LAUSR

Frailty indicators were proposed to be significant predictors of falls. However, their comparative and additional value to other simple fall predictors has not been evaluated. This study aims to compare the clinical value of three frailty indicators to fall history in predicting recurrent falls in older adults. Four thousand community-dwelling elderly (2000 female) aged 65 years or above were recruited in Hong Kong. The three candidate frailty indicators were the FRAIL scale, the Cardiovascular Health Study Criteria (CHS), and the Study for Osteoporosis and Fracture Criteria (SOF). The comparative predictor was fall history. The area under the receiver operating characteristic curve (AUC) was used to assess their ability to predict recurrent falls in the second year after baseline. Classification and Regression Tree (CART) analysis was used to evaluate the clinical potential of significant predictors on identifying older recurrent fallers in a simulated clinical pathway. Fall history is better than all frailty indicators in predicting recurrent falls in men and women (Fall History: AUC: men=0.681; women=0.645; Frailty indicators: AUC≤0.641). Upon adjusting for fall history, FRAIL and SOF remained as significant predictors for women but no frailty indicator remained significant for men. FRAIL could separate older women into two groups with different proportion of recurrent fallers in those with history of recurrent falls (16.0% vs 30.8%), single fall history (9.5% vs 37.5%), and no fall history (3.8% vs 7.5%). SOF provides limited value in people with a single fall history (history of recurrent falls: 16.5% vs 20.6%; single fall history: 10.2% vs 10.9%; no fall history: 3.9% vs 8.6%). Adding FRAIL or SOF on top of fall history improves the prediction of recurrent falls in older women but not men. FRAIL could be a useful tool for identifying community-dwelling older women at high risk of recurrent falls.