Aim: We documented the number of falls and falls risk profile over two years to derive a falls risks prediction score.
Background: Simple falls risk assessment tools not requiring equipment or trained personnel may be used as a first step in the primary care setting to identify older people at risk who may be referred for further falls risk assessment in special clinics.
Design: Survey.
Method: Men (n = 1941) and 1949 women aged 65 years and over living in the community were followed up for two years to document the number of falls. Information was collected regarding demography, socioeconomic status, medical history, functional limitations, lifestyle factors and psychosocial functioning. Measurements include body mass index, grip strength and stride length. Logistic regression was used to determine significant predictions of falls and to calculate predictive scores.
Result: Twelve factors in men and nine factors in women were used to construct a risk score. The AUC of the receiver operating characteristic curve was >0.70 for both men and women and a cut off score of >or=8 gave sensitivity and specificity values between 60-78%. The factors included chronic disease, drugs, functional limitation, lifestyle, education and psychosocial factors. When applied to future predictions, only low energy level and clumsiness in both hands in men and feeling downhearted in women, were significant factors.
Conclusions: A risk assessment tool with a cut off score of >or=8 developed from a two-year prospective study of falls may be used in the community setting as an initial first step for screening out those at low risk of falls.
Relevance to clinical practice: A simple tool may be used in the community to screen out those at risk for falls, concentrating trained healthcare professionals' time on detailed falls assessment and intervention for those classified as being at risk.