Analysis of ultrasound-related medical incidents in Japan: A SHELL model-based study using a national database

Incident reporting systems are increasingly recognized not only as surveillance mechanisms but also as essential tools for organizational learning and healthcare risk management. While ultrasound is widely regarded as a safe and noninvasive diagnostic modality, the expanding scope of its use has introduced new patient safety risks that remain insufficiently characterized. We conducted a retrospective analysis of 160 ultrasound-related incident reports submitted to the national database of the Japan Council for Quality Health Care between 2014 and 2023. Incident characteristics were analyzed descriptively, and contributing factors were classified using the SHELL model framework to identify system-level vulnerabilities relevant to healthcare risk management. Most incidents occurred during weekday daytime hours (84% on weekdays; 78% between 08:00–17:59). Although the majority required no or minor intervention, 16% involved significant medical treatment. SHELL model analysis revealed that Liveware (human factors) accounted for 68.1% of contributing factors, followed by Software (21.3%) and Hardware (10.6%). No statistically significant association was observed between SHELL categories and severity of harm. Ultrasound-related incidents are primarily driven by human and system interface factors rather than device failure alone. These findings highlight the importance of risk management strategies that emphasize human factors engineering, standardized protocols, and nonpunitive learning-oriented incident reporting systems. National incident databases can serve as valuable resources for identifying latent risks and informing proactive patient safety interventions.