# Artificial Intelligence Readiness Among Young Family Doctors in Europe *[Seyma Handan Akyon](https://orcid.org/0000-0002-2288-8915), MD, Family Medicine Specialist1* *[Ikbal Humay Arman,]( https://orcid.org/0000-0003-1492-6763 ) MD, Family Medicine Specialist2* *[Gül¸sah Onur](https://orcid.org/0000-0003-1024-5258), MD, Family Medicine Specialist3* *[Rita Abecasis](https://orcid.org/0000-0001-9360-9293), MD4,5* *[Pemra Cobek Unalan](https://orcid.org/0000-0003-3446-7082), MSc, MD6* Sincan Education and Research Hospital, Ankara, Türkiye Istanbul Medipol University, School of Medicine, Department of Family Medicine, Istanbul, Türkiye 3 Hacettepe University, Faculty of Medicine, Department of Family Medicine, Ankara, Türkiye 4 Primary Health Care Area, Unidade de Saúde Familiar da Baixa, Unidade Local de Saúde São José, Lisbon, Portugal Centro Clínico Académico de Lisboa, Lisbon, Portugal Marmara University, School of Medicine, Department of Family Medicine, Istanbul, Türkiye, The European General Practice Research Network (EGPRN) ![](https://assets.reachrx.ai/images/10.1370%E2%98%8Dafm.250045/_page_0_Picture_12.jpeg) *Conflicts of interest: authors report none.* #### **CORRESPONDING AUTHOR** Seyma Handan Akyon Sincan Education and Research Hospital Gökçek, 250 Cadde 2/A, 06949 Ankara, Türkiye [drseymahandan@gmail.com](mailto:drseymahandan%40gmail.com?subject=) #### **ABSTRACT** We aimed to assess family doctors' readiness toward artificial intelligence (AI) in health care. We conducted a descriptive study using snowball sampling, targeting young family doctors. We administered a web-based questionnaire incorporating the validated medical AI readiness scale (MAIRS) via digital networks. A total of 134 participants from 20 European countries were included. MAIRS scores reflected variable readiness levels, with a median total of 69 of 110. Participants' readiness for AI increased significantly with their knowledge of its current applications and knowledge of AI usage in health care (*P*<.001, *P*=.013, respectively). The limited AI readiness highlights the need for tailored training and collaboration to support effective integration into primary care. Limitations of this study include sampling bias, small sample size, and lack of formal power analysis. *Ann Fam Med* 2025;23:542-545. ## **INTRODUCTION** I ntegrating artificial intelligence (AI) in primary care can improve efficiency, support personalized care, and optimize resource allocation.1 However, AI adoption requires training of primary care physicians (PCPs), ethical oversight, and measures to prevent worsening health inequalities.2 Understanding PCPs' AI readiness is crucial for effective integration.3 This study examined AI readiness among young European family physicians using the validated medical AI readiness scale (MAIRS),4 providing the first standardized evaluation across multiple countries. Our aim was to inform curriculum development, ensuring that future PCPs acquire essential AI skills. # **METHODS** In this quantitative, observational, analytical study, participants were young PCPs from the World Organization of Family Doctors (WONCA) Europe member countries who were trainees or specialists in their first 5 years of practice. We recruited participants via European Young Family Doctor Movement (EYFDM) network country representatives and digital channels. Given that the total number of young PCPs active in the EYFDM is unknown, we did not conduct a power analysis to determine the minimum required sample size; therefore, we used snowball sampling. Inclusion criteria included proficiency in English and active practice in family medicine across Europe. Participants were asked to complete a web-based survey, and we gathered sociodemographic data and explored participants' knowledge and usage of AI in health care. We used the 22-item MAIRS, a 5-point Likert-type scale measuring readiness across 4 dimensions (cognition, ability, vision, and ethics), was used. The Marmara University School of Medicine Clinical Research Ethics Committee approved the study (22.04.2024.368) in June 2024. Data were collected via digital networks from August to September 2024. Participation was voluntary, and informed consent was obtained. We analyzed data using IBM SPSS Statistics 27. Descriptive statistics included frequencies, percentages, medians, and interquartile ranges (IQRs). We used the Kolmogorov-Smirnov test to assess normality and conducted Mann-Whitney *U*, Kruskal-Wallis, and Spearman ρ correlation analyses. Statistical significance was determined using a significance level of .05 and a power level of 80%. ![](https://assets.reachrx.ai/images/10.1370%E2%98%8Dafm.250045/_page_0_Picture_25.jpeg) ## **RESULTS** The study included 134 participants from 20 European countries. The median age was 31 years (IQR=8.25 years), and the median graduation year was 2018 (IQR=6 years). Participants were from Türkiye (19.4%), Romania (14.9%), Slovakia (13.5%), Portugal (11.9%), and Spain (10.5%), with other countries, including the United Kingdom, Poland, Belgium, Croatia, Israel, Malta, Bosnia and Herzegovina, France, Italy, Netherlands, Czechia, Estonia, Ireland, Lithuania, and Luxembourg, contributing <5% of participants each. Median MAIRS subcategory scores were as follows: cognition: 23 (IQR=11)/40, ability: 26 (IQR=10)/40, vision: 10 (IQR=5)/15, ethics: 9.5 (IQR=4.25)/15. The total median MAIRS score was 69 (IQR=23)/110. **[Figure 1](#page-1-0)** presents the 22-item MAIRS score visual data. **[Table 1](#page-2-0)** shows associations and correlations between MAIRS scores and participants' characteristics, knowledge, and usage of AI. Participants' readiness for AI increased significantly with their knowledge of its current applications in health care (*P*<.001) and knowledge of AI usage in health care (*P*=.013) (**[Table 1](#page-2-0)**). We found greater AI readiness among male participants (*P*=.013) but no significant differences in terms of age, training status, or experience with AI. # **DISCUSSION** Participants from 20 European countries reflected diverse primary care perspectives among internationally engaged young PCPs. Contrary to prior studies showing no gender differences, our findings indicated greater AI readiness among male participants (*P*=.013), consistent with research studying medical students.5,6 Age, training level, and experience with AI did not significantly influence this population's AI readiness. Our findings indicated moderate AI readiness, with a median total MAIRS score of 69/110, emphasizing variability across demographic and knowledge-based factors. Participants' MAIRS scores aligned with prior studies involving medical students, indicating similar AI readiness levels among family doctors and undergraduate medical students.6,7 No previous studies evaluating PCPs' AI readiness were identified. Participants with greater knowledge of current AI applications in health care showed significantly greater AI readiness. A notable correlation existed between readiness and AI usage in family medicine practice. Although medical students worldwide show enthusiasm toward AI integration, they often report insufficient learning during their education, leading to a lack of confidence and understanding of the need to use these tools effectively.7 Therefore, there is a clear need for targeted education and training to equip PCPs with the necessary skills to use AI effectively. We also found that AI usage in family medicine remains low; approximately one-quarter did not use AI, one-third rarely used it, and frequent use was minimal (**[Table 1](#page-2-0)**).8 In the literature, barriers include insufficient training, accuracy concerns, and effects on patient-doctor relationships, underscoring the need for structured AI education.9 More than one-half of participants lacked formal AI exposure; those with prior AI training or greater knowledge of AI showed significantly greater readiness (**[Table 1](#page-2-0)**). Attending AI-related events also boosted scores, highlighting the importance of tailored programs on accountability, ethics, legality, and patient safety to ensure responsible, efficient use of AI in practice.3,10 This study has several limitations. The convenient snowball sampling method likely introduced selection bias, favoring digitally and internationally engaged young PCPs and potentially overestimating AI readiness. Nonresponse bias is also a concern because those less interested in AI might have opted out. Country representation was uneven, with 70% of participants from 5 countries, generating country-level bias. ![](https://assets.reachrx.ai/images/10.1370%E2%98%8Dafm.250045/_page_1_Figure_11.jpeg) However, we did not aim for country-specific comparisons, given that small sample sizes per country limited statistical power. Language barriers also limited participation, favoring respondents fluent in English. Interpretation of AI readiness was limited because potential confounding factors, such as prior technical history and clinical workload, were not assessed. Finally, self-reported data might have introduced bias, given that participants could overstate their familiarity **Table 1. Medical AI Readiness Score Association and Correlation With Participant Characteristics, Knowledge, and Usage of AI** | | No. (%) | Median | IQR | P valuea-d | Correlation | P valuee | |---------------------------------------------------------------------------------------------------------|-----------|--------|-------|--------------|--------------------|----------| | Sex | | | | | | | | Female | 94 (70.1) | 66.5 | 23.5 | .013a | −0.216
(female) | .012 | | Male | 40 (29.9) | 75 | 19.75 | | | | | Age, y | | | | | | | | <30 | 49 (36.6) | 67 | 20.5 | .693b | 0.015 | .86 | | 30-39 | 65 (48.5) | 67 | 27.5 | | | | | ≥40 | 20 (14.9) | 73 | 31.25 | | | | | Family medicine training status | | | | | | | | Trainee | 65 (48.5) | 72 | 21.5 | .755b; .478c | −0.047 | .591 | | Specialist | 64 (47.8) | 67 | 25.25 | | | | | Working in family medicine/general practice settings | 5 (3.7) | 67 | 34.50 | | | | | Level of knowledge of current applications of AI in health care | | | | | | | | No knowledge | 11 (8.2) | 59 | 24 | <.001b | 0.504 | <.001 | | Insufficient | 49 (36.6) | 57 | 23.5 | | | | | Neutral | 37 (27.6) | 75 | 17 | | | | | Somewhat sufficient | 34 (25.4) | 78 | 18 | | | | | Very sufficient | 3 (2.2) | 93 | 9.5 | | | | | Level of knowledge of AI usage in health care | | | | | | | | No knowledge | 14 (10.4) | 56 | 23.75 | .013b | 0.307 | <.001 | | Only follow developments (television, newspapers,
magazines, internet, etc) | 56 (41.8) | 65 | 24.5 | | | | | Attended ≥1 class (and/or presentations, congresses,
conferences, panels, symposiums, seminars, etc) | 52 (38.8) | 74.5 | 23.5 | | | | | Involved in ≥1 research project | 7 (5.2) | 76 | 14 | | | | | Presented papers, publications, etc | 5 (3.7) | 78 | 24.50 | | | | | Experience with AI usage in family medicine practice | | | | | | | | Never | 35 (26.1) | 59 | 29 | .192b | 0.200 | .02 | | Rarely | 51 (38.1) | 69 | 24 | | | | | Occasionally | 31 (23.1) | 72 | 15 | | | | | Frequently | 12 (9.0) | 75.5 | 24 | | | | | Very frequently | 5 (3.7) | 84 | 36.5 | | | | | Attended any training, course, seminar, or presentation on use of AI in health services | | | | | | | | Yes | 53 (39.6) | 74 | 20 | .065b; .02d | −0.170 | .049 | | No | 75 (56.0) | 65 | 27 | | | | | Not sure (n=6) | 6 (4.5) | 68 | 33 | | | | AI=artificial intelligence; IQR=interquartile range. Independent samples Mann-Whitney *U* test. b Independent samples Kruskal-Wallis test. Mann-Whitney *U* test for only trainees and specialists. d Mann-Whitney *U* test for only Yes or No. Spearman ρ correlation. with AI. These factors might affect the generalizability of the findings, primarily reflecting the perspectives of young family doctors actively engaged in professional networks rather than the entire community. Nevertheless, this study offers novel insights into AI readiness among young European family doctors using a standardized measurement scale with established validity and reliability. As the first standardized evaluation using MAIRS, the findings reveal limited readiness and adoption, underscoring the urgency for comprehensive AI training, collaborative initiatives, and tailored curricula. Equipping future PCPs with essential skills for ethical and effective AI integration is crucial to meet the evolving demands of health care. ![](https://assets.reachrx.ai/images/10.1370%E2%98%8Dafm.250045/_page_3_Picture_3.jpeg) #### **[Read or post commentaries in response to this article.](https://www.annfammed.org/content/23/6/542/tab-e-letters)** **Key words:** primary care; artificial intelligence; health services; digital technology Submitted January 19, 2025; submitted, revised, March 22, 2025; accepted July 28, 2025. **Previous presentation:** 29th WONCA (World Organization of Family Doctors) Europe Conference; September 25-28, 2024; Dublin, Ireland. **Acknowledgments:** This study was presented at the WONCA Europe conference in Dublin, Ireland, in September 2024; we appreciate the platform to share our findings. We thank European Young Family Doctors Movement (EYFDM) members for their active participation and valuable contributions. #### **References** 1. Lin SY, Mahoney MR, Sinsky CA. Ten ways artificial intelligence will transform primary care. *J Gen Intern Med*. 2019;34(8):1626-1630. doi:[10.1007/](http://doi.org/10.1007/s11606-019-05035-1) [s11606-019-05035-1](http://doi.org/10.1007/s11606-019-05035-1) - 2. Gerke S, Minssen T, Cohen G. Chapter 12 Ethical and legal challenges of artificial intelligence-driven healthcare. In: Bohr A, Memarzadeh K, eds. *Artificial Intelligence in Healthcare*. Academic Press; 2020:295-336. doi[:10.1016/](http://doi.org/10.1016/B978-0-12-818438-7.00012-5) [B978-0-12-818438-7.00012-5](http://doi.org/10.1016/B978-0-12-818438-7.00012-5) - 3. Liaw W, Kueper JK, Lin S, Bazemore A, Kakadiaris I. Competencies for the use of artificial intelligence in primary care. *Ann Fam Med*. 2022;20(6):559-563. doi[:10.1370/afm.2887](http://doi.org/10.1370/afm.2887) - 4. Karaca O, Çalı*¸s*kan SA, Demir K. Medical artificial intelligence readiness scale for medical students (MAIRS-MS) – development, validity and reliability study. *BMC Med Educ*. 2021;21(1):112. doi:[10.1186/s12909-021-02546-6](http://doi.org/10.1186/s12909-021-02546-6) - 5. Waheed MA, Liu L. Perceptions of family physicians about applying AI in primary health care: case study from a premier health care organization. *JMIR AI*. 2024;3:e40781. doi[:10.2196/40781](http://doi.org/10.2196/40781) - 6. Lugito NPH, Cucunawangsih C, Suryadinata N, et al. Readiness, knowledge, and perception towards artificial intelligence of medical students at faculty of medicine, Pelita Harapan University, Indonesia: a cross sectional study. *BMC Med Educ*. 2024;24(1):1044. doi[:10.1186/s12909-024-06058-x](http://doi.org/10.1186/s12909-024-06058-x) - 7. Tung AYZ, Dong LW. Malaysian medical students' attitudes and readiness toward AI (artificial intelligence): a cross-sectional study. *J Med Educ Curric Dev*. 2023;10:23821205231201164. doi[:10.1177/23821205231201164](http://doi.org/10.1177/23821205231201164) - 8. Blease CR, Locher C, Gaab J, Hägglund M, Mandl KD. Generative artificial intelligence in primary care: an online survey of UK general practitioners. *BMJ Health Care Inform*. 2024;31(1):e101102. doi[:10.1136/bmjhci-2024-101102](http://doi.org/10.1136/bmjhci-2024-101102) - 9. Hanna K, Chartash D, Liaw W, et al. Family medicine must prepare for artificial intelligence. *J Am Board Fam Med*. 2024;37(4):520-524. doi:[10.3122/](http://doi.org/10.3122/jabfm.2023.230360R1) [jabfm.2023.230360R1](http://doi.org/10.3122/jabfm.2023.230360R1) - 10. AlZaabi A, AlMaskari S, AalAbdulsalam A. Are physicians and medical students ready for artificial intelligence applications in healthcare? *Digit Health*. 2023;9:20552076231152167. doi[:10.1177/20552076231152167](http://doi.org/10.1177/20552076231152167)