A model for analysis of antibiotic usage in low-income settings.

OBJECTIVES: Quantification of antibiotic usage is an important component of antimicrobial stewardship programmes. We aimed to estimate institutional antibiotic usage and costs using methodology and metrics applicable to low-income settings without electronic health records. METHODS: The DDD per 100 patient-days (DDDs/100 PDs) of antibiotics used in a calendar year was calculated retrospectively from ward registers and inpatient drug records in general surgical wards of a tertiary hospital. The antibiotics were categorized using the Access, Watch, Reserve classification. The annual expenditure on antibiotics was estimated from price lists of the state medication procurer. RESULTS: Annual usage of IV co-amoxiclav, cefuroxime and metronidazole was significantly higher than other antibiotics and certain wards showed outlier use of the same. The IV formulations of co-amoxiclav (5-fold), metronidazole (3-fold) and ciprofloxacin (2-fold) were used in excess of the oral formulation. Proportionate antibiotic usage based on the AWaRe category did not vary significantly between wards. Two wards were outliers for annual expenditure/100 PDs. IV clindamycin and meropenem combined accounted for 43.8% of expenditure on antibiotics. CONCLUSIONS: This study demonstrated intra-institutional variations of annual antibiotic usage and related costs. The metric DDD/100 PDs and the methodology used here are suitable for intra- and inter-institutional analyses of antibiotic usage, particularly in low-income settings.