Evaluating the Perceptions and Adoption of ChatGPT in Medical Education and Clinical Practice: A Cross-Sectional Study in Khyber Pakhtunkhwa, Pakistan

Abstract

Background: Artificial intelligence (AI) has increasingly transformed healthcare, particularly in diagnostics, predictive analytics, and medical education. Recently, ChatGPT has emerged as a widely accessible AI-driven tool, gaining rapid adoption among medical students and healthcare professionals. While it offers potential benefits such as academic support and clinical assistance, concerns persist regarding over-reliance, diminished critical thinking, and accuracy of clinical recommendations. Objective: This study aimed to evaluate the perceptions and adoption of ChatGPT in medical education and clinical practice across medical and dental institutions in Peshawar and Swat, Khyber Pakhtunkhwa (KPK), Pakistan. Methodology: An analytical cross-sectional study was conducted between January and April 2025, recruiting 370 participants through stratified convenience sampling from 12 medical colleges and teaching hospitals. Data were collected using a validated, standardized 19-item questionnaire (Cronbach’s α = 0.78; CVI = 0.89). Associations between demographic variables and ChatGPT adoption were assessed using chi-square tests with Bonferroni correction, and effect sizes were reported using Cramer’s V. Results: The study revealed that early-year medical students demonstrated greater reliance on ChatGPT for academic purposes, while senior clinicians expressed caution regarding its role in clinical practice. Significant associations were found between academic year and perceptions of ChatGPT’s importance for academic success (χ²=25.31, p<0.001) and reliability (χ²=18.64, p=0.001), as well as between field of study and its use for self-diagnosis (χ²=22.47, p<0.001) and academic recommendations (χ²=19.76, p<0.001). Weak-to-moderate effect sizes were observed. Conclusion: ChatGPT is increasingly recognized as a supplementary educational and clinical resource, particularly by early-stage students. However, its limitations in nuanced clinical judgment raise concerns about over-reliance. AI tools like ChatGPT should therefore be integrated thoughtfully into medical education, complementing rather than replacing critical thinking and professional expertise. Keywords: Adoption, Artificial Intelligence, ChatGPT, Clinical practice, Medical education, Perceptions.

Ähnliche Arbeiten