# Artificial Intelligence for Medicines Information: Scoping Review of Clinical Applications and Digital Health Inequalities

## TL;DR

This scoping review explores how AI can support medicines information services, highlighting pharmacist roles and challenges like digital health inequalities.

## Contribution

The study provides a comprehensive map of AI applications in medicines information and identifies gaps in addressing digital health inequalities.

## Key findings

- AI tools show promise in medicines information but struggle with complex clinical queries.
- Pharmacists are most involved in evaluating AI-generated content.
- Digital health inequalities and misinformation risks hinder AI adoption.

## Abstract

Artificial intelligence (AI) has the potential to support medicines information services. However, a comprehensive mapping of its use, particularly within pharmacy practice and in the context of digital health inequalities, is lacking.

This scoping review mapped existing evidence on AI-driven medicines information, focusing on the accuracy and completeness of AI-generated content, the role of health care professionals (HCPs), particularly pharmacists, and the impact of digital health inequalities on AI adoption.

This scoping review was informed by the methodological framework proposed by Levac et al, which includes modifications to the original Arksey and O’Malley scoping review framework. A systematic search was conducted across MEDLINE (Ovid), PubMed Central, Cochrane Library, CINAHL Plus (EBSCOhost), International Pharmaceutical Abstracts (IPA), Web of Science, and Google Scholar from inception to January 2025, which served as the search cutoff date. Peer-reviewed studies in English evaluating the role of AI in medicines information across any health care settings (including patient homes) were included. The results are reported in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines.

A total of 1911 citations were identified, with 14 studies meeting the inclusion criteria. AI tools showed promise in supporting medicines information services but were found to have limitations in accuracy, particularly when applied to complex clinical queries. Pharmacists were the most engaged HCPs in the evaluation of AI-generated content. Only 3 studies explored digital health inequalities in the context of AI and access to medicines information. Reported barriers included misinformation risks, regulatory gaps, and digital health inequalities, particularly infrastructure limitations and disparities in digital literacy, which affected AI adoption.

AI-driven tools show promise in supporting medicines information services, but concerns remain. HCPs, particularly pharmacists, play a critical role in AI evaluation and validation, yet their involvement remains ill-defined. Addressing digital health inequalities is essential for effective AI integration. Future research should focus on identifying and minimizing digital health inequalities, as well as evidence-informed AI implementation in medicines information services.