Isolated pathogens and clinical outcomes of adult bacteremia in the emergency department: a retrospective study in a tertiary Referral Center.

BACKGROUND Approximately two-thirds of the patients with severe sepsis or septic shock are first encountered in the emergency departments (EDs) of western countries, in which bacteremia is present in about 50% of patients with severe sepsis. The situation of bacteremia presenting to the EDs in Taiwan is not well documented. The objective of this study was to examine the epidemiology and microbiology of bacteremia in adult patients who visited the ED of a medical center in southern Taiwan. METHODS A retrospective observational study of the epidemiology and microbiology of bacteremia was conducted in the ED of a medical center involving 6,137 adult patients and 13,903 blood cultures. RESULTS A total of 831 consecutive patients with 890 episodes of bacteremia were obtained from January 1 to December 31, 2004, indicating a positive culture rate of 13.5% (1,872/13,903). Among these episodes, 525 (59%) were defined as true community-acquired infections followed by 263 (29.5%) as health care-associated infections and 102 (11.5%) as nosocomial infections. Of the 972 isolates, 289 (29.7%) were gram-positive species and 683 (70.3%) were gram-negative species. Urinary tract infections (32.2%, 287/890) were most common in these patients, with Escherichia coli (30.8%, 299/972) being the most common pathogen. Bacteremia caused by Staphylococcus aureus was more common in nosocomial than true community-acquired infections (31.3% vs. 12%) and had significantly higher possibility of resistance to methicillin in infections not purely acquired from community (odds ratio = 24.92; 95% confidence interval, 9.88-62.87). The overall crude mortality rate was 21% and nearly half of the mortalities occurred within 3 days of visiting the ED. All patients discharged inadvertently were uneventful (n = 65, two lost at follow-up). CONCLUSIONS Categories of bacteremia acquisition was associated with different distribution of pathogens, antimicrobial resistance, and clinical outcome. Traditional classification might overestimate the problem of drug resistance in community-acquired infections. The concept of health care-associated infection should be introduced to avoid overemphasis of drug-resistant problem in true community-acquired infection.