Performance of a clinical decision-support tool on empirical antibiotic treatment in the emergency department: a retrospective study.

BACKGROUND AND OBJECTIVES: Antimicrobial resistance remains a global health challenge despite the WHO AWaRe framework promoting responsible antibiotic use. Adherence to clinical guidelines, particularly during initial empirical prescribing in the emergency department (ED), is crucial. The clinical decision-support system (CDSS), integrating clinical and laboratory data, aims to improve evidence-based therapy and guideline adherence. To evaluate whether a CDSS can improve adherence to antibiotic guidelines in ED patients with suspected infections and its effect on prescribing according to WHO AWaRe classification. METHODS: We retrospectively analysed 756 adults with confirmed infection from a multicentre cohort study in Southern Denmark. Expert diagnoses were mapped to the CDSS TREAT-Essential, which recommended empirical antibiotics based on illness severity, local guidelines and patient-specific factors. Empirical therapy was defined as antibiotics administered within 4 hours of ED arrival. Outcomes included guideline adherence and division in relation to WHO AWaRe classification. Adherence was defined as either CDSS-recommended or physician-prescribed antibiotics according to each patient's diagnosis. Data were compared using McNemar's χ 2 and Wilcoxon signed-rank tests. RESULTS: Clinicians prescribed empiric antibiotics to fewer patients than recommended by the CDSS or guidelines (P < 0.001), fewer combination therapies (P < 0.001) and more broad-spectrum antibiotics (P < 0.001). For the four most common diagnoses, the CDSS showed higher full guideline adherence (P < 0.001) and more frequently recommended Access agents. CONCLUSION: The CDSS demonstrated higher adherence to antibiotic guidelines than clinicians and greater alignment with WHO AWaRe classification. Further research is needed to evaluate prospective implementation and clinical impact in the ED.