Healthcare systems are increasingly adopting artificial intelligence (AI) to improve clinical decision-making, yet many implementations fall short in practice because learning, decision support, and clinical workflows remain disconnected. Clinicians are often expected to use complex AI tools without continuous, context-specific training, leading to underutilization, mistrust, and inconsistent outcomes. This study presents a critical integrative review of the literature on AI-enhanced digital health education platform designed to improve clinical decision-making and workforce competency in the U.S. healthcare systems. Rather than treating these domains separately, the review synthesizes evidence to examine how they can be meaningfully integrated within routine clinical workflows. Four key evidence domains are analyzed: AI-driven clinical decision support systems, AI literacy and competency development, explainability and ethical governance, and implementation challenges in real-world healthcare settings. The findings reveal consistent agreement that AI improves diagnostic accuracy, efficiency, and clinical outcomes, but its effectiveness is limited by fragmented system design, lack of continuous learning mechanisms, and insufficient attention to trust and explainability. This study therefore proposes an implementation framework for workflow-integrated AI-enhanced digital health education platform. The framework explains how clinicians can receive guidance during decision-making, learn from immediate feedback, and build confidence and competence over time. In addition, practical design principles are outlined to guide the development of AI-enabled systems that are both effective and trustworthy. This study contributes to existing body of knowledge by reframing AI adoption in healthcare as a learning-centered process rather than pure technology. As such, AI moves from being a standalone tool to becoming a continuous learning partner, ultimately improving clinical decis