Forty-four patients aged 65 years and over who fell whilst in an acute
hospital and 44 patients who did not fall during their hospital stay
underwent structured medical examinations to identify factors associat
ed with falling. The control subjects were matched for age(+/- 3 years
), sex, patient type, and primary diagnosis. The examination was based
on established assessments of posture, balance and gait, the musculos
keletal system, vision, cardiovascular status, and neurological functi
on. Bivariate analyses revealed seven assessment measures that were si
gnificantly associated with falls: cognitive impairment, particularly
impaired orientation; evidence of previous cerebrovascular accident; i
ncoordination as measured clinically; inability to perform the 'Get-up
-and-go' test, especially an inability to turn around after a 5-metre
walk, and the use of psyche-active medications. Of these variables, im
paired orientation, psyche-active drug use, evidence of stroke, and im
paired performance in the 'Get-up-and-go' test were included in a step
wise logistic regression which correctly classified 80% of the patient
s into faller and non-faller groups. Falling was also related to the n
umber of these identified risk factors. These findings suggest that a
simple screening protocol, taking about 5 min to complete, can assist
in the identification of patients at risk of falls whilst in hospital.