Monitoring community antibiotic consumption in Belgium: reimbursement versus retail data (2013-22)

In Belgium, monitoring antibiotic consumption relies on reimbursement data, which is obtained with a time delay and does not account for over-the-counter or nonreimbursed products. This study aims to bridge this gap by comparing reimbursement and retail data for primary care to understand variations and assess the accuracy of current surveillance methods. Reimbursement data were obtained from the National Institute for Health and Disability Insurance, and retail data were obtained from IQVIA for the period 2013-22. The community consumption of systemic antibiotics was expressed in defined daily doses (DDD-WHO ATC/DDD Index 2023) per inhabitants per day (DID). Relative differences in DID (RDs) based on the two data sets were computed and validated through Bland-Altman plots and correlation analysis. The sales of antibiotics declined from 22.89 DID (2013) to 20.50 (2022), with a steep drop during the COVID-19 pandemic-from 21.31 DID in 2019 to 16.55 DID in 2020-and a subsequent rebound. Reimbursement data slightly underestimated consumption compared to retail data, with RDs ranging from 2% (2013) to 9% (2022) when including quinolones and from 2% to 4% when excluding them. Bland-Altman plots showed high agreement between reimbursement and retail estimates, identifying quinolones as outliers. Our findings suggest that reimbursement data are generally reliable for monitoring antibiotic consumption, but incorporating retail data is crucial for accurate assessments. The use of retail data can facilitate timely interventions and inform public health strategies to effectively address antimicrobial resistance.