RESUMO
BACKGROUND: Assessment of antimicrobial use places an emphasis on therapeutic aspects of infected patients. Our aim was to determine the risk factors for unnecessary antibiotic therapy (UAT). METHODS: This was a prospective, multicenter study evaluating all curative antibiotic therapies prescribed over 2 consecutive days through the same electronic medical records. Each item that could participate in these prescriptions was collected from the computerized file (reason for hospitalization, comorbid conditions, suspected or definitive diagnosis of infection, microbial analyses). UAT was defined as the recognition of noninfectious sydromes (NIS), nonbacterial infections, use of redundant antimicrobials, and continuation of empirical broad-spectrum antimicrobials. RESULTS: Four hundred fifty-three antibiotic therapies were analyzed at 17 institutions. An infectious disease was the reason for hospitalization in 201 cases (44%). An unspecified diagnosis of infection was observed in 104 cases (23%). Microbial samples were taken in 296 cases (65%), allowing isolation of a pathogen in 156 cases (53%). Unspecified diagnosis was associated with the absence of a microbial sample compared to patients with a diagnosis: (56/104 [54%] vs 240/349 [69%]; P = .005). A total of 158 NIS were observed (35%). UAT was observed in 169 cases (37%), due to NIS in 106 cases. In multivariate analysis, the modifiable risk factors for UAT were unspecified diagnosis (adjusted odds ratio [AOR], 1.83; 95% confidence interval [CI], 1.04-3.20) and absence of a blood culture (AOR, 5.26; 95% CI, 2.56-10.00). CONCLUSIONS: UAT is associated with an unspecified diagnosis and the absence of microbial testing. Antimicrobial stewardship programs should focus on diagnostic difficulties and microbial testing, the latter facilitating antibiotic reassessment and therapeutic interruption.