Work

Knowledge, attitudes and practices regarding artificial intelligence among pharmacy students in Guinea and Malawi: a cross-national, multicentre cross-sectional study

Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Background Generative artificial intelligence (AI) has entered pharmacy education faster than curricula have adapted, yet evidence on how pharmacy students in sub-Saharan Africa understand and use it is almost absent. We assessed knowledge, attitudes and practices (KAP) regarding AI among pharmacy students in a Francophone (Guinea) and an Anglophone (Malawi) country, and identified their determinants. Methods A multicentre online cross-sectional survey (29 March–12 June 2026) included pharmacy students from three universities in Conakry, Guinea, and from Kamuzu University of Health Sciences, Malawi. A bilingual self-administered questionnaire measured knowledge (9 items), attitudes (8 Likert items) and practices (7 items, score 0–11). Prespecified cut-offs defined adequate knowledge (≥ 70%), favourable attitude (≥ 75%) and adequate practice (≥ 8/11). Multivariable logistic regression models, structured along the KAP pathway, estimated adjusted odds ratios (aOR) with 95% confidence intervals (CI). Results Of 322 participants (Guinea 115, Malawi 207; median age 23 years), 99.7% had heard of AI and 96.2% had used AI tools, but only 27.9% had received formal AI training. Adequate knowledge was found in 32.0% (95% CI 27.1–37.3), favourable attitudes in 30.2% (25.4–35.4) and adequate practices in 64.3% (58.9–69.3). Half of the students (49.8%) believed that AI always provides correct results, and only 18.9% always verified AI-generated information. The two countries showed mirror-image profiles: Malawian students knew AI applications better and were more favourable to AI (aOR 3.77; 1.79–7.97), whereas Guinean students more often recognised AI's limitations and verified its outputs. Adequate practice was independently associated with adequate knowledge (aOR 2.28; 1.26–4.13) and favourable attitude (aOR 2.57; 1.40–4.73), and inversely with age > 23 years (aOR 0.39; 0.21–0.74) and studying in Malawi (aOR 0.39; 0.19–0.81). Advanced years of study predicted adequate knowledge (≥ 5th year: aOR 4.62; 1.27–16.87), while self-reported formal AI training was not associated with better knowledge. Conclusions Pharmacy students in both countries use AI almost universally but lack the critical knowledge needed to use it safely, with a high risk of automation bias. Curricula should move beyond tool familiarity towards structured, competency-based AI literacy centred on output verification, limitations, ethics and pharmacist accountability, adapted to each national context.
Language: English
Field: Medicine
Open Access status: green
OpenAlex ID: W7236576624
Cites: 29
Cited by:
Related to: 10