Abstract
In this study, we analyzed the prediction accuracy and characteristics of the assessment items of inpatient fall risk assessment tools published in Japan, to investigate the feasibility of their clinical application and the associated problems. Regarding the tools published in articles meeting the criteria, the quality was assessed, and the sensitivity, specificity, positive likelihood ratio and negative likelihood ratio were calculated, and the assessment items were classified. Analysis of the tools published in 8 articles revealed a good prediction accuracy, with a sensitivity of 0.6 to 0.76, a specificity of 0.6 to 0.91, a positive likelihood ratio of 1.89 to 8.21, and a negative likelihood ratio of 0.29 to 0.99. The results suggested that two of the tools can be utilized in hospital wards as general fall risk assessment tools for inpatients. It was clarified that the items included in fall risk assessment tools assessed the causal factors of inpatient falls from multiple aspects. It is necessary to ensure their reliability and validity by expressing the items more clearly.
