A research group led by Professor Yoshimasa Nagao of Nagoya University Hospital uses AI technology to analyze near-miss incident* and incident report data collected at the hospital to prevent medical accidents. developed a model to measure hospital safety.
The incident report system is a tool that captures incidents that occur at hospitals through voluntary reports from on-site staff and tries to pick up latent risks in the organization. There is also great variation in the quality ofHowever, since the free-text field that details the content of the incident reflects a variety of valuable information, the research group attempted to develop a ruler for measuring safety in hospitals using AI technology.
The research group calculated a severity score based on the frequency of occurrence of each word from text data of incident reports that experts judged to be "severe" or "non-severe."Each word was quantified, and then summarized for each report and group (department) of the medical staff who submitted the report. 48,041 words were finally digitized from 1,802 incident reports.As a result of verification, a significant difference was found between the severity score calculated for each report and the experts' judgment of severity or non-severity.The scores in the population also showed a high correlation with conventional expert analysis.
In the future, by adding other factors related to hospital safety and improving the accuracy of the ruler, we will be able to measure hospital safety in real time and compare safety with other medical institutions. is expected to become possible.
*A dangerous event that fortunately did not lead to an accident.
Paper information:[Journal of Medical Systems] Development of a Novel Scoring System to Quantify the Severity of Incident Reports: An Exploratory Research Study
