Eighteen studies (n=10,634) were included in the review: 14 RCTs (n=9,241), one RCT with a 2x2 factorial design (n=54), one multifactorial RCT (n=1,090), one longitudinal study (n=225) and one clinical case study (n=24).
Fourteen studies used allocation concealment (randomisation method described), two were randomised but not described, and two did not support concealment. The outcome assessors were blinded in only 7 studies, while the participants were blinded in one. The participants were clearly comparable at baseline in 12 studies. The outcome measures were defined in all but 2 studies.
Institutional settings (7 studies).
Three studies reported a reduction in the number of falls, fallers or recurrent fallers, while 3 studies found no significant effect on the number of falls. One study reported more injurious falls on vinyl than on carpet, but the data were unclear. Four studies reported no effect of the intervention on injury, one study found an unclear effect on injury, and one did not report the effects on injury.
Community settings (11 studies).
Six studies reported a significant positive effect of the intervention on falls or the risk of falls, while 5 studies reported no significant effect on falls. Six studies reported no significant effect of the intervention on injury, 2 studies did not measure the effects on injury, and 3 studies did not report the effects of the interventions on injury. One study reported a significant effect of the intervention on home hazards.