OBJECTIVE: To analyse the power to predict risk and verify the validity of the Morse Fall Scale - Brazilian version (MFS-B).
METHOD: This is a methodological, longitudinal study with 1487 adult patients of two university hospitals of Rio Grande do Sul, Brazil conducted from November 2013 to March 2014. The MFS-B was used to assess the risk of falls. Statistical analysis comprised multivariate methods (discriminant function analysis and ROC curve). The research was approved by the ethics committees of the institutions.
RESULTS: The best estimate to predict falls was at the cutoff point 44.78 of the average MFS-B score, with a sensitivity of 95.2% and a specificity of 64%. The occurrence of falls and the high-risk classification were significant (p RESULTS: < RESULTS: 0.00001).
CONCLUSIONS: The results show that the MFS-B can appropriately predict the risk of falls at the cutoff point for the high-risk classification, according to the original classification. The MFS-B had adequate validation test results and maintained the six items of the original scale.