Artificial Intelligence–Based Patient Safety Incident Reporting System for Nurses: Content Validity Study (Preprint)

BACKGROUND Patient safety is a primary priority in healthcare and serves as a central focus of hospital risk management systems. Patient safety incident reporting plays a crucial role in these efforts. However, despite various global programs and policies aimed at strengthening safety culture, incident reporting among nurses remains low, particularly in low- and middle-income countries, due to administrative burdens, fear of blame, and limited feedback mechanisms. Advances in artificial intelligence (AI) offer opportunities to improve reporting efficiency, data accuracy, and safety culture. The SAFE-Nurse (Smart Assisted Feedback & Education for Nurses on Reporting Safety Events) system was developed as an AI-based digital reporting platform to address these challenges. OBJECTIVE This study aimed to evaluate the content validity of the SAFE-Nurse system to ensure the relevance, clarity, and appropriateness of its features in supporting patient safety incident reporting. METHODS This methodological study applied a content validity approach. Six experts in nursing informatics, nursing management, and patient safety evaluated 43 items across five domains: content appropriateness, clarity and usefulness, reporting workflow, technology and security, and benefit and impact. Each item was rated using a 4-point Likert scale. The Item Content Validity Index (I-CVI) and Scale Content Validity Index (S-CVI/Ave and S-CVI/UA) were calculated to assess the level of expert agreement. RESULTS The findings showed that I-CVI values ranged from 0.83 to 1.00. The overall S-CVI/Ave was 0.98, and S-CVI/UA was 0.91, indicating excellent expert agreement. The domains of reporting workflow and technology and security achieved perfect validity (S-CVI = 1.00), while the benefit and impact domain demonstrated high validity (S-CVI/Ave = 0.97). CONCLUSIONS Overall, these findings confirm that the SAFE-Nurse system demonstrates excellent content validity, representing key domains of patient safety incident reporting. The system is feasible for further reliability testing and usability studies, with the potential to enhance incident reporting, risk management, and patient safety culture in hospitals.