•  
  •  
 

Article Type

Article

Abstract

Background: The misuse of antibiotics in the community is a significant health issue of concern since it directly leads to antimicrobial resistance, treatment failure, and health care burden. Aim: determine the trend of antibiotic misuse in community pharmacies and to use analytical models based on artificial intelligence. Methods: The current research was carried out in a sample of the community pharmacies in the city of Hilla, Babylon, Iraq, between November 2025 and February 2026. There were 200 encounters of antibiotics dispensing. The structured recording form was used as the data collection instrument to record the sociodemographic data, presenting symptoms, the presence of prescriptions, the class of antibiotics, the duration of treatment and other applicable dispensing variables. The classification of each encounter was based on predetermined rational-use criteria as appropriate or inappropriate. Results: The findings had indicated that 64.0% of antibiotic dispensing encounters were classified as inappropriate whereas 61.0% were dispensed without prescription. The most common pattern of misuse was dispensing without prescription and was the strongest independent predictor of inappropriate antibiotic use. Overuse was considerably linked with the presentation of symptoms in upper respiratory and the past history of self-medication. In logistic regression analysis, the non-prescription, self-medication history, and upper respiratory symptoms were also important predictors of misuse. All the models tested in the artificial intelligence analysis demonstrated an acceptable performance, with the majority of the models, though, demonstrating the best classification results, with the highest model being the Random Forest model, with an accuracy of 89, F1-score of 0.91, and ROC-AUC of 0.93, community pharmacies have a very high rate of antibiotic misuse that is largely affected by non-prescription dispensing, self-medication and demand based on symptoms. Conclusions: The results suggest that community pharmacies are an important address to the antimicrobial stewardship intervention. This study also reveals that artificial intelligence models have the potential to serve as a helpful tool in the recognition of irrational patterns of antibiotic dispensing.

Keywords

Antimicrobial misuse, Community pharmacies, Antimicrobial resistance, Artificial intelligence, Machine learning, Antimicrobial stewardship

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Download

Share

COinS