Antibiotic Stewardship Gaps at adult-ED: Determinants and Predictors of Inappropriate Prescribing at Discharge (I-PAD Study)

BACKGROUND The appropriateness of antibiotic prescribing at Emergency Department (ED) discharge is a major concern, with reported inappropriateness rates exceeding 50% and implications for patient safety and antimicrobial resistance. Inappropriate prescribing is common in urinary tract infections (UTIs), community-acquired pneumonia (CAP), and skin and skin structure infections (ABSSSIs), and is driven by patient-, clinician-, and system-level factors. This study aimed to assess determinants and predictors of inappropriate prescriptions at ED discharge.METHODS A retrospective observational study was conducted in the general EDs of the University Hospital of Modena between 1 January 2024 and 21 January 2025. All adult patients discharged home with at least one oral antibiotic prescription were included. The analysis focused on CAP, mild non- purulent ABSSSIs and uncomplicated UTIs. Predictors of inappropriate prescribing were identified using multivariate logistic regression. RESULTS During the study period, 30,715 ED discharges were recorded, 13% of which resulted in an antibiotic prescription at discharge. Overall, 3,993 antibiotic prescriptions were reviewed; 2,156 (54%) were inappropriate. Among the 1,837 prescriptions deemed appropriate, 606 (33%) concerned infections outside the predefined target syndromes, while 1,230 (67%) were prescribed for CAP, cystitis, or mild erysipelas/cellulitis. In this target cohort, 728 prescriptions (59.2%) were considered inappropriate according to GL, whereas 502 (40.8%) were guideline-concordant . Cystitis was the most frequent diagnosis, followed by CAP and erysipelas/cellulitis. Guideline-concordant prescribing differed across syndromes, being lowest in cystitis (31.3%), intermediate in CAP (39.5%), and highest in erysipelas/cellulitis (72.0%). The population had a mean age about 59 years and a balanced sex distribution. Adherence to guideline-recommended antibiotic selection was 62.11%, whereas compliance with recommended duration was lower (57.7%), confirming duration as a major source of error. In multivariate analysis, male sex, higher CRP, cephalosporin use, and combination therapy were associated with increased odds of inappropriate prescribing. Conversely, CAP, erysipelas/cellulitis, and amoxicillin-clavulanate use were associated with lower odds. Cephalosporins were the strongest predictor of non-guideline-concordant therapy. During follow-up, ED reattendance occurred in 11.7% and infection-related mortality in 5.4%. Overall appropriateness was not significantly associated with ED reattendance or mortality, although cephalosporin use was independently associated with ED readmission. CONCLUSION Inappropriate antibiotic prescribing at ED discharge remains a substantial AMS challenge, affecting more than half of prescriptions. CAP, cystitis, and erysipelas/cellulitis were the target syndromes managed with oral antibiotics and represent key areas for intervention. Prescribing gaps were mainly related to syndrome-specific patterns, antibiotic class selection, and treatment duration rather than baseline instability. Cystitis emerged as the most critical syndrome, particularly because of cephalosporin and fluoroquinolone use, while erysipelas/cellulitis showed the highest guideline adherence. Future AMS initiatives should prioritize duration optimization, reduction of inappropriate cephalosporin and fluoroquinolone prescribing, assessment of combination therapy, and decision support at discharge.