6ER-006 Fact-finding survey on the use of artificial intelligence in hospital pharmacy practice

Abstract

<h3>Background and Importance</h3> Artificial intelligence (AI) represents a revolutionary new technology with a wide scope of application, including healthcare and hospital pharmacy. However, AI models use in clinical and pharmaceutical practice is facing critical challenges, i.e. data privacy, cybersecurity threats, potential algorithm biases, and ethical concerns. <h3>Aim and Objectives</h3> The aim of the study was to investigate whether and how AI models are currently employed in hospital pharmacy practice, analysing also educational aspects and hospital pharmacists’ (HPs) attitudes to more consistent AI use in pharmacy practice. <h3>Material and Methods</h3> An anonymous questionnaire, consisting of 11 multiple-choice items, was developed and made available online for HPs and resident HPs in hospital pharmacy. A total of 65 respondents answered the questionnaire in a 3 month period, from June 2025 to August 2025; the data were extrapolated and analysed using a spreadsheet. <h3>Results</h3> Results showed that 56/65 of respondents use AI models in daily hospital pharmacy practice (41/65 use generative AI; 14/65 use predictive AI); however, only 4/65 attended educational activities concerning AI before use. A relevant percentage of respondents have knowledge of two critical issues concerning AI models, i.e. ‘AI hallucination’ phenomenon (59/65) and incomplete reliability of AI outputs due to inaccuracies, biases, or inconsistencies in training datasets (51/65). The survey revealed that 55/64 of respondents verify AI conclusions or predictions. According to the survey, the respondents, who principally employ AI tools that use text as input (like ChatGPT; 53/63), defined their AI use as: ‘occasional’ (31/65), ‘moderate’ (14/65), ‘regular’ (13/65). The survey revealed the principal uses of AI tools include: data analysis and elaboration (25/62), search for characteristics of pharmaceutical products (25/62) and medical device (13/62), translation of scientific texts (21/62). Moreover, 44/62 of respondents are not worried about the risk of decline in healthcare employment due to increase of AI use and express confidence in the regulatory processes. <h3>Conclusion and Relevance</h3> A substantial percentage of respondents (61/65) are favourably disposed to educational activities that aim to guide and support pharmacists on the effective and ethical use of AI technologies. Although the critical challenges, the integration of AI systems into healthcare could represent a huge opportunity to improve clinical and pharmaceutical practice. <h3>Conflict of Interest</h3> No conflict of interest

Ähnliche Arbeiten