Objective The development of countermeasures for preventing childhood injuries requires prioritizing injury situations to be prevented based on situation data. Conventionally, a R-Map is well known as a method for prioritizing injuries from the viewpoint of both frequency and severity. However, it is difficult to apply the method to a large amount of text data describing injury situations. This study proposes a situational R-Map analysis, which is a new method for prioritizing injury situations by integrating a R-Map analysis method and a text mining method.

Methods In this study, a situational R-Map analysis was applied to the data on the ten types (e.g., bars, slides, sandboxes, and jungle gyms) of playground-equipment-related injuries that occurred in elementary schools. The data were collected by the Japan Sport Council (JSC) from schools across Japan in 2018. The authors selected playground-related cases (about 25,000 cases) from the 1 million cases.

Results As a result of the analysis, injury situations with high priority were found in ten types of equipment. For example, ‘failing to land and sticking his/her hand to the ground during the long jump in physical exercise classes’ was an example of sandbox-related situations with a high priority for preventing bone fractures. This result shows that the situational R-Map analysis can be used to automatically extract high-priority injury situations from big data, which is conventionally difficult to analyze manually. JSC created a brochure on injury situations and preventive measures clarified by the proposed method and disseminated it to more than 2,000 municipalities throughout Japan.