P2). Results: There were 100 EB out of 3060 (3.3%) total bacteraemias. Median age was 71.5 y (range, 21 97), 65% were males; 27% were seen in P1 and 73% in P2 (p = 0.0006, CI 0.036 0.01) that represented 2% and 4.2% of all bacteraemias, respectively. 54% were due to E. faecalis and 23% to E. faecium. EB had a community-onset (C-O) in 56% of cases, polymicrobial in 32%. Nosocomial EB was due to urinarytract-infection (UTI) (39.5%), unknown-origin (U-O) 16.3% and 11.6% catheter-related. In C-O cases, 37% were UTI, 32% were biliary tract infection and 16% U-O. Endocarditis was diagnosed in 4% cases, all of C-O. 8% presented shock. LOS was 18 days and mortality 24%, 5/24 (21%) patients died within <72 h. Charlson score was 2.13 vs 3.38 (p = 0.027), respectively. There were no significant differences between both periods in terms of origin of EB, severity, LOS or mortality. Conclusions: The incidence of EB has doubled in the last 4 years in our hospital. A higher presence of co-morbidities in admitted patients could probably explain this finding. Enterococcal endocarditis remains a C-O entity. EB is rarely associated with shock or severity and it presents with a high late mortality rate.

016 HEALTH CARE-ASSOCIATED INFECTIVE ENDOCARDITIS

Background: Infective endocarditis (IE) from different causative pathogens differ with regards to prognosis. We looked at predictive factors for different pathogens in a series of 386 patients admitted to two Copenhagen tertiary centres from 2002 to 2008, including differences in patients with native valve (NVE) or prosthetic IE, stratified by time from surgery (<3 months = early PVE or 3 months = late PVE). Methods: The study population consisted of 439 patients, excluding IV drug users (N = 22), PM endocarditis (N = 20), and recurrent IE (N = 11). Data is presented as number (%) and differences were tested by c2test. Predictive factors for individual pathogens were analysed by multivariate logistic regression modelling. Results: Median age (25th and 75th percentile) was 65 (56 75) and 113 (29%) were female. The prevalence of known risk factors for IE were: early PVE 18 (5%), late PVE 79 (20%), diabetes 34 (9%), history of cancer 38 (10%), renal dysfunction 57 (15%) and immunosuppression 36 (10%). Causative pathogen were Viridans group Streptococcus 120 (31%), Staphylococcus aureus 80 (21%), enterococci 66 (17%), Coagulase-negative staphylococci 32 (8%), other 44 (11%) and culture negative 44 (11%). Viridans group Streptococcus IE was associated with NVE (OR = 3.8, 95% CI: 1.9 7.4) and was infrequent in patients with renal dysfunction (OR = 0.33, 0.13 0.79). Enterococcus IE was related to age (OR = 1.3, 1.2 1.5 per 5 years) and frequent in patients with renal dysfunction (OR = 4.2, 2.0 8.9). Coagulase-negative Staphylococcus IE was frequent in PVE (OR = 4.5, 2.0 10.1). Culture Negative IE was frequently associated with PVE (OR = 2.6, CI: 1.2 5.4) and younger age (OR = 0.8, 0.7 0.9). Rare pathogens, grouped as ‘other’ was associated with younger age (OR = 0.8, 0.7 0.9 per 5 years) and less frequent in patients with renal dysfunction (OR = 0.45, 0.23 0.90). No independent predictors of Staphylococcus aureus IE were identified (IV drug users excluded). No differences in the pathogens associated with early PVE and late PVE could be identified. Conclusions: Viridans group streptococci are not common in patients with prosthetic heart valves, whereas coagulase-negative staphylococci and culture negative IE were more common in this group. Enterococcus IE are associated with increasing age. Considerable variation in the clinical appearance of IE and risk factors associated with causal pathogens underline the continuing diagnostic challenge of IE.