Patients with bacteraemia constitute an useful population for an audit of antibiotic treatments. Empirical antibiotic therapy (EAT) and its reassessment must take into account clinical data and microbiological results. Our aim was to determine the impact of these sequential steps of the therapy on survival. This was a retrospective multicentre study which included patients admitted to emergency departments (ED) for whom blood cultures were positive over a 4-month period. Microbial results were compiled from the database of the laboratories. The relevant information was extracted from the computerized patient's chart. An efficient EAT was based on antibiotic susceptibility of the bacteria. An effective antibiotic reassessment (AR) was defined as any modification of the EAT. Unfavorable outcome was defined as death of the patient during in-hospital care. Three hospitals and two clinics took part in this study, 169 patients with bacteraemia being included. The diagnosis in ED was undetermined in 21 cases (12%), 35 patients (21%) required intensive care, and 23 died (14%). One hundred and thirty-six patients (80%) received an EAT, the latter being efficient in 107 cases (63%). An effective AR was performed in 116 cases (69%). In multivariate analysis, risks factors for death were: ongoing cancer AOR (adjusted odds ratio) 3.34, undetermined diagnosis in ED: AOR 9.34 and severe sepsis or shock: AOR 6.98. Effective AR was a protective factor: AOR 0.28 [0.09-0.81]. One third of bacteraemic patients in ED did not benefit from AR. Improvement of antimicrobial stewardship should be associated with a higher rate of survival.
Keywords: Antibiotic reassessment; Bacteraemia; Emergency department; Empirical antibiotic therapy; Healthcare-associated infections.