The prevention of falls in the elderly could be based on risk assessme
nt and the prediction of when an individual is likely to sustain an in
jury. A telecare system should allow the dynamic assessment of risk to
be produced by a control centre computer, based on data transmitted f
rom local sensors in the elderly person's home. A fall risk index coul
d be calculated from mobility, activities of daily living and medicati
on. However, substantial data-sets will be required before such fall r
isk indices attain statistical significance. As they are developed, fa
ll prediction algorithms could be constantly reviewed and modified in
order to reach the point where an acceptable level of accuracy is reac
hed.