RESUMEN
BACKGROUND: Dengue is an important mosquito-borne tropical viral disease and dual infection, though rare, has been regarded as a risk factor for severe disease and mortality. However, few studies focused on bloodstream infections (BSIs) and empirical antibiotic therapy rarely addressed. METHODS: Dengue patients with concurrent or subsequent BSIs between July 1 and December 31, 2015 were included. Clinical information, laboratory data, and drug susceptibility data were collected. RESULTS: Totally 80 patients, with an in-hospital mortality rate of 32.5%, were included and categorized into three groups. 32 patients in Group I (BSI onset within 48 h after admission), 32 in Group II (between 48 h and one week), and 16 in Group III (more than one week). Patients in Group I were older (mean age: 75.6 vs. 72.6 or 69.6 years; P = 0.01) and had a higher Charlson comorbidity index (3.1 vs. 1.8 or 1.9; P = 0.02) than those in Group II or III. Streptococcus species (28.9%, 11/38) and Escherichia coli (23.7%, 9/38) were major pathogens in Group I. Enterobacteriaceae (38.2%, 13/34) isolates predominated in Group II. Fatal patients more often received inappropriate empirical antibiotic than the survivors (61.5% vs. 35.2%; P = 0.03). According to susceptibility data, pathogens in Group I and II shared similar susceptibility profiles, and levofloxacin, cefepime, or piperacillin/tazobactam, can be empirically prescribed for those hospitalized within one week. CONCLUSIONS: BSI pathogens vary among dengue patients. For adults with dengue and suspected BSI hospitalized within one week, empirical antimicrobial agents are recommended.