Sendekie Ashenafi Kibret, Limenh Liknaw Workie, Abate Biruk Beletew, Chanie Gashaw Sisay, Kassaw Abebe Tarekegn, Tamene Fasil Bayafers, Gete Kalab Yigermal, Dagnew Ephrem Mebratu
Department of Clinical Pharmacy, School of Pharmacy, College of Medicine and Health Sciences, University of Gondar, Gondar, Ethiopia.
School of Pharmacy, Curtin Medical School, Faculty of Health Sciences, Curtin University, Bentley, WA, Australia.
Explor Res Clin Soc Pharm. 2024 Nov 19;16:100542. doi: 10.1016/j.rcsop.2024.100542. eCollection 2024 Dec.
Artificial intelligence (AI) has a significant potential to impact pharmacy practices worldwide. This study investigates pharmacists' perceptions of AI's role in pharmacy practices, their willingness to adopt it, and perceived barriers to its implementation at community pharmacies in Ethiopia.
A cross-sectional study was conducted among community pharmacists in Ethiopia. Data were collected using a self-administered questionnaire. Independent samples -test, one-way ANOVA, and post-hoc analyses were used to compare pharmacists' perception and willingness scores. A linear regression analysis examined the association of independent variables with pharmacists' perception of AI and willingness to utilize AI. A -value <0.05 was considered statistically significant.
Of 241 pharmacists approached, 225 (93.3 %) completed the survey. Overall, about two-thirds (67.1 % and 66.2 %) of community pharmacists had a high level of perception and willingness to use AI applications in pharmacy, respectively. Pharmacists with bachelor's degrees and above (β = 2.76: 95 % CI: 0.09, 5.01 vs. β = 1.79: 95 % CI: 0.05, 4.21), those who utilized scientific drug information sources (β = 2.45, 95 %: 0.17, 4.45 vs. β = 1.76, 95 % CI: 0.91, 3.89), pharmacists who had a previous exposure of AI (β = 1.02, 95 %: 0.03, 3.24 vs. β =1.13, 95 % CI: 0.07, 2.93), and those who with higher perceived AI knowledge (β =1.09, 95 % CI: 0.02, 2.46 vs. β = 1.14, 95 %CI: 0.17, 3.11) had significantly higher perception of AI and willingness to utilize it, respectively compared to their counterparts. Lack of internet availability (89.3 %), lack of AI-related software/hardware (88.2 %), and limited training (80.9 %) were the most frequently reported barriers by pharmacists to AI adoption. Over 90 % of pharmacists agreed on the importance of internet availability (93.3 %), policies/frameworks (91.6 %), and research/learning from others (89.3 %) for successful AI integration.
Despite positive perceptions and willingness from pharmacists, AI implementation in community pharmacies could be hindered by resource limitations, training gaps, skill constraints, and infrastructure issues. To facilitate adoption, enhancing knowledge and skills, and developing policies/frameworks are crucial.
人工智能（AI）在全球范围内对药学实践具有重大影响潜力。本研究调查了埃塞俄比亚社区药房药师对AI在药学实践中作用的看法、采用AI的意愿以及实施AI的感知障碍。
在埃塞俄比亚社区药师中开展了一项横断面研究。使用自填式问卷收集数据。采用独立样本t检验、单因素方差分析和事后分析来比较药师的认知和意愿得分。线性回归分析检验自变量与药师对AI的认知以及使用AI意愿之间的关联。P值<0.05被认为具有统计学意义。
在接触的241名药师中，225名（93.3%）完成了调查。总体而言，约三分之二（分别为67.1%和66.2%）的社区药师对在药学中使用AI应用有较高的认知度和意愿。拥有学士学位及以上的药师（β = 2.76：95%CI：0.09，5.01；相比之下β = 1.79：95%CI：0.05，4.21）、使用科学药物信息来源的药师（β = 2.45，95%：0.17，4.45；相比之下β = 1.76，95%CI：0.91，3.89）、曾接触过AI的药师（β = 1.02，95%：0.03，3.24；相比之下β = 1.13，95%CI：0.07，2.93）以及对AI知识认知较高的药师（β = 1.09，95%CI：0.02，2.46；相比之下β = 1.14，95%CI：0.17，3.11），分别比其同行对AI的认知度和使用意愿显著更高。药师报告的采用AI的最常见障碍是缺乏互联网接入（89.3%）、缺乏与AI相关的软件/硬件（88.2%）以及培训有限（80.9%）。超过90%的药师认同互联网接入（93.3%）、政策/框架（91.6%）以及向他人学习/研究（89.3%）对于成功整合AI的重要性。
尽管药师有积极的看法和意愿，但社区药房中AI的实施可能会受到资源限制、培训差距、技能限制和基础设施问题的阻碍。为促进采用，增强知识和技能以及制定政策/框架至关重要。
