Revision of Fall Assessment and Classification of Risk in Acute Care Wards Using Latent Rank Theory

[Purpose] To revise a fall assessment method to create a valid and simple assessment for use in acute care wards. [Participants and Methods] This retrospective study included 1309 inpatients. Risk factors were extracted and weighted based on the results of a fall assessment sheet and the presence or absence of falls. Then, the revised fall score and prediction accuracy were calculated. For the revised score, risk was classified according to the latent rank. [Results] The number of fall factors was reduced from 39 to seven. The sensitivity and specificity of prediction and the area under the curve for the conventional and revised scores were 88.5%, 43.0%, and 0.700 vs. 65.6%, 71.0%, and 0.718, respectively. The number of falls significantly differed for each potential rank according to the revised scores. [Conclusion] Fall prediction based on revised scores and risk classification using latent rank theory is appropriate.