Pharmacists versus artificial intelligence: Bridging the clinical evidence gap

Artificial intelligence (AI) is increasingly being used in healthcare practice to support diagnostic interpretation, triage, and clinical documentation. However, evidence demonstrating its effect on clinical outcomes remains limited. A randomized clinical trial published in 2024 found that access to a large language model did not significantly improve physicians’ diagnostic accuracy compared with standard tools,1 while other studies have shown measurable gains in efficiency, such as faster and higher-quality radiology reporting.2 These findings suggest that, although AI can streamline workflow, its contribution to clinical accuracy and patient outcomes is uncertain. For pharmacists, the evidence base is even more limited. Pharmacists routinely make high-stakes decisions that directly influence patient safety, including performing antimicrobial stewardship, chemotherapy dose adjustment, and management of drug-drug interactions. Despite growing interest in AI to assist in these tasks, very few studies have evaluated its real-world impact. A 2025 review summarized healthcare applications of AI, including early experiences in pharmacy practice.3 However, nearly all studies included were observational or descriptive, and few measured patient-oriented outcomes such as error reduction, therapeutic appropriateness, and time efficiency.