Knowledge, attitude, and practice of artificial intelligence applications in medicine among physicians in Sudan: a national cross-sectional survey.

Abstract

Artificial intelligence (AI) has emerged as a rapidly developing tool within the medical landscape, globally aiding in diagnosis and healthcare management. However, its integration within healthcare systems remains varied across different regions. In Sudan, there exists a burgeoning interest in AI potential applications within medicine. This study aims to evaluate the knowledge, attitudes, and practices of AI applications in medicine among physicians in Sudan. The authors conducted a web-based survey cross-sectional analytical study using an online questionnaire-based survey regarding demographic details, knowledge, attitudes, and practice of AI distributing through various e-mail listings and social media platforms. A sample of 825 Physicians including doctors in Sudan with different ranks and specialties were selected using the convenient non-probability sampling technique. Out of 825 Physicians, 666 (80.7%) of Physicians have previous knowledge about AI. However, only a small number 123 (14.9%) were taught about AI during their time in medical school, even fewer, just 120 (14.5%) had AI-related lessons in their training program. Regarding attitude, 675 (81.8%) agree that AI is very important in medicine, almost the same number, 681 (82.6%) support the idea of teaching AI in medical schools. Practically, 535 (64.8%) of doctors, think that should get special training in using AI tools in healthcare. Excitingly 651 (78.9%) of physicians are interested in working with AI in future. Based on different ranks of doctors toward AI; Medical Officers exhibited the highest proportion at (32.7%) of knowledge and understanding of AI concepts, followed by House Officers at (16.7%) (p=0.076); regarding attitude, Medical Officers demonstrated the highest (31.6%) favorable attitude, followed by House Officers at (17.5%) (p=0.229); In practice also, Medical Officer showed the highest portion (28.0%) among participants (p=0.129). While there is a positive attitude and some level of AI practice, there remains a considerable gap in knowledge that needs addressing.