Inappropriate antibiotic prescribing in Ethiopian hospitals: a systematic review

INTRODUCTION: Inappropriate antibiotic prescribing is a major driver of antimicrobial resistance globally, and its consequences are more severe in low- and middle-income countries such as Ethiopia. Although antibiotics are commonly prescribed for inpatients in Ethiopian hospitals, evidence on their appropriateness remains limited and inconsistent. AIM: This systematic review aimed to assess the prevalence, prescribing characteristics, and factors associated with inappropriate antibiotic prescribing in Ethiopian hospitals. METHOD: We searched MEDLINE, Embase, Scopus, Web of Science from inception to 4 May 2025, and also searched Google Scholar and manually reviewed the reference lists of eligible studies. Primary studies conducted in Ethiopian hospitals that reported on the appropriateness of inpatient antibiotic prescribing were included. Two authors independently screened studies and extracted data. Methodological quality was assessed using Joanna Briggs Institute critical appraisal tools. Because of the heterogeneity in study population, definitions and outcome measures, findings were synthesised descriptively. The review was registered in PROSPERO (CRD420251051881). RESULTS: Fifty-three studies were included from seven Ethiopian regions, most from Amhara (n = 16). Overall, study quality was variable, with common limitations including unclear sampling and incomplete methodological reporting. Most studies focused on adult patients (n = 24) and used the Ethiopian Standard Treatment Guideline (n = 32). Cephalosporins were most frequently assessed (n = 45), ceftriaxone the most commonly evaluated antibiotic (n = 44). Prophylactic and empiric prescribing were commonly reported (n = 29 and n = 28) respectively. The prevalence of inappropriate prescribing reported per patient was a median of 56.8% (range, 10.2% to 91.4%) in 43 studies, and per prescription was a median of 40.7% (range, 19.2% to 66.9%) in nine studies. Indication, dose, dosing frequency, and duration were the prescribing characteristics most frequently reported. Patient-related factors, including comorbidity and polypharmacy were frequently associated with inappropriate prescribing. CONCLUSION: Inappropriate antibiotic prescribing was common in Ethiopian hospitals, although prevalence varied widely. Inconsistencies in guideline use and prescribing criteria likely contribute to this variability. Frequent empiric and prophylactic prescribing and reliance on ceftriaxone may reflect limited diagnostic capacity and raise concerns about antimicrobial resistance. Strengthening antimicrobial stewardship through standardised prescribing audits and improved diagnostic services is urgently needed to support appropriate antibiotic use in Ethiopia.