4CPS-320 The role of artificial intelligence in the control of antibiotic resistance

Abstract

<h3>Background and Importance</h3> Antibiotic resistance is a phenomenon associated with the inappropriate use of antibiotics, whereby microorganisms acquire resistance to the pharmacological mechanism of action, making the resulting infections more difficult to treat. To address this issue, the World Health Organization (WHO) has released the AWARE manual, which classifies antibiotics into three categories: Access (antibiotics with a narrow-spectrum of activity and a favourable safety profile, recommended for the treatment of common infections), Watch (broad-spectrum antibiotics reserved for specific clinical scenarios) and Reserve (antibiotics used as a last-resort for infections caused by multidrug-resistant organisms). <h3>Aim and Objectives</h3> The general objective of this study is to counteract antibiotic resistance by promoting appropriate prescribing practices through the adoption of targeted models and the implementation of effective control measures. The specific objective is to reduce bacterial resistance and healthcare-associated infections (HAIs) by encouraging the use of artificial intelligence (AI) by clinicians and pharmacists as a decision support tool aimed at minimising prescribing errors. <h3>Material and Methods</h3> In the last quarter of 2025, the prescribing trends of antibiotics such as piperacillin/tazobactam and ceftriaxone were evaluated in the general surgery and internal medicine departments during the period prior to the implementation of antibiotic request forms. The analysis revealed an increase in prescriptions for pneumonia and sepsis, with a total of 45 requests for piperacillin/tazobactam and 121 for ceftriaxone. The objective was to assess prescribing appropriateness and the corresponding antibiogram in collaboration with clinicians, in order to prevent resistance phenomena and inappropriate prescriptions. <h3>Results</h3> Following the implementation of these prescription request protocols, a significant reduction of approximately 50% in prescriptions for the two antibiotics evaluated was observed relative to the preceding period. In a near-future scenario, the integration of artificial intelligence in diagnostic processes is projected to result in a 25.4% decrease in antibiotic prescriptions, thereby mitigating the issue of overprescribing. <h3>Conclusion and Relevance</h3> Digital platforms and mobile applications represent the future for enhancing therapeutic decision-making; however, they will never replace the decisions made by the Infection Control Team (CIO) in ensuring the most appropriate therapeutic approach. <h3>Conflict of Interest</h3> No conflict of interest

Ähnliche Arbeiten