Unplanned extubation (UE)—the unintended removal of an airway tube—remains a critical patient safety concern in intensive care units (ICUs), often leading to severe complications or death. Real-time UE detection has been limited, largely due to the ethical and privacy challenges of obtaining annotated ICU video data. We propose AURA (Augmented Unplanned Removal Alert), a vision-based risk detection system developed and validated entirely on a fully synthetic ICU video dataset. By leveraging text-to-video diffusion, we generated diverse and clinically realistic scenarios capturing a range of patient behaviors and care contexts. The system applies pose estimation to identify two high-risk movement patterns: collision, defined as hand entry into spatial zones near airway tubes, and agitation, quantified by the velocity of tracked anatomical keypoints. Expert assessment validated the realism of the synthetic data, while quantitative evaluation demonstrated high collision detection accuracy and moderate agitation recognition performance. This work demonstrates a novel pathway for developing privacy-preserving, reproducible patient safety monitoring systems with potential for deployment in intensive care settings.