What prevents falls: the evidence

The evidence

Across 14 trials, exercise probably reduced the rate of falls by about a third (rate ratio 0.68, 95% CI 0.51 to 0.91) and the risk of falling slightly (risk ratio 0.86, 95% CI 0.75 to 1.00), moderate-certainty evidence. Once programs ended there was no remaining effect on the rate of falls (rate ratio 1.02, high-certainty evidence). Exercise had little or no effect on fractures in the three trials that reported them ². In the Australian Sunbeam trial of 221 residents in 16 facilities, progressive resistance and balance training for 50 hours over 25 weeks, followed by maintenance, reduced the rate of falls by 55% (incidence rate ratio 0.45, 95% CI 0.17 to 0.74) with no serious adverse events ⁶.

In practice

The national guidelines recommend tailored, supervised exercise for every resident who chooses to take part, and that it continues ¹. Sunbeam excluded residents who were bed-bound or wheelchair-bound, or who could not take part in a group class, so its result applies to people able to exercise ⁶.

The evidence

Pooled across 12 trials, multifactorial interventions probably had little or no effect on the rate of falls (rate ratio 0.87, 95% CI 0.68 to 1.12), with very inconsistent results between trials. A subgroup analysis found that programs tailored to each resident’s circumstances and delivered with facility staff engagement probably produced a large reduction (rate ratio 0.61, 95% CI 0.54 to 0.69; 7 trials) ². That distinction came from a subgroup analysis, which is weaker evidence than a comparison planned in advance. The largest of these trials, in 84 UK care homes with 1,657 residents, found fewer falls in the period three to six months after randomisation (unadjusted incidence rate ratio 0.57, 95% CI 0.45 to 0.71), with no fall in activity and no rise in dependency ⁷.

In practice

All seven of the tailored trials combined assessment of personal and environmental risk, medication optimisation, attention to walking aids, and exercise ². International guidelines likewise recommend a multifactorial assessment leading to personalised interventions for older people at high risk ³.

The evidence

A two-year cluster randomised trial in 60 Australian aged care facilities (7,195 residents, mean age 86) provided extra dairy foods, bringing intake to about 1,142 mg of calcium and 69 g of protein a day. All fractures fell by 33% (hazard ratio 0.67, 95% CI 0.48 to 0.93), hip fractures by 46% (hazard ratio 0.54, 95% CI 0.35 to 0.83) and falls by 11% (hazard ratio 0.89, 95% CI 0.78 to 0.98). Mortality was unchanged ⁸. This is a single trial, and the Cochrane reviewers rated the certainty of the evidence as low ².

In practice

The intervention was a change to the menu costing about 66 cents per resident per day, and a later economic analysis found it cost-saving ⁹. The national guidelines recommend at least 3.5 servings of dairy foods a day on residential care menus, with a dietitian involved ¹. Residents in the trial already had adequate vitamin D ⁸.

The evidence

Across five trials, vitamin D supplementation probably reduced the rate of falls (rate ratio 0.63, 95% CI 0.46 to 0.86) but made little or no difference to the risk of falling (risk ratio 0.99, 95% CI 0.90 to 1.08). Participants in these trials had low vitamin D levels ². In an Australian trial of 625 residents who were not classically deficient, all of whom received calcium, adding vitamin D reduced the rate of falls (incidence rate ratio 0.73, 95% CI 0.57 to 0.95) without a significant effect on fractures ¹⁰.

In practice

The national guidelines recommend daily or weekly doses and advise against monthly or yearly large doses ¹.

The evidence

A single yearly dose of 500,000 IU given to 2,256 Australian women aged 70 and over increased falls (incidence rate ratio 1.15, 95% CI 1.02 to 1.30) and fractures (1.26, 95% CI 1.00 to 1.59) compared with placebo ¹². In a Swiss trial of 200 people aged 70 and over who had already fallen, monthly doses of 60,000 IU, or of 24,000 IU with calcifediol, were followed by a higher proportion of people falling than 24,000 IU monthly (about 66% against 48%), with no gain in leg function ¹³. Both trials were in people living at home, not in residential care.

The evidence

A Cochrane review pooled 14 trials in nursing or residential care (11,808 participants) and found a small reduction in hip fracture (risk ratio 0.82, 95% CI 0.67 to 1.00), about 11 fewer people with a hip fracture per 1,000, with no effect on falls and possibly a slight increase in pelvic fractures. Adherence was poor, particularly over time ¹⁴. A later US trial gave 1,042 residents a protector on one hip only, so that each person acted as their own comparison. It was stopped for lack of effect: 3.1% of protected hips fractured, against 2.5% of unprotected hips ¹⁵.

In practice

The national guidelines suggest that hip protectors be considered ¹. They are an option for a resident at high risk who is willing to wear them. Some people find them uncomfortable, and wearing them is the resident’s choice.

The evidence

Medication optimisation interventions were varied, and overall may make little or no difference to the rate of falls (rate ratio 0.92, 95% CI 0.75 to 1.13; 13 trials). For medication review and deprescribing specifically, the reviewers were very uncertain of the effect ².

In practice

This does not make a sedative safe. The association between psychotropic medicines and falls is consistent in observational studies ²¹, and medication optimisation was part of all seven tailored multifactorial programs that reduced falls ². The medicines page in this section goes through the detail.

The evidence

In a cluster randomised trial across 16 units of a US hospital (27,672 inpatients), an intervention that greatly increased bed alarm use had no effect on falls, injurious falls or the use of physical restraint (risk ratio for falls 1.09, 95% CI 0.85 to 1.53) ¹⁶. This was a hospital study, so how far it applies to residential care is uncertain.

In practice

An alarm tells staff that a resident is already moving. It helps only if someone can reach the person in time, and a loud alarm can startle or distress a person living with dementia.

The evidence

A pragmatic cluster randomised trial in 18 Queensland public hospital wards provided one low-low bed for every 12 beds. There was no significant difference in falls, falls with injury or fractures, although the study was too small to judge fractures ¹⁷. This was also a hospital study.

In practice

A low bed may still shorten the distance of a fall from bed for a particular resident. If a bed is set so low that the person cannot get up from it, it is restricting their movement, and the same questions apply as for any restraint.

The evidence

A Cochrane review of 11 studies in residential care (19,003 participants) describes clear evidence that physical restraints lack effectiveness and safety. Programs that put a least-restraint policy into practice probably reduced the number of residents restrained (risk ratio 0.86, 95% CI 0.78 to 0.94) and may have little or no effect on falls or fall-related injuries ¹⁸. In a German cluster randomised trial across 36 nursing homes, restraint use fell from 31.5% to 22.6% of residents with no significant difference in falls or fall-related fractures ¹⁹. In an older observational study of 322 residents, restraint was not associated with a lower risk of falls or injuries among the residents most likely to be restrained, and among confused residents who could walk it was associated with more recurrent falls ²⁰.

In practice

In Australia, restraint is regulated as a restrictive practice and may be used only as a last resort to prevent harm ⁴⁶. The page on restraint, bed rails and alarms explains the rules.