GENERATIVE AI-ASSISTED DRUG-DRUG INTERACTION CASE SUPPORT AND PHARMACY STUDENTS' COMPETENCE: A MIXED-METHODS STUDY.

BACKGROUND

Drug-drug interactions (DDIs) significantly contribute to preventable adverse drug events and constitute a vital skill domain in pharmacy education. Recent advancements in generative artificial intelligence (GenAI) have created novel opportunities for enhancing clinical reasoning and medication safety training; however, evidence regarding their educational effectiveness remains limited.

OBJECTIVE

To evaluate the effect of a generative AI-assisted DDI alert tool on pharmacy students' competence in managing complex DDI cases and to explore students' perceptions of AI-supported clinical reasoning.

METHODS

A comparative mixed-methods study was conducted among advanced-level undergraduate pharmacy students in Saudi Arabia during the 2025 academic year. Participants were allocated to either an AI-assisted group (n=96) or a non-AI group (n=88). Both groups completed identical case-based assessments involving high-risk DDIs. Objective competence was assessed using a standardized analytic rubric evaluating DDI identification, mechanism assessment, clinical management, and therapeutic justification. Quantitative data were analyzed using descriptive and inferential statistics, while qualitative responses were examined using thematic analysis.

RESULTS

A total of 184 students completed the study. Compared with the non-AI group, students in the AI-assisted group reported significantly higher scores in understanding DDI mechanisms (p=0.0089), confidence in clinical decision-making (p=0.0028), perceived appropriateness of challenge level (p=0.0050), and overall satisfaction (p=0.0004). Objective competence scores were significantly higher among AI-assisted students across all assessed domains, including DDI identification (p=0.008), mechanism assessment (p=0.002), clinical management (p=0.005), and therapeutic justification (p=0.004). The composite competence score was also significantly higher in the AI-assisted group (8.75±1.09 vs. 8.10±1.46; p=0.001). Qualitative analysis identified three overarching themes: clinical relevance and case realism, generative AI as a cognitive support tool, and the need for clearer assessment guidance and additional clinical detail.

CONCLUSION

Generative AI-assisted DDI support was associated with improved competence, confidence, and overall assessment experience among pharmacy students managing complex clinical cases. The findings suggest that generative AI can serve as an effective educational scaffold that enhances clinical reasoning and therapeutic decision-making when integrated within structured learning environments and supported by appropriate faculty oversight.