Tehran University of Medical Sciences Journal of Pharmaceutical Care 2322-4630 13 4 2025 12 31 260 268 Mehrshad Ebrahimpour Department of Clinical Pharmacy, School of Pharmacy, Mashhad University of Medical Sciences, Mashhad, Iran Ayoub Tavakolian Department of Emergency Medicine, School of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran. Amirhossein Malaekeh-Nikouei Department of Clinical Pharmacy, School of Pharmacy, Mashhad University of Medical Sciences, Mashhad, Iran Zahra Ebnehoseini Department of Medical Informatics, Psychiatry and Behavioral Sciences Research Center, Mashhad University of Medical Sciences, Mashhad, Iran. Hassan Vakili Arki Department of Medical Informatics, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran Sepideh Elyasi 2025 12 07 2026 05 31 Background: Drug interactions present significant challenges in emergency departments due to the diverse patient population and the complexity of medical conditions. These interactions threaten patient safety and increase healthcare costs. Strategies such as Clinical Decision Support Systems (CDSS) and audit-and-feedback mechanisms are essential for alerting physicians to potential medication risks. Methods: A quasi-experimental before-and-after study was conducted in the emergency department of Ghaem Hospital in Mashhad. Phase 1 (June–August 2024) retrospectively analyzed drug interactions for 482 patients using Lexidrug 2024. To ensure comparability and minimize ascertainment bias, the Phase 1 analysis was restricted to interactions involving the 150 most commonly used medications in the emergency ward, which were included in the Phase 2 software. Phase 2 (October–December 2024) assessed 482 patients using a locally developed software (Idea Pardazan Asr Hoshmand Etminan) combined with a weekly audit-and-feedback intervention sent to physicians. Major interactions, classified as Category X (life-threatening) and Category D (clinically significant), were compared using the Chi-square test in SPSS version 12. Results: Phase 1 recorded 42 Category X interactions (8.7%) and 118 Category D interactions (24.5%). In Phase 2, Category X interactions decreased to 32 (6.6%), but this reduction was not statistically significant (P = 0.276). However, Category D interactions significantly decreased to 45 (9.3%) in Phase 2 (P < 0.01). There were no significant differences in interaction rates between academic and non-academic specialists. Conclusion: The implementation of the software, combined with weekly feedback reports, significantly reduced clinically significant (Category D) interactions. However, the lack of a statistically significant reduction in life-threatening (Category X) interactions suggests that weekly feedback may be insufficient for critical alerts requiring immediate action. Future iterations should focus on real-time integration into clinical workflows to address urgent patient safety threats. https://jpc.tums.ac.ir/index.php/jpc/article/view/922 https://jpc.tums.ac.ir/index.php/jpc/article/download/922/439