Environmental interventions for preventing falls in older people living in the community

Lindy Clemson; Susan Stark; Alison C Pighills; Nicola J Fairhall; Sarah E Lamb; Jinnat Ali; Catherine Sherrington

doi:10.1002/14651858.CD013258.pub2

. 2023 Mar 10;2023(3):CD013258. doi: 10.1002/14651858.CD013258.pub2

Environmental interventions for preventing falls in older people living in the community

Lindy Clemson ^1,^✉, Susan Stark ², Alison C Pighills ^3,⁴, Nicola J Fairhall ⁵, Sarah E Lamb ⁶, Jinnat Ali ⁷, Catherine Sherrington ⁸

Editor: Cochrane Bone, Joint and Muscle Trauma Group

PMCID: PMC9998238 PMID: 36893804

Abstract

Background

Falls and fall‐related injuries are common. A third of community‐dwelling people aged over 65 years fall each year. Falls can have serious consequences including restricting activity or institutionalisation. This review updates the previous evidence for environmental interventions in fall prevention.

Objectives

To assess the effects (benefits and harms) of environmental interventions (such as fall‐hazard reduction, assistive technology, home modifications, and education) for preventing falls in older people living in the community.

Search methods

We searched CENTRAL, MEDLINE, Embase, other databases, trial registers, and reference lists of systematic reviews to January 2021. We contacted researchers in the field to identify additional studies.

Selection criteria

We included randomised controlled trials evaluating the effects of environmental interventions (such as reduction of fall hazards in the home, assistive devices) on falls in community‐residing people aged 60 years and over.

Data collection and analysis

We used standard methodological procedures expected by Cochrane. Our primary outcome was rate of falls.

Main results

We included 22 studies from 10 countries involving 8463 community‐residing older people. Participants were on average 78 years old, and 65% were women. For fall outcomes, five studies had high risk of bias and most studies had unclear risk of bias for one or more risk of bias domains. For other outcomes (e.g. fractures), most studies were at high risk of detection bias. We downgraded the certainty of the evidence for high risk of bias, imprecision, and/or inconsistency.

Home fall‐hazard reduction (14 studies, 5830 participants)

These interventions aim to reduce falls by assessing fall hazards and making environmental safety adaptations (e.g. non‐slip strips on steps) or behavioural strategies (e.g. avoiding clutter).

Home fall‐hazard interventions probably reduce the overall rate of falls by 26% (rate ratio (RaR) 0.74, 95% confidence interval (CI) 0.61 to 0.91; 12 studies, 5293 participants; moderate‐certainty evidence); based on a control group risk of 1319 falls per 1000 people a year, this is 343 (95% CI 118 to 514) fewer falls. However, these interventions were more effective in people who are selected for higher risk of falling, with a reduction of 38% (RaR 0.62, 95% CI 0.56 to 0.70; 9 studies, 1513 participants; 702 (95% CI 554 to 812) fewer falls based on a control risk of 1847 falls per 1000 people; high‐certainty evidence). We found no evidence of a reduction in rate of falls when people were not selected for fall risk (RaR 1.05, 95% CI 0.96 to 1.16; 6 studies, 3780 participants; high‐certainty evidence).

Findings were similar for the number of people experiencing one or more falls. These interventions probably reduce the overall risk by 11% (risk ratio (RR) 0.89, 95% CI 0.82 to 0.97; 12 studies, 5253 participants; moderate‐certainty evidence); based on a risk of 519 per 1000 people per year, this is 57 (95% CI 15 to 93) fewer fallers. However, for people at higher risk of falling, we found a 26% decrease in risk (RR 0.74, 95% CI 0.65 to 0.85; 9 studies, 1473 participants), but no decrease for unselected populations (RR 0.99, 95% CI 0.92 to 1.07; 6 studies, 3780 participants) (high‐certainty evidence).

These interventions probably make little or no important difference to health‐related quality of life (HRQoL) (standardised mean difference 0.09, 95% CI −0.10 to 0.27; 5 studies, 1848 participants; moderate‐certainty evidence). They may make little or no difference to the risk of fall‐related fractures (RR 1.00, 95% 0.98 to 1.02; 2 studies, 1668 participants), fall‐related hospitalisations (RR 0.96, 95% CI 0.87 to 1.06; 3 studies, 325 participants), or in the rate of falls requiring medical attention (RaR 0.91, 95% CI 0.58 to 1.43; 3 studies, 946 participants) (low‐certainty evidence). The evidence for number of fallers requiring medical attention was unclear (2 studies, 216 participants; very low‐certainty evidence). Two studies reported no adverse events.

Assistive technology

Vision improvement interventions may make little or no difference to the rate of falls (RaR 1.12, 95% CI 0.84 to 1.50; 3 studies, 1489 participants) or people experiencing one or more falls (RR 1.09, 95% CI 0.79 to 1.50) (low‐certainty evidence). We are unsure of the evidence for fall‐related fractures (2 studies, 976 participants) and falls requiring medical attention (1 study, 276 participants) because the certainty of the evidence is very low. There may be little or no difference in HRQoL (mean difference 0.40, 95% CI −1.12 to 1.92) or adverse events (falls while switching glasses; RR 1.00, 95% CI 0.98 to 1.02) (1 study, 597 participants; low‐certainty evidence).

Results for other assistive technology ‐ footwear and foot devices, and self‐care and assistive devices (5 studies, 651 participants) ‐ were not pooled due to the diversity of interventions and contexts.

Education

We are uncertain whether an education intervention to reduce home fall hazards reduces the rate of falls or the number of people experiencing one or more falls (1 study; very low‐certainty evidence). These interventions may make little or no difference to the risk of fall‐related fractures (RR 1.02, 95% CI 0.96 to 1.08; 1 study, 110 participants; low‐certainty evidence).

Home modifications

We found no trials of home modifications that measured falls as an outcome for task enablement and functional independence.

Authors' conclusions

We found high‐certainty evidence that home fall‐hazard interventions are effective in reducing the rate of falls and the number of fallers when targeted to people at higher risk of falling, such as having had a fall in the past year and recently hospitalised or needing support with daily activities. There was evidence of no effect when interventions were targeted to people not selected for risk of falling. Further research is needed to examine the impact of intervention components, the effect of awareness raising, and participant‐interventionist engagement on decision‐making and adherence.

Vision improvement interventions may or may not impact the rate of falls. Further research is needed to answer clinical questions such as whether people should be given advice or take additional precautions when changing eye prescriptions, or whether the intervention is more effective when targeting people at higher risk of falls.

There was insufficient evidence to determine whether education interventions impact falls.

Plain language summary

Reducing fall hazards within the environment

Key messages

For older adults at a higher risk of having a fall, such as having had a fall in the past year and recently hospitalised or needing support with daily activities, removing environmental fall hazards in the home can reduce the number of falls by 38%. Examples of environmental fall hazards are a stairway without railings, a slippery pathway, or poor lighting.

Why is it important to consider environmental hazards as part of a fall prevention programme?

Falls are common and can be deadly, but they are preventable. Approximately one‐third of people age 65 years and older fall each year. Most falls occur in the home, and more than 30% of all falls are caused by environmental hazards. Environmental fall‐hazard removal programmes are interventions delivered by professionals that identify and remove environmental fall hazards.

What did we want to find out?

We wanted to find out:

• which types of environmental programmes work best to prevent falls.

We also wanted to find out for these programmes that prevent falls:

• the best ways to deliver programmes that reduce environmental fall hazards; and

• if such programmes can prevent falls that result in serious injury.

We examined four types of programmes, including those that:

• focused on removing fall hazards in and about the home;

• only provided assistive devices such as up‐to‐date glasses or special footwear;

• only provided education about environment‐related fall risk; and

• focused on home modifications to enable independence and performance of daily tasks in the home.

What did we do?

We searched for studies that investigated different types of environmental fall‐hazard programmes for older adults who live in the community. We compared and summarised the results of the studies and rated our confidence in the evidence, based on factors such as study methods and sample sizes.

What did we find?

We found 22 studies involving 8463 older adults that lived in the community. The study participants were on average 78 years of age, and 65% were women. The studies were conducted in 10 countries. Most studies followed up participants for 12 months.

Main results

A programme that removes fall hazards in the home given to older adults that live in the community can reduce the number of falls.

We are not certain if assistive devices (such as checking prescription for glasses, special footwear, or bed alarm systems) can reduce the risk of a fall.

We are not certain if just giving an older adult who lives in the community education about fall risk has any impact on reducing risk of falls.

We found no completed studies that helped us answer our question about fall reduction for home modification programmes targeting independence in daily activity performance.

We found little evidence to determine if environmental fall‐hazard removal programmes of any kind reduce the risk of serious injury.

How up‐to‐date is the evidence?

The evidence is current to January 2021.

Summary of findings

Summary of findings 1. Home fall‐hazard reduction compared with control (e.g. usual care) for preventing falls in older people living in the community .

Population: older people living in the community (included trials conducting pre‐ and/or post‐hospital discharge home visits focusing on fall prevention) Setting: community (home or places of residence that do not provide residential health‐related care) Intervention: home fall‐hazard reduction interventions aim to reduce the risk of falling in and about the home environment. This intervention includes the assessment of fall hazards in and about the home and recommendations for modifying risk by adapting or changing the home environment. They may also include behavioural strategies to reduce risk in negotiating the home environment, joint problem solving, relevant assistive devices, provision of and/or follow‐up support for safety adaptations and changes. Comparison: usual care or attention control that is thought not to reduce falls, such as general health education or social visits
Outcomes	Illustrative comparative risks (95% CI)*		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Control	Home fall‐hazard reduction
Rate of falls (falls per person‐years) Follow‐up range: 3 to 18 months	Study population^a		Rate ratio 0.74 (0.61 to 0.91)	5293 (12 studies)	⨁⨁⨁◯ MODERATE^b	Overall, there was a reduction of 26% (95% CI 9% to 39%) in the number of falls. Guide to the data: If 1000 people were followed over 1 year, the number of falls would be 976 (95% CI 805 to 1201) compared with 1319 falls in the group receiving usual care or attention control.  In the subgroup selected for higher risk, there was a reduction of 38% (95% CI 30% to 44%) in the number of falls. Guide to the data: If 1000 people were followed over 1 year, the number of falls would be 1145 (95% CI 1035 to 1293) compared with 1847 in the control group. In the subgroup not selected for higher risk of falls, there was an increase of 5% (95% CI 4% reduction to 16% increase) in the number of falls. Guide to the data: If 1000 people were followed over 1 year, the number of falls would be 1094 compared with 1042 in the control group.
	1319 per 1000	976 per 1000  (805 to 1201)
	Selected for high‐risk population		Rate ratio 0.62 (0.56 to 0.70)^c	1513 (9 studies)	⨁⨁⨁⨁ HIGH^d
	1847 per 1000	1145 per 1000  (1035 to 1293)
	Not selected for high‐risk population		Rate ratio 1.05 (0.96 to 1.16)^c	3780 (6 studies)	⨁⨁⨁⨁ HIGH^d
	1042 per 1000	1094 per 1000  (1000 to 1209)
Number of people who experienced 1 or more falls Follow‐up range: 3 to 18 months	Study population^e		RR 0.89 (0.82 to 0.97)^f	5253 (12 studies)	⨁⨁⨁◯ MODERATE^g	Overall, there was a reduction of 11% (95% CI 3% to 18%) in the number of people who experienced 1 or more falls. In the subgroup selected for higher risk, there was a reduction of 26% (95% CI 15% to 35%) in the number of people who experienced 1 or more falls. In the subgroup not selected for higher risk, there was a reduction of 1% (95% CI reduction of 8% to increase of 7%) in the number of people who experienced 1 or more falls.
	519 per 1000	462 per 1000 (426 to 504)
	Selected for high‐risk population		RR 0.74 (0.65 to 0.85)	1473 (9 studies)	⨁⨁⨁⨁ HIGH^g
	550 per 1000	461 per 1000 (426 to 503)
	Not selected for high‐risk population		RR 0.99 (0.92 to 1.07)	3780 (6 studies)	⨁⨁⨁⨁ HIGH^d
	533 per 1000	528 per 1000 (491 to 571)
Number of people who experienced 1 or more fall‐related fractures Follow‐up range: 12 months	Study population^h		RR 1.00 (0.98 to 1.02)	1668 (2 studies)	⨁◯◯◯ VERY LOWⁱ	Overall, there was no difference 0% (95% CI 2% reduction to 2% increase) in the number of people who experienced a fall‐related fracture.
	205 per 1000	205 per 1000 (201 to 209)
Number of people who experienced 1 or more falls that resulted in hospital admission Follow‐up range: 6 to 12 months	Study population^h		RR 0.96 (0.87 to 1.06)	325 (3 studies)	⨁⨁◯◯ LOW^j	Overall, there was a 4% reduction (95% CI 23% reduction to 6% increase) in the number of people who experienced a fall resulting in hospital admission.
	60 per 1000	58 per 1000 (52 to 64)
Falls requiring medical attention Rate of falls requiring medical attention Follow‐up range: 12 to 36 months Number of people who required medical attention for a fall Follow‐up: 12 months	Study population^k		Rate ratio 0.91 (0.58 to 1.43)	946 (3 studies)	⨁⨁◯◯ LOW^l	Overall, there was a 9% reduction (95% CI 42% reduction to 42% increase) in the rate of falls requiring medical attention. Overall, there was an 11% reduction (95% CI 44% reduction to 42% increase) in the number of people who experienced a fall requiring medical attention.
	170 per 1000	155 per 1000 (98 to 251)
	Study population^h		RR 0.89 (0.56 to 1.42)	216 (2 studies)	⨁◯◯◯ VERY LOWⁿ
	‐	Not estimable^m
Health‐related quality of life SMD was calculated from 4 studies with EQ‐5D (range 0 to 1), 1 study with SF‐36, and 1 study with WHOQOL‐BREF (range 0 to 100). Higher scores indicate better quality of life. Follow‐up time points in the included studies were 6 to 12 months.	Study population		SMD 0.09 (−0.10 to 0.27)	1848 (6 studies)	⨁⨁⨁◯ MODERATE^o	Re‐expressing the results using the EQ‐5D (0‐to‐1 scale), there was probably an increase in quality of life of 0.02 in the intervention group (95% CI −0.04 to 0.05). MID for the EQ‐5D is typically 0.074 (Walters 2005).
	In the control group, the mean EQ‐5D score ranged from 0.56 to 0.7; mean SF‐36 score was 43.74; and mean WHOQOL‐BREF score was 65.1.	SMD 0.09 higher (0.10 lower to 0.27 higher)
Adverse events Follow‐up: 6 to 12 months	Study population		‐	73 (2 studies)	⨁⨁◯◯ LOW^p	2 studies reported this outcome for both intervention and control, finding no adverse events.
		Not estimable^m
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MID: minimal important difference; RR: risk ratio; SF‐36: 36‐item Short Form Health Survey; SMD: standardised mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

^aThe all‐studies population was based on the number of events and the number of participants in the control group for this outcome over the 12 studies. We calculated the risk in the control group using the weighted mean of falls per person‐year for the subgroups of trials which were a) selected for higher risk of falls at baseline (9 studies, 1513 participants) or b) not selected for higher risk of falls at baseline (6 studies, 3780 participants). ^bWe downgraded the overall analysis one level for inconsistency due to significant heterogeneity (I² = 83%). ^cSubgroup analysis found a significant difference between those selected for higher risk at baseline and those not selected for higher risk at baseline in the rate of falls (test of subgroup differences P < 0.001). ^dNo downgrading, which included no studies at high risk of bias for any of the specified domains. ^eThe all‐studies population was based on the number of events and the number of participants in the control group for this outcome over the 12 studies. We calculated the risk in the control group using the median falls per person‐year for the subgroups of trials which were a) selected for higher risk of falls at baseline (9 studies, 1513 participants) or b) not selected for higher risk of falls at baseline (6 studies, 3780 participants). ^fSubgroup analysis found a significant difference between those selected for higher risk at baseline and those not selected for higher risk at baseline in the number of falls (test for subgroup difference P < 0.01). ^gWe downgraded one level for inconsistency. Although the overall analysis had low levels of statistical heterogeneity (I² = 18%), subgroup analysis revealed differences according to baseline risk of falling. We did not downgrade for risk of bias, as there was one study with high risk of bias which did not impact the result. ^hWe calculated the risk in the control group based on the number of events and the number of participants in the control group for that outcome. ⁱWe downgraded one level due to imprecision (low event rates) and two levels for risk of bias, as both included trials had one domain with high risk of (detection) bias.  ^jWe downgraded one level due to imprecision (low event rates) and one level for risk of bias, as results were changed with removal of the two trials with high risk of bias on specified domains (detection bias). ^kThe all‐studies population was based on the number of events and the number of participants in the control group for this outcome. ^lWe downgraded two levels due to imprecision (wide confidence interval, event rates less than proposed 300 cut‐off (n = 274)). ^mNumber of events not reported in one trial. ⁿWe downgraded two levels due to imprecision (wide confidence interval crosses both appreciable benefit and harm and few events), and one level for risk of bias, as both trials had high risk of detection bias. ^oWe downgraded for imprecision, as wide confidence interval crosses both appreciable benefit and harm. ^pWe downgraded two levels for imprecision as there were no events in either trial.

Summary of findings 2. Assistive technology for vision improvement interventions compared with usual care or attention control for preventing falls in older people living in the community.

Population: older people living in the community Setting: community (home or places of residence that do not provide residential health‐related care) Intervention: assistive technology or devices as a strategy to improve vision and reduce fall risk (such as vision assessment and provision of glasses) Comparison: usual care or attention control, the latter being one which is thought not to reduce falls, such as general health education or social visits
Outcomes	Illustrative comparative risks (95% CI)*		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Control	Assistive technology  interventions
Rate of falls (falls per person‐years) Follow‐up range: 3 to 18 months	Study population^a		Rate ratio 1.12 (0.84 to 1.50)	1489 (3 studies)	⨁⨁◯◯ LOW^b	Overall, there was an increase of 12% (95% CI 50% increase to 16% decrease) in the rate of falls in the intervention group.
	869 per 1000	973 per 1000 (730 to 1304)
Number of people who experienced 1 or more falls Follow‐up range: 3 to 18 months	Study population		RR 1.09 (0.79 to 1.50)	1489 (3 studies)	⨁⨁◯◯ LOW^c	Overall, there was an increase of 9% (95% CI 50% increase to 21% reduction) in the number of people experiencing 1 or more falls.
	559 per 1000	609 per 1000 (482 to 915)
Number of people who experienced 1 or more fall‐related fractures Follow‐up range: 12 to 13 months	Study population		RR 1.22 (0.75 to 2.00)	976 (2 studies)	⨁◯◯◯ VERY LOW^d	Overall, there was an increase of 22% (95% CI 100% increase to 25% reduction) in the number of people experiencing 1 or more fall‐related fractures.
	43 per 1000	53 per 1000 (40 to 105)
Number of people who experienced 1 or more hospitalisations following a fall	‐		‐	‐	‐	No studies reported this outcome.
Rate of falls requiring medical attention Follow‐up: 18 months	Study population		Rate ratio 0.93 (0.48 to 1.82)	276 (1 study)	⨁◯◯◯ VERY LOW^e	1 trial reported a 7% reduction (95% CI 52% reduction to 82% increase) in the number of people who experienced a fall resulting in hospital admission.
	650 per 1000	410 per 1000 (42 to 159)
Health‐related quality of life 1 trial of single‐lens glasses compared with usual glasses (multifocals) reported this outcome using the SF‐12 (range 0 to 100). Higher scores indicate better quality of life. Follow up: 13 months	Study population		MD 0.40 (−1.12 to 1.92)	597 (1 study)	⨁⨁◯◯ LOW^f	MID for the SF‐12 is typically 2.0 to 4.0 (Díaz‐Arribas 2017).
	The mean SF‐12 score was 43.	MD 0.40 (1.12 lower to 1.92 higher)
Adverse events Falls while switching glasses Follow‐up: 13 months	Study population		RR 1.00 (0.98 to 1.02)	597 (1 study)	⨁⨁◯◯ LOW^g	1 trial reported adverse events for both intervention and control (intervention n = 2; control n = 1).
	‐	‐
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; MID: minimal important difference; RR risk ratio; SF‐12: 12‐item Short Form Health Survey
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

^aThe all‐studies population was based on the number of events and the number of participants in the control group for this outcome over the three studies in this comparison. We calculated the risk in the control group using the weighted mean of falls per person‐year for the subgroups of trials which were a) selected for higher risk of falls at baseline (2 studies, 595 participants) or b) not selected for higher risk of falls at baseline (3 studies, 1228 participants). ^bWe downgraded one level due to inconsistency (heterogeneity, I² = 79%) and one level due to imprecision (confidence interval crosses both appreciable benefit and appreciable harm). ^cWe downgraded two levels due to inconsistency (heterogeneity, I² = 88%) and little overlap of confidence intervals. ^dWe downgraded two levels due to imprecision (few events, confidence interval crosses appreciative harm and benefit) and one level because both studies had high risk of bias for one domain (detection of fractures).  ^eWe downgraded two levels due to imprecision (few events, wide confidence interval that covers appreciable harm and benefit) and one level due to high risk of bias (detection of medical fall outcomes) in the one included trial. ^fWe downgraded two levels due to imprecision (wide confidence interval and small number of participants).  ^gWe downgraded two levels due to imprecision (no events and small number of participants).

Summary of findings 3. Education interventions compared with control (e.g. usual care) for preventing falls in older people living in the community .

Population: older people living in the community Setting: community (home or places of residence that do not provide residential health‐related care) Intervention: education interventions provide information about environmental fall risks, environmental hazards, or assistive devices with the aim of reducing the risk of falls Comparison: usual care or attention control, the latter being one that is thought not to reduce falls, such as general health education or social visits
Outcomes	Illustrative comparative risks (95% CI)*		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Control	Education
Rate of falls (falls per person‐years) Follow‐up: 12 months	Study population^a		Rate ratio 0.63 (0.37 to 1.07)	110 (1 study)	⨁◯◯◯ VERY LOW^b	1 trial reported a 37% reduction (95% CI 7% increase to 63% reduction) in the rate of falls.
	650 per 1000	410 per 1000 (241 to 696)
Number of people who experienced 1 or more falls Follow‐up: 12 months	Study population		RR 0.84 (0.64 to 1.11)	110 (1 study)	⨁◯◯◯ VERY LOW^b	1 trial reported a 16% reduction (95% CI 11% increase to 36% reduction) in the number of people who experienced a fall.
	388	327 per 1000 (249 to 432)
Number of people who experienced 1 or more fall‐related fractures Follow‐up: 12 months	Study population		RR 1.02 (0.96 to 1.08)	110 (1 study)	⨁⨁◯◯ LOW^c	1 trial reported a 2% increase (95% CI 8% increase to 4% reduction) in the number of people who experienced a fall‐related fracture.
	19 per 1000	19 per 1000 (18 to 20)
Number of people who experienced 1 or more falls requiring hospitalisation	‐	‐	‐	‐	‐	Outcome not reported.
Number of people who experienced 1 or more falls requiring medical attention	‐	‐	‐	‐	‐	Outcome not reported.
Health‐related quality of life	‐	‐	‐	‐	‐	Outcome not reported.
Adverse events	‐	‐	‐	‐	‐	Outcome not reported.
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

^aThe all‐studies population was based on the number of events and the number of participants in the control group for this outcome.  ^bWe downgraded two levels due to imprecision (small number of events, confidence interval crossing appreciable benefit and no effect) and one level due to high risk of bias for one or more selected domains.  ^cWe downgraded one level due to imprecision (small number of events) and one level due to high risk of bias for one or more selected domains.

Background

The environmental interventions evaluated in this review were those aimed at reducing the risk of falls of older people living in the community from the interaction with their physical environment, whether indoors or outdoors. These may be single interventions, such as non‐slip walking shoes or assistive devices, or more complex interventions, such as adaptations of the home environment to reduce fall risk combined with education on fall hazards following an environmental fall‐risk assessment of the home environment.

Description of the condition

Falls and fall‐related injuries are common, with a third of community‐dwelling people over 65 years of age falling each year (Campbell 1990; Moreland 2020). The rate of fall‐related injuries increases with age (Verma 2016), and falls can have serious consequences. Ten per cent to 20% of falls result in a fracture (Campbell 1990; Moreland 2020); fall‐associated fractures in older people are a significant source of morbidity and mortality (Moreland 2020), and are an independent predictor of admission to a nursing home (Prabhakaran 2020). Most fall‐related injuries are minor, such as bruising, abrasions, lacerations, strains, and sprains, but they are still associated with pain and reduced function, reduced activity, and can negatively affect quality of life.

Fear of falling and loss of confidence can result in other consequences including self‐restricted activity levels. This can impact on strength and balance, contributing to frailty where the person does not have physical reserves, and leading to a reduction in social interactions (Makino 2018). Both injurious and non‐injurious falls can have these psychological and subsequent physical effects.

Epidemiological studies have identified a number of risk factors for falling in community‐dwelling older people (Deandrea 2010; Lusardi 2017). These risk characteristics can be broadly classified as medical, mobility, sensory, psychological, medication‐related, environmental, or behavioural. Environmental fall‐related risks include environmental factors such as slippery or uneven surfaces, clutter, poor lighting, poor footwear, step hazards, unsafe rails, and loose mats (Keglovits 2020; Stevens 2014). Behavioural fall‐related risk factors include, for example, rushing, not paying attention to the path ahead, or poor ladder/climbing safety (Clemson 2003; Ferrari 2012). Other fall‐related risk factors include increasing age, previous falls, walking aid use, gait problems, difficulty with activities of daily living, slower gait speeds, dizziness, fear of falling, urge incontinence, comorbidity, vision impairment, chronic diseases such as Parkinson's disease, depression, and dementia, and history of stroke. These factors can modify the way a person negotiates their physical environment. It is estimated that 60% of falls have multiple causal factors (Karlsson 2013), and that more than 30% are attributed to environmental causes (Leavy 2015; Rubenstein 2006), with women having a higher risk of indoor falls, and men having higher odds of falling in the presence of outdoor environmental hazards (Lee 2021). Multiple contributing and interacting factors include, for example, slower gait, poor vision, and tripping on a loose mat resulting in a fall (Pighills 2021).

Interventions have largely been conducted in higher‐income countries, and there is a lack of research on how translatable these interventions are cross‐culturally where environments can differ greatly (Hill 2018; Shi 2014; Williams 2015). This is particularly so for low‐income countries, where the determinants and conditions associated with fall‐related injuries are complex and poorly understood (e.g. narrow steps with poor lighting or the water source being outside the home).

Despite early attempts to achieve a consensus definition of 'a fall' (Kellog International Work Group 1987), many definitions still exist in the literature. It is particularly important to have a clear, simple definition for studies in which older people record their own falls, as their concept of a fall may differ from that of researchers or healthcare professionals (Zecevic 2006). A consensus statement defines a fall as “an unexpected event in which the participant comes to rest on the ground, floor, or lower level” (Lamb 2005). The recommended wording when asking study participants is: 'In the past month, have you had any fall including a slip or trip in which you lost your balance and landed on the floor or ground or lower level?' (Lamb 2005).

Description of the intervention

Environmental approaches are aimed at improving individual safety at home, outdoors, and in community and public places to reduce the risk of falls in older people living in the community. This may include assessment for and the provision of an assistive device (e.g. walking aids, grab bars, orthotic footwear, eye glasses), material adaptations (e.g. clearing pathways, fastening carpets, non‐slip strips on step edges), behavioural adaptations (e.g. avoiding ladder use), or structural modifications (e.g. installing a skylight to improve visibility) (Clemson 2021; Gitlin 2009; Stevens 2014).

The intervention descriptors draw on the Prevention of Falls Network Europe (ProFaNE) taxonomy (Lamb 2005; Lamb 2007), with refinement of descriptors to capture emerging literature over the last decade. We covered the following categories of environmental interventions in this review. Please note that the description of 'home', which can be indoors or outdoors, reflects the primary location of the intervention but does not exclude extension to public places.

Home fall‐hazard reduction intervention as a single‐ or multiple‐strategy intervention. This is a package of strategies that aims to reduce falls.
Assistive technology as a single or stand‐alone intervention to reduce falls.
Education on environmental fall risks as a stand‐alone intervention to reduce falls.
Home modifications as a stand‐alone intervention that aims to enable functional independence.

The following is a description of these interventions. A table showing the links to the ProFaNE fall prevention taxonomy and further examples can be found in Appendix 1.

Home fall‐hazard reduction: environmental assessment, adaptation, and modification aims to reduce fall hazards in and about the home or outdoors, or both. In addition to identifying and assessing fall hazards, this intervention category may include any of the following: awareness raising of fall risks within the person's environment, joint problem‐solving of fall hazard management strategies, environmental adaptations, relevant assistive technologies or equipment, and environmental and behavioural safety strategies to reduce fall risk (Clemson 2021; Iwarsson 2009; Peterson 2008; Pighills 2016). Although the environmental intervention is oriented towards the home, the outdoors, or both venues, extension to public places is possible. The focus of this intervention is on fall prevention. Solutions that focus on hazard reduction usually do not involve major structural changes. Criteria have been developed to determine the quality of the intervention and the degree to which it is tailored to falls; these include "(a) a comprehensive evaluation process of environmental hazard identification and priority setting taking into account both personal risk, individual capacity and assessment of the person’s environment, (b) the use of an assessment tool validated for the broad range of potential fall hazards, (c) inclusion of formal or observational evaluation of the functional capacity (physical capacity, behaviour, functional vision, habits) of the person within the context of their environment, (d) the provision of adequate follow‐up by the health professional and support for adaptations and modifications", and (e) the active involvement of the older person in the assessment and priority setting of fall hazards (Clemson 2008).
Assistive technology refers to devices, equipment, products, or systems that support a person to increase or maintain their ability to perform a task or increases the ease or safety (or both) with which a task can be performed (WHO 2004). The aim of fall prevention is safe performance of mobility or tasks. Examples of single‐strategy assistive technologies for fall prevention include personal mobility devices (e.g. walking aids); body‐worn aids (e.g. antislip devices for shoes, orthotic footwear); communication and sensory (e.g. eyeglasses, hearing aids); protection (e.g. alarm sensors and systems); and self‐care aids and equipment (e.g. grab bars or other self‐care aids). These interventions typically include information/education on reducing risk of falls specific to the intervention.
Education to reduce environmental fall risks in the home as a stand‐alone intervention. While often provided within an intervention, if given as a single environmental approach, education involves providing generalised information and/or learning activities to increase knowledge about environmental fall risks and/or enable self‐assessment audits of environmental hazards and no active intervention (Horowitz 2013). The intervention can be delivered in various ways, such as through booklets and other written materials, videos, lectures, and checklists including via mobile apps.
Home modifications to enable task performance, access, and functional independence. This stand‐alone intervention makes changes and adaptations to the permanent physical features of the home to meet the needs of people with activity limitations to performing daily living activities so that they can continue to live independently. The aim of these home visits is to remove barriers to function, enable ease of task performance, and improve accessibility (Aliberti 2019; Fange 2005; Wellecke 2022). The focus is primarily to improve task performance and independence of individuals with functional impairments or to reduce the demands on caregivers, or both (Carnemolla 2019; Stark 2009). Home modification interventions differ from a fall‐hazard reduction intervention which focuses primarily on fall prevention (Peterson 2008). Studies that include a component that is considered a fall‐hazard assessment or that includes specific fall‐hazard management strategies to reduce falls would be included in the first category of home fall‐hazards as a ‘package of strategies’.

Environmental interventions are delivered by various individuals, ranging from health professionals (occupational therapists, ergotherapists, optometrists, and nurses) to healthcare workers (care or support workers without a professional qualification). They can be conducted as part of a consultation (e.g. aids or education provided in hospital), tele‐rehabilitation (e.g. Sanford 2004), or, more often, conducted via a home visit.

We considered all types of environmental interventions and all delivery modalities.

Environmental interventions may be delivered as part of a multicomponent intervention (e.g. along with exercise or medication review). The impact of such programmes is investigated in a separate Cochrane Review (Hopewell 2018).

How the intervention might work

The quality, tailoring, and consequent outcomes of environmental assessment and intervention have been conceptualised in terms of person‐environment fit models (Gitlin 2003; Pighills 2016). Law's Person, Environment, Occupation (PEO) model proposes that it is the interaction and fit between the person’s capacity (e.g. strength, vision), their environment (e.g. the layout), and the occupation (e.g. task or activity) that is important, and when one of these three elements is affected or altered, this then impacts the others (Law 1997). Competence‐environmental press models elaborate further on the relationship between the person’s competencies and the demand placed on the individual by the physical environment (Nahemow 2000). These models support the idea that the person’s capacity and their task performance should be assessed and considered when determining environmental solutions. Observational studies show that the mere presence of a hazard is not associated with falling (Lord 2006; Pighills 2021); rather it is the interaction between capacity, environmental barriers, and tasks that results in falls. This further supports the relevance of assessment of all three elements to reduce fall risk.

Environmental interventions are likely to modify risk by adapting or changing the environment, removing fall hazards, or providing an assistive device to afford protection from risk of falling; by enabling people to mobilise and engage in activity in a safer way; by compensating for specific risk factors known to be predictive of falls (such as age‐related changes to motor and sensory systems, gait or vision impairment, cognitive impairment, and impairment or limitations related to chronic illness); by modifying risky behaviours; or by avoidance of hazardous situations. Alternatively, some interventions (new glasses, non‐slip modifications to outdoor shoes) might increase outdoor activity and exposure to risk and increase falls. Home modification interventions that primarily aim to improve independence and task performance of individuals with functional impairment or reduce the burden on caregivers may not include a sufficient focus on fall prevention.

Why it is important to do this review

This review covers a key set of interventions arising from the splitting of the scope of the previous Cochrane Review of falls, which covered a broad range of interventions (Gillespie 2012). Gillespie 2012 concluded that "Home safety interventions reduce rate of falls and risk of falling" and that an "anti‐slip shoe device for icy conditions significantly reduced winter outside falls in one study". Gillespie 2012 found several studies that tested other assistive devices as single interventions, with mixed reports of effectiveness.

An update of the effects of environmental interventions is warranted given that evidence is emerging and new trials have been published. With projected demographic changes and people living longer (He 2016), the number of older people at risk for falls living in the community is increasing, which will magnify the consequences associated with falls and fall‐related injuries to both the individual and to society.

Different environmental intervention types, different delivery approaches, or for whom, may have different effects on falls and fall‐related injuries, so careful analysis of the impact of these differences is crucial. This evidence is important for healthcare professionals, policymakers, consumers, researchers, and others with an interest in this topic. The formal evaluation of cost‐effectiveness, cost benefit, or cost utilisation of environmental interventions for falls prevention is also important for making informed decisions about application.

Objectives

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials, either individual or cluster randomised, that evaluated the effects of environmental interventions on the incidence of falls in older people living in the community. We excluded quasi‐randomised trials (e.g. allocation to groups by alternation or date of birth).

Types of participants

We included trials if they specified an inclusion criterion of 60 years of age or over. We included trials involving younger participants if the mean age minus one standard deviation (SD) was greater than 60 years. We included trials where the majority of participants were living in the community, either at home or in places of residence that, on the whole, do not provide residential health‐related care or rehabilitative services, for example hostels (in Australia), retirement villages, or sheltered housing. Trials with mixed populations (community and higher‐dependency places of residence) were eligible for inclusion if data were provided for subgroups based on setting, or the numbers in higher‐dependency residences were very few and balanced in the comparison groups. We included trials recruiting participants in hospital if the majority were discharged to the community and this was where most of the intervention was delivered and falls were recorded, or where the in‐hospital intervention or consultation was aimed at providing advice or education about home hazards postdischarge (and not about fall prevention in‐hospital).

We excluded studies that tested interventions for preventing falls in people after stroke and with Parkinson's disease, as these topic areas are covered in other Cochrane Reviews (Canning 2015; Verheyden 2013). We acknowledge that some individuals with these (and other) health conditions may be included in studies of the general community.

Types of interventions

This review included all trials of environmental interventions that measured falls in older people. We excluded multicomponent interventions.

The descriptions of interventions used in individual trials were examined and the intervention categorised according to the following primary categories. Appendix 1 provides a summary outline of the descriptors and their links to the ProFaNE taxonomy (Lamb 2005; Lamb 2007).

Home fall‐hazard reduction as a single strategy or package of strategies to reduce fall risk (indoors or outdoors environmental adaptations such as non‐slip strips on stairs, way‐finding lighting at night, removing clutter in traffic ways; handrails and lighting at entrances; behavioural changes such as using a step ladder instead of climbing on chairs to reach high places or holding a rail when descending stairs).
Assistive technology or devices as a single strategy to reduce fall risk (such as special footwear, sensory aids (e.g. prescription glasses), self‐care aid (e.g. grab rails)).
Education related to providing information, in paper or other formats, on fall risks related to the environment as a sole intervention to reduce fall risk.
Home modifications to enable task performance and functional independence.

Intervention scope, uptake, duration, degree of tailoring (comprehensive assessment of fall hazards; assessment of person and environment related to fall risk; priority setting with participant, follow‐up), and additional information or support were expected to vary in the trials and would be noted and reported in our review.

For our main comparisons, we included trials where the intervention was compared with 'usual care' (i.e. no change in usual activities) or a control intervention (attention control, i.e. an intervention that is not thought to reduce falls but provides attention (e.g. general health education or social visits)). We planned the following comparisons.

Home fall‐hazard reduction interventions (as a single intervention or package of components) versus control.
Assistive technology (as a single component intervention) (e.g. footwear modifications, vision modifications) versus control.
Education (as a stand‐alone intervention) versus control.
Home modifications that aim to enable function (e.g. access ramps, toilet or shower modifications for access) versus control.

Where appropriate, we compared interventions from the different categories. We selected the least intensive or complex intervention for the control group, for example information/education where compared with home fall‐hazard reduction.

We also considered comparisons of how an intervention was delivered, that is high‐ versus limited‐tailoring of intervention to falls, and of who delivered the intervention (e.g. occupational therapists versus other personnel who do not have qualifications and training in person‐environment fit). An intervention of higher tailoring to falls was defined as meeting most of the recognised criteria for best practice in conducting effective home‐fall hazard interventions, Clemson 2008; Clemson 2021; Pighills 2019, as described in the Description of the intervention section (e.g. use of comprehensive fall‐hazard assessment, evaluation of person and environment related to fall risk, and hazard identification and priority setting engaging the person). Limited tailoring to fall risk may include a checklist of limited hazards and little or no assessment of personal fall risk related to hazards.

Types of outcome measures

We included the following outcomes in this review.

Primary outcomes

Rate of falls (number of falls; falls per person‐year).

Secondary outcomes

Number of fallers (i.e. number of people experiencing one or more falls; risk of falling).
Number of participants experiencing one of more fall‐related fractures.
Number of participants experiencing one or more falls that resulted in hospital admission.
Number of participants experiencing one or more falls that required medical attention.
Health‐related quality of life (HRQoL) (measured using a validated scale, e.g. EQ‐5D (EuroQol Research Foundation; Herdman 2011) or similar).
Number of participants experiencing one or more adverse events as a consequence of the intervention (such as a fall as a result of a faulty grab rail).

We chose the rate of falls as the single primary outcome for ease of interpretation of the results of this review. Furthermore, rate of falls is likely to be more sensitive to change than the proportion of fallers, especially in samples with high fall rates. As falls are count data, dichotomisation to falling versus not falling represents a loss of information, therefore many trials use rate of falls as the primary outcome and negative binomial regression to compare the rates between intervention and control groups, as recommended by Robertson 2005.

We recorded and reported intervention adherence data where available for use in the interpretation of trial and review findings.

We extracted health economic data (cost utilisation, cost benefit, and cost‐effectiveness) where this information was available.

Timing of outcome measurement

We planned to assess outcomes at short‐term (18 months or less) and long‐term (longer than 18 months) follow‐up. For studies with 18 months or less of follow‐up, we used the longest duration reported and included the duration (from and to) in the summary of findings tables.

Search methods for identification of studies

Electronic searches

Our search extended the searches performed up to February 2012 in Gillespie 2012.

We searched the Cochrane Central Register of Controlled Trials (CENTRAL, CRS Web, to 29 January 2021, Issue 1), MEDLINE (Ovid MEDLINE(R) 1946 to 28 January 2021), Embase (1980 to 28 January 2021), CINAHL (Cumulative Index to Nursing and Allied Health Literature) (1937 to 29 January 2021), and OTseeker (Occupational Therapy Systematic Evaluation of Evidence) (to 28 January 2021), using tailored search strategies. We did not apply any language restrictions.

At the time of the search, CENTRAL was fully up‐to‐date with all records from the Cochrane Bone, Joint and Muscle Trauma Group’s Specialised Register, therefore it was not necessary to search this separately.

In MEDLINE, we combined subject‐specific search terms with the sensitivity‐ and precision‐maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2019). The search strategies are shown in Appendix 2.

We also searched the following trial registries for ongoing and recently completed trials: World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (apps.who.int/trialsearch/) (29 January 2021) and the US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (clinicaltrials.gov) (29 January 2021).

Searching other resources

We checked the reference lists of other systematic reviews. We also contacted researchers in the field to identify ongoing and unpublished trials.

Data collection and analysis

In order to reduce bias, we ensured that any review author who was a study author, or who has had an advisory role on any potentially relevant study, remained independent of study selection decisions, risk of bias assessment, and data extraction for their study.

Selection of studies

Pairs of review authors (of LC, AP, SS) independently screened the titles and abstracts of the citations retrieved by the searches for relevance. After this initial assessment, we obtained the full‐text copies of all studies deemed potentially relevant. Two review authors independently checked the full papers for eligibility, resolving any disagreements by consensus and the input of a third review author. We attempted to contact study authors where the eligibility of a study was unclear. We recorded the reasons for exclusion of studies obtained as full text. We summarised this process in a PRISMA flowchart. Where there were several reports of a study, we attempted to obtain all reports.

Data extraction and management

Pairs of review authors (of LC, SS, AP) independently performed data extraction using a data extraction form based on the one used in Gillespie 2012. We first piloted the data extraction form using a representative sample of studies in order to identify any missing items or unclear coding instructions. Any disagreements were resolved by consensus or third‐party adjudication. Review authors were not blinded to authors and sources.

We used the standardised data extraction form to record the following items.

General information: review author’s name, date of data extraction, study ID, first author of study, author’s contact address (if available), citation of paper, study dates, and trial objectives.
Trial details: trial design, location, setting, timeframe from start to completion, sample size, inclusion and exclusion criteria, comparability of groups, length of follow‐up, stratification, stopping rules, and funding source.
Risk of bias assessment: sequence generation, allocation concealment, blinding (participants, personnel, outcome assessors), incomplete outcome data, selective outcome reporting, and other bias (recall bias).
Characteristics of participants: age, gender, ethnicity, and number randomised, analysed, lost to follow‐up, and dropouts in each arm (with reasons).
Interventions: experimental and control interventions, timing of intervention, whether studies assessed adherence (compliance) with experimental and control interventions and associated data, who delivered the intervention, and additional co‐interventions (such as motivational strategies). We collected as much information as possible on control interventions, including assessing what 'usual care' comprised.
Type of intervention: intervention scope, uptake, duration, degree of tailoring (comprehensive evaluation of hazard identification and priority setting; validated assessment tool for broad range of hazards; assessment of person and environment related to fall risk; priority setting with participant, follow‐up).
Outcomes measured: rate of falls, number of fallers, number of people experiencing one or more fall‐related fractures, number of people who experienced one or more falls that resulted in hospital admission, number of participants who experienced one or more falls that required medical attention, HRQoL, and number of people experiencing one or more adverse events.
Other details: cost and cost‐effectiveness information.

We retrieved data from both full‐text and abstract reports of studies, including those with multiple reports. Where information was insufficient, we contacted the study authors for additional details.

Assessment of risk of bias in included studies

Pairs of review authors (of LC, AP, SS) independently assessed risk of bias using Cochrane's risk of bias tool (Higgins 2011). Any disagreements were resolved by consensus or third‐party adjudication (LC, CS).

As outlined in Appendix 3, we assessed the following risk of bias domains: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias, for falls and fractures separately), incomplete outcome data (attrition bias, for falls and fractures separately), and selective outcome reporting (reporting bias). We also assessed bias in the recall of falls due to unreliable methods of ascertainment (Hannan 2010). This is a form of detection bias, more often used in cohort studies, which has been used for evidence of accurate self‐reporting of falls in previous Cochrane Reviews (e.g. Gillespie 2012).

Specifically for trials using cluster randomisation, we planned to consider the risk of additional bias relating to recruitment, baseline imbalance, loss of clusters, incorrect analysis, and comparability with individually randomised trials, as described in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

We rated risk of bias as low, high, or unclear for each domain.

Measures of treatment effect

We reported the treatment effects for rate of falls, fractures, and falls requiring medical attention as a rate ratio (RaR) with 95% confidence intervals (CIs). For number of fallers, number of participants sustaining fall‐related fractures, and number of participants experiencing falls that required medical attention, we have reported risk ratios (RR) and 95% CIs.

The rate of falls is the total number of falls per unit of person‐time that falls were monitored (e.g. falls per person‐year). The RaR compares the rate of falls in any two groups during each trial. We used an RaR (e.g. incidence RaR or hazard ratio for all falls) with a 95% CI if these were reported in the paper. If both adjusted and unadjusted RaRs were reported, we used the unadjusted estimate, unless the adjustment was for clustering. If an RaR was not reported but appropriate raw data were available, we used a Microsoft Excel spreadsheet to calculate an RaR and 95% CI (Microsoft Excel). We used the reported rate of falls (falls per person‐year) in each group and the total number of falls for participants contributing data, or we calculated the rate of falls in each group from the total number of falls and the actual total length of time falls were monitored (person‐years) for participants contributing data. In cases where data were only available for people who had completed the study, or where the trial authors reported no losses to follow‐up, we assumed that these participants had been followed up for the maximum possible period.

For number of fallers, a dichotomous outcome, we used the RR as the treatment effect. The RR compares the number of people who fell once or more (fallers) between groups. We used a reported estimate of RR (hazard ratio for first fall, risk ratio (relative risk), or odds ratio) and 95% CI if available. If both adjusted and unadjusted estimates were reported, we used the unadjusted estimate, unless the adjustment was for clustering. If an odds ratio was reported, or an effect estimate and 95% CI was not reported, and appropriate data were available, we calculated an RR and 95% CI using SPSS software (IBM Corp 2019). For the calculations we used the number of participants contributing data in each group if this was known; if this was not reported, we used the number randomised to each group. We used the same approach for the number of people sustaining fractures, the number of people experiencing falls requiring medical attention, and the number of people experiencing adverse events.

For continuous outcomes (HRQoL), we presented the mean difference (MD) with 95% CIs where the same outcome measure was used, or standardised mean difference (SMD) with 95% CIs for outcomes measured using different scales. We only used results based on change scores if final values were unavailable.

Unit of analysis issues

For trials that were cluster randomised (e.g. by medical practice), we planned to perform adjustments for clustering as described in Higgins 2011 if this was not done in the published report, and to use an intraclass correlation coefficient (ICC) of 0.01 as reported in Smeeth 2002. We anticipated that trials would be unlikely to report details of clustering by household, and that clustering effects by household would be very small. Two trials in this review were cluster randomised; this was household clustering, and included a small portion of their sample where more than one older person from the household was included. Both of these trials also reported their main results adjusted for clustering.

For trials with multiple arms, we included multiple pair‐wise comparisons (intervention versus control) in analyses, but in order to avoid the same group of participants being included twice, we have 'split' the control group by distributing the number of control group participants to each analysis in proportion to the number of participants in each intervention group.

Dealing with missing data

It is inevitable that data will be missing for some participants in studies of older people given the increased risk of ill health, institutionalisation, and death. We attempted to contact study investigators regarding any unclear data or information on their trial; we only sought clarification of outcome data for number of falls or number of people who experienced falls. We conducted sensitivity analysis to explore the effects of missing data on trials with more than 20% loss to follow‐up. Where studies did not report SDs for continuous outcomes, we calculated these from standard errors and CIs. We did not impute missing SDs.

Assessment of heterogeneity

The decision about whether or not to combine the results of individual studies depends on an assessment of clinical and methodological heterogeneity. For studies that we considered to be sufficiently homogeneous in their study design and intervention, we carried out meta‐analyses and assessed the statistical heterogeneity. We assessed statistical heterogeneity of treatment effects between trials by visual inspection of the graphs, Chi² test with a significance level at P < 0.10, and the I² statistic. We based our interpretation of the I² results on that suggested by Higgins 2011: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% may represent very substantial ('considerable') heterogeneity.

Assessment of reporting biases

To explore the possibility of publication and other reporting biases, we constructed funnel plots for analyses that contained 10 or more data points. We also investigated the possibility of small‐study effects by excluding studies with fewer than 20 participants in each arm.

Data synthesis

We grouped similar environmental interventions for analysis using the classification system proposed, which has built upon and broadened the scope of the earlier ProFaNE taxonomy (Lamb 2005). For example, we only grouped assistive technology interventions if the technology or devices were similar.

Where appropriate, we pooled the results of comparable studies using random‐effects models. We used 95% CIs throughout. Where pooling of data was inappropriate, due to the diversity of interventions and contexts, we presented trial data in the analyses for illustrative purposes and reported these in the text.

Where appropriate, we pooled data using the generic inverse variance method in Review Manager 5 (Review Manager 2020). This method enables pooling of the adjusted and unadjusted treatment effect estimates (rate ratios or risk ratios) reported in the individual studies or that can be calculated from data presented in the published article (see Measures of treatment effect). The generic inverse variance option in Review Manager 5 requires entering the natural logarithm of the rate ratio or risk ratio and its standard error for each trial; we calculated these in a Microsoft Excel spreadsheet. For continuous measures (HRQoL), we presented MDs with 95% CIs where the same outcome measure was used, or SMDs with 95% CIs for outcomes measured using different scales.

To facilitate interpretation of the HRQoL outcome, we expressed SMD in the units of one of the measurement instruments used by the included studies, the EQ‐5D. Using the approach suggested by Schünemann 2022, we calculated the absolute difference in means by multiplying the SMD by an estimate of the SD associated with the EQ‐5D. We obtained this SD by calculating a weighted pre‐intervention average across all intervention groups of all studies that used this instrument. The summary effect, re‐expressed in the original units of the EQ‐5D, was compared with the minimally important difference reported by Walters 2005.

Where there was household clustering, we did not adjust where the study authors had adjusted for clustering in their analysis.

Subgroup analysis and investigation of heterogeneity

Where sufficient data were available, within all outcomes and categories of environmental interventions as outlined, we explored heterogeneity by carrying out the following prespecified subgroup analyses.

Higher versus lower falls risk at enrolment (i.e. comparing trials with participants selected for inclusion, or when available a pre‐planned analysis of higher risk at baseline, based on history of falling, recent hospitalisation, or other specific risk factors for falling, versus unselected participants).
For the home fall‐hazard intervention: high versus limited tailoring of intervention to falls (high‐tailoring interventions meet 75% of the identified criteria described in the Description of the intervention section and as reported by Clemson 2008).
Delivery by people with different qualifications (e.g. for the home fall‐hazard reduction intervention, trials that used an occupational therapist, ergotherapist, or equivalent versus those trials that were delivered by a healthcare worker who did not possess specific training in evaluating person‐environment fit).

We used the test for subgroup differences available in Review Manager 5 to determine whether there was evidence of a difference in treatment effect between subgroups (Review Manager 2020).

Sensitivity analysis

We assessed the robustness of our findings by conducting the following sensitivity analyses.

Risk of bias in included trials

To explore the possible impact of risk of bias on pooled estimates of treatment effects, we examined the impact on the results of the removal of studies that had a high risk of bias for the following specified domains:
- selection bias from inadequate generation of randomised sequence;
- selection bias from inadequate concealment of allocation;
- attrition bias due to incomplete outcome data;
- recall (detection) bias due to method of ascertaining fall outcomes. For fracture, medical‐attention falls, and falls requiring hospitalisation, we used blinding of outcome assessors as the detection bias domain.
If cluster‐randomised trials did not adequately adjust for clustering, we used sensitivity analyses to assess the effects of including/excluding these studies from analyses.
We planned to conduct sensitivity analysis for data reported at each time point, i.e. up to 18 months and greater than 18 months to 3 years.

Meta‐analysis decisions

We also examined the impact on the primary pooled estimates of treatment effect of the use of fixed‐effect rather than random‐effects models for data pooling.

Summary of findings and assessment of the certainty of the evidence

We used the GRADE approach to assess the certainty of the evidence as it relates to the primary and secondary outcomes listed in the Types of outcome measures section (Schünemann 2017). Using the GRADE approach, we assessed the certainty of the evidence as ‘high’, ‘moderate’, ‘low’, or ‘very low’ depending on the presence and extent of five factors: risk of bias, inconsistency of effect, indirectness, imprecision, and publication bias, as described below.

Risk of bias. We downgraded by one level when removing studies with high risk of bias for one or more of the selected four domains had a marked impact on results (if effect estimate changed by ≥ 25%). We did not downgrade if the point estimate stayed the same and the CIs became wider. We did not consider risk of bias items that were not related to the outcome of interest (e.g. items related to falls when evaluating GRADE for HRQoL).
Inconsistency. We downgraded where there was significant heterogeneity (I² > 75%) that could not be explained by emergence of credible subgroups. We also examined variation in effect estimates across studies and Chi² tests.
Indirectness. We downgraded where trials examined a limited version of the main review question with regard to population, intervention, comparison, or outcomes.
Imprecision. We downgraded when there were fewer than 400 participants for continuous outcomes, fewer than 300 events for dichotomous outcomes and the CI was wide, or the CI crossed no effect (1.0) and an estimate of clinically important effect (appreciable harm (1.25) or appreciable benefit (0.75)).
Publication bias. For outcomes that included 10 or more data points, we constructed and visually inspected funnel plots. In the case of fewer than 10 studies, we performed sensitivity analysis excluding studies with fewer than 20 participants in each arm.

We prepared summary of findings tables for the primary outcome and secondary outcomes for individual comparisons as described in the Types of interventions section. We prepared separate summary of findings tables for home fall‐hazard reduction interventions, assistive technology interventions, and education interventions. Regarding assistive technology interventions, we included data in the summary of findings table only for the vision improvement group, chosen as the interventions in this grouping were sufficiently cohesive to provide clinical implications. We used standardised qualitative statements to describe the different combinations of effect size and the certainty of evidence (Cochrane Norway 2017).

Results

Description of studies

Results of the search

We screened a total of 15,992 records from the following databases: CENTRAL (4199), MEDLINE (2096), Embase (4438), CINAHL Plus (1958), OTseeker (64), the WHO ICTRP (1184), and ClinicalTrials.gov (2053). We identified 45 potentially eligible studies from other sources.

After de‐duplication, 12,698 records remained. We excluded 12,659 records based on title and abstract and obtained the full texts of the remaining 39 records. We excluded 15 studies (see Characteristics of excluded studies). We also added one record to Characteristics of studies awaiting classification and identified one ongoing study. Excluded studies included environmental interventions that were delivered as part of a multicomponent intervention (e.g. along with exercise or medication review). The impact of such programmes is investigated in a separate Cochrane Review (Hopewell 2018).

We included 22 studies reported in 42 references. For a detailed description of our screening process, see Figure 1.

Included studies

This review includes 22 studies with 8463 randomised participants, contributing data for analysis from 7943 participants at follow‐up. For details of the included studies, see Characteristics of included studies; Table 4; Table 5. One study had two intervention arms and was included in the analysis of both assistive technology interventions and home fall‐hazard interventions (Day 2002). Fourteen studies assessed home fall‐hazard reduction interventions (Campbell 2005; Chu 2017; Cockayne 2021a; Cumming 1999; Day 2002; Keall 2015; Lannin 2007; Lin 2007; Lockwood 2019; Nikolaus 2003; Pardessus 2002; Pighills 2011; Stark 2021; Stevens 2001); eight studies assessed assistive technology interventions (Cumming 2007; Day 2002; Haran 2010; McKiernan 2005; Perry 2008; Tchalla 2013; Thomas 2018; Wang 2019); and one study assessed an education intervention (Kamei 2015). There were no studies that assessed home modifications for functional independence and reported falls as an outcome.

1. Study design, length of follow‐up, setting, and trial size.

Study ID	Study design	Length of follow‐up (months)	Setting	No. randomised	No. analysed (falls outcome)	% lost to follow‐up (falls outcome)
Home fall‐hazard reduction interventions
Campbell 2005	Parallel	12	New Zealand	196	196	0
Chu 2017	Parallel	12	Hong Kong	204	198	3%
Cockayne 2021a	Parallel	12	UK	1331	1308	2%
Cumming 1999	Parallel	12	Australia	530	530	0
Day 2002	Parallel (HH)	18	Australia	273	273	0
Keall 2015	Cluster by household^a	36	New Zealand	477	477	0
Lannin 2007	Parallel	3	Australia	10	10	10%
Lin 2007	Parallel	6	Taiwan	100	100	0
Lockwood 2019	Parallel	6	Australia	77	73	6%
Nikolaus 2003	Parallel	12	Germany	360	360	0
Pardessus 2002	Parallel	12	France	60	60	0
Pighills 2011	Parallel	12	UK	238	233	2%
Stark 2021	Parallel	12	USA	310	275	11%
Stevens 2001	Cluster by household^b	12	Australia	1879	1737	8%
Assistive technology interventions
Cumming 2007	Parallel	12	Australia	616	616	0
Day 2002	Parallel (V)	18	Australia	278	276	0
Haran 2010	Parallel	13	Australia	606	597	.02%
McKiernan 2005	Parallel	12	USA	113	109	4%
Perry 2008	Parallel	3	Canada	46	40	13%
Tchalla 2013	Parallel	12	France	96	96	0
Thomas 2018	Parallel	3.75	USA	626	371	41%
Wang 2019	Parallel	12	USA	44	35	20%
Education interventions
Kamei 2015	Parallel	12	Japan	130	110	15%

Open in a new tab

HH Home hazard data from study  V Vision data from study  ^aHouseholds of all ages were recruited; data reported in review where houses included one or more individuals aged 70 or over. ^bIncluded clustering by household (20% included spouses).

2. Key characteristics of participants and intervention scope.

Study ID	Age  (mean)	% women	High risk of falls	Intervention approach	Intervention delivered by relevant health professional	Tailoring of intervention to falls^a
Home fall‐hazard reduction interventions
Campbell 2005	84	68%	Yes	Home visit, follow‐up	Occupational therapist	High
Chu 2017	78	71%	Yes	Home visit, follow‐up	Occupational therapist	High
Cockayne 2021a	80	66%	No^b	Home visit, follow‐up	Occupational therapist	High
Cumming 1999	77	57%	Yes^c	Home visit, follow‐up	Occupational therapist	High
Day 2002 (HH)	76	60%	No	Home assessment, mailed recommendations/referral	No, trained assessor	Limited
Keall 2015	‐	‐	No	Home modified if prespecified hazard present.	Qualified builder	Limited
Lannin 2007	81	80%	No	Pre‐discharge home visit	Occupational therapist	Limited
Lin 2007	77	51%	Yes	Home visit, support with adaptations	No, public health worker	Limited
Lockwood 2019	82	71%	Yes	Pre‐discharge home visit, follow‐up	Occupational therapist	Limited
Nikolaus 2003	82	73%	Yes^d	Home visits (mean 2.6), follow‐up	Occupational therapist, accompanied by nurse or physical therapist	High
Pardessus 2002	83	78%	Yes	Home visit	Occupational therapist	High
Pighills 2011	79	67%	Yes	Home visit, follow‐up	Yes, occupational therapist group/no, domiciliary support worker group	High
Stark 2021	75	79%	Yes	2 sessions: assessment/intervention with offer of additional visit if needed for facilitating adaptations/training; 6‐month booster visit	Occupational therapist	High
Stevens 2001	76	53%	No	Home assessment	No, nurse	Limited
Assistive technology interventions
Cumming 2007	80	68%	No^c	Assessment, provision of glasses/vision intervention, follow‐up	Optometrist	Active
Day 2002 (V)	76	60%	No	Assessment, referral	No, trained assessor	Low
Haran 2010	80	65%	Yes^e	Provision of glasses, education	Optometrist	Active
McKiernan 2005	74	60%	Yes	Device provided, information.	‐	Active
Perry 2008	69	48%	Yes	Insole provided, intervention and control provided with walking shoes.	No	Active
Tchalla 2013	87	77%	Yes	Device installed, tele‐support.	‐	Active
Thomas 2018	76	‐	Yes^c	Daily fresh or weekly frozen meals over 15 weeks	‐	Active
Wang 2019	75	70%	No	Custom‐fitted orthosis provided.	Podiatrist	Active
Education interventions
Kamei 2015	‐	85%	No	1 education interactive session with self‐home audit	No, public health researcher	Limited

Open in a new tab

HH: home hazard data from Day 2002 V: vision data from Day 2002  ^aTailoring of intervention to falls characterised by the following:  Assistive technology interventions: Low – assessment and referral only. Active ‐ includes active intervention. Home fall‐hazard reduction interventions: High tailoring – provision and support with aids or equipment. High if met ≥ 75% of following criteria: a comprehensive evaluation process of environmental hazard identification and priority setting taking into account both personal fall risk, individual capacity, and assessment of the person's environment; use of an assessment tool validated for the broad range of potential fall hazards; inclusion of formal or observational evaluation of the functional capacity (physical capacity, behaviour, functional vision, habits) of the person within the context of their environment; and the provision of adequate follow‐up by the health professional with support for adaptations and modifications. Limited tailoring ‐ met less than 75% criteria. ^bStudy reported subgroup data for participants who had a higher or lower falls risk at enrolment defined as recent hospitalisation.  ^cStudy reported subgroup data for participants who had a higher or lower falls risk at enrolment defined as a history of falls. ^dStudy reported subgroup data for participants who had a higher or lower falls risk at enrolment defined as a history of multiple falls. ^eStudy reported subgroup data for participants who had a higher or lower falls risk at enrolment defined as exposure to outdoor mobility.

Two studies had three arms which were included in secondary analysis (Pighills 2011; Thomas 2018). One home fall‐hazard study contributed data only for the secondary analysis of falls requiring medical attention (n = 477 participants) (Keall 2015). This study was also the only one to report fall data for a period longer than 18 months. We contacted the authors of five studies for additional information regarding study design, Cockayne 2021a; Lockwood 2019; Pighills 2011, and confirmation of outcome data used in the review (Cockayne 2021a; Kamei 2015; Wang 2019). This does not include those studies in Gillespie 2012 that were contacted for additional information.

Funding sources were largely national or state government (14) and/or philanthropic organisations or foundations (7). Two studies were fully funded from hospital grants (Kamei 2015; Lannin 2007), and two were partially funded through higher degree research fellowship support (Lockwood 2019; Pighills 2011). One study had partial funding from an industry partner (Wang 2019), which was reported not to have been involved in the study at any stage, and one study reported no funding source (Pardessus 2002).

Home fall‐hazard interventions

Study design

All 14 included studies were randomised controlled trials with a parallel‐group design, with one study including a percentage of the sample (20%) from cluster households (older person living with the index recruit) (Stevens 2001), and one study using a cluster randomisation by household (Keall 2015). See Table 4.

Study size

The included studies involved a total of 5830 participants, ranging from 10 participants in Lannin 2007 to 1879 in Stevens 2001. The median number of participants was 256 (interquartile range (IQR) 124 to 448). See Table 4.

Study setting

The included studies were carried out in eight countries, the most common being Australia (five studies). Two studies each were conducted in the UK and New Zealand, and one study was conducted in each of France, Germany, Hong Kong, Taiwan, and the USA. See Table 4.

Participants

The included studies randomised a total of 6045 participants, with fall data at follow‐up available for 5830 participants, which included one study contributing data only for medical‐attention fall outcomes (Keall 2015, n = 477). Overall, 67% of included participants were women. The average age of participants was 79 years. Nine studies included participants judged to be at higher risk of falls at enrolment (i.e. participants were selected for inclusion based on history of falling or other specific risk factors for falling) (Campbell 2005; Chu 2017; Cockayne 2021a; Cumming 1999; Lin 2007; Lockwood 2019; Nikolaus 2003; Pighills 2011; Stark 2021), and six studies included participants not judged to be a higher risk of falls (i.e. participants were not selected for inclusion based on history of falling or other specific risk factors for falling) (Cockayne 2021a; Cumming 1999; Day 2002; Lannin 2007; Nikolaus 2003; Stevens 2001). Of these studies, three contributed data to both high and not‐high risk of falls at enrolment (Cockayne 2021a; Cumming 1999; Nikolaus 2003). The study reporting only medical‐attention falls recruited people who lived in older housing (prior to 1978) and were on social benefits. The inclusion/exclusion criteria and other participant details are listed in Characteristics of included studies and Table 5.

Interventions

Most fall‐hazard reduction studies conducted one home visit with varying elements of the environmental descriptors (Appendix 1), with or without follow‐up for adaptations or modifications. In two studies, there was the option of a second visit if required (Campbell 2005; Cockayne 2021a), and two studies offered additional follow‐up home visits (Nikolaus 2003, mean 2.6 home visits; Stark 2021, two assessment/intervention visits with offer of additional visits for training/facilitating modifications plus a six‐month booster). In seven studies, participants were recruited in the community (Campbell 2005; Cockayne 2021a; Day 2002; Lin 2007; Pighills 2011; Stark 2021; Stevens 2001). In five studies, participants were recruited in hospital, and one in the emergency department (Chu 2017). Of those in hospital, one study conducted fall prevention interventions while in the community posthospitalisation and also supplemented recruitment from outpatient clinics and day centres (Cumming 1999); three studies were conducted as a pre‐discharge home visit with the focus on transition home from hospital/independence in daily activities (Lannin 2007; Lockwood 2019; Pardessus 2002); and one study included both pre‐discharge and follow‐up fall prevention home visits (Nikolaus 2003). In the Keall 2015 study, the focus of the intervention was entirely on the house, where builders modified specific hazards when present. We identified no studies that investigated fall‐hazard reduction in public places.

Outcomes

The source of data used for calculating outcomes for each study for generic inverse variance analysis is shown Appendix 4, and the source of raw data for rate and risk of falls in Appendix 5. Rate of falls was reported in seven studies, and could be calculated in a further five. Data on risk of falling (number of fallers) were available for four studies and could be calculated in a further eight.

Two studies reported fracture outcome (Cockayne 2021a; Nikolaus 2003); three studies reported number of people who experienced one or more falls requiring hospital admissions (Chu 2017; Lockwood 2019; Pardessus 2002); three studies reported the number of falls requiring medical attention (Campbell 2005; Day 2002; Keall 2015); and two studies reported the number of people requiring medical attention (Cumming 1999; Lannin 2007). Five studies reported an HRQoL outcome using a validated outcome measure (Cockayne 2021a; Lin 2007; Lockwood 2019; Pighills 2011; Stark 2021).

Ten studies reported adherence to intervention (Campbell 2005; Chu 2017; Cockayne 2021a; Cumming 1999; Day 2002; Lockwood 2019; Nikolaus 2003; Pighills 2011; Stark 2021; Stevens 2001).

Two studies reported adverse events related to the intervention only (Campbell 2005; Stark 2021), and two studies reported this information for both intervention and control (Cockayne 2021a; Lockwood 2019).

All studies reported data within 18 months of the intervention, except Keall 2015, which contributed data at 36 months and only for falls requiring medical attention.

Assistive technology interventions

Study design

All eight included studies were randomised controlled trials with a parallel‐group design. See Table 4.

Study size

The studies included a total of 2425 participants, ranging from 44 in Wang 2019 to 626 in Thomas 2018. The median number of participants was 196 (IQR 84 to 609). See Table 4.

Study setting

The included studies were carried out in four countries, Australia and the USA with three each, and Canada and France with one study each.

Participants

The included studies randomised a total of 2425 participants, with fall data at follow‐up available for 2140 participants. Overall, 64% of included participants were women. The average age of participants was 77 years. Five studies included participants judged to be at higher risk of falls at enrolment (i.e. participants were selected for inclusion based on history of falling or other specific risk factors for falling) (Haran 2010; McKiernan 2005; Perry 2008; Tchalla 2013; Thomas 2018), and three studies included participants not judged to be a higher risk of falls (i.e. participants were not selected for inclusion based on history of falling or other specific risk factors for falling) (Cumming 2007; Day 2002; Wang 2019). Three studies included a subgroup of higher risk of falling, history of falling, (Cumming 2007; Thomas 2018), or exposure to falling due to frequency of outdoor activity (Haran 2010). The inclusion/exclusion criteria and other participant details are listed in Characteristics of included studies and Table 5.

Interventions

Assistive technology interventions were a diverse group of interventions that focused on three themes, as follows.

Vision improvement (3 studies): in one study vision was tested by an optometrist and glasses provided or referred for a vision intervention (e.g. cataract referral) (Cumming 2007); one study used a trained assessor to test vision and provided a written referral for usual eye care provider or general practitioner (Day 2002); and in one study an optometrist assessed vision and provided participants with a pair of single‐lens distance glasses and, if needed, updated multifocal lens, with engagement to increase follow through (Haran 2010).
Footwear and foot devices (3 studies): one study provided custom‐fitted bilateral ankle‐foot orthosis (Wang 2019), and another study provided insoles designed to improve lateral stability during walking (Perry 2008). The third study provided an 'over‐shoe' device to reduce risk of slipping while walking outdoors in winter (McKiernan 2005).
Self‐care and assistive devices (2 studies): one study installed an automatic nightlight path activated by a floor sensor near the bed in participant homes coupled with a tele‐assistance service (Tchalla 2013), while the other study tested two interventions of delivering i) daily meals or ii) once‐weekly frozen meals to participant homes (Thomas 2018).

Outcomes

The source of data used for calculating outcomes for each study for generic inverse variance analysis is shown in Appendix 4. Rate of falls was reported in three studies, and could be calculated in a further three studies. Data on risk of falling (number of fallers) were available for two studies and could be calculated for a further five. Three studies reported fracture outcome (Cumming 2007; Haran 2010; McKiernan 2005); two studies reported the number of people requiring medical attention (Day 2002; Lockwood 2019); and one study reported an HRQoL outcome (Haran 2010).

All studies but one reported adherence to intervention (Wang 2019). Two studies reported adverse events related to the intervention only, and two studies reported this information for both intervention and control.

All studies reported data within 18 months of the intervention.

Education interventions

Study design, size, and setting

The one included study of an education intervention to reduce environmental fall risk was a randomised controlled trial with a parallel‐group design (Kamei 2015). The study included 130 community‐residing participants aged 65 years and over, of whom 85% were women. The study was conducted in Tokyo, Japan, between 2008 and 2011.

Intervention and outcomes

In this intervention participants received one educational interactive session using a model mock‐up of a typical Japanese home to raise awareness of home hazards followed by a self‐home audit. We were able to calculate rate of falls, risk of falling, and risk of fracture. The study reported data within 12 months of the intervention.

Excluded studies

We excluded 15 reports on full‐text review (for details, see Characteristics of excluded studies). Nine studies were either multifactorial or multicomponent interventions (Ferrer 2014; Greenberg 2020; Hill 2019; Johansson 2018; Lee 2013; Luck 2013; Moller 2014; Tannenbaum 2019; Wesson 2013); three studies were not randomised study designs (Casteel 2020; Oddsson 2020; Tiefenbachova 2019); two studies did not meet the review intervention descriptors (Liu 2020; Yamashita 2019); and one study was a younger age cohort (Deems‐Dluhy 2021).

Studies awaiting classification

There is one study awaiting classification (Stark 2016). This study assesses a home modification intervention that includes training in assistive technology and daily living activities with the aim of improving functional independence. An abstract was published in 2016 and a paper outlining the process outcomes in 2019; the main results paper is now submitted pending revision.

Ongoing studies

There is one ongoing study (Hatton 2019). This published protocol outlines a study testing textured shoe insoles to improve sensory input and balance and reduce falls in adults with diabetic peripheral neuropathy. Falls will be measured.

Risk of bias in included studies

Details of the risk of bias assessment across all included studies and each individual item are shown in the Characteristics of included studies and are visually represented in Figure 2 and Figure 3.

Risk of bias summary: review authors’ judgements about each risk of bias item for each included study. Blank spaces indicate that risk of bias assessment was not completed because the study did not report this outcome.

Overall risk of bias: review authors’ judgements about each risk of bias item presented as percentages across all included studies.

Allocation

Of the 22 included studies, we judged risk of bias for generation of the allocation sequence as low in 64% (n = 14) of trials; unclear in 23% (n = 5); and high in three trials. We assessed risk of bias for methods of concealment of the allocation prior to group assignment as low in 46% (n = 10); unclear in 50% (n = 11); and high in one study.

Blinding

Blinding of participants and personnel

Due to the nature of the interventions, it was not possible to blind participants and personnel to the allocated group. As the likelihood of awareness of group allocation introducing performance bias was not clear, we assessed risk of bias for this domain as unclear for all included studies.

Blinding of outcome assessment

We assessed risk of detection bias separately for rate of falls and risk of falling, risk of fractures, falls requiring hospital admission or medical attention, and HRQoL.

Rate of falls and risk of falling

We judged the risk of detection bias in relation to the methods of ascertainment of the rate or risk of falls, or both, to be low in 45% (n = 10); unclear in 45% (n = 10); and high in 9% (n = 1) of the included studies.

Risk of fractures

We judged the risk of detection bias in relation to the methods of ascertainment of fractures in trials to be high in 100% (n = 6/6) of the included studies that measured fractures.

Requiring hospital admission/medical attention

In studies of home fall‐hazard interventions reporting on the risk of hospital admission or requiring medical attention, or both, we judged the risk of detection bias in relation to the method of ascertainment of these outcomes to be low in 33% (3/9) of studies and high in 66% (6/9) of studies.

Health‐related quality of life

In studies that reported on HRQoL, we judged the risk of detection bias in relation to the method of ascertainment of HRQoL to be low in 4/6 studies and unclear in 2/6 studies.

Incomplete outcome data

We judged risk of attrition bias (fall outcomes) to be low in 64% (n = 14); unclear in 27% (n = 6); and high in 9% (n = 2) of the included studies.

Selective reporting

We judged risk of reporting bias to be low in 36% (n = 8) and unclear in 64% (n = 14) of the included studies.

Other potential sources of bias

Bias in the recall of falls due to less reliable methods of ascertainment

We assessed 68% of the included studies (n = 15) as being at low risk of bias for the recall of falls (i.e. falls were recorded concurrently using recommended methods of monthly diaries or postcards). We judged the risk of bias to be unclear in four studies (18%), as retrospective recall was for a short period only. We judged the risk of bias to be high in 14% of studies (n = 3), in that ascertainment of falling episodes was by participant recall, at intervals longer than one month during the study.

Cluster‐randomised trials

We judged the two cluster‐randomised trials (by household) as at low risk of bias due to clustering. Recruitment occurred prior to randomisation; comparability of cluster recruit to index recruit (relevant to one study) was reported; there were no loss of cluster or missing outcomes; and in both studies the analysis included adjusting for clustering. Results were comparable with other trials.

Effects of interventions

See: Table 1; Table 2; Table 3

Overview of results

Home fall‐hazard reduction interventions versus control

Fourteen studies compared home fall‐hazard reduction with usual care (i.e. no change in usual activities or an attention control intervention which is not thought to reduce falls such as social visits or general health education). One study provided data only for rate of falls (Lin 2007), and one study provided data only for risk of falling (Pardessus 2002). One study provided data for only one comparison, that is falls at home requiring medical attention (Keall 2015).

Primary outcome

See Table 1.

Rate of falls (falls per person‐year)

Home fall‐hazard reduction interventions probably reduce the rate of falls by 26% compared with control (rate ratio (RaR) 0.74, 95% confidence interval (CI) 0.61 to 0.91; I² = 83%; 12 studies, 5293 participants; moderate‐certainty evidence; Analysis 1.1). We downgraded the certainty of the evidence one level for inconsistency due to significant heterogeneity.

1.1 — Comparison 1: Home fall‐hazard reduction versus control: primary outcome: rate of falls, Outcome 1: Rate of falls ‐ overall analysis

Subgroup analysis by risk of falling at baseline found a reduced rate of falls where participants were at an increased risk of falling at baseline (RaR 0.62, 95% CI 0.56 to 0.70; I² = 38%; 9 studies, 1513 participants; high‐certainty evidence) compared with not selecting participants for increased risk of falling at baseline (RaR 1.05, 95% CI 0.96 to 1.16; I² = 0%; 6 studies, 3780 participants; high‐certainty evidence); test for subgroup differences P < 0.001 (Analysis 1.2). The evidence was of high certainty. The subgroup analyses included a mix of between‐study subgroup data and within‐study subgroup data (the latter from three studies; Cockayne 2021a; Cumming 1999; Nikolaus 2003), with details for all studies included in footnotes. This analysis provides credible support for explaining the inconsistency in heterogeneity found in the overall analysis.

1.2 — Comparison 1: Home fall‐hazard reduction versus control: primary outcome: rate of falls, Outcome 2: Rate of falls ‐ subgrouped by risk of falls at baseline

Other subgroup analysis

Subgroup analysis of difference between tailoring interventions demonstrated no significant interaction effect in formal tests (P = 0.06 when the intervention was highly tailored to falls (RaR 0.69, 95% CI 0.52 to 0.91; I² = 86%; 8 studies, 3100 participants; moderate‐certainty evidence, downgraded due to inconsistency) compared to those with limited tailoring of interventions (RaR 0.95, 95% CI 0.78 to 1.15; I² = 24%; 5 studies, 2193 participants; high‐certainty evidence) (Analysis 1.3).
Subgroup analysis of personnel delivering the intervention demonstrated no significant interaction effect in formal tests (P = 0.11) when the intervention was delivered by an occupational therapist (RaR 0.68, 95% CI 0.51 to 0.91; I² = 84%; 9 studies, 3074 participants; moderate‐certainty evidence, downgraded due to inconsistency) compared to when the intervention was delivered by other personnel (RaR 0.91, 95% CI 0.74 to 1.12; I² = 44%; 4 studies, 2219 participants; high‐certainty evidence) (Analysis 1.4).

1.3 — Comparison 1: Home fall‐hazard reduction versus control: primary outcome: rate of falls, Outcome 3: Rate of falls ‐ subgrouped by tailoring of intervention to falls

1.4 — Comparison 1: Home fall‐hazard reduction versus control: primary outcome: rate of falls, Outcome 4: Rate of falls ‐ subgrouped by personnel

Secondary outcomes

See Table 1.

Number of participants who experienced one or more falls (risk of falling)

Home fall‐hazard reduction interventions probably reduce the number of people experiencing one or more falls by 11% compared with control (risk ratio (RR) 0.89, 95% CI 0.82 to 0.97; I² = 18%; 12 studies, 5253 participants; moderate‐certainty evidence; Analysis 2.1). We downgraded the certainty of the evidence one level for inconsistency. While the overall analysis had low levels of statistical heterogeneity (I² = 18%), subgroup analysis revealed differences according to baseline risk of falling. There was no further downgrading after removal of one study with high risk of bias for one domain, as this did not impact the effect estimate.

2.1 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 1: Number of fallers ‐ overall analysis

Subgroup analysis

Subgroup analysis by risk of falling at baseline found a reduced rate of falls where participants were at an increased risk of falling (RR 0.74, 95% CI 0.65 to 0.85; I² = 13%; 9 studies, 1473 participants) compared with studies not selecting participants for increased risk of falling at baseline (RR 0.99, 95% CI 0.92 to 1.07; I² = 0%; 6 studies, 3780 participants); test for subgroup difference P < 0.001. The evidence is of high certainty (Analysis 2.2).
Subgroup analysis of difference between high tailoring of interventions and limited tailoring of interventions demonstrated no significant interaction effect in formal tests (P = 0.27) when the intervention was highly tailored to falls prevention (RR 0.83, 95% CI 0.73 to 0.94; I² = 21%; 8 studies, 3100 participants; high‐certainty evidence) compared to studies where the intervention had limited tailoring to falls (RR 0.91, 95% CI 0.81 to 1.03; I² = 0%; 5 studies, 2153 participants; high‐certainty evidence) (Analysis 2.3).
Subgroup analysis of difference for personnel delivering the interventions demonstrated no significant interaction effect in formal tests (P = 0.12) when the intervention was delivered by an occupational therapist (RR 0.80, 95% CI 0.69 to 0.93; I² = 48%; 10 studies, 2881 participants; high‐certainty evidence) compared to studies where the intervention was delivered by other personnel (RR 0.93, 95% CI 0.82 to 1.05; I² = 0%; 3 studies, 2118 participants; high‐certainty evidence) (Analysis 2.4).

2.2 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 2: Number of fallers ‐ subgrouped by risk of falls at baseline

2.3 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 3: Number of fallers ‐ subgrouped by tailoring of intervention to falls

2.4 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 4: Number of fallers ‐ subgrouped by personnel

Number of participants who experienced one or more fall‐related fractures

Only two studies reported this outcome. We are uncertain of the finding that home fall‐hazard interventions make little or no difference to the risk of people experiencing a fall‐related fracture compared to usual care (RR 1.00, 95% CI 0.98 to 1.02; I² = 50%; 2 studies, 1668 participants; Analysis 2.5); the evidence was of very low certainty, downgraded one level due to imprecision (low event rates) and two levels for risk of bias, as both the included studies had high risk of (detection) bias of fractures. See Appendix 6 for a summary of reported data.

2.5 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 5: Number of people experiencing a fall‐related fracture

Number of participants who experienced one or more falls that resulted in hospitalisation

Three studies reported this outcome. Home fall‐hazard interventions may make little or no difference to the number of people who experience one or more falls requiring hospital admission compared with control (RR 0.96, 95% CI 0.87 to 1.06; I² = 48%; 3 studies, 325 participants; Analysis 2.6); the evidence was of low certainty, downgraded due to high risk of bias (detection) in two of the studies (changing effect estimate to be significant) and imprecision (few events). See Appendix 6 for a summary of reported data.

2.6 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 6: Number of people who experienced 1 or more falls requiring hospital admission

Rate of falls requiring medical attention

Three studies reported this outcome. Home fall‐hazard interventions may make little or no difference to requiring medical attention following a fall compared with control (RaR 0.91, 95% CI 0.58 to 1.43; I² = 63%; 3 studies, 946 participants; Analysis 2.7); low‐certainty evidence, downgraded two levels due to imprecision (wide CI crosses both appreciable benefit and harm, low number of events). One study had a three‐year follow‐up period (Keall 2015). Sensitivity analysis by removing this study increased heterogeneity (I² = 81%). We did not downgrade for inconsistency related to cluster randomisation, as this study was judged to be at low risk of bias (Keall 2015). See Appendix 6 for a summary of reported data.

2.7 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 7: Rate of falls requiring medical attention

Number of participants who experienced one or more falls that required medical attention

Only two studies reported this outcome. We are uncertain of the finding that home fall‐hazard interventions make little or no difference in the number of people who experience one or more falls requiring medical attention compared to control (RR 0.89, 95% CI 0.56 to 1.42; I² = 0%; 2 studies, 216 participants; Analysis 2.8); very low‐certainty evidence, downgraded two levels due to imprecision (wide CI crosses both appreciable benefit and harm and few events) and a further level for high risk of detection bias in both studies. See Appendix 6 for a summary of reported data.

2.8 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 8: Number of people who experienced 1 or more falls requiring medical attention

Health‐related quality of life

Five studies reported HRQoL. Home fall‐hazard reduction interventions probably result in little important difference to HRQoL compared to control (standardised mean difference 0.09, 95% CI −0.10 to 0.27; I² = 53%; 5 studies, 1848 participants); moderate‐certainty evidence, downgraded one level for imprecision as wide CI crosses both appreciable benefit and harm (Analysis 2.9). Re‐expressing the results using the EQ‐5D, there was probably no meaningful difference to HRQoL compared to control, with an increase in EQ‐5D score of 0.02 on the 0‐to‐1 scale in the intervention group, where minimally important difference for the EQ‐5D is typically 0.074 (Walters 2005). The 95% CI includes both worse HRQoL and improved HRQoL (95% CI −0.04 to 0.05). Appendix 7 provides a summary of HRQoL reported in studies.

2.9 — Comparison 2: Home fall‐hazard reduction versus control: secondary outcomes, Outcome 9: Health‐related quality of life ‐ overall analysis

Number of participants who experienced one or more adverse events

Ten studies did not report adverse events. Two studies reported this outcome for intervention and control, finding no adverse events (Cockayne 2021a; Lockwood 2019); two studies reported this outcome for intervention only and found no adverse events.

Adherence

Of the 10 studies reporting adherence, most (n = 7) reported partial or complete adherence with one or more of the recommendations ranging from 76% to 91%, except for Stevens 2001, which averaged 27% of the six hazards in their limited checklist. Three studies largely limited reporting to receiving the intervention or not (Cockayne 2021a; Day 2002; Lockwood 2019), averaging 88% to 91% (Appendix 8).

Assistive technology interventions versus control

Eight studies compared assistive technologies with usual care (i.e. no change in usual activities, or an attention control intervention which is not thought to reduce falls such as social visits or general health education). We pooled data for three studies of vision improvement, as these contribute meaningful data; see Table 2. We did not pool data for the remaining five studies (three footwear/foot devices; two self‐care) due to the diversity of these interventions, and have presented these results as a narrative summary.

Assistive technology: vision improvement

Primary outcomes

Rate of falls (falls per person‐year)

Three studies reported this outcome. Vision improvement interventions may make little or no difference to the rate of falls (RaR 1.12, 95% CI 0.84 to 1.50; I² = 79%; 3 studies, 1489 participants; Analysis 3.1); low‐certainty evidence, downgraded due to inconsistency with substantial heterogeneity and imprecision with CIs crossing both no effect and appreciable harm. Heterogeneity could not be explained by subgroup analysis according to baseline risk of falling (formal test of interaction effect P = 0.70) (Analysis 3.2).

3.1 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 1: Rate of falls ‐ overall analysis

3.2 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 2: Rate of falls ‐ subgrouped by risk of falls at baseline

Secondary outcomes

Number of participants who experienced one or more falls (risk of falling)

Vision improvement interventions may make little or no difference to the number of people experiencing one or more falls compared to control (RR 1.09, 95% CI 0.79 to 1.50; I² = 88%; 3 studies, 1489 participants; Analysis 3.3); low‐certainty evidence, downgraded two levels due to inconsistency (I² = 88%; little overlap of CIs).

3.3 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 3: Number of fallers ‐ overall analysis

Number of participants who experienced one or more falls that resulted in fracture

Two studies reported this outcome. We are uncertain of the findings for vision improvement interventions due to imprecision (few events, and CI crosses appreciable benefit and appreciable harm) and high risk of (detection) bias of fractures in both studies (RR 1.22, 95% CI 0.75 to 2.00; 2 studies, 976 participants; very low‐certainty evidence; Analysis 3.4). See Appendix 6 for a summary of reported data.

3.4 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 4: Number of people experiencing 1 or more fall‐related fractures

Number of participants experiencing one or more falls that resulted in hospital admission

No studies reported this outcome.

Rate of falls (number of falls) that required medical attention

One study reported this outcome. The evidence for whether vision improvement makes little or no difference in the rate of falls requiring medical attention versus control is very uncertain (RaR 0.93, 95% CI 0.48 to 1.82; 1 study, 276 participants; Analysis 3.5); very low‐certainty evidence due to imprecision (few events; wide CI covers appreciable harm and benefit) and high risk of (detection) bias. See Appendix 6 for a summary of reported data.

3.5 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 5: Rate of falls requiring medical attention

Health‐related quality of life

One study reported HRQoL. Vision assessment and provision of single‐lens glasses may make little or no difference to HRQoL compared to control (multifocals) (mean difference 0.40, 95% CI −1.12 to 1.92; 1 study, 597 participants; Analysis 3.6); low‐certainty evidence due to imprecision (wide CI and small number of participants). See Appendix 7 for a summary of reported data.

3.6 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 6: Health‐related quality of life

Number of participants who experienced one or more adverse events

One study reported falls while switching glasses (intervention n = 2; control n = 1; 597 participants). Switching glasses between single‐lens and mutifocal lens may make little or no difference to fall outcomes (RR 1.00, 95% CI 0.98 to 1.02; Analysis 3.7); low‐certainty evidence due to imprecision (small number of events) (Haran 2010).

3.7 — Comparison 3: Assistive technology versus control ‐ vision improvement, Outcome 7: Number of people who experienced 1 or more adverse events (fall after switching glasses)

Adherence

Haran 2010 reported that adherence to wearing single‐lens glasses when walking outdoors was 41% at 12 months, and Day 2002 reported 52% follow‐through with recommended referral for vision test. Cumming 2007 reported that 44% of participants received some form of vision treatment. See Appendix 8.

Assistive technology: footwear and foot devices

The following is a narrative of the results for this thematic intervention group.

Primary outcome

Rate of falls (falls per person‐year)

Two studies provided data for the primary analysis of rate of falls. An 'over‐shoe' device for winter intervention may reduce the rate of falls compared to control (RaR 0.42, 95% CI 0.22 to 0.78; 1 study, 109 participants; Analysis 4.1); low‐certainty evidence due to imprecision (small number of events). We are uncertain as to whether a bilateral foot orthosis intervention makes little or no difference to the rate of falls (RaR 0.89, 95% CI 0.42 to 1.87; 1 study, 35 participants; Analysis 4.1); very low‐certainty evidence due to imprecision (small number of events and wide CI crossing both appreciative benefit and harm) and high risk of (selection, attrition, recall) bias in this single study.

4.1 — Comparison 4: Assistive technology versus control ‐ footwear, self‐care and assistive devices, Outcome 1: Rate of falls ‐ overall analysis

Secondary outcomes

Number of participants who experienced one or more falls (risk of falling)

Two studies provided data for the secondary analysis of fall risk. We are uncertain as to whether balance‐enhancing insoles reduce the number of people experiencing a fall (RR 0.73, 95% CI 0.46 to 1.17; 1 study, 40 participants; Analysis 4.2); very low‐certainty evidence due to imprecision (low number of events; wide CI crosses appreciable benefit and no effect). We are uncertain as to whether a bilateral foot orthosis intervention makes little or no difference to the number of people falling (RR 0.76, 95% CI 0.42 to 1.40; 1 study, 35 participants; Analysis 4.2); very low‐certainty evidence due to imprecision (small number of events and wide CI crossing both appreciative benefit and harm) and high risk of (selection, attrition, recall) bias in this single study.

4.2 — Comparison 4: Assistive technology versus control ‐ footwear, self‐care and assistive devices, Outcome 2: Number of fallers ‐ overall analysis

Number of participants who experienced one or more falls that resulted in fracture

We are uncertain as to whether an ‘over‐shoe’ device makes little or no difference to the risk of fracture compared to control (RR 0.98, 95% CI 0.02 to 48.63; 1 study, 109 participants; Analysis 4.3); very low‐certainty evidence due to imprecision (no events reported; wide CI crossing appreciable benefit and no effect) and high risk of (detection) bias in this single study.

4.3 — Comparison 4: Assistive technology versus control ‐ footwear, self‐care and assistive devices, Outcome 3: Number of people experiencing 1 or more fall‐related fractures

Number of participants experiencing one or more falls that resulted in hospital admission

No studies reported this outcome.

Rate of falls (number of falls) or risk of falls (number of people) that required medical attention

An ‘over‐shoe’ device in winter may make little or no difference to the number of people who experience a fall (RR 0.96, 95% CI 0.91 to 1.02; 1 study, 109 participants; Analysis 4.4); low‐certainty evidence due to imprecision (small number of events) and high risk of (detection) bias in this single study.

4.4 — Comparison 4: Assistive technology versus control ‐ footwear, self‐care and assistive devices, Outcome 4: Number of people requiring medical attention

Health‐related quality of life

No studies reported this outcome.

Number of participants who experienced one or more adverse events

One study reported no adverse events in the intervention group, with minor discomfort from shoe inserts initially (n = 20 participants).

Adherence

The three studies reporting adherence used variable measures, though were reportedly high, for example 78% wearing the device (McKiernan 2005), to 96% acceptance of the new technology (Tchalla 2013) (Appendix 8).

Assistive technology: self‐care and assistive devices

The following is a narrative of the results for this thematic intervention group.

Primary outcome

Rate of falls (falls per person‐year)

One study provided data for the primary analysis of rate of falls. A bed transfer safety system intervention may make little or no difference to the rate of falls (RaR 0.67, 95% CI 0.42 to 1.07; 1 study, 96 participants; Analysis 4.1); low‐certainty evidence due to imprecision (small number of events) and high risk of (recall) bias in this single study.

Secondary outcomes

Number of participants who experienced one or more falls (risk of falling)

One study provided data for the secondary analysis of risk of falling only, with two intervention arms. We are uncertain as to whether delivering meals daily reduces the number of people experiencing a fall compared to control (RR 0.55, 95% CI 0.46 to 0.65; 202 participants) or whether delivering frozen meals weekly makes little or no difference to the number of people experiencing a fall (RR 0.98, 95% CI 0.81 to 1.19; 169 participants); very low‐certainty evidence due to imprecision (small number of events) and high risk of (selection, attrition, recall) bias in this single study (Analysis 4.2).

Number of participants experiencing one of more fall‐related fractures

No studies reported this outcome.

Number of participants experiencing one or more falls that resulted in hospital admission

No studies reported this outcome.

Number of participants experiencing one or more falls that required medical attention

No studies reported this outcome.

Health‐related quality of life

No studies reported this outcome.

Number of participants who experienced one or more adverse events

No studies reported this outcome.

Adherence

One study reported adherence, which was high for both intervention arms with 96% of meals being delivered (Thomas 2018).

Education interventions versus control

See Table 3.

Primary outcomes

Rate of falls (falls per person‐year)

We are uncertain as to whether an education intervention to reduce home fall hazards reduces the rate of falls (RaR 0.63, 95% CI 0.37 to 1.07; 1 study, 110 participants; Analysis 5.1); very low‐certainty evidence due to imprecision (small number of events; CI crossing appreciable benefit and no effect) and high risk of (selection) bias in this single trial.

5.1 — Comparison 5: Education versus control , Outcome 1: Rate of falls ‐ overall analysis

Secondary outcomes

Number of participants who experienced one or more falls (risk of falling)

We are uncertain as to whether an education intervention to reduce home fall hazards reduces the number of people experiencing one or more falls (RR 0.84, 95% CI 0.64 to 1.11; 1 study, 110 participants; Analysis 5.2); very low‐certainty evidence due to imprecision (small number of events; CI crossing appreciable benefit and no effect) and high risk of (selection) bias in this single trial.

5.2 — Comparison 5: Education versus control , Outcome 2: Number of fallers ‐ overall analysis

Number of participants who experienced one or more falls that resulted in fracture

An education intervention to reduce home fall hazards may make no or little difference to the rate of fractures (RR 1.02, 95% CI 0.96 to 1.08; 1 study, 110 participants; Analysis 5.3); low‐certainty evidence due to imprecision (small number of events) and high risk of (selection, detection of fractures) bias in this single study. See Appendix 6 for a summary of reported data.

5.3 — Comparison 5: Education versus control , Outcome 3: Number of people experiencing 1 or more fall‐related fractures

No studies reported on the outcomes number of participants experiencing one or more falls requiring hospitalisation or medical attention; health‐related quality of life; or adverse events.

Economic data

We identified four studies that reported economic data, all of which evaluated home fall‐hazard interventions. The studies reported cost of intervention and/or health service use and/or the results of trial‐based cost‐effectiveness or cost‐utility analyses (Appendix 9).

As in Gillespie 2012, the perspectives taken, the cost items measured and valued, and the type of healthcare resources included in the calculation of incremental cost‐effectiveness ratios (ICERS) varied, so that comparison remains difficult.

Nonetheless, the results from three studies demonstrate the potential cost‐effectiveness of home fall‐hazard interventions: Campbell 2005 reported an incremental cost per fall prevented for delivering intervention NZD 650. Salkeld 2000 (from Cumming 1999) reported an ICER incorporating health, home, and community care costs to be AUD 4986. A subgroup analysis by Salkeld 2000 of people recently discharged from hospital with a history of a fall in the past year (n = 203) was found to be cost saving (Cumming 1999). Stark 2021, in a cost utilisation study with older at‐risk community residents, reported a per‐person decline in healthcare utilisation valued at USD 1613.63. In contrast, Cockayne 2021a, which did not recruit people at high risk of falls, did not find a reduced rate of falls, reporting a small quality‐adjusted life‐year (QALY) difference between the two groups (reduction of 0.0015 QALYs; P = 0.814 for the intervention group) (Cockayne 2021b).

Sensitivity analysis

The following are results of the sensitivity analysis for the pooled estimates of treatment effects. We have reported data where the certainty of evidence was downgraded due to a marked impact on results (i.e. if effect estimate changed ≥ 25%).

Home fall‐hazard interventions

Sensitivity analyses revealed little difference in the results when we excluded studies in one or more of the four selected risk of bias domains for the following outcomes: rate of falls (overall, selected or not selected for high risk at baseline, tailoring of intervention, personnel delivering the intervention), risk of falling (overall, selected or not selected for high risk at baseline, tailoring of intervention, personnel delivering the intervention), rate of falls requiring medical attention, or HRQoL.

For the outcomes number of fall‐related fractures and number of people requiring medical attention following a fall, all studies were excluded due to high risk of detection bias, as events were self‐report and not confirmed by primary care case records. The main results remained unchanged, as the evidence for little or no difference between groups for both outcomes was reported as uncertain (very low‐certainty evidence).

There was a marked impact on the result for fall‐related hospital admissions when two of the three studies included in the outcome were excluded. The one remaining trial provided moderate‐certainty evidence of a 17% reduction of falls (RR 0.83, 95% CI 0.88 to 0.99; 67 participants); the certainty of the evidence was downgraded one level for low number of events. This differed to the result that there may be little or no difference between groups when all three trials were included (low‐certainty evidence).

Only one study had more than 18 months analysis (Keall 2015), and they only reported rate of falls for the outcome requiring medical support (3‐year follow‐up). There was no important change in the effect estimate when excluding this study; however, heterogeneity increased (I² = 88%).

Two studies in this review were cluster randomised; this was household clustering, and only a small portion of the study samples included more than one participant from the household (Keall 2015; Stevens 2001). Both studies also reported their main results adjusting for the clustering, so we conducted no further adjustment.

Sensitivity analysis revealed little difference in the results when using fixed‐effect (rather than random‐effects) meta‐analysis in any of the pooled‐analysis comparisons.

Assistive technology ‐ vision improvement interventions

For the outcomes fall‐related fractures (two studies) and rate of people requiring medical attention following a fall (one study), all studies were excluded due to high risk of detection bias, as events were self‐report and not confirmed by primary care case records. The main results remained unchanged, as the evidence for little or no difference between groups for both outcomes was reported as uncertain (very low‐certainty evidence).

No studies had more than 18 months analysis, and there were no cluster‐randomised trials.

Sensitivity analysis revealed little difference in the results when using fixed‐effect (rather than random‐effects) meta‐analysis in any of the pooled‐analysis comparisons.

Education interventions

One study was included in this comparison, with evidence rated as very low certainty for reducing the rate and number of falls; this study would be excluded in sensitivity analysis due to high risk of bias for selection bias (both randomisation sequence and concealment of allocation). This same study, for which evidence was rated as low certainty for little or no difference between groups in fall‐related fractures, would also be excluded due to high risk of selection bias (both randomisation sequence and concealment of allocation) and detection bias.

Funnel plots

Where there were 10 or more studies included in a pooled meta‐analysis, we explored potential publication bias by generating a funnel plot. The funnel plots for the primary analysis home fall‐hazard reduction versus control, rate of falls (Figure 4); and the secondary analysis of rate of fallers (Figure 5) do show some asymmetry, but we did not consider this sufficient to downgrade the level of evidence. We conducted an additional sensitivity test removing the one study with fewer than 20 participants, which did not impact the result.

Home fall‐hazard reduction versus control (1.1 Rate of falls ‐ overall analysis).

Funnel plot (2.1 Number of fallers ‐ overall analysis).

Discussion

Summary of main results

Home fall‐hazard reduction interventions

The evidence for this comparison was evaluated by 14 studies with 5830 participants as summarised in Table 1. There is moderate‐certainty evidence that home fall‐hazard reduction interventions probably reduce the rate of falls (12 studies) and reduce the number of people experiencing falls (12 studies).

For the overall risk category, based on an illustrative risk of 1319 falls per 1000 person‐years in the control group, there were 343 (26%) fewer falls per 1000 person‐years in the home fall‐hazard group (95% CI 118 (9%) to 514 (39%) fewer falls). Based on an illustrative risk of 519 people experiencing a fall per 1000 older people in the control group, there were 57 (11%) fewer fallers per 1000 older people in the home fall‐hazard group (95% CI 15 (3%) to 93 (18%) fewer fallers).

Subgroup analysis provides high‐certainty evidence that for those participants selected at higher risk at baseline, home fall‐hazard reduction interventions reduce the rate of falls, and for those participants not selected at higher risk at baseline, home fall‐hazard reduction interventions make little or no difference to the rate of falls. This finding provides a credible explanation for the heterogeneity found in the all‐risk analysis above.

The absolute numbers of falls or number of fallers prevented are greater in the group selected for high risk at baseline, as described below.

For the selected higher risk category, based on an illustrative risk of 1847 falls per 1000 person‐years in the control group, there were 702 (38%) fewer falls per 1000 person‐years in the home fall‐hazard group (95% CI 554 (30%) to 812 (44%) fewer falls). Based on an illustrative risk of 550 people experiencing a fall per 1000 older people in the control group, there were 89 (26%) fewer fallers per 1000 older people in the home fall‐hazard group (95% CI 47 (15%) to 124 (35%) fewer fallers).
For the non‐selected lower risk category, based on an illustrative risk of 1042 falls per 1000 person‐years in the control group, there were 52 (5%) more falls per 1000 person‐years in the home fall‐hazard group (95% CI 42 (4%) fewer to 167 (16%) more falls). Based on an illustrative risk of 533 people experiencing a fall per 1000 older people in the control group, there were 5 (1%) fewer people experiencing a fall per 1000 older people in the home fall‐hazard group (95% CI 42 (8%) fewer to 38 (7%) more fallers).

In other prespecified analysis of subgroups, formal test for subgroup interactions did not reach significant difference.

There is moderate‐certainty evidence that home fall‐hazard reduction interventions probably make little or no important difference to HRQoL. There is low‐certainty evidence that home fall‐hazard interventions may make little or no difference to the risk of experiencing one or more falls that require hospitalisation, or to the rate of falls requiring medical attention. We are uncertain if home fall‐hazard interventions make little or no difference to the number of people experiencing a fall‐related fracture or the number of fallers that require medical attention. Two studies reported no adverse events for both groups (low‐certainty evidence).

Assistive technology interventions

Assistive technology interventions were compared with control in eight trials and were characterised within three intervention themes: vision improvement (3 trials, 1489 participants), footwear and foot devices (3 trials, 184 participants), and self‐care and assistive devices (2 trials, 467 participants).

As summarised in Table 2, assistive technology interventions for vision impairment (3 trials, 1489 participants) may make little or no difference to the rate or falls or the number of fallers. We are uncertain of the effects of vision improvement interventions on an increased risk of fracture (2 trials) and on the rate of falls requiring medical attention (1 trial) compared with control, based on very low‐certainty evidence. We found low‐certainty evidence that these interventions may make little or no difference to HRQoL. No studies reported whether people experienced falls requiring hospitalisations.

Because of the diversity of interventions and contexts, as well as limited data, we judged that pooling of data would not be meaningful for the other two assistive technology interventions: footwear and foot devices, and self‐care/assistive devices. The majority of trials provided low‐ to very low‐certainty evidence of little or no effect for rate of falls or risk of falling. In one small study providing low‐certainty evidence, an ‘over‐shoe’ device to prevent slipping in icy conditions reduced the rate of falls compared to usual winter footwear (Analysis 4.1). We are uncertain of the evidence from a trial showing a reduction of the number of people falling with daily home‐delivery of meals to homebound people (Analysis 4.2).

Two trials reported adverse events: falling while switching glasses (Haran 2010, n = 2/299 intervention; n = 1/298 control) and discomfort from shoe inserts (Wang 2019, 10/20 in the intervention group).

Education interventions

The evidence for this comparison, which is summarised in Table 3, was from one small trial that evaluated an education session to reduce home fall hazards conducted by lecture and interactive practice in a mock‐up Japanese home. It is uncertain if education interventions make little or no difference to the rate of falls or the risk of people sustaining falls compared to an attention control (very low‐certainty evidence). This intervention may make little or no difference to the risk of people sustaining a fracture compared to an attention control (low‐certainty evidence).

Home modification interventions

We do not know if home modifications as a stand‐alone intervention with the aim of task enablement and functional independence reduce falls, as no studies were found in this category.

Overall completeness and applicability of evidence

Trial design and participants

We included 14 trials with a total of 5830 participants assessing the effects of home fall‐hazard interventions. Most trials were relatively small (median 238 participants, range of number of participants from 10 to 1737) with a mean age of 79 years (range from a mean age of 75 to a maximum mean age of 83 years), and 67% of participants were women. The rate of falls in the control arm (1392 falls per 1000 person‐years) and the older age indicated that trial participants were more often at high risk of falls. Two trials recruited healthy volunteers from electoral rolls. The majority of trials (77%) excluded older people who were cognitively impaired, therefore these results may not be applicable in that particular high‐risk group. We included no trials from countries considered of low or middle income, thereby reducing the generalisability of these results. Trials were undertaken over 22 years, from 1997 to 2019. More recent trials tended to provide more information about how they were tailored to falls, giving better insights for translation to practice. Control groups were comparable, with the majority being usual care, and the few others attention control (social visit). Several control groups additionally provided a fall prevention pamphlet.
We included eight trials with a total of 2140 participants assessing the effects of home assistive technology interventions. Most trials were small (median 195 participants, range of number of participants from 44 to 62) with a mean age of 77 years (ranging from a mean age of 69 to a maximum mean age of 87 years), and 64% of participants were women. One trial recruited healthy volunteers from electoral rolls. The majority of participants were recruited from aged‐care organisations, care services, specific registers, or had inclusion criteria of fall risk. Three trials excluded people screened to have cognitive impairment, and one trial only included older people with mild or moderate Alzheimer’s disease (Tchalla 2013). All trials were conducted in high‐income countries and were undertaken over 17 years from 2001 to 2018. Control groups were usual care or usual devices.
The trial of education to reduce falls was undertaken in Japan during 2008 to 2011 with 130 participants and included 85% women. The control group participated in the same community fall prevention programme, but which excluded the home safety eduction component.
We found no trials that evaluated home modifications as a stand‐alone intervention with the aim of task enablement and independence.

Intervention

We classified interventions based on the Prevention of Falls Network Europe (ProFaNE) taxonomy, Lamb 2005; Lamb 2007, and refined descriptors from emerging literature over the past decade. These are described in Appendix 1, and the descriptors were mapped to the ProFaNE taxonomy in the protocol (Clemson 2019). For the home fall‐hazard interventions, the degree of tailoring of the intervention to falls was determined using criteria developed by Clemson and colleagues (Clemson 2008), which included a comprehensive and validated evaluation process, consideration of environment‐person fit, and adequate follow‐up and support. Judging was conducted by two review authors who were blinded to each other’s results in order to minimise bias. This subgroup analysis provided three categories: interventions that were highly tailored to falls meeting at least 75% of the criteria; pre‐discharge interventions that focused on transition home goals as the primary aim; and interventions that had low tailoring to falls. The last group were mostly interventions that included a limited set of potential fall hazards without evaluation of the person‐environment interaction. Assistive technology included a very diverse group of interventions and were broadly classified into vision improvement, footwear and foot devices, and self‐care and assistive device interventions. The education intervention included a lecture with a home mock‐up of hazards/equipment.

Outcomes

We sought data for rate of falls, number of people falling, number of people sustaining a fall‐related fracture, number of people who experienced falls leading to medical attention, number of people who had a fall‐related hospital admission, HRQoL, and number of people who experienced adverse events. However, few studies provided fracture, medical attention, hospital admission, HRQoL, or adverse events data.

As the analyses in Appendix 6 and Appendix 7 demonstrate, some studies provided data for both falls and fallers, as recommended in Lamb 2005, and others provided data for one or other falls outcomes. One study provided data only for falls at home requiring medical attention.

The outcome of interest, falling, was not always clearly defined, which is a source of concern. Comparability of future research findings would be enhanced by the adoption of the consensus definition of a fall developed for trials in community‐dwelling populations by ProFaNE (Lamb 2005).

The included studies also varied in the methods used for falls ascertainment, recording, analysing, and reporting. Studies should use accepted protocols for recording of falls data, including daily recording of falls with monthly or more frequent follow‐up by the researchers who are blind to group allocation (Lamb 2005). While all but one of the home fall‐hazard interventions (92%) used these protocols, only 50% of the assistive technologies complied. This is despite evidence of a 25% underreporting of falls when data were collected retrospectively by telephone at the end of a three‐month period, compared with data collected daily and returned monthly over the same period (Hannan 2010). There are difficulties in using fall diaries over long time periods however, with trial dropouts due to overburden of paperwork, as reported by Iliffe 2015.

There was minimal reporting of adverse events across trials. Only four trials of home fall‐hazard interventions reported data, and only two studies included both control and intervention, one of these with no events, and the other reporting readmissions from falls. It is likely that adverse events may be related to diverse circumstances (such as a faulty handrail installation or poor slip‐resistant materials) (Aplin 2015). However, monitoring to link the circumstances of all falls or other potential adverse events to the intervention is likely resource intensive, difficult to monitor, and likely small number of occurrences. Two of the assistive technology trials reported adverse event data; one of these trials reported events in both control and intervention (Haran 2010), which was a fall while switching glasses (n = 3 participants).

Quality of the evidence

We have summarised the certainty of the evidence according to the GRADE approach for the following comparisons: home fall‐hazard reduction interventions versus control (Table 1); assistive technology interventions ‐ vision improvement versus control (Table 2); and education interventions versus control (Table 3).

This review, containing 13 trials (5353 participants), provides moderate‐certainty evidence of the effectiveness of home fall‐hazard reduction interventions for preventing falls among community dwelling people aged 65 years and over; however, there is high‐certainty evidence of its effectiveness when targeted to people at higher risk of falling. Three trials (1489 participants) of assistive technology interventions for vision improvement showed that they may make little or no difference in preventing falls. We did not pool data for other assistive technology interventions (5 trials, 651 participants), as we deemed these not appropriate to be combined in a meta‐analysis. We identified only one trial that conducted an education intervention; we assessed the evidence for preventing falls as of very low certainty, and the evidence for reducing the number of fall‐related fractures as of low certainty.

Overall, the certainty of the evidence ranged from high to very low. We downgraded the certainty of the evidence for risk of bias if the results changed with removal of the trials with a high risk of bias in selected domains. We further downgraded for inconsistency where there was significant heterogeneity (greater than 75%) and where the heterogeneity could not be explained by credible subgroup analysis. Furthermore, we downgraded the certainty of the evidence for imprecision for small number of events (fewer than 300 for fall or other dichotomous events, and 400 participants for the continuous outcome of HRQoL) and due to wide CIs. We additionally evaluated publication bias by visually inspecting funnel plots, though we note this is not always a conclusive method (Guyatt 2011).

Potential biases in the review process

We updated GRADE assessment criteria, outlined in the 'Summary of findings and assessment of the certainty of the evidence' section in the Methods, to illustrate how we made judgements of the certainty of the evidence. We are not aware of any obvious biases within the review process. We conducted a comprehensive search that was not restricted by language or by full‐text publication in order to optimise the chances of identifying all relevant trials. Despite this thorough search strategy, we acknowledge the possibility that some relevant trials may have been missed, especially if they were published in languages other than English. Two review authors who were blinded to each other’s results performed screening and data extraction in duplicate to minimise bias. Four studies included authors in our review team, and these authors were not involved in data extraction for these trials. Publication bias was by visual inspection, though we note this is not always a conclusive method (Guyatt 2011)

Agreements and disagreements with other studies or reviews

This review adds to the existing body of evidence and supports and extends the findings of Gillespie 2012, where home safety interventions were effective in reducing rate of falls (RR 0.81, 95% CI 0.68 to 0.97; 6 trials, 4208 participants) and risk of falling (RR 0.88, 95% CI 0.80 to 0.96; 7 trials, 4051 participants), with our review concluding that these interventions were more effective in people at higher risk of falling, which is also supported in a previous review (Clemson 2008). The Gillespie 2012 review concluded that these interventions appeared to be more effective when delivered by an occupational therapist, and we recommend further research to explore the effects of intervention components and qualitative inquiry to better understand how awareness raising and participant‐interventionist engagement impacts decision‐making and adherence.

Gillespie 2012 did not pool data for the assistive technology interventions, finding that individual trials provided the following evidence: an intervention to treat vision problems (616 participants) resulted in a significant increase in the rate of falls (RaR 1.57, 95% CI 1.19 to 2.06) and risk of falling (RR 1.54, 95% CI 1.24 to 1.91). When regular wearers of multifocal glasses (597 participants) were given single‐lens glasses, all falls and outside falls were significantly reduced in the subgroup that regularly took part in outside activities. Conversely, there was a significant increase in outside falls in intervention group participants who took part in little outside activity. An anti‐slip shoe device reduced rate of falls in icy conditions (RaR 0.42, 95% CI 0.22 to 0.78; 1 study, 109 participants). They did not report any trials that tested education of the environment to reduce falls or established home modifications that aim for task enablement and independence as a comparison.

Authors' conclusions

Implications for practice.

There is high‐certainty evidence that home fall‐hazard interventions are effective in reducing the rate of falls and the number of fallers when targeted to people at higher risk of falling. There was evidence of no effect when interventions were targeted to people not selected for risk of falling. It makes sense that this intervention that focuses on the home environment and interaction of the person within their environment would benefit those at higher risk rather than the general community of older people.

The effects of home fall‐hazard reduction interventions on other fall outcomes are uncertain, reflecting the non‐reporting of these outcomes in most trials. We are thus uncertain if there is little or no effect on the number of people experiencing one or more fall‐related fractures or falls requiring medical attention or hospitalisation for a fall. Home fall‐hazard reduction interventions probably make little important difference to health‐related quality of life (moderate‐certainty evidence).

There is currently evidence that assistive technology interventions for vision improvement may or may not reduce falls. There is uncertain evidence from two trials that precautions when changing eye prescriptions may or may not increase risk of fractures. Pooling data was not considered meaningful for the other assistive technology interventions, footwear and foot devices, or self‐care/assistive devices due to the diversity of interventions and limited data.

There is insufficient evidence from one trial to conclude whether education interventions reduce the risk of falls or rate of falling.

We do not know if home modification interventions that aim to improve task enablement and independence also reduce falls, as we found no studies for this category.

Implications for research.

This review sets the stage for next steps in more fully understanding the mechanisms underlying home fall‐hazard interventions and to provide insight into implementation of the intervention. The first set of implications for future research are focused on more fully understanding the essential elements of a home fall‐hazard reduction intervention. Standardising and defining the elements will improve rigour, fidelity, and permit comparisons between studies. Explicitly describing the intervention components (including follow‐up) will be important to understand differences in effects between approaches. One important element to better understand is the effect of raising awareness, participant‐interventionist engagement in decision‐making, and adherence and self‐management of fall risk as part of tailoring a home hazard removal plan. We recommend that further research should stratify according to risk of falls at baseline as well as tailoring of intervention to falls and delivery by personnel to fully explore the impact of these factors on fall outcomes.

The uncertainty and unexplained heterogeneity in the assistive technology ‐ vision improvement interventions warrants additional research to determine the effects of interventions. Research is needed to clarify if the intensity or components of intervention rather than participant risk impacts outcome. It would have been plausible, for example, that an intervention that provided assessment of vision and referral that relied on participant follow‐through would show no effect, but concerning that an active intervention that provided both assessment of vision and support for provision may cause harm. In contrast, another study showed considerable benefit when participant risk was higher when wearers of multifocal lenses who regularly participate in outdoor activities were given single‐lens glasses. There are also challenges to further research in trials of vision prescription due to contamination with trial authors suggesting the need to be replicated in countries where prescribed spectacles are not part of usual care.

There is a need for further research to understand the effectiveness of other assistive technologies, and these need to be of more rigorous design and be adequately powered. These studies as a group present with the lowest quality of evidence.

A qualitative approach to understanding the important elements of the interventions and how they influence behaviours is warranted and may inform future reporting methods. There is limited information available on adverse events from the environmental changes. This information will be critical for understanding the risk and benefit of the intervention.

The next set of implications for future research are related to understanding the implementation of home environmental interventions. Implementation in general has not been fully addressed. It is unknown which methods and procedures will support the uptake of these interventions in usual care. The evidence would benefit from further research related to costs and cost‐effectiveness, and there is a gap in this evidence for assistive technologies. This information will be critical for policy and public health planning. Understanding if hazard removal in the community is effective at reducing falls is another unanswered question. Home fall‐hazard reduction studies were able to achieve a 40% reduction in fall rates. It is unknown if addressing barriers beyond the home can further reduce falls.

Future studies should use the consensus definition of a fall developed for trials in community‐dwelling populations by Prevention of Falls Network Europe (ProFaNE) and consistent methods of falls ascertainment (Lamb 2005). Monitoring of falls independently of self‐report could be considered if wearable sensor devices improve in terms of feasibility and acceptability.

A majority of the included studies were conducted in high‐income countries. Future studies should focus on low‐ and middle‐income countries where the nuances of context in environment differ (e.g. lack of lighting, outdoor toileting) and the complexities between the person and how they negotiate their environment can be very different and, in addition, the burden of disease is higher and increasing.

The dearth of national or local policy plans is, unfortunately, evident, and there is a need for advocacy to call for policy, recognition, and investment in community‐based fall prevention. This review can contribute to the evidence base and act as an impetus to include environment considerations in any such policy.

History

Protocol first published: Issue 2, 2019

Acknowledgements

We wish to thank Maria Clarke and Sharon Lewis for their helpful comments at editorial review, and Clemens Becker and Lianne McInally for their external review. We thank Joanne Elliott for her assistance in developing the search strategy and editorial support. We also thank Elaine Tam for her assistance in our search review. We are also grateful to the authors of Gillespie 2012, particularly Lesley Gillespie and Clare Robertson, for the development of methods and procedures.

This project was supported by the National Institute for Health Research (NIHR) via Cochrane Infrastructure funding to the Cochrane Bone, Joint and Muscle Trauma Group. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, National Health Service (NHS), or the Department of Health.

Editorial and peer‐reviewer contributions

Cochrane BJMT supported the authors in the development of this review.

The following people conducted the editorial process for this article.

Sign‐off Editor (final editorial decision): Xavier Griffin, Co‐ordinating Editor, Cochrane BJMT
Managing Editor (selected peer reviewers, collated peer‐reviewer comments, provided editorial guidance to authors, edited the article): Joanne Elliott, Managing Editor, Cochrane BJMT
Editor (advised on methodology and review content): Sharon Lewis, Deputy Co‐ordinating Editor, Cochrane BJMT
Information Specialist (ran searches and provided guidance on search methods): Maria Clarke, Cochrane BJMT
Copy Editor (copy‐editing and production): Lisa Winer, Cochrane Copy Edit Support
Peer reviewers (provided comments and recommended an editorial decision): Clemens Becker, Lianne McInally

Appendices

Appendix 1. Mapping of ProFaNE fall prevention classification taxonomy (Lamb 2007) to the Cochrane Environmental Intervention Descriptors

ProFaNE classification	Assistive technology (ISO9999, 2002^a)	Knowledge/education intervention	‐	Home hazard management (C300)
Cochrane intervention classification	Assistive technology (as a single strategy) (WHO, ISO9999, 2016^a)	Information/ education on environmental fall risks (as a single strategy)	Home modifications (as a single strategy)	Home fall‐hazard reduction (may be single or package of strategies)
Main intervention aim	Reduce falls	Reduce falls	Improve task performance, independence, and accessibility for people with disability (activity limitations)	Reduce falls and environmental and behavioural fall risk
ProFaNE taxonomy descriptors	Illustrative examples from classification used in review
Assessment
Environment assessment (dwelling unit) C207	_	e.g. self‐completed checklist as a single strategy	e.g. performance‐based functional assessments/ assessment of environmental barriers to function	e.g. formal assessment of fall hazards in and about the home; environmental risk assessment taking into account individual capacity
Environment assessment (public) C208	_	_	_	Extension to public places is possible.
Environment/assistive technology^a
Furnishings and adaptations to homes D700	_	_	e.g. features of the house designed to assist a disabled person to function independently such as grab rails, stair lifts, task lighting, remove shower recess hob	e.g. features of the home that can be adapted to reduce fall risk such as lighting for safe mobility, grab rails, safety equipment such as non‐slip materials for floors and stairs
Aids for personal mobility D710	e.g. mobility aids	_	e.g. mobility aids, wheelchairs, lifting aids	e.g. mobility aids
Aids for communication, information, and signalling (optical) D720	e.g. eyeglasses to aid or improve vision	_	_	_
Aids for communication, information, and signalling (hearing) D721	e.g. hearing aids to amplify sound	_	_	_
Aids for protection ‐ alarm systems D723	e.g. falls monitor (personally worn devices)	_	_	_
Body‐worn aids for personal care and protection: protective aids D730	e.g. hip protectors^b	_	_	_
Body‐worn aids for personal care and protection: clothes and shoes D731	e.g. antislip devices for shoes; orthotic footwear	_	_	e.g. safe footwear; clothing (such as clothing that is not trippable when climbing stairs)
Other environmental interventions
Behavioural adaptations (ProFaNE ‐ see under psychological)	_	_	_	Behavioural safety strategies to reduce fall risk (e.g. activity avoidance, cognitive adaptations, practical strategies)
Self‐management (ProFaNE – see B600)	_	_	_	e.g. self‐risk evaluation (jointly with therapist); strategy instruction to prevent future falls
Environmental ‐ social
Caregiver training D804	_	_	_	e.g. target important aspects of fall prevention and safety at home
Knowledge
e.g. pamphlets, information, booklets, videos, lectures D900	_	Information/education about reducing environmental fall risks in the home as a stand‐alone intervention	_	_
Postintervention follow‐up
Others D999	_	_	Follow up to ensure modifications have been installed	Follow through to ensure recommendations are completed and appropriate and that older person perceives they have reduced falls risk

Open in a new tab

^aThese ProFaNE intervention categories were based on the 2002 International Standard ISO 9999 'Technical aids for persons with disabilities – classification and terminology'. This is now updated and published as WHO Assistive products for persons with disability – 2016 WHO International Standard ISO 9999 'Classification and terminology for persons with disability' 6th edition. ^bHip protectors as a sole intervention are covered in a separate Cochrane Review, which includes rate of falls among the types of outcomes (Santesso 2014). 

Appendix 2. Search strategies

CENTRAL (CRS Web)

#1 (MESH DESCRIPTOR Accidental Falls) AND CENTRAL:TARGET (1527) #2 (falls or faller*):AB,EH,KW,KY,MC,MH,TI,TO AND CENTRAL:TARGET (7569) #3 #1 OR #2 (7569) #4 MESH DESCRIPTOR Aged EXPLODE ALL AND CENTRAL:TARGET (214547) #5 ((senior* or elder* or old* or aged or ag?ing or postmenopausal or community dwelling)):AB,EH,KW,KY,MC,MH,TI,TO AND CENTRAL:TARGET (612915) #6 #4 OR #5 (612922) #7 #3 AND #6 (5609) #8 01/01/2012_TO_29/01/2021:CRSCREATED AND CENTRAL:TARGET (1073352) #9 #8 AND #7 (4199)

MEDLINE (Ovid Interface)

1 Accidental Falls/ (25030) 2 (falls or faller*1).tw. (48469) 3 1 or 2 (59317) 4 exp Aged/ (3195618) 5 (senior*1 or elder* or old* or aged or ag?ing or postmenopausal or community dwelling).tw. (2316780) 6 4 or 5 (4844705) 7 3 and 6 (31171) 8 Randomized controlled trial.pt. (521766) 9 Controlled clinical trial.pt. (94043) 10 randomized.ab. (509822) 11 placebo.ab. (215459) 12 Clinical trials as topic/ (194466) 13 randomly.ab. (351188) 14 trial.ti. (235051) 15 8 or 9 or 10 or 11 or 12 or 13 or 14 (1346774) 16 exp Animals/ not Humans/ (4781739) 17 15 not 16 (1240365) 18 7 and 17 (3622) 19 (2012* or 2013* or 2014* or 2015* or 2016* or 2017* or 2018* or 2019* or 2020* or 2021*).ed,dt. (11374813) 20 18 and 19 (2096)

Embase (Ovid Interface)

1 Falling/ (42012) 2 (falls or fallers).tw. (65791) 3 1 or 2 (89203) 4 exp Aged/ (3045727) 5 (senior*1 or elder* or old* or aged or ag?ing or postmenopausal or community dwelling).tw. (3128244) 6 4 or 5 (5315006) 7 3 and 6 (46631) 8 exp Randomized Controlled Trial/ or exp Single Blind Procedure/ or exp Double Blind Procedure/ or Crossover Procedure/ (709392) 9 (random* or RCT or placebo or allocat* or crossover* or cross over or trial or (doubl* adj1 blind*) or (singl* adj1 blind*)).ti,ab. (2243082) 10 8 or 9 (2336988) 11 (exp Animal/ or animal.hw. or Nonhuman/) not (exp Human/ or Human cell/ or (human or humans).ti.) (6139199) 12 10 not 11 (2073800) 13 7 and 12 (6777) 14 (2012* or 2013* or 2014* or 2015* or 2016* or 2017* or 2018* or 2019* or 2020* or 2021*).dc,yr. (14616568) 15 13 and 14 (4438)

CINAHL (EBSCO)

S1 (MH “Accidental Falls”) (23,503) S2 TI ( falls or faller* ) OR AB ( falls or faller* ) (38,536) S3 S1 OR S2 (48,238) S4 (MH “Aged+”) (858,391) S5 TI ( senior* or elder* or old* or aged or ag?ing or postmenopausal or community dwelling ) OR AB ( senior* or elder* or old* or aged or ag?ing or postmenopausal or community dwelling) (544,566) S6 S4 OR S5 (1,168,579) S7 S3 AND S6 (23,822) S8 PT Clinical Trial (108,838) S9 (MH “Clinical Trials+”) (313,464) S10 TI clinical trial* OR AB clinical trial* (119,715) S11 TI ( (single blind* or double blind*) ) OR AB ( (single blind* or double blind*) ) (50,111) S12 TI random* OR AB random* (365,824) S13 S8 OR S9 OR S10 OR S11 OR S12 (563,408) S14 S7 AND S13 (3,522) S15 EM 20120101‐20210129 (3,759,605) S16 S14 AND S15 (1,958)

OTseeker

Advanced search option selected Abstract and Title: Fall Age group: gerontology

Total = 64

ClinicalTrials.gov

(prevent OR reduce OR reduction OR risk) AND (fall OR fallers) | Interventional Studies | Older Adult | First posted after 01/01/2012 (2503)

WHO ICTRP

prevent* AND fall* OR reduc* AND fall* OR risk* AND fall* (1184)

Appendix 3. Risk of bias assessment tool*

Domain	Criteria for judging risk of bias
Random sequence generation relating to selection bias (biased allocation to interventions) due to inadequate generation of a randomised sequence	Judgement of 'low risk' if the trial authors described a random component in the sequence generation, e.g. referring to a random number table; using a computer random number generator; coin tossing; shuffling cards or envelopes; throwing dice; drawing of lots; minimisation. Judgement of 'high risk' if the trial used a systematic non‐random method, e.g. date of admission; odd or even date of birth; case record number; clinician judgement; participant preference; patient risk factor score or test results; availability of intervention. Judgement of 'unclear risk' if there is insufficient information about the sequence generation process to permit judgement of 'low risk' or 'high risk'.
Allocation concealment relating to selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment	Judgement of 'low risk' in studies using: individual randomisation if the trial described allocation concealment as by central allocation (telephone, internet‐based, or pharmacy‐controlled randomisation); sequentially numbered, identical drug containers; sequentially numbered, opaque, sealed envelopes; cluster randomisation if allocation of all cluster units performed at the start of the study and individual participant recruitment was completed prior to assignment of the cluster, and the same participants were followed up over time, or individual participants were recruited after cluster assignment, but recruitment was carried out by a person unaware of group allocation and participant characteristics (e.g. fall history), or individual participants in intervention and control arms were invited by mail questionnaire with identical information. Judgement of 'high risk' in studies using: individual randomisation if investigators enrolling participants could possibly have foreseen assignments thus introducing selection bias, e.g. using an open random allocation schedule (e.g. a list of random numbers); assignment envelopes unsealed, non‐opaque, or not sequentially numbered; alternation or rotation; date of birth; case record number; or any other explicitly unconcealed procedure; cluster randomisation if individual participant recruitment was undertaken after group allocation by a person who was unblinded and may have had knowledge of participant characteristics. Judgement of 'unclear risk' if insufficient information to permit judgement of 'low risk' or 'high risk'. This is usually the case if the method of concealment is not described or not described in sufficient detail to allow a definitive judgement, e.g. if the use of assignment envelopes is described, but it is unclear whether envelopes were sequentially numbered, opaque, and sealed.
Blinding of participants and personnel relating to performance bias due to knowledge of the allocated interventions by participants and personnel carrying out the interventions	Judgement of 'low risk' if blinding of participants and personnel implementing the interventions was ensured, and it is unlikely that the blinding could have been broken, or blinding was not done but the review authors judge that the outcomes (falls and fractures) are unlikely to be influenced by lack of blinding. Judgement of 'high risk' if participants or intervention delivery personnel, or both, were not blinded to group allocation (e.g. exercise intervention), and the outcomes (falls and fractures) are likely to be influenced by lack of blinding. Judgement of 'unclear risk' if there is insufficient information to permit a judgement of 'low risk' or 'high risk'.
Blinding of outcome assessment relating to detection bias due to knowledge of the allocated interventions by outcome assessors	Falls, fallers: judgement of 'low risk' if outcomes were recorded/confirmed in all allocated groups using the same method, and the personnel recording/confirming outcomes were blind to group allocation; judgement of 'high risk' if outcomes were not recorded/confirmed in all allocated groups using the same method, or the personnel recording/confirming outcomes were not blind to group allocation; judgement of 'unclear' if there is insufficient information to permit a judgement of 'low risk' or 'high risk'. Fractures, hospital admission, medical attention: judgement of 'low risk' if events were recorded/confirmed in all allocated groups using the same method, and events were confirmed (e.g. by results of radiological examination or from primary care case records), and the personnel recording/confirming events were blind to group allocation; judgement of 'high risk' if events were not recorded/confirmed in all allocated groups using the same method, or the only evidence for events was from self‐reports from participants or caregivers; judgement of 'unclear risk' if there is insufficient information to permit a judgement of 'low risk' or 'high risk'.
Incomplete outcome data relating to attrition bias due to the amount, nature, or handling of incomplete outcome data	Judgement of 'low risk' if there are no missing outcome data, or less than 20% of outcome data are missing and losses are balanced in numbers across intervention groups with similar reasons for missing data across groups, or missing data have been imputed using appropriate methods. Judgement of 'high risk' if greater than 20% of outcome data are missing, or reasons for missing outcome data are likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups, or ‘as‐treated’ analysis done with substantial departure of the intervention received from that assigned at randomisation or potentially inappropriate application of simple imputation. Judgement of 'unclear risk' if there is insufficient information to permit a judgement of 'low risk' or 'high risk'.
Selective outcome reporting relating to bias due to the selective reporting or non‐reporting of findings	Judgement of 'low risk' if the study protocol is available, was published prior to study commencement or registered prospectively in a clinical trial registry and all prespecified study outcomes are reported in the prespecified way, or the study protocol is unavailable, but it is clear that the published report includes all expected outcomes. Judgement of 'high risk' if not all prespecified study outcomes are reported, or one or more primary outcomes are reported in ways that were not prespecified, or one or more outcomes are reported incompletely, or the study fails to include results for a key outcome that would be expected to have been reported. Judgement of 'unclear risk' if there is insufficient information to permit a judgement of 'low risk' or 'high risk'. This includes if the study protocol is unavailable, but it is clear that the published report includes all expected outcomes, or the study protocol is published after study commenced or the trial is registered retrospectively.
Method of ascertaining falls relating to bias in the recall of falls due to unreliable methods of ascertainment	Judgement of 'low risk' if the study used some form of concurrent collection of data about falling, e.g. participants were given postcards to fill in daily and mail back monthly, calendar to mark monthly, or, more frequently, follow‐up by the researchers. Judgement of 'high risk' if ascertainment relied on participant recall at longer intervals than one month during the study or at its conclusion. Judgement of 'unclear risk' if there was retrospective recall over a short period only, or if the trial authors did not describe details of ascertainment, i.e. insufficient information was provided to permit a judgement of 'low risk' or 'high risk'.
Cluster‐randomised trials relating to bias due to factors particular to cluster‐randomised trials	Judgement of 'low risk’ if the study predominantly had the following characteristics: individuals were recruited to the trial prior to randomisation of the clusters; baseline comparability of clusters was reported, or there was statistical adjustment for baseline characteristics; no loss of clusters or missing outcomes for individuals within specific clusters; clustering is accounted for in the analyses; results are comparable with individually randomised trials. Judgement of 'high risk’ if the study predominantly had the following characteristics: individuals were recruited to the trial after randomisation of the clusters; baseline comparability of clusters was not reported, and there was no statistical adjustment for baseline characteristics; loss of entire clusters or missing outcomes for individuals within clusters; no accounting for clustering in analyses; results not comparable with individually randomised trials. Judgement of ’unclear risk’ if there is insufficient information to permit a judgement of ’low risk’ or ’high risk’.

Open in a new tab

*We adapted this from Table 8.5.a 'The Cochrane Collaboration's tool for assessing risk of bias' and Table 8.5.d 'Criteria for judging risk of bias in the 'Risk of bias' assessment tool' (Higgins 2011).

Appendix 4. Source of data for generic inverse variance analysis (see bottom of table for explanation of codes)

Study ID	Source for rate ratio: falls	Source for risk ratio: fallers	Source for risk ratio: number with fractures	Source for rate or risk ratio: falls, or number of people requiring medical attention	Source for risk ratio: hospitalisation following a fall
Campbell 2005	1	7	‐	3	‐
Chu 2017	3	7	‐	‐	7
Cockayne 2021a	1	7	7	ND	ND
Cumming 1999	2	4	‐	5	‐
Cumming 2007	1	4	4	‐	‐
Day 2002 (HH)	3	4	‐	3	‐
Day 2002 (V)	3	4	‐	3	‐
Haran 2010	1	7	7	ND	ND
Kamei 2015	3	7	7	‐	‐
Keall 2015	‐	‐	‐	7	‐
Lannin 2007	3	7	‐	7	ND
Lin 2007	3	‐	‐	‐	‐
Lockwood 2019	1	7	‐	‐	7
McKiernan 2005	1	‐	7	7	‐
Nikolaus 2003	1	7	7	‐	‐
Pardessus 2002	NA	7	‐	‐	7
Perry 2008	NA	7	‐	‐	‐
Pighills 2011	3	7	‐	‐	‐
Stevens 2001	1^a	6^a	ND	ND	‐
Stark 2021	1	4	‐	‐	‐
Tchalla 2013	3	7	‐	‐	‐
Thomas 2018	NA	7	‐	‐	‐
Wang 2019	3	7	‐	‐	‐
Codes for source of rate ratio: 1: incidence rate ratio reported by trial authors 2: hazard ratio/relative hazard (multiple events) reported by trial authors 3: incidence rate ratio calculated by review authors Codes for source of risk ratio: 4: hazard ratio/relative hazard (first fall only) reported by trial authors 5: risk ratio reported by trial authors 6: odds ratio reported by trial authors 7. risk ratio reported by review authors

Open in a new tab

Abbreviations: HH: home hazard data from paper; NA: data not available for this outcome; ND: outcomes relating to falls, fallers, fractures or falls requiring medical attention or hospital admission were reported, but there were no usable data; V: vision data from paper ^aAdjusted for clustering by trial authors. 

Appendix 5. Raw data for rate of falls (falls per person‐year) and number of fallers, where available

Study ID	Intervention group: falls per person‐year	Control group: falls per person‐year	Intervention group: number of fallers	Intervention group: number in analysis	Control group: number of fallers	Control group: number in analysis	Follow‐up: months
Campbell 2005	0.65	1.65	36	100	59	96	12
Chu 2017	0.17	0.30	13	95	21	103	12
Cockayne 2021a All/recent hospitalisation/no recent hospitalisation	1.9/ 1.88/1.93	1.6/ 2.21/1.53	245/ 26/219	420/ 41/389	506 /54/452	888/ 83/818	12
Cumming 1999 All/history of falls/no history of falls	0.86/1.25/0.79	1.22/2.24/0.80	96/43/53	264/103/161	119/67/52	266/103/163	12
Cumming 2007	2.45	1.68	201	309	153	307	12
Day 2002 Home safety/AT: vision improvement	1.14/ 1.03	1.20	78/ 84	136/ 139	87	137	18
Haran 2010 All/increased exposure outdoors/low exposure	1.42/ 1.17/1.68	1.54/ 1.99/1.26	170/ 77/93	299/ 148/151	175/ 68/107	298/ 113/185	13
Kamei 2015	0.41	0.65	15	56	21	54	12
Lannin 2007	0.80	2.40	1	5	2	5	3
Lin 2007	0.40	0.88	‐	50	‐	50	6
Lockwood 2019	1.41	1.44	12	37	15	36	6
McKiernan 2005	1.28	2.94	‐	55	‐	54	3.5
Nikolaus 2003 All/history of multiple falls/history of no falls or 1 fall	0.965/1.19/0.78	1.243/1.62/0.92	51/21/30	181/53/128	61/36/25	179/55/125	12
Pardessus 2002	0.68	0.82	13	30	15	30	7
Perry 2008	‐	‐	5	20	9	20	3
Pighills 2011 All/OT/domiciliary health worker	2.42/ 2.06/2.86	3.77	100/ 50/50	156/ 85/71	54	77	12
Stevens 2001	0.69	‐	0.72	570	‐	1167	12
Stark 2021	1.5	2.3	67	140	74	135	12
Tchalla 2013	0.61	0.91	16	49	30	47	12
Thomas 2018 Daily meals/frozen weekly meals	‐	‐	33/29	139/106	36	126	3.75
Wang 2019	0.75	0.84	6	16	10	19	12

Open in a new tab

Abbreviations: AT: assistive technology; OT: occupational therapist

Appendix 6. Raw data for fall‐related fractures, fall‐related hospital admissions, and falls requiring medical attention, where available

Study ID	Intervention group: fall‐related fracture	Control group: fall‐related fracture	Intervention group: number reporting a fall requiring hospital admission	Control group: number reporting a fall requiring hospital admission	Intervention group: falls requiring medical attention	Control group: falls requiring medical attention	Intervention group: number in analysis	Control group: number in analysis	Follow‐up: months
Campbell 2005	‐	‐	‐	‐	19 falls	32 falls	100	96	12
Chu 2017	‐	‐	4	6	‐	‐	95	103	12
Cockayne 2021a	16	38	‐	‐	‐	‐	420	888	12
Cumming 1999	‐	‐	‐	‐
Cumming 2007	31	18	‐	‐	RR reported^a	RR reported^a	103	103	12
Day 2002 (HH)	‐	‐	‐	‐	27 falls	18 falls	136	137	18
Day 2002 (V)	‐	‐	‐	‐	17 falls	18 falls	139	137	18
Haran 2010	16	10	‐	‐	‐	‐	299	298	13
Kamei 2015	2	1	‐	‐	‐	‐	56	54	12
Lannin 2007	‐	‐	‐	‐	0	0	5	5	3
Lin 2007	‐	‐	‐	‐	‐	‐	‐	‐	‐
Lockwood 2019	‐	‐	1	7	‐	‐	32	35	6
McKiernan 2005	0	0	‐	‐	0	2	55	54	14
Nikolaus 2003	4	3	‐	‐	‐	‐	181	179
Pardessus 2002	‐	‐	4	3	‐	‐	30	30	7
Perry 2008	‐	‐	‐	‐	‐	‐	‐	‐	‐
Pighills 2011	‐	‐	‐	‐	‐	‐	‐	‐	‐
Stevens 2001	‐	‐	‐	‐	‐	‐	‐	‐	‐
Stark 2021	‐	‐	‐	‐	‐	‐	‐	‐	‐
Tchalla 2013	‐	‐	‐	‐	‐	‐	‐	‐	‐
Thomas 2018	‐	‐	‐	‐	‐	‐	‐	‐	‐
Wang 2019	‐	‐	‐	‐	‐	‐	‐	‐	‐

Open in a new tab

Abbreviations: HH: home hazard data from paper; RR: risk ratio; V: vision data from paper ^aRR outcome used in analysis, but no data on number of falls reported. RR 0.89 (95% confidence interval 0.56 to 1.42). 

Appendix 7. Raw data for health‐related quality of life outcome, where available

Study ID	Outcome measure	Outcome format	Intervention group:  quality of life	Intervention  group: number in analysis	Control group: quality of life	Control  group: number in analysis	Data included in review analysis
Cockayne 2021a	EQ‐5D‐5L Utility score	Mean (SD) Baseline 12 month	0.68 (0.19) 0.64 (0.23)	383	0.69 (0.20) 0.65 (0.22)	834	EQ‐5D‐5L, 12 month means (SD)
Haran 2010	SF‐12 Physical	Mean (SD) Baseline 13 month	44.3 (9.1) 43.4 (9.2)	299	43.4 (9.4) 43.0 (9.7)	298	SF‐12 Physical domain, 13 month mean (SD)
Lin 2007	WHOQOL‐BREF: Physical domain Psychological domain Social domain Environmental domain	Means (SD) Baseline (Phys) 6 month 8 month Baseline (Psych) 6 month 8 month Baseline (Soc) 6 month 8 month Baseline (Env) 6 month 8 month	52.6 (15.1) 60.9 (9.5) 60.9 (14.5) 58.1 (13.2) 59.3 (13.5) 61.5 (13.1) 70.9 (11.9) 71.9 (11.2) 73.6 (10.2) 65.8 (10.5) 69.0 (10.7) 70.2 (9.6)	46	48.9 (17.3) 52.6 (13.8) 55.5 (15.3) 55.7 (16.0) 53.8 (17.0) 56.3 (17.6) 68.8 (10.6) 63.8 (14.8) 66.3 (13.3) 62.5 (9.8) 62.1 (14.4) 65.1 (14.3)	40	WHOQOL‐BREF Environmental domain, 8 month mean (SD)
Lockwood 2019	EQ‐5D	Baseline 30 days 6 month	0.4 (0.2) 0.6 (0.2) 0.7 (0.2)	29	0.5 (0.2) 0.7 (0.2) 0.7 (0.3)	30	EQ‐5D, 6 month mean (SD)
Pighills 2011 (occupational therapy group)	SF‐12 Physical SF‐12 Mental EuroQol	SF‐12 Baseline means (SD) 12 month means (95% CI)	Baseline 33 (14) 49 (11) 0.6 (0.3) 12 month 35 (34, 37) 50 (48, 51) 0.58 (0.55, 0.62)	85	Baseline 33 (12) 47 (11) 0.6 (0.3) 12 month 34 (32, 36) 49 (48, 50) 0.56 (0.53, 0.6)	77	SF‐12 Physical domain, 12 month mean (CI converted to SD in Review Manager 5)
Pighills 2011 (trained assessor group)	SF‐12 Physical SF‐12 Mental EuroQol	SF‐12 Baseline means (SD) 12 month means (95% CI)	Baseline 37 (13) 51 (12) 0.6 (0.3) 12 month 36 (33, 37) 50 (49, 51) 0.63 (0.59, 67)	71	Baseline 33 (12) 47 (11) 0.6 (0.3) 12 month 34 (32, 36) 49 (48, 50) 0.56 (0.53, 0.6)	77	SF‐12 Physical domain, 12 month mean (CI converted to SD in Review Manager 5)
Stark 2021	SF‐36	Mean (SD) Baseline 12 month	Baseline 41.81 (10.98) 12 month 42.26 (11.01)	127	Baseline 44.44 (10.39) 12 month 43.74 (11.10)	126	SF‐36, 12 month means (SD)

Open in a new tab

Abbreviations: CI: confidence interval; SD: standard deviation; SF‐12: 12‐item Short Form Health Survey; SF‐36: 36‐item Short Form Health Survey

Appendix 8. Adherence

Study ID	Adherence was measured	Adherence data reported	Measurement of adherence	Reported adherence results
Home fall‐hazard interventions
Campbell 2005 in LaGrow 2008	Yes	Yes	Telephoned at 6 months by OT. Recommendations by OT and agreed upon by participant	Mean (SD) number of recommendations agreed to: 2.5 (2.2); number of recommendations actioned or partially actioned: 2.2 (1.8) (home safety only) 90% (152 of 169) reported complying partially or completely with 1 or more of the recommendations made by the OT (this was combined exercise + home safety and home safety‐alone group data).
Chu 2017	Yes	Yes	Participant adherence rates to OT recommendations, recorded by telephone follow‐up at 2 months	An average of 4.6 fall hazards identified per case. Adherence to recommendations: 76.3% for environmental hazards and daily life routine, 38.9% for educational advice on fall reduction care plans given to participants or caregivers, 68% for assistive device recommendations, and 44.4% for receipt of other community services
Cockayne 2021a	Yes	Yes (full report to be published elsewhere)	Number receiving the home visit	Only if received home visit (381, 88.6%). 41 withdrew from treatment. Stated "recommendations were followed in varying degrees by participants".
Cumming 1999	Yes	Yes	Follow‐up home visit by research assistant at 12 months	At least 1 home modification was recommended in 150 of the 178 homes visited. The most common recommendations were to remove mats and throw rugs (48%), change footwear (24%), and use a non‐slip bathmat (21%). In the 121 homes revisited after 12 months, 419 home modifications had been recommended, and 216 (52%) were met with partial or complete adherence.
Day 2002 (2)	Yes	Yes	Number of referrals for help with home modifications; at 18 months, 442 were randomly selected to assess risk factors (number of hazards in home)	Of the 543 participants receiving the home fall‐hazard management intervention, 478 participants were advised to have modifications in their homes; 363 (67%) of these received help to do these modifications, which included handrails fitted (275 participants), modifications to floor coverings (72), contrast edging fitted to steps (72), and maintenance to steps or ramps (66). The mean average number of hazards among participants receiving the home fall‐hazards intervention decreased from 10.2 to 7.4, compared with a decrease of 9.1 to 7.9 in the control group (F = 42.87 (df = 1440), P < 0.001).
Keall 2015	Yes	No Not reported for older age group	Assessed hazards were fixed.	For all age groups (0 to > 70 years): 5% of houses did not provide fixable hazards, and 8% did not receive the intended intervention.
Lannin 2007	No	No	‐	‐
Lin 2007	No	No	‐	‐
Lockwood 2019	Yes	Yes	Adherence to protocol assessed by weekly review of medical records by the lead investigator to confirm that each participant had received the intervention as allocated.	7 participants did not receive the intervention as allocated (intervention: n = 3 discharged to residential care, n = 1 declined visit; control: n = 1 died, n = 1 withdrew, n = 1 received home visit). More participants in the intervention group (n = 7) appeared to access community‐based occupational therapy after discharge compared to the control group (n = 1).
Nikolaus 2003	Yes	Yes	Number/per cent of recommendations completed at 12 months	There were 137/222 homes with home modification recommendations. Compliance at 12 months (105 homes) where at least 1 recommended change was completed was 75.5%. Compliance of individual recommendations averaged 63%, ranging from 42% (rugs/carpets) to 83% (shower seat).
Pardessus 2002	No	No	‐	‐
Pighills 2011	Yes	Yes	Level of adherence at 12 months reported for trained assessor and occupational therapy groups.	The trained assessors made an average of 1.6 (± 1.7) recommendations per visit; 45% were partially adhered to, and 25% were fully adhered to. An average of 0.9 recommendations per person were resolved with intervention. The mean number of recommendations made by the OTs per visit was 3.6 (± 2.3), of which 60% were partially adhered to, and 28% were fully adhered to. An average of 2.3 recommendations per person were resolved with intervention. There was a statistically significant difference of 1.9 recommendations between groups, OT minus trained assessor group (95% CI 1.3 to 2.6, z = 5.73, P < 0.001). There was significantly greater adherence to recommendations for participants in the OT group, adjusting for the number of recommendations made (P = 0.04). For participants who had the same number of recommendations, the odds of greater adherence for those in the OT group were 2.44 (95% CI 51.03 to 5.77).
Stark 2021	Yes	Yes	Participation in at least 2 sessions and adherence to at least 80% of the hazard removal recommendations. Self‐reported at 6 and 12 months.	140 of the intervention group (90%) completed initial intervention, and 130 (93%) of these completers finished the booster session. Adherence to recommendations was 92% at 6 months and 91% at 12 months. The most frequent hazards addressed were lack of hand support, lack of adaptive equipment, tripping hazards, clutter, and lighting. The mean (SD) number of hazards for the intervention group was 10.0 (6.7) at baseline, reduced to 4.5 (5.0) at the 6‐month booster session, and at this time 54 (44%) reported implementing additional changes to their homes or routines to reduce fall risk.
Stevens 2001a Stevens 2001b	Yes	Yes	Self‐report postal survey sent at 11 months plus home visit conducted in sample of 51 homes at 11 months.	Proportion of participants who reported acting on 1 or more of the 6 hazard recommendations ranged from 12.9% (repair damaged floor) to 79% (improved poor lighting), averaging 27% compliance. A higher proportion reported safer behaviours rather than hazard change (e.g. 72% wearing safer shoes; 88% avoiding climbing).
Assistive technology interventions
Cumming 2007	Yes	Yes	Time frame received glasses, cataract surgery completed	146 were referred for treatment, of which 11 declined, leaving 44% (135) with some form of treatment. 27% of new eye glasses delivered within 30 days and 77% delivered within 60 days; 7 out of 15 participants had cataract surgery within 12‐month follow‐up.
Haran 2010	Yes	Yes	Number of months reporting satisfactory adherence. Monthly reporting on calendars	The median number of months intervention participants reported satisfactory adherence with the recommendations for wearing single‐lens glasses when walking in the street or in shopping centres was 7 (2 to 11) months. 96 (32%) of 299 participants adhered for 0 to 3 months, 41 (14%) for 4 to 6 months, 40 (13%) for 7 to 9 months, and 122 (41%) for 10 to 12 months.
Day 2002	Yes	Yes	Number participants taking up the referral	287 recommended for referral, 186 recently visited or about to visit; 97/186 (52%) took up referral, and 26 had new or modified prescription glasses (20) or surgery (6). At 12 months, the reported time since seeing an eye care professional was 5.0 months in control and 6.5 months in intervention group.
McKiernan 2005	Yes	Yes	Self‐report of wearing the device as primary winter footwear under appropriate environmental conditions	78% of intervention participants confirmed primary use of device; 19% of control participants reported having used the device or a similar gait‐stabilising device during the study.
Perry 2008	Yes	Yes	Weekly report of hours of use	Intervention: 37.6 (SD 5.4) hours of wear; control 38.1 (SD 5.44) hours of wear
Thomas 2018	Yes	Yes	Received meals or not/no longer wished to receive meals	Daily meals delivered 95% (n = 204); weekly frozen meals delivered 95% (n = 191)
Wang 2019	No	Not reported	‐	‐
Tchalla 2013	Yes	Yes	Rate of acceptance	New technology 95.9% acceptance (47/49)
Education intervention
Kamei 2015	Yes	Yes	10‐item fall prevention awareness questionnaire; self‐report of home hazard changes	Increase in fall prevention awareness F(2) = 3.25, P = 0.05; higher rates of home hazard adaptations in 13/33 items (self‐decision on what to change, as participants conducted own hazard assessment)

Open in a new tab

Abbreviations: CI: confidence interval; OT: occupational therapist; SD: standard deviation

Appendix 9. Studies reporting cost‐effectiveness, cost‐utility, or costs (intervention and/or healthcare resource use) related to fall outcomes

Study ID (source if not primary reference), sample, efficacy analyses, type of evaluation	Intervention(s) and comparator (N in analysis)	Perspective(s), type of currency, price year, time horizon	Cost items measured	Mean (SD) intervention cost per person	Healthcare service costs	Incremental cost per fall prevented/per Quality‐adjusted life years (QALY) gained
Campbell 2005 Severe vision impairment, > 75 years, recruited from RNZ Foundation for the Blind and low‐vision clinics Analysis Cost‐effectiveness analysis	Home safety assessment and modification (n = 100) vs no intervention (n = 96)	Societal perspective. New Zealand dollars, 2004, 1 year	Intervention costs (training costs, recruitment, occupational therapists’ time and transport, administration, services and equipment installed in homes, overhead costs)	Intervention costs NZD 325 (± 292) per participant	Not reported	Incremental cost per fall prevented for delivering intervention NZD 650
Cockayne 2021a Community residing, concerned with falling or previous fall Cost utility Cost‐effectiveness	Home fall‐hazard reduction (n = 121, complete data) vs usual care (n = 291, complete data)	Pound sterling, 2018 Healthcare resource use (cost utility) and secondary analysis (cost‐effectiveness) from societal perspective. 2018, 1 year	Falls‐related healthcare services (general practitioner, nurse, physiotherapist outpatient visits, inpatient overnight stays). The intervention cost comprised (1) the cost of occupational therapists’ time spent on the home visit, including travel for the visit, and follow‐up telephone call(s); (2) the cost of training the OTs; and (3) the cost of the equipment provided to the participant following recommendations from the OTs at the home visit.	The base‐case cost‐effectiveness analysis from a National Health Service and Personal Social Services perspective found that, on average per participant, the intervention was associated with additional costs (GBP 18.78, 95% confidence interval 16.33 to 21.24), compared to usual care. The intervention cost was GBP 137, but was less effective (mean QALY loss −0.0042, 95% confidence interval −0.0041 to −0.0043). Sensitivity analyses demonstrated that the findings were uncertain.	‐	Small cost and QALY differences between groups: QALYs were found to be marginally lower for the intervention group (reduction of 0.0015 QALYs; P = 0.814) when adjusting for baseline utility and all covariates. Incremental cost‐effectiveness ratio (ICER) was not calculated.
Cumming 1999 Source: Salkeld 2000 Recently hospitalised (n = 444) plus from outpatients (n = 26) and day‐care centres (n = 60), 65 years and over, mean age 80 Analysis: Cost‐effectiveness	Home fall‐hazard reduction vs no intervention Subsample of last 212 recruited (N = 103 intervention; N = 109 control), which was extrapolated to all 530 participants to derive cost‐effectiveness estimates (and 203 participants with a history of fall)	Societal perspective. Australian dollar 1997, 328 days	Intervention costs: home modification costs, OT costs. Hospitalisation, other healthcare costs: provided in institutional setting (e.g. outpatients), other healthcare costs provided in the home (e.g. home nursing), informal care costs (e.g. personal care provided by a relative or friend and help around the home)	Mean occupational therapist cost AUD 116; mean home modification costs AUD 7	Mean health, home, and community care costs: control AUD 8279; intervention: AUD 10,084. Mean costs for participants with history of fall: control: AUD 8279; intervention: AUD 11,457	Incremental cost per fall prevented: AUD 4986; participants with a previous fall cost per fall prevented: AUD < 0 (sensitivity analysis, outliers removed)
Stark 2021 Analysis Cost utilisation	Home fall‐hazard reduction (n = 155 intervention; n = 155 control) Receiving services from aged‐care agency, past fall or fear of falling	US dollar, 2015, 1 year	Emergency department use, inpatient hospitalisation, physician visits, OT and PT visits as inpatient and outpatient. Intervention costs: OT and contractor time, equipment and materials	A mean (SD) of USD 765.83 (298.27) per participant was spent on delivering the intervention.	The per‐person declines in healthcare utilisation were valued at USD 1613.63 based on medical expenditure data, for a 111% ROC.	‐
Abbreviations: OT: occupational therapist; PT: physical therapist; ROC: receiver operating characteristic curve (graph showing the performance of a classification model at all classification thresholds); SD: standard deviation

Open in a new tab

Data and analyses

Comparison 1. Home fall‐hazard reduction versus control: primary outcome: rate of falls.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Rate of falls ‐ overall analysis	12	5293	Rate Ratio (IV, Random, 95% CI)	0.74 [0.61, 0.91]
1.2 Rate of falls ‐ subgrouped by risk of falls at baseline	12		Rate Ratio (IV, Random, 95% CI)	Subtotals only
1.2.1 Selected for high risk of falling at baseline	9	1513	Rate Ratio (IV, Random, 95% CI)	0.62 [0.56, 0.70]
1.2.2 Not selected for high risk of falling at baseline	6	3780	Rate Ratio (IV, Random, 95% CI)	1.05 [0.96, 1.16]
1.3 Rate of falls ‐ subgrouped by tailoring of intervention to falls	12		Rate Ratio (IV, Random, 95% CI)	Subtotals only
1.3.1 Higher tailoring of intervention to falls	7	3100	Rate Ratio (IV, Random, 95% CI)	0.69 [0.52, 0.91]
1.3.2 Limited tailoring of intervention to falls	5	2193	Rate Ratio (IV, Random, 95% CI)	0.95 [0.78, 1.15]
1.4 Rate of falls ‐ subgrouped by personnel	12		Rate Ratio (IV, Random, 95% CI)	Subtotals only
1.4.1 Intervention delivered by occupational therapist	9	3074	Rate Ratio (IV, Random, 95% CI)	0.68 [0.51, 0.91]
1.4.2 Intervention delivered by other personnel (not occupational therapist)	4	2219	Rate Ratio (IV, Random, 95% CI)	0.91 [0.74, 1.12]

Open in a new tab

Comparison 2. Home fall‐hazard reduction versus control: secondary outcomes.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Number of fallers ‐ overall analysis	12	5253	Risk Ratio (IV, Random, 95% CI)	0.89 [0.82, 0.97]
2.2 Number of fallers ‐ subgrouped by risk of falls at baseline	12		Risk Ratio (IV, Random, 95% CI)	Subtotals only
2.2.1 Selected for high risk of falls at baseline	9	1473	Risk Ratio (IV, Random, 95% CI)	0.74 [0.65, 0.85]
2.2.2 Not selected for high risk of falls at baseline	6	3780	Risk Ratio (IV, Random, 95% CI)	0.99 [0.92, 1.07]
2.3 Number of fallers ‐ subgrouped by tailoring of intervention to falls	12		Risk Ratio (IV, Random, 95% CI)	Subtotals only
2.3.1 Higher tailoring of intervention to falls	7	3100	Risk Ratio (IV, Random, 95% CI)	0.83 [0.73, 0.94]
2.3.2 Limited tailoring of intervention to falls	5	2153	Risk Ratio (IV, Random, 95% CI)	0.91 [0.81, 1.03]
2.4 Number of fallers ‐ subgrouped by personnel	12		Risk Ratio (IV, Random, 95% CI)	Subtotals only
2.4.1 Intervention delivered by occupational therapist	10	2881	Risk Ratio (IV, Random, 95% CI)	0.80 [0.69, 0.93]
2.4.2 Intervention delivered by other personnel (not occupational therapist)	3	2118	Risk Ratio (IV, Random, 95% CI)	0.93 [0.82, 1.05]
2.5 Number of people experiencing a fall‐related fracture	2	1668	Risk Ratio (IV, Random, 95% CI)	1.00 [0.98, 1.02]
2.6 Number of people who experienced 1 or more falls requiring hospital admission	3	325	Risk Ratio (IV, Random, 95% CI)	0.96 [0.87, 1.06]
2.7 Rate of falls requiring medical attention	3	946	Rate Ratio (IV, Random, 95% CI)	0.91 [0.58, 1.43]
2.8 Number of people who experienced 1 or more falls requiring medical attention	2	216	Risk Ratio (IV, Random, 95% CI)	0.89 [0.56, 1.42]
2.9 Health‐related quality of life ‐ overall analysis	5	1848	Std. Mean Difference (IV, Random, 95% CI)	0.09 [‐0.10, 0.27]

Open in a new tab

Comparison 3. Assistive technology versus control ‐ vision improvement.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Rate of falls ‐ overall analysis	3	1489	Rate Ratio (IV, Random, 95% CI)	1.12 [0.84, 1.50]
3.2 Rate of falls ‐ subgrouped by risk of falls at baseline	3		Rate Ratio (IV, Random, 95% CI)	Subtotals only
3.2.1 Selected for higher risk of falls at baseline	2	595	Rate Ratio (IV, Random, 95% CI)	0.96 [0.39, 2.39]
3.2.2 Not selected for higher risk of falls at baseline	3	894	Rate Ratio (IV, Random, 95% CI)	1.16 [0.87, 1.55]
3.3 Number of fallers ‐ overall analysis	3		Risk Ratio (IV, Random, 95% CI)	Subtotals only
3.3.1 Vision improvement	3	1489	Risk Ratio (IV, Random, 95% CI)	1.09 [0.79, 1.50]
3.4 Number of people experiencing 1 or more fall‐related fractures	2	976	Risk Ratio (IV, Random, 95% CI)	1.22 [0.75, 2.00]
3.5 Rate of falls requiring medical attention	1		Rate Ratio (IV, Fixed, 95% CI)	Totals not selected
3.6 Health‐related quality of life	1	597	Mean Difference (IV, Random, 95% CI)	0.40 [‐1.12, 1.92]
3.7 Number of people who experienced 1 or more adverse events (fall after switching glasses)	1	597	Risk Ratio (IV, Fixed, 95% CI)	1.00 [0.98, 1.02]

Open in a new tab

Comparison 4. Assistive technology versus control ‐ footwear, self‐care and assistive devices.

Outcome or subgroup title	No. of studies	Statistical method	Effect size
4.1 Rate of falls ‐ overall analysis	3	Rate Ratio (IV, Random, 95% CI)	Totals not selected
4.1.1 Footwear and foot devices	2	Rate Ratio (IV, Random, 95% CI)	Totals not selected
4.1.2 Self‐care and assistive devices	1	Rate Ratio (IV, Random, 95% CI)	Totals not selected
4.2 Number of fallers ‐ overall analysis	4	Risk Ratio (IV, Random, 95% CI)	Totals not selected
4.2.1 Footwear and foot devices	2	Risk Ratio (IV, Random, 95% CI)	Totals not selected
4.2.2 Self‐care and assistive devices	2	Risk Ratio (IV, Random, 95% CI)	Totals not selected
4.3 Number of people experiencing 1 or more fall‐related fractures	1	Risk Ratio (IV, Fixed, 95% CI)	Subtotals only
4.4 Number of people requiring medical attention	1	Risk Ratio (IV, Fixed, 95% CI)	Totals not selected

Open in a new tab

Comparison 5. Education versus control .

Outcome or subgroup title	No. of studies	Statistical method	Effect size
5.1 Rate of falls ‐ overall analysis	1	Rate Ratio (IV, Fixed, 95% CI)	Totals not selected
5.2 Number of fallers ‐ overall analysis	1	Risk Ratio (IV, Random, 95% CI)	Totals not selected
5.3 Number of people experiencing 1 or more fall‐related fractures	1	Risk Ratio (IV, Fixed, 95% CI)	Totals not selected

Open in a new tab

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Campbell 2005.

*Study characteristics*
Methods	Study design: RCT (factorial) Number of study arms in this review: 2 Length of follow‐up: 12 months
Participants	Setting: New Zealand Number of participants: 196 Number analysed: 196 Number lost to follow‐up: 0 Sample: men and women with severe visual impairment (visual acuity 6/24 or worse) identified in blind register, university and hospital outpatient clinics, and private ophthalmology practice Study period: October 2002 to October 2004  Age (years): mean 83.6 (SD 4.7), range 75 to 96 Sex: 68% women Inclusion criteria: vision worse than 6/24 in better eye; age ≥ 75 years Exclusion criteria: unable to walk around home
Interventions	Randomised into 4 groups. 2 are included in this review. Home safety programme Control: 2 x 1‐hour social visits during the first 6 months of the trial
Outcomes	Rate of falls Number of people falling
Adherence	Mean number of recommendations agreed to = 2.5 (SD 2.2) and number actioned or partially actioned = 2.2 (SD 1.8). 90% complied partially or completely with 1 or more recommendations by occupational therapist (latter data from combined home safety alone and home/safety exercise intervention)
Notes	Cost‐effectiveness analysis reported in the primary reference. Funding source: Health Research Council of Australia
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random numbers
Allocation concealment (selection bias)	Low risk	Schedule held by independent person at separate site, telephone access.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Falls reported by participants who were aware of their group allocation. Phoned by independent assessor blind to allocation. Person classifying fall events also blind to allocation.
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	Low risk	Medical attention: circumstances of serious injuries were confirmed from hospital and general practice records
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	No missing fall data
Selective reporting (reporting bias)	Unclear risk	Clinical trial registration retrospective
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective. Falls recorded daily on monthly prepaid postcard calendars, telephone follow‐up.

Open in a new tab

Chu 2017.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 12 months
Participants	Setting: Hong Kong Number of participants: 204 Number analysed: 198 Number lost to follow‐up: 6 Sample: men and women who had fallen, admitted to emergency department, and discharged directly home Age (years): mean 78.4 (SD 6.1) Sex: 71.1% female Inclusion criteria: age ≥ 65 years, community dwelling, ambulatory with or without a walking aid Exclusion criteria: fell because of excess alcohol intake; sustained a sudden blow or loss of consciousness; sudden onset of paralysis due to stroke or an epileptic seizure; cognition measured by Mini‐Mental State Examination score less than 15; unable to speak Cantonese
Interventions	Occupational therapy fall reduction home visit Control: 1.5‐hour social "well‐wishing" visit
Outcomes	Rate of falls Number of fallers
Adherence
Notes	Source of funding: Medical Research fund, Government of Hong Kong
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used block randomisation (blocks of 4)
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Falls recorded monthly by participants returning postcard calendar. Person classifying fall events also blind to allocation.
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	High risk	Self‐report of falls reported requiring medical attention.
Incomplete outcome data (attrition bias) Fall outcomes	Unclear risk	Less than 20% of fall data missing (< 1%). 6 participants in intervention group only had case files missing and were excluded.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall outcomes reported. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective. Falls reported in fortnightly telephone follow‐up.

Open in a new tab

Cockayne 2021a.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Number of participants: 1331 Number analysed: 1308 Number lost to falls follow‐up: 23 Sample: community‐dwelling adults recruited from trial and GP databases, advertising, and opportunistic screening Study period: July 2016 to December 2019 Age (years): 80.01 (SD 6.3) Sex: 65.5% female Inclusion criteria: aged 65 years or over; community dwelling, i.e. not living in a nursing/residential home; and reported a fear of falling or a fall in the previous 12 months Exclusion criteria: unable to: give informed consent (e.g. due to dementia or Alzheimer’s disease); speak English and had no relative or friend to translate/interpret for them; walk 10 feet even using a walking aid; or if they already had an occupational therapy assessment for falls prevention in the past 12 months or were waiting for an assessment
Interventions	Home visit by occupational therapist to identify personal fall‐related hazards and risk‐taking behaviours, when walking through the home with provision or follow‐up of recommended home modifications. Control group received usual care and a fall prevention leaflet.
Outcomes	Rate of falls Number of people experiencing 1 or more falls (risk of falling) Number of people reporting a fall‐related fracture
Adherence	Reported if received home visit or not, other information not included
Notes	National Institute for Health Research (NIHR) Health Technology Assessment (HTA) Programme (grant 14/49/149). Health service utilisation, health‐related quality of life, and treatment fidelity to be published elsewhere.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The allocation sequence was generated by an independent data systems manager not involved in recruitment.
Allocation concealment (selection bias)	Low risk	Used a secure, remote web‐based randomisation service
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Data entry (fall calendar) staff were blinded. Research staff who contacted participants for fall data when calendars were not returned were not blinded.
Blinding of outcome assessment (detection bias) Fractures Fracture outcome	High risk	Self‐reported fractures, reported by telephone follow‐up by personnel not blinded to group allocation
Blinding of outcome assessment (detection bias)  Health‐related quality of life	Unclear risk	Self‐report questionnaires were mailed to study team. Data entry staff were blind to group allocation.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	Less than 20% (.02%) loss to fall outcomes. Numbers were balanced between groups (10 intervention and 14 controls), and these were imputed as zero falls over a minimal time period of 0.1 month.
Incomplete outcome data (attrition bias)  Health‐related quality of life	Unclear risk	Less than 20% of outcome data missing (11%). There were some imbalances in the intervention (n = 59, 13.7%) and control groups (n = 86, 9.5%); reasons for loss were balanced.
Selective reporting (reporting bias)	Low risk	Prospectively registered in ISRCTN trial registry
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective. Falls recorded daily on monthly calendars, mailed prompt, or telephone follow‐up.

Open in a new tab

Cumming 1999.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 12 months
Participants	Setting: Sydney, Australia Number of participants: 530 Number analysed: 530 Number lost to follow‐up: 0 Sample: community‐dwelling people recruited primarily in selected hospital wards with a high proportion of older people, supplemented by outpatient clinics, and day‐care centres for older clients Study period: June 2095 to January 2098 Age: mean 77 (SD 7.2) Sex: 57% women Inclusion criteria: aged ≥ 65; community dwelling within study area Exclusion criteria: cognitively impaired if not living with someone who could give informed consent and report falls; occupational therapy home visit already planned
Interventions	1 home visit by experienced occupational therapist assessing environmental hazards (validated hazard checklist) and supervision of home modifications. Telephone follow‐up after 2 weeks Control: usual care
Outcomes	Rate of falls Number of people falling Number of people reporting medical attention after a fall (participants with a history of fall)
Adherence	At 12 months, 52% (216/419) recommendations were met with partial or complete adherence. Adherence reported in detail in Cumming 2001.
Notes	Cost‐effectiveness analysis reported in Salkeld 2000. Funding source: Public Health Research and Development Committee grant of the National Health & Medical Research Council
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Stratified block randomisation using random numbers table
Allocation concealment (selection bias)	Low risk	Randomised offsite by person not involved in recruitment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Falls self‐reported by participants. Quote: “Subjects who had not returned a calendar within 10 days of the end of the month were telephoned and asked about falls in the previous month. If one or more falls were reported, a telephone‐administered questionnaire was used to elicit details of each fall” Quote: “Follow‐up interviews were blind to group allocation”
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	High risk	Falls requiring medical attention ascertained by self‐report.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	All participants reported fall data. Data collected for 97% of the person‐months of follow‐up.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall outcomes reported. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective. Falls ascertained using monthly falls calendars.

Open in a new tab

Cumming 2007.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Setting: Sydney, Australia Number of participants: 616 Number analysed: 616 Number lost to falls follow‐up: 0 Sample: men and women from outpatient aged‐care services, some volunteers recruited by advertisement Study period: August 2002 to July 2005 Age (years): mean 80.6 (SD 6) Sex: 68% women Inclusion criteria: age 70 and older; living independently in the community; no cataract surgery or new eye glass prescription in previous 3 months; participant or caregiver able to complete monthly falls calendar Exclusion criteria: none noted
Interventions	Vision tests and eye examinations. Dispensing of new spectacles if required. Referral for expedited ophthalmology treatment if appropriate occular pathology identified. Mobility training and canes if required Control: usual care
Outcomes	Rate of falls Number of people falling Number sustaining a fracture
Adherence	Number declining prescribed vision‐related intervention (11/174, 7.5%). A further 7/15 referred for cataract surgery had not had surgery before end of follow‐up.
Notes	Funding source: National Health & Medical Research Council of Australia, project grant
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described
Allocation concealment (selection bias)	Low risk	Randomised offsite by person not involved in recruitment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	All participants used monthly falls diary, with telephone contact in 2 weeks. Blinding of study personnel recording phoning or recording data from the calendars not described.
Blinding of outcome assessment (detection bias) Fractures Fracture outcome	High risk	Injury including fractures were self‐reported in monthly fall diaries.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	All participants contributed fall data.
Selective reporting (reporting bias)	Low risk	Prespecified fall and fracture outcomes reported. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Self‐report falls calendar. Participants asked to record on each day an "N" if they did not fall and an "F" if they had a fall. If a fall occurred, the participant completed an additional postcard about fall‐related injuries (including fractures).

Open in a new tab

Day 2002.

*Study characteristics*
Methods	Study design: RCT (factorial) Number of study arms included in review: 3 Length of follow‐up: 18 months
Participants	Setting: Melbourne, Australia Number of participants: 412 Number analysed: (Home safety only: N = 136, Vision improvement only: N = 139, No intervention: N = 137) Number lost to follow‐up: 0 Sample: community‐dwelling men and women identified from electoral roll (60% women) Age (years): mean 76.1 (SD 5.0) Inclusion criteria: aged ≥ 70; community dwelling and able to make modifications; expected to remain in area for 2 years (except for short absences); have approval of family physician Exclusion criteria: undertaken regular to moderate exercise with a balance component in previous 2 months; unable to walk 10 to 20 m without rest or help or having angina; severe respiratory or cardiac disease; psychiatric illness prohibiting participation; dysphasia; recent major home modifications; education and language adjusted score > 4 on the short portable mental status questionnaire
Interventions	Randomised into 8 groups. 3 included in this review.Home hazard management only: home assessed by "trained assessor", hazards removed or modified by participants or City of Whitehorse's home maintenance programme. Staff visited home, provided quote for work including free labour and materials up to AUD 100. Checklist focused on steps and stairs, floor surfaces, lighting, bathroom fittings, furniture. Vision improvement only: assessed at baseline using dual visual acuity chart. Referred to usual eye care provider, general practitioner, or local optometrist if not already receiving treatment for identified impairment No intervention. Received brochure on eye care for over 40‐year‐olds
Outcomes	Rate of falls Number of people who experienced 1 or more falls (risk of falling)
Adherence	Home hazards. Type and number of modifications completed by Whitehorse home maintenance reported but not participant changes. The mean number of hazards at baseline and follow‐up reported for intervention and control groups. Vision correction. Number receiving referrals and number with prescribed glasses changes or cataract surgery. Visual acuity changes measured in intervention and control groups.
Notes	Source of funding: Australian National Health and Medical Research Council, Victorian Department of Human Services (Aged Care), City of Whitehorse, Victorian Health Promotion Foundation, Rotary, National Safety Council
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised by "adaptive biased coin" technique to ensure balanced group numbers
Allocation concealment (selection bias)	Low risk	Computer generated by an independent third party contacted by telephone
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	All participants used monthly falls diary, with telephone contact from a researcher blinded to group allocation if not returned in 5 days.
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	High risk	Falls requiring medical attention: obtained through self‐report of participants
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	No missing fall data
Selective reporting (reporting bias)	Unclear risk	Prespecified fall and fall requiring medical care outcomes reported. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Blinded assessor telephoned every 2 weeks.

Open in a new tab

Haran 2010.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 13 months
Participants	Setting: Sydney and Illawarra regions, New South Wales, Australia Number of participants: 606 Number analysed: 597 Number lost to follow‐up: 9 Sample: sample from electoral roll; residents of retirement villages; outpatients and inpatients discharged from rehabilitation and orthopaedic ward; responders to advertising, etc. Study period: June 2005 to 2008 Age: mean 80 (SD 6.6) Sex: 65% women Inclusion criteria: community dwelling; at a relatively high risk of falls (80 years, or 65 to 79 and Timed Up and Go Test ≥ 15 s and/or 1 fall in past 12 months); using bifocal, trifocal, or progressive lens glasses 3 x per week when walking outdoors; reviewed by an optometrist or ophthalmologist in previous 24 months; “quite or very confident” that they could comply with the study recommendations Exclusion criteria: using single‐lens distance glasses; residing in high‐care residential facility; cognitive impairment (Mini‐Mental State Examination < 24); severe visual impairment; insufficient English language skills; ophthalmic surgery planned in the next 12 months; unstable medical condition
Interventions	Optometrist examination; prescribed single‐lens distance glasses for use in most walking or standing activities and given advice on use of their glasses Control: used their multifocal glasses in their usual manner (no advice) All participants received an optometry assessment and updated multifocal glasses (if required) at baseline.
Outcomes	Rate of falls Number of people falling Number sustaining fall‐related fractures Health‐related quality of life Number of adverse events (non‐fall injuries and falls as a result of switching between single‐lens and multifocal glasses)
Adherence	Adherence, reported monthly, to wearing single‐lens distance glasses was most or all of the time walking in the street or in shopping centres.
Notes	Funding source: Australian National Health and Medical Research Council grant
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Each stratum was randomly allocated in permuted blocks of 10 generated externally (by JS)" a professor of statistics
Allocation concealment (selection bias)	Low risk	Quote: "by using sequentially numbered opaque sealed envelopes containing group assignment"
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Both groups received an intervention, i.e. an optometrist examination and updated multifocal lens prescription if required. The intervention group were prescribed a pair of single‐lens distance glasses and advice.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Falls reported in monthly calendars and follow‐up telephone calls as required. Research personnel who received the calendars and made the calls were blinded to group allocation.
Blinding of outcome assessment (detection bias) Fractures Fracture outcome	High risk	Fall‐related injuries were reported on monthly fall calendar with a follow‐up phone call. They were not verified from the primary sources.
Blinding of outcome assessment (detection bias)  Health‐related quality of life	Unclear risk	Blinding not described.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	Less than 20% of fall outcome data are missing (2%). Minor imbalance between groups (intervention n = 6, control n = 3). Reasons for fall data loss were unclear.
Incomplete outcome data (attrition bias)  Health‐related quality of life	Low risk	Less than 20% of outcome data are missing (8%). Loss of data was balanced between intervention (n = 27) and control (n = 21), reasons for loss were similar.
Selective reporting (reporting bias)	Low risk	Prespecified fall outcomes reported. Trial protocol published.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Daily diaries returned monthly with a follow‐up phone call if not returned.

Open in a new tab

Kamei 2015.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 12 months
Participants	Setting: Japan Number of participants: 130 Number analysed: 110 Number lost to follow‐up: 20 Sample: Tokyo metropolitan area, recruited by flyers, posters and website Age: not reported (65 or over) Sex: 85% female Inclusion criteria: aged 65 and over, first time in programme, lives in own residence, physician approved to exercise Exclusion criteria: low cognitive function, dementia and/or poor physical condition such as inability to walk by themselves
Interventions	Intervention: home hazard modification programme of home hazard awareness education using combination of lecture with residential mock set‐up and equipment, participants provided a self‐report home hazard assessment. This was in addition to a fall prevention programme including fall risk, exercise, food nutrition and foot self‐care. Controls: no home hazard modification programme. Attended usual fall prevention programme as above
Outcomes	Falls risk reduction
Adherence	Home safety awareness questionnaire (10 items) and hazard modification checklist (33 item) self‐reported at 12 weeks and 52 weeks
Notes	Number of falls at 12 months clarified by authors in email. Funding source: St. Luke's and TERUMO collaborative fund
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Research assistants allocated participants randomly. No details provided.
Allocation concealment (selection bias)	High risk	No details or evidence of concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Trained nurses collected the fall diaries, but not reported if blinded.
Blinding of outcome assessment (detection bias) Fractures Fracture outcome	High risk	Self‐report of outcome
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	Less than 20% of fall outcome data are missing (15%). Loss of fall data was balanced in the intervention (n = 11) and control (n = 9) groups, reasons for missing data were unclear.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall and fracture outcomes reported. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	Unclear risk	Daily calendar completed by participants, but was handed to researchers at 12 weeks and 1 year.

Open in a new tab

Keall 2015.

*Study characteristics*
Methods	Study design: RCT (cluster by household) Number of study arms: 2  Length of follow‐up: 36 months
Participants	Setting: New Zealand Number of participants: 477 Number analysed: 477 Number lost to falls requiring medical attention follow‐up: 0 Sample: occupants of Taranaki Region housing constructed before 1978 and owned by a Community Study period: March 2010 to May 2014 Services card holder (entitled to government assistance due to low income or pension) Age (years): all ages included, data used in review were for participants 70 years and over Sex: no data on numbers of males and females Inclusion criteria: occupants of Taranaki Region housing constructed before 1978 and owned by a Community Services Card holder (entitled to government assistance due to low income or pension) Exclusion criteria: not intending to stay in house for next 3 years
Interventions	Qualified builders assessed the house using a standard checklist of hazards in the home that were within the scope of the home modifications intervention (e.g. grab rails, outside lighting, non‐slip surfacing).
Outcomes	Injurious falls that were reported to the national no‐faults accident insurer ACC
Adherence	A small sample (10%) was contacted by phone to ascertain if the house was modified as specified.
Notes	Funding source: Health Research Council of New Zealand. Cost‐benefit analysis published, but no data available for older age groups.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Statistician generated a randomisation schedule with sR version 2.10.0 using electronic coin toss.
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	House owners were aware if allocated to intervention, but impact on outcomes unclear.
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	Low risk	Data for primary outcome obtained from ACC by matching participant name; date of birth and address. Text descriptions used to code if a fall (fall, trip or slip). Ambiguous text referred to the committee, coders and the committee were unaware of the group status of individual whose injury text descriptions were being assessed.
Incomplete outcome data (attrition bias) Fall outcomes	Unclear risk	The authors state that the coverage of medically treated injuries by the ACC scheme is high. All injurious falls that were treated medically were reported to the ACC and included in intention‐to‐treat analysis. There may have been some participants that moved house during the 3‐year follow‐up, and this would not have been known.
Selective reporting (reporting bias)	Low risk	Trial registered prospectively with the ANZ Clinical Trial Registry.
Method of ascertaining falls (recall bias) Fall outcomes	Unclear risk	Data were obtained from the national no‐fault accident insurer ACC. The authors state a limitation is "the dependence on text‐field reporting of the circumstances of the injury, with unknown validity ‐ particularly, whether or not a fall might have been the cause". Further the injury was coded as a fall (fall, slip or trip).
Cluster‐randomised trials	Low risk	Randomisation occurred after consent was obtained for individuals within the household. No missing outcomes in the cluster, clustering was accounted for in the analysis and results reasonably comparable with other 2 trials.

Open in a new tab

Lannin 2007.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 3 months
Participants	Setting: Sydney, Australia Number of participants: 10 Number analysed: 10 Number lost to falls follow‐up: 0 Sample: patients admitted to a rehabilitation facility and referred to occupational therapy Age: mean 81 (SD 7) Sex: 80% female Inclusion criteria: mild or no cognitive impairment; community dwelling (non‐institutional); aged 65 or older; no medical contraindications that would require strict adherence to equipment recommendations Exclusion criteria: none
Interventions	Best‐practice occupational therapy home visit intervention to maximise functional capacity and confidence in returning home Control: standard practice in‐hospital assessment and education
Outcomes	Number of falls Number of people falling Number of people reporting a fall requiring medical attention
Adherence	Not reported
Notes	Pilot study. Funding source: Royal Ryde Rehabilitation Centre
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Allocation schedule computer generated.
Allocation concealment (selection bias)	Low risk	Quote: "Concealed in opaque, consecutively numbered envelopes by a person not involved in the study."
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Falls reported by participants. Outcome assessor collecting data was blinded to group allocation.
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	High risk	Medical attention self‐reported.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	All participants contributed fall data.
Selective reporting (reporting bias)	Low risk	Prespecified fall outcomes reported. Trial protocol registered prospectively (Australian Clinical Trials Register 1206000213549).
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Data collection at first fortnight, then monthly by phone

Open in a new tab

Lin 2007.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 6 months
Participants	Setting: Taiwan Number of participants: 100 Number lost to follow‐up: 0 Sample: residents of rural agricultural area  Age (years): mean 76.8 Sex: 51% female Inclusion criteria: medical attention for a fall in previous 4 weeks, ≥ 65 years Exclusion criteria: none described
Interventions	Randomised into 3 groups; 2 are included in this review. Home safety assessment and modification (public health worker) Control: "education". 1 social visit 30 to 40 min every 2 weeks for 4 months with fall prevention pamphlets provided (public health worker)
Outcomes	Rate of falls Health‐related quality of life
Adherence	Not reported
Notes	Source of funding: Bureau of Health Promotion, Department of Health, National Science Council
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Block randomised in groups of 6. Insufficient information to permit judgement
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Quote: “Participants were asked to report their falls by telephone or postcard; they were also contacted by telephone every 2 weeks to ascertain the occurrence of falling”. The method of ascertaining falls was the same in all groups. Blinding of assessors not reported.
Blinding of outcome assessment (detection bias)  Health‐related quality of life	Low risk	Assessors collecting baseline and 2 follow‐up assessments were blinded to group allocation. Validated quality of life tool used (WHOQOL).
Incomplete outcome data (attrition bias) Fall outcomes	Unclear risk	Unclear if all participants contributed fall data, as fall data collected fortnightly. Less than 20% of outcome data missing (14%) in other (quality of life) outcomes reported (2‐ and 4‐month data collection) with 4 from intervention and 10 from control lost.
Incomplete outcome data (attrition bias)  Health‐related quality of life	Unclear risk	Less than 20% of outcome data are missing (14%). Loss of data was unbalanced between intervention (n = 4) and control (n = 10), reasons were unclear.
Selective reporting (reporting bias)	Unclear risk	Health‐related quality of life reported as primary outcome, and falls reported as secondary outcome in paper. Protocol not published.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective. Reported falls by telephone or postcard when they occurred. Phoned every 2 weeks to ascertain occurrence of falls

Open in a new tab

Lockwood 2019.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 6 months
Participants	Setting: Melbourne, Australia Number of participants: 77 Number analysed: 73 Number lost to falls follow‐up: 4 Sample: hospitalised for hip fracture and expected to return to private residence (90% discharged to a private residence) Study period: May 2016 to August 2017  Age (years): 82.2 (SD 7.2) Sex: 71.4% female Inclusion criteria: admitted to a public hospital within the region with a primary diagnosis of hip fracture, aged 50 years or older, discharge plan with an expected return to a private residence Exclusion criteria: patients with comorbidities or cognitive impairment, or both, were eligible unless, after assessment by the treating healthcare team, the patient was deemed to have unusually complex medical or social circumstances likely resulting in an expected length of stay greater than 28 days; lived more than 1‐hour drive from the hospital
Interventions	Occupational therapy home visit with participant 1 to 5 days prior to discharge; with a family member if moderate cognitive impairment. The home assessment visit aimed to maximise functional capacity and confidence in returning home. Assessment of mobility, self‐care, and household safety during observation of task performance in the home environment including 6 specific fall‐hazard items. Provision of education, advice and recommendations on equipment, home adaptations, and community support services The control group received hospital‐based discharge planning only, which included an occupational therapy interview and use of the Home Falls and Accidents Screening Tool as a basis for discussion of managing day‐to‐day home activities, and education on safe use of adaptive equipment.
Outcomes	Rate of falls Number of people who experienced 1 or more falls (risk of falling) Number of unplanned hospital readmissions Health‐related quality of life
Adherence	6 participants did not receive the home visit, and 1 did not receive the home visit intervention as allocated.
Notes	The study main outcome was at 30 days. 6‐month outcomes used in the review. Funding source: Australian Government Research Training Program Scholarship and a grant from the Eastern Health Foundation and Research Focus Area of Sport, Exercise and Rehabilitation at La Trobe University
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used a computer‐generated block random numbers
Allocation concealment (selection bias)	Low risk	Sequentially numbered, sealed, opaque envelopes prepared by a team member not involved in recruitment or data collection.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Falls reported by participants aware of group allocation, same method for both groups. Diary calendar and phone calls by blinded assessor
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	Low risk	Medical records used to ascertain readmissions in publicly funded hospitals within the region.
Blinding of outcome assessment (detection bias)  Health‐related quality of life	Low risk	Blinded assessors collected secondary data.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	Less than 20% of fall outcome data (6%) and hospital readmission for a fall data (13%) are missing (6%). Missing data are mostly related to death and unlikely to be linked to the intervention.
Incomplete outcome data (attrition bias)  Health‐related quality of life	High risk	More than 20% of outcome data are missing (22%). Groups were balanced and reasons for loss were similar.
Selective reporting (reporting bias)	Low risk	Prespecified fall outcomes reported. Trial registered prospectively.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Fall calendars and monthly phone calls

Open in a new tab

McKiernan 2005.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Setting: Wisconsin, USA Number of participants: 113 Number analysed: 109 Number lost to falls follow‐up: 4 Sample: recruited from fall registry and by single media release Study period: October 2003 to April 2004  Age (years): mean 74.2, range 65 to 96 Sex: 60% women Inclusion criteria: aged ≥ 65 years; community dwelling; ≥ 1 falls in previous year; independently ambulatory Exclusion criteria: not capable of applying Yaktrax walker correctly or of discerning correct outdoor conditions to wear them
Interventions	Yaktrax walker (netting applied over usual footwear with wire coils to increase grip in winter outdoor conditions) Control: usual winter footwear
Outcomes	Rate of falls (outdoor falls) Number of fall‐related fractures Number of falls requiring medical attention
Adherence	Footwear satisfaction survey
Notes	Funding source: Marshfield Clinic Research Foundation
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "randomised"
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group. 19% of the control group admitted to wearing a similar device during the course of the study. Unclear if this was a bias within study
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Falls reported by participants who were aware of their group allocation. Blinding of research staff not described.
Blinding of outcome assessment (detection bias) Fractures Fracture outcome	High risk	Self‐report of fracture, not verified
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	High risk	Medical attention: self‐report and not verified
Incomplete outcome data (attrition bias) Fall outcomes	Unclear risk	Less than 20% of outcome data are missing (4%). It is unclear which groups had the participants (n = 4) missing fall data.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall and injurious fall outcomes reported. No data included on number of fallers. Protocol not published.
Method of ascertaining falls (recall bias) Fall outcomes	Unclear risk	Prospective fall diary returned by post, but not clear if returned on a monthly basis.

Open in a new tab

Nikolaus 2003.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Setting: enrolled in hospital, but community‐based intervention, Germany Number of participants: 360 Number analysed: 360 Number lost to falls follow‐up: 0 Sample: frail "older people" admitted to a geriatric clinic who normally lived at home. Included a subgroup who had a history of multiple falls Age (years): mean 81.5 (SD 6.4) Sex: 73% women Inclusion criteria: lived at home before admission and able to be discharged home; with at least 2 chronic conditions (e.g. osteoarthritis or chronic cardiac failure, stroke, hip fracture, parkinsonism, chronic pain, urinary incontinence, malnutrition) or functional decline (unable to reach normal range on at least 1 assessment test of activities of daily living or mobility) Exclusion criteria: terminal illness; severe cognitive decline; living > 15 km from clinic
Interventions	Home assessment and intervention (from interdisciplinary team (HIT). 1 home visit prior to discharge to identify home hazards and prescribe technical aids if necessary (occupational therapist plus either nurse or physiotherapist depending on anticipated needs). At least 1 more visit (mean 2.6) to inform about possible fall risks in home, advice and follow‐up on changes to home environment, facilitate changes, and teach use of technical and mobility aids. Comprehensive geriatric assessment in hospital Control: no intervention home visit. Comprehensive geriatric assessment in hospital. Usual care postdischarge, GP managed
Outcomes	Rate of falls Number sustaining a fracture
Adherence	Compliance with home safety recommendations reviewed at 12 months. In 105 homes at least 1 recommended change had been implemented (compliance rate 75.5%).
Notes	Methods paper described a third arm receiving usual hospital and home care. Funding source: Sozialministerium Baden‐Württemberg
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used a random number sequence
Allocation concealment (selection bias)	Unclear risk	Insufficient information: sealed envelopes containing group assignments
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Interviewer was blinded to group allocation.
Blinding of outcome assessment (detection bias) Fractures Fracture outcome	High risk	Self‐report of fractures
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	No missing outcome data for falls
Selective reporting (reporting bias)	Unclear risk	Prespecified fall outcomes reported. Early protocol published with subsequent changes to scope and focus of study.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective. Participants completed a diary to record falls and injuries and were also contacted monthly by phone to obtain information on falls, fall injury and circumstances.

Open in a new tab

Pardessus 2002.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2 Length of follow‐up: 12 months
Participants	Setting: France Number of participants: 60 Number analysed: 60 Number lost to falls follow‐up: 0 Sample: individuals hospitalised for a fall Study period: December 1997 to September 1999 Age (years): mean 83.2 (SD 7.7) Sex: 78% female Inclusion criteria: aged ≥ 65; hospitalised for falling; able to return home; able to give consent Exclusion criteria: cognitive impairment (MMSE < 24); falls due to cardiac, neurologic, vascular, or therapeutic problems; without a phone; lived > 30 km from hospital
Interventions	Comprehensive 2‐hour home visit prior to discharge with ergotherapist (occupational therapist) and "physical medicine and rehabilitation doctor" to evaluate abilities in real‐life environment. Assessment of activities of daily living, instrumental activities of daily living, transfers, mobility inside and outside, use of stairs. Environmental hazards identified and modified where possible; if not, advice given. Discussion of social support. Referrals for social assistance Control: usual care
Outcomes	Number of people falling Mean number of falls per person reported, but unable to calculate a rate of falls.
Adherence	Not reported
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised using random numbers table
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	High risk	Blinding not reported. Authors conducted all follow‐up phone calls.
Blinding of outcome assessment (detection bias)  Hospital admission or Medical attention	High risk	Self‐reported hospitalisations
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	No missing data
Selective reporting (reporting bias)	Unclear risk	Prespecified data reported. Did not include data for calculating rate of falls. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	Unclear risk	Interval recall, but short interval. Falls identified by monthly phone calls.

Open in a new tab

Perry 2008.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 3 months
Participants	Setting: Ontario, Canada Number of participants: 46 Number analysed: 40 Number lost to falls follow‐up: 6 Sample: healthy volunteers aged 65 to 75 years Age (years): mean 69 (SD 3.4) Sex: 48% women Inclusion criteria: moderate insensitivity of the foot soles, as compared to published norms for young adults Exclusion criteria: a clinical diagnosis of diabetes or peripheral neuropathy
Interventions	Intervention: balance‐enhancing insole ("Sole‐Sensor") ‐ raised edge around perimeter of insole Control: normal insole All participants fitted with standard‐pattern shoe.
Outcomes	Number of people falling Other study outcomes reported but not included in this review.
Adherence	Weekly self‐reported hours of wear per week
Notes	Falls are a secondary outcome in this study. Funding source: Canadian Institutes of Health Research, Canadian Foundation for Innovation, Ontario Innovation Trust and Wilfrid Laurier University
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "Each participant performed the gait‐perturbation protocol … with both types of insoles, and was then randomly assigned to either the test or control group". Insufficient information to permit judgement
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants were not blinded to which allocated group they were in, facilitatory or conventional insole; however this is unlikely to have introduced performance bias.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Falls reported by participants who were aware of their group allocation. Not clear if outcome assessors were blind to allocation
Incomplete outcome data (attrition bias) Fall outcomes	Unclear risk	Less than 20% of fall outcome data are missing (13%). It is unclear to which groups the 6 participants with missing data were allocated.
Selective reporting (reporting bias)	Unclear risk	Data collected for rate of falls not included.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Weekly postcards of fall information, 100% compliance in reporting

Open in a new tab

Pighills 2011.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 3 Length of follow‐up: 12 months
Participants	Setting: affluent rural and deprived urban areas, Yorkshire, UK Number of participants: 238 Number analysed: 233 Number lost to falls follow‐up: 5 Sample: recruited from 13 GP lists in the Airedale NHS Trust Study period: February 2006 to September 2007 Age (years): mean 79 (SD 6) Sex: 67% women Inclusion criteria: community dwelling; aged ≥ 70; history of ≥ 1 fall in previous 12 months, interviewed by OT for eligibility Exclusion criteria: living in nursing or residential care homes; receiving OT services; had fall‐specific OT intervention in past year
Interventions	Fall‐hazard reduction assessment and intervention provided by occupational therapist, recommendations sent to participant and referrals made for equipment and other input As above, but provided by a trained non‐professionally qualified domiciliary support worker Control: usual care from GP
Outcomes	Rate of falls Number of people falling
Adherence	Adherence to recommendations were reviewed at 4 weeks and 12 months by telephone.
Notes	Funding source: National Institute for Health Research, National Coordinating Centre for Research Capacity Development programme fellowship
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Independent and remotely conducted simple web‐based computer‐generated randomisation
Allocation concealment (selection bias)	Low risk	The computer‐generated outcome of randomisation was automatically emailed to an independent person who passed the participant's case notes on to the contact person for the group to which participant had been randomised.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Fall diary data was mailed monthly and followed up in 2 weeks by a blinded researcher if not returned. All diary data imputed by independent staff of the York Trials Unit.
Blinding of outcome assessment (detection bias)  Health‐related quality of life	Low risk	Used validated health‐related quality of life measure
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	Less than 20% lost to fall follow‐up (2%). Minor imbalance of loss between groups (1 died in control group; 2 withdrew in intervention delivered by domiciliary care worker; 2 withdrew in intervention delivered by occupational therapist)
Incomplete outcome data (attrition bias)  Health‐related quality of life	Low risk	Less than 20% of outcome data are missing (10%). There was some imbalance between intervention (OT n = 7, 8%; AT n = 4, 5%) and control (n = 10, 13%); imbalances largely due to deaths and not likely to impact outcome.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall and health‐related quality of life outcomes reported. Trial registered retrospectively.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Used postcard diary, mailed monthly; followed up in 2 weeks if not sent. Participant could also ring toll‐free phone number if they wished to report if fall or multiple falls had occurred.

Open in a new tab

Stark 2021.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Number of participants: 310 Number analysed: 275 Number lost to falls follow‐up: 35  Sample: recruited older adults receiving services from an Area Agency on Aging Study period: January 2015 to October 2017 Age (years): 75 (SD 7.4) Sex: 74% female Inclusion criteria: adults aged 65 or older; did not have dementia; high risk of falling (fall/s in past month; receiving services from aged‐care agency); community residing Exclusion criteria: living in institutional care or severely cognitively impaired
Interventions	Home hazard identification and removal: comprehensive assessment of person, behaviours, and environment; home hazard removal plan; remediation of hazards; 6‐month booster visit Control group: usual care
Outcomes	Rate of falls Number of people experiencing one or more falls (risk of falling) Health‐related quality of life
Adherence	Home visits completed (at least 2); adherence to at least 80% of hazard removal recommendations
Notes	Healthcare utilisation costs reported with a return‐on‐investment calculation.  Funding source: US Department of Housing and Urban Development (grant No. MOHHU0024‐14)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation sequences were generated by a computerised formal probability model.
Allocation concealment (selection bias)	Low risk	Allocation concealed using an electronic interface.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Low risk	Fall outcomes were recorded and confirmed using the same method in both groups. Raters recording and confirming outcomes were blind to group allocation.
Blinding of outcome assessment (detection bias)  Health‐related quality of life	Low risk	Data collected by raters blinded to group allocation.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	Less than 20% fall data are missing (11%). Minor imbalance in loss for intervention (n = 20) and controls (n = 15); reasons for loss are similar
Incomplete outcome data (attrition bias)  Health‐related quality of life	Low risk	Less than 20% of outcome data are missing (18%). Groups were balanced in number with similar reasons for loss.
Selective reporting (reporting bias)	Low risk	Prespecified fall outcomes reported in published protocol paper and prospectively in trial registry.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Daily calendar/journal with family birthdays and anniversary dates for enhanced accuracy mailed monthly with follow‐up call if not returned.

Open in a new tab

Stevens 2001.

*Study characteristics*
Methods	Study design: RCT. Some household clusters (n = 318): co‐inhabitants assigned to same group as index recruit. Study population stratified by age (< 80 years and > 80 years) and sex. Within strata index, recruits were allocated in 2:1 ratio to control or intervention. Number of study arms: 2  Length of follow‐up: 12 months
Participants	Setting: Perth, Australia Number of participants: 1879 Number analysed: 1737 Number lost to falls follow‐up: 142 Sample: randomly selected from State Electoral Roll and telephone directory Study period: July 1995 to July 1996  Age (years): mean 76 Sex: 53% female Inclusion criteria: aged ≥ 70; living independently; cognitively intact and able to speak and write in English; anticipated living at home for 10/12 coming months; no ramps or rails installed Exclusion criteria: if living with more than 2 other older people
Interventions	Intervention: home hazard assessment by registered nurse. Hazard list (n = 11) designed with occupational therapy input to include most frequent hazards identified from literature and existing checklists. All identified hazards discussed with participants, but only the 3 most conspicuous or remediable selected to give specific advice on their removal or modification. Safety devices offered at no cost and installed by tradesman within 2 weeks of visit. Control: usual care Co‐inhabitants assigned to same group as index recruit.
Outcomes	Rate of falls Number of people falling
Adherence	Change in hazard prevalence assessed in sample of 51 homes at 11 months postrandomisation.
Notes	Source of funding: Western Australian Health Promotion Foundation, Lotteries Commission of Western Australia, and Health Department of Western Australia
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Study population divided into 4 strata defined by age (< 80 years and > 80 years) and sex. Within these strata, index recruits were allocated in 2:1 ratio to control or intervention. Co‐inhabitants assigned to same group as index recruit.
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Falls were recorded in a diary and mailed monthly. Reported falls were confirmed by phone, blinding of interviewer not described.
Incomplete outcome data (attrition bias) Fall outcomes	Unclear risk	Less than 20% of fall outcome are missing (8%). Recruitment in ratio of 2 control: 1 intervention. Unbalanced losses between intervention group (n = 65, 10%) and control group (n = 77, 6%). Reasons for loss were moved, became ill or died, but unclear between groups.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall outcomes reported. Protocol not published.
Method of ascertaining falls (recall bias) Fall outcomes	Low risk	Prospective, falls recorded in diary
Cluster‐randomised trials	Low risk	Individuals were recruited to the trial prior to randomisation, no cluster was lost, clustering was accounted for in the analysis, and results are comparable with other trials.

Open in a new tab

Tchalla 2013.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Stetting: Corrèze district of Limousin, France Number of participants: 96 Number analysed: 96 Number lost to falls follow‐up: 0 Sample: frail elderly people register Study period: July 2009 to June 2010 Age (years): mean 86.6 (SD 6.5) Sex: 77% women Inclusion criteria: mild to moderate diagnosis of Alzheimer's disease, 65 years and older, living at home, on frail elderly people register Exclusion criteria: severe dementia (MMSE < 10) or already in a falls prevention programme
Interventions	Intervention: automatic night light near the bed which senses person's foot on the floor coupled with tele‐assistance service Control: no automatic night light system
Outcomes	Number of people falling at home Cumulative incidence of falls at home
Adherence	Rate of acceptance of the new technology
Notes	Funding source: Fondation Caisse d’Épargne
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Dynamic random allocation using minimisation criteria
Allocation concealment (selection bias)	Unclear risk	Method not described.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Clinical research associate undertook follow‐up. Blinding not described.
Incomplete outcome data (attrition bias) Fall outcomes	Low risk	All participants included in intention‐to‐treat analysis.
Selective reporting (reporting bias)	Unclear risk	Prespecified fall outcomes reported. No published protocol
Method of ascertaining falls (recall bias) Fall outcomes	High risk	Monthly calls to GP, unclear how information obtained from family/caregiver

Open in a new tab

Thomas 2018.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 3  Length of follow up: 15 weeks
Participants	Setting: USA (Texas, Florida, Georgia, North Carolina, New Jersey, Rhode Island) Number of participants: 626 Number analysed: 459 Number lost to falls follow‐up: 167 Sample: homebound older adults on waiting lists for Meals on Wheels programmes Study period: September 2013 to May 2014 Age (years): 76.3 (SD 9.7) Sex: not reported Inclusion criteria: nil Exclusion criteria: nil
Interventions	Participants in the daily delivery group received daily delivery of hot/chilled meals by Meals on Wheels staff or volunteers during weekdays. Participants in the frozen meal group received once‐weekly deliveries of 5 days of frozen meals by Meals on Wheels staff or volunteers. The control group were on a waiting list for the service.
Outcomes	Number of people who experienced 1 or more falls (risk of falling)
Adherence	Adherence measured by receipt of meals or no longer wishing to receive the meals. In the 'daily meals' group 10/214 declined, in the 'frozen meals' group 11/202 declined.
Notes	Funding: supported by Meals on Wheels America and grant from the AARP Foundation and the Agency for Healthcare Research and Quality
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Used a systematic non‐random method: alphabetised participant names assigned sequentially
Allocation concealment (selection bias)	Unclear risk	Based on sequentially numbered list; however, it was not clear if concealed from Meals on Wheels staff at sites
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Outcomes for all groups used same method, but data collected by trained Meals on Wheels staff, unclear if blinded to group allocation.
Incomplete outcome data (attrition bias) Fall outcomes	High risk	More than 20% loss to fall follow‐up (41%)
Selective reporting (reporting bias)	Unclear risk	Did not include data for calculating rate of falls
Method of ascertaining falls (recall bias) Fall outcomes	High risk	Retrospective reporting at 15‐week follow‐up survey

Open in a new tab

Wang 2019.

*Study characteristics*
Methods	Study design: RCT Number of study arms: 2  Length of follow‐up: 12 months
Participants	Setting: Houston, Texas, USA Number of participants: 44 Number analysed: 35 Number lost to falls follow‐up: 9 Sample: community dwelling from outpatient clinics or education centres for senior/older adults Study period: 2016 to 2019 Age (years): mean 74.7 (SD 6.4) Sex: 70.5% women Inclusion criteria: being ambulatory and aged 65 years or older, with self‐reported concern about falling or at risk for falling (confirmed by either 13 seconds or more in the Timed Up and Go Test or a fall in the past 6 months) Exclusion criteria: (i) wound on either the feet or the ankles; (ii) taking medication that might impact on gait and balance; (iii) major foot amputation or lower‐extremity fracture; (iv) unstable medical conditions affecting gait and balance; (v) participating in other interventional study; (vi) unable to stand without help, non‐ambulatory, or unable to walk 1.8 m without assistance; (vii) cognitive impairment; or (viii) unable or unwilling to participate in all procedures and follow‐up evaluations
Interventions	Intervention: fitted with walking shoes and bilateral custom‐made ankle‐foot orthoses Control: fitted with walking shoes and no ankle‐foot orthoses
Outcomes	Fall incidence rate Number of fallers
Adherence	Self‐report; hours per day wearing the shoes and orthoses
Notes	Funding source: Othotic Holdings Inc in part, and National Institute on Aging
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Eligible participants were allocated using a random number sequence generated by a computer software program.
Allocation concealment (selection bias)	Unclear risk	Method of allocation not described.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants and personnel not blind to allocated group, but impact of non‐blinding unclear.
Blinding of outcome assessment (detection bias) Fall outcomes	Unclear risk	Not reported if assessors were blinded
Incomplete outcome data (attrition bias) Fall outcomes	High risk	More than 20% loss to fall outcome (20.5%). There was some imbalance in fall data (3 in intervention group and 6 in control group), with reasons varying between groups.
Selective reporting (reporting bias)	Low risk	Fall outcomes reported as expected for study aim. Published protocol in trial registry.
Method of ascertaining falls (recall bias) Fall outcomes	High risk	Retrospective recall at 3, 6, and 12 months

Open in a new tab

dB: decibel GP: general practitioner RCT: randomised controlled trial SD: standard deviation

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Casteel 2020	Evaluation of a safety programme in the homes of older people; not a randomised trial, and information was collected on fall perceptions, not fall outcomes
Deems‐Dluhy 2021	Small study of a keen ankle foot orthosis, randomised cross‐over design: mean age was less than 60 years (−1 SD)
Ferrer 2014	RCT recruiting people aged 80 years or older including those with comorbidities and cognitive impairment. We excluded this study because the intervention had multiple components (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Greenberg 2020	RCT recruiting people 65 and over in hospital. We excluded this study because it was a multifactorial education programme (multifactorial interventions are evaluated in another Cochrane Review, Hopewell 2018).
Hill 2019	RCT recruiting older people at high risk of falls after hospital discharge. We excluded this study because the education intervention had multiple components (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Johansson 2018	RCT recruiting people aged 65 or older who had fallen in the past year or had a fear of falling. We excluded this study because the intervention had multiple components (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Lee 2013	RCT recruiting people who had fallen in the previous year or who were at risk of falling. We excluded this study because the intervention had multiple components (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Liu 2020	RCT of low‐income adults aged 65 years or over. We excluded this study because the aim was to reduce fall‐related self‐efficacy and life‐space use.
Luck 2013	RCT recruiting people aged 80 years or more with functional impairment. We excluded this study because the intervention had multiple components (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Moller 2014	RCT of older people at high risk of falls. We excluded this study because it was a multifactorial intervention (multifactorial interventions are evaluated in another Cochrane Review, Hopewell 2018).
Oddsson 2020	Study recruited people with a high risk of falls; not a randomised trial.
Tannenbaum 2019	A pragmatic cluster‐RCT of women aged 65 or over with urinary incontinence. We excluded this study because the education intervention had multiple components including exercise (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Tiefenbachova 2019	This was a non‐randomised, single‐arm, pre‐post intervention study of people aged 65 and above. We excluded this study because it was not a randomised trial, and no fall outcome data had been collected.
Wesson 2013	RCT of older people with dementia. We excluded this study because it was a multifactorial intervention with home safety and exercise (multicomponent interventions are evaluated in another Cochrane Review, Hopewell 2018).
Yamashita 2019	This RCT was conducted in older participants with foot problems. We excluded this study because the intervention descriptors were not applicable, and the focus was on fall risk factors (e.g. balance, strength).

Open in a new tab

RCT: randomised controlled trial SD: standard deviation

Characteristics of studies awaiting classification [ordered by study ID]

Stark 2016.

Methods	Blinded, randomised, sham‐controlled trial
Participants	92 community‐dwelling older adults
Interventions	Home modifications to improve task enablement and function
Outcomes	Process evaluation published: the treatment group improved daily activity performance over 12 months compared to the sham control group (F = 4.13; P = 0.024). The intervention elements and dose were delivered with greater than 90% accuracy. Participants reported a 91% adherence rate at 12 months.
Notes	Awaiting publication of final results, trial completed

Open in a new tab

Characteristics of ongoing studies [ordered by study ID]

Hatton 2019.

Study name	Textured shoe insoles to improve balance performance in adults with diabetic peripheral neuropathy: study protocol
Methods	Prospective, parallel, randomised control trial
Participants	70 adults
Interventions	Wearing textured insoles for 4 weeks alters balance performance in adults with diabetic peripheral neuropathy
Outcomes	Standing balance (foam/firm surface; eyes open/closed) and walking tasks will be completed barefoot, wearing standard shoes only, and 2 different insoles (smooth, textured). The primary outcome measure will be centre of pressure (CoP) velocity, with higher values indicating poorer balance. Secondary outcome measures include walking quality (gait velocity, base of support, stride length, and double‐limb support time), physical activity levels, foot sensation (light‐touch pressure, vibration) and proprioception (ankle joint position sense), and other balance parameters (CoP path length, anteroposterior and mediolateral excursion). Patient‐reported outcomes will be completed evaluating foot health, frequency of falls, and fear of falling.
Starting date	16 November 2017
Contact information	Dr Anna L Hatton; a.hatton1@uq.edu.au
Notes	Trial registered ClinicalTrials.gov (NCT01833182).

Open in a new tab

Differences between protocol and review

Types of outcomes

For falls requiring medical attention, study authors also reported number and rate. We analysed both types of data in order to capture the potential sensitivity in rate data which can be lost in the data when reported as number of events. While we collected all reports of adverse events, these were not routinely collected for most trials, and seldom collected for both intervention and control. These were summarised in text but not included in pooled analysis or summary of results tables.

Dealing with missing data

We attempted to contact study investigators regarding any unclear data or information on their trial; we only sought clarification of outcome data for number of falls or number of people who experienced falls. We used a cut‐off of 20% when exploring missing data.

Risk of bias assessment

We applied risk of bias assessments for the primary outcome (rate of falls). In addition, we reported blinding of outcome assessment (detection bias) separately for four groups of outcomes (falls, fractures, medical attention, hospital admission, and health‐related quality of life).

We have added an assessment of risk of bias specifically for trials using cluster randomisation. We assessed the risk of additional bias relating to recruitment, baseline imbalance, loss of clusters, incorrect analysis, comparability with individually randomised trials, and whether or not clustering was accounted for in the analysis, as described in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We updated selective outcome reporting to include retrospective and prospective reporting in trial registers.

Data synthesis

We removed the requirement for "not pooling of data where there is considerable heterogeneity (I² > 75%) that cannot be credibly explained", as this is now dealt with in the updated GRADE criteria for judging when to downgrade the certainty of the evidence. We did not include outcomes in the summary of findings tables where insufficient information precluded data pooling and there were limited clinical implications due to the diversity of interventions.

Sensitivity analysis

We conducted sensitivity analysis by removing trials with a high risk of bias in one or more of the selected four domains. We added an additional risk of bias domain to the planned sensitivity analysis, selection bias from inadequate generation of randomised sequence, and also decided that it was more relevant to our primary fall outcomes that we include recall (detection) bias ‐ method of ascertaining falls as the selected detection bias domain. Regarding the evidence for number of people experiencing a fracture, a fall requiring medical attention, or a fall requiring hospitalisation, we excluded trials with a high risk of detection bias (blinding of outcome assessment) rather than recall detection bias. For publication bias, where outcomes included fewer than 10 studies, we explored the possibility of small‐study effects by excluding studies with fewer than 20 participants in each arm.

GRADE assessment

We used the updated GRADE assessment criteria, which expresses judgements of the quality of the evidence in terms of ’certainty’ rather than ’quality’. We also provided additional and detailed criteria for the five GRADE considerations: risk of bias, inconsistency, indirectness, imprecision, and publication bias. For risk of bias judgements, we removed trials with a high risk of bias in one or more of the four specified domains.

Author changes

David Torgerson stepped down as an author, and Nicola Fairhall and Jinnat Ali joined the review team.

Contributions of authors

Lindy Clemson (LC) was involved in writing the protocol, study screening, data extraction, data analysis, led the writing of the review, and acted as guarantor of the review. Susan Stark (SS) was involved in writing the protocol, study screening, data extraction, data analysis, and contributed to writing the review. Alison C Pighills (AP) was involved in writing the protocol, study screening, data extraction, data analysis, and contributed to writing the review. Nicola J Fairhall (NF) contributed to data extraction and commented on drafts of the review. Sarah E Lamb (SL) contributed to writing the protocol and commented on drafts of the review. Catherine Sherrington (CS) was involved in writing the protocol, data analysis, and contributed to writing the review. Jinnat Ali (JA) contributed to study screening and commented on drafts of the review.

Sources of support

Internal sources

Librarian, University of Sydney, Other

Assistance with searches

External sources

NIHR, UK

National Institute for Health Research (NIHR) via Cochrane Infrastructure funding to the Cochrane Bone, Joint and Muscle Trauma Group

Declarations of interest

Lindy Clemson was an investigator in one trial of an environmental intervention to reduce falls which she did not review, and developed an assessment to identify home fall hazards that has been used in several trials. Susan Stark ran a trial of an environmental intervention to reduce falls and did not review her own trial. Alison C Pighills ran a trial of an environmental intervention to reduce falls and was an investigator on another trial; she did not review these trials. Nicola J Fairhall has no known conflicts of interest. Sarah E Lamb was an investigator on a trial of an environmental intervention to reduce falls and did not review this trial. She is on the Data and Safety Monitoring Committee for a pharmaceutical company, but this company is not involved in the development of environmental interventions for preventing falls. Catherine Sherrington has no known conflicts of interest. Jinnat Ali has no known conflicts of interest.

New

References

References to studies included in this review

Campbell 2005 {published data only}

Campbell AJ, Robertson MC, La Grow SJ, Kerse NM, Sanderson GF, Jacobs RJ, et al. Randomised controlled trial of prevention of falls in people aged > or =75 with severe visual impairment: the VIP trial. BMJ 2005;331(817):1-8. [DOI: 10.1136/bmj.38601.447731.55] [DOI] [PMC free article] [PubMed] [Google Scholar]
ISRCTN15342873. VIP falls prevention trial. Interventions to prevent falls and injury in elderly people with impaired vision. www.isrctn.com/ISRCTN15342873 (first received 2 March 2005). [DOI: 10.1186/ISRCTN15342873] [DOI]
Jacobs R, Campbell AJ, Robertson MC. Randomized controlled trial of falls prevention in people 75 years and older with severe visual impairment [abstract]. American Academy of Optometry Meeting, 2005 Dec 8-12; San Diego (CA) 2005. [CENTRAL: CN-00634854]
Kiata L, Kerse NM, Hughes WE, Hayman KJ, Robertson MC, La Grow SJ, et al. Agreement and compliance with advice on removing mats or rugs by older people with visual impairments. Journal of Visual Impairment & Blindness 2008;102(3):167-72. [Google Scholar]
La Grow SJ, Robertson MC, Campbell AJ, Clarke GA, Kerse NM. Reducing hazard related falls in people 75 years and older with significant visual impairment: how did a successful program work? Injury Prevention 2006;12(5):296-301. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

Chu 2017 {published data only}

Chu MM, Fong KN, Lit AC, Rainer TH, Cheng SW, Au FL, et al. An occupational therapy fall reduction home visit program for community-dwelling older adults in Hong Kong after an emergency department visit for a fall. Journal of the American Geriatrics Society 2017;37(1):5-13. [DOI] [PubMed] [Google Scholar]

Cockayne 2021a {published data only}

Cockayne S, Pighills A, Adamson J, Fairhurst C, Crossland S, Drummond A, et al. Home environmental assessments and modification delivered by occupational therapists to reduce falls in people aged 65 years and over: the OTIS RCT. Health Technology Assessment 2021;25(46):1-117. [DOI: 10.3310/hta2546] [DOI] [PMC free article] [PubMed] [Google Scholar]
Cockayne S, Pighills A, Adamson J, Fairhurst C, Drummond A, Hewitt C, et al. Can occupational therapist-led home environmental assessment prevent falls in older people? A modified cohort randomised controlled trial protocol. BMJ Open 2018;8:e022488. [DOI: 10.1136/ bmjopen-2018-022488] [DOI] [PMC free article] [PubMed] [Google Scholar]
Cockayne S, Pighills A, Fairhurst C, Adamson J, Crossland S, Drummond A, et al. Home hazard assessment and environmental modification to prevent falls in older people: the OTIS trial [version 1; peer review: 1 approved, 1 approved with reservations]. F1000Research 2021;10(500):1-18. [DOI: 10.12688/f1000research.52313.1] [DOI] [Google Scholar]
ISRCTN22202133. Occupational therapist home assessment and modification for prevention of falls. www.isrctn.com/ISRCTN22202133 (first received 20 June 2016). [DOI: 10.1186/ISRCTN22202133] [DOI]

Cumming 1999 {published data only}

Cumming RG, Thomas M, Szonyi G, Frampton G, Salkeld G, Clemson L. Adherence to occupational therapist recommendations for home modifications for falls prevention. American Journal of Occupational Therapy 2001;55(6):641-8. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]
Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: a randomized trial of falls prevention. Journal of the American Geriatrics Society 1999;47(12):1397-402. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]
Salkeld G, Cumming RG, O'Neill E, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Cumming 2007 {published data only}

Cumming RG, Ivers R, Clemson L, Cullen J, Hayes MF, Tanzer M, et al. Improving vision to prevent falls in frail older people: a randomized trial. Journal of the American Geriatrics Society 2007;55(2):175-81. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Day 2002 {published and unpublished data}

Day L, Fildes B, Gordon I, Fitzharris M, Flamer H, Lord S. Randomised factorial trial of falls prevention among older people living in their own homes. BMJ 2002;325(7356):128-31. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
Fitzharris MP, Day L, Lord SR, Gordon I, Fildes B. The Whitehorse NoFalls trial: effects on fall rates and injurious fall rates. Age & Ageing 2010;39(6):728-33. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Haran 2010 {published data only}

Haran MJ, Cameron ID, Ivers RQ, Simpson JM, Lee BB, Tanzer M, et al. Effect on falls of providing single lens distance vision glasses to multifocal glasses wearers: VISIBLE randomised controlled trial. BMJ 2010;340:c2265. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
Haran MJ, Lord SR, Cameron ID, Ivers RQ, Simpson JM, Lee BB, et al. Preventing falls in older multifocal glasses wearers by providing single-lens distance glasses: the protocol for the VISIBLE randomised controlled trial. BMC Geriatrics 2009;9:10. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
NCT00350389. VISIBLE study (Visual Intervention Strategy Incorporating Bifocal & Long-Distance Eyewear) (VISIBLE) [Randomised controlled falls prevention trial of plain distance glasses in elderly multifocal wearers]. clinicaltrials.gov/show/NCT00350389 (first received 10 July 2006).

Kamei 2015 {published and unpublished data}

Kamei T, Fumiko K, Yuko Y, Yukako I, Rumi K, Tomoko S, et al. Effectiveness of a home hazard modification program for reducing falls in urban community-dwelling older adults: a randomized controlled trial. Japan Journal of Nursing Science 2015;12(3):184-97. [DOI] [PubMed] [Google Scholar]

Keall 2015 {published data only}

Keall MD, Howden-Chapman P, Baker MG, Kamalesh V, Cunningham M, Cunningham C, et al. Formulating a programme of repairs to structural home injury hazards in New Zealand. Accident Analysis & Prevention 2013;57:124-30. [DOI: 10.1016/j.aap.2013.04.011] [DOI] [PubMed] [Google Scholar]
Keall MD, Pierse N, Howden-Chapman P, Cunningham C, Cunningham M, Guria J, et al. Home modifications to reduce injuries from falls in the home injury prevention intervention (HIPI) study: a cluster-randomised controlled trial. Lancet 2015;385(9964):231-8. [DOI: 10.1016/S0140-6736(14)61006-0] [DOI] [PubMed] [Google Scholar]

Lannin 2007 {published data only}

Lannin NA, Clemson L, McCluskey A, Lin CW, Cameron ID, Barras S. Feasibility and results of a randomised pilot-study of pre-discharge occupational therapy home visits. BMC Health Services Research 2007;7:42. [DOI: 10.1186/1472-6963-7-42] [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

Lin 2007 {published and unpublished data}

Lin MR, Wolf SL, Hwang HF, Gong SY, Chen CY. A randomized, controlled trial of fall prevention programs and quality of life in older fallers. Journal of the American Geriatrics Society 2007;55(4):499-506. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Lockwood 2019 {published data only}

Lockwood KJ, Harding KE, Boyd JN, Taylor NF. Predischarge home visits after hip fracture: a randomized controlled trial. Clinical Rehabilitation 2019;33(4):681-92. [DOI: 10.1177/0269215518823256] [DOI] [PubMed] [Google Scholar]

McKiernan 2005 {published data only}

McKiernan FE. A simple gait-stabilizing device reduces outdoor falls and nonserious injurious falls in fall-prone older people during the winter. Journal of the American Geriatrics Society 2005;53(6):943-7. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Nikolaus 2003 {published data only}

Nikolaus T, Bach M. Preventing falls in community-dwelling frail older people using a home intervention team (HIT): results from the randomized falls-HIT trial. Journal of the American Geriatrics Society 2003;51(3):300-5. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]
Nikolaus T, Specht-Leible N, Bach M, Wittmann-Jennewein C, Oster P, Schlierf G. Effectiveness of hospital-based geriatric evaluation and management and home intervention team (GEM-HIT). Rationale and design of a 5-year randomized trial. Zeitschrift fur Gerontologie und Geriatrie 1995;28(1):47-53. [MEDLINE: ] [PubMed] [Google Scholar]

Pardessus 2002 {published data only}

Pardessus V, Puisieux F, Di P, Gaudefroy C, Thevenon A, Dewailly P. Benefits of home visits for falls and autonomy in the elderly: a randomized trial study. American Journal of Physical Medicine and Rehabilitation 2002;81(4):247-52. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Perry 2008 {published data only}

Perry SD, Radtke A, McIlroy WE, Fernie GR, Maki BE. Efficacy and effectiveness of a balance-enhancing insole. Journals of Gerontology Series A: Biological Sciences & Medical Sciences 2008;63(6):595-602. [DOI] [PubMed] [Google Scholar]

Pighills 2011 {published and unpublished data}07575807

ISRCTN07575807. Reducing home hazards to prevent falls in older people [Environmental assessment and modification to prevent falls in older people: a randomised controlled single centre trial]. controlled-trials.com/ISRCTN07575807 (first received 10 May 2009).
Pighills A, Torgerson DJ, Sheldon T. Publicity does not increase recruitment to falls prevention trials: the results of two quasi-randomized trials. Journal of Clinical Epidemiology 2009;62(12):1332-5. [DOI] [PubMed] [Google Scholar]
Pighills AC, Torgerson DJ, Sheldon TA, Drummond AE, Bland JM. Environmental assessment and modification to prevent falls in older people. Journal of the American Geriatrics Society 2011;59(1):26-33. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Stark 2021 {published data only}

Stark S, Keglovits M, Somerville E, Hu Y-L, Barker A, Sykora S, et al. Home hazard removal to reduce falls among community-dwelling older adults. A randomized clinical trial. JAMA Network Open 2021;4(8):e2122044. [DOI: 10.1001/jamanetworkopen.2021.22044] [DOI] [PMC free article] [PubMed] [Google Scholar]
Stark S, Somerville E, Keglovits M, Conte J, Li M, Hu Y-L, et al. Protocol for the home hazards removal program (HARP) study: a pragmatic, randomized clinical trial and implementation study. BMC Geriatrics 2017;17(90):1-9. [DOI: 10.1186/s12877-017-0478-4] [DOI] [PMC free article] [PubMed] [Google Scholar]

Stevens 2001 {published data only}

Stevens M, Holman CD, Bennett N, Klerk N. Preventing falls in older people: outcome evaluation of a randomized controlled trial. Journal of the American Geriatrics Society 2001;49(11):1448-55. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]
Stevens M, Holman CD, Bennett N. Preventing falls in older people: impact of an intervention to reduce environmental hazards in the home. Journal of the American Geriatrics Society 2001;49(11):1442-7. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

Tchalla 2013 {published data only}

Tchalla AE, Lachal F, Cardinaud N, Saulnier I, Rialle V, Preux PM, et al. Preventing and managing indoor falls with home-based technologies in mild and moderate Alzheimer's disease patients: pilot study in a community dwelling. Dementia & Geriatric Cognitive Disorders 2013;36(3/4):251-61. [DOI] [PubMed] [Google Scholar]

Thomas 2018 {published data only}

Thomas KS, Akobundu U, Dosa D. More than a meal? A randomized control trial comparing the effects of home-delivered meals programs on participants' feelings of loneliness. Journals of Gerontology: Social Sciences 2016;71(6):1049-58. [DOI: 10.1093/geronb/gbv111] [DOI] [PubMed] [Google Scholar]
Thomas KS, Rarikh RB, Zullo AR, Dosa D. Home-delivered meals and risk of self-reported falls: results from a randomized trial. Journal of Applied Gerontology 2018;37(1):41-57. [DOI] [PMC free article] [PubMed] [Google Scholar]

Wang 2019 {published and unpublished data}

NCT02819011. OHI - randomized control trial to evaluate efficacy, acceptability, and perception of benefit of an innovative custom AFO. clinicaltrials.gov/ct2/show/NCT02819011 (first received 30 June 2016).
Wang C, Goel R, Zhang Q, Lepow B, Najafi B. Daily use of bilateral custom-made ankle‐foot orthoses for fall prevention in older adults: a randomized controlled trial. Journal of the American Geriatrics Society 2019;67(6):1656-61. [DOI] [PMC free article] [PubMed] [Google Scholar]

References to studies excluded from this review

Casteel 2020 {published data only}

Casteel C, Bruening R, Carson M, Berard-Reed K, Ashida S. Evaluation of a falls and fire safety program for community‑dwelling older adults. Journal of Community Health 2021;45:717-27. [DOI: 10.1007/s10900-019-00786-8] [DOI] [PubMed] [Google Scholar]

Deems‐Dluhy 2021 {published data only}

Deems-Dluhy S, Hoppe-Ludwig S, Mummidisetty CK, Semik P, Heinemann AW, Jayaraman A. Microprocessor controlled knee ankle foot orthosis (KAFO) vs stance control vs locked KAFO: a randomized controlled trial. Archives of Physical Medicine and Rehabilitation 2021;102:233-44. [DOI: 10.1016/j.apmr.2020.08.013] [DOI] [PubMed] [Google Scholar]

Ferrer 2014 {published data only}

Ferrer A, Formiga F, Sanz H, Vries OJ, Badia T, Pujol R, et al. Multifactorial assessment and targeted intervention to reduce falls among the oldest-old: a randomized controlled trial. Clinical Interventions in Aging 2014;9:PMC3940644. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

Greenberg 2020 {published data only}

Greenberg MR, Goodheart V, Jacoby JL, Barraco RD, Crowley LM, Day R, et al. Emergency Department Stopping Elderly Accidents, Deaths and Injuries (ED STEADI) program. Journal of Emergency Medicine 2020;59(1):1-11. [DOI: 10.1016/j.jemermed.2020.04.019] [DOI] [PubMed] [Google Scholar]

Hill 2019 {published data only}

Hill A, McPhail SM, Haines TP, Morris ME, Etherton-Beer C, Shorr R. Reducing falls after hospital discharge using patient education. Journal of the American Geriatrics Society 2019;67(S1):S50. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hill AM, McPhail SM, Haines TP, Morris ME, Etherton-Beer C, Shorr R, et al. Falls after hospital discharge: a randomized clinical trial of individualized multimodal falls prevention education. Journals of Gerontology Series A: Biological Sciences & Medical Sciences 2019;74(9):1511-7. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

Johansson 2018 {published data only}

Johansson E, Jonsson H, Dahlberg R, Patomella AH. The efficacy of a multifactorial falls-prevention programme, implemented in primary health care. British Journal of Occupational Therapy 2018;81(8):474-81. [DOI: 10.1177/0308022618756303] [DOI] [Google Scholar]

Lee 2013 {published data only}

Lee H-C, Chang K-C, Tsauo J-Y, Huang Y-C, Lin SI. Effects of a multifactorial fall prevention program on fall incidence and physical function in community-dwelling older adults with risk of falls. Archives Physical Medicine & Rehabilitation 2013;94(4):606-15. [PMID: ] [DOI] [PubMed] [Google Scholar]

Liu 2020 {published data only}

Liu M, Xue Q-L, Gitlin L, Wolff J, Guralnik J, Leff B, et al. Disability prevention program improves life-space and falls efficacy: a randomized controlled trial. American Geriatrics Society 2021;69:85-90. [DOI: 10.1111/jgs.16808] [DOI] [PMC free article] [PubMed] [Google Scholar]

Luck 2013 {published data only}

Luck T, Motzek T, Luppa M, Matschinger H, Fleischer S, Sesselmann Y, et al. Effectiveness of preventive home visits in reducing the risk of falls in old age: a randomized controlled trial. Clinical Interventions in Aging 2013;8:697-702. [DOI: 10.2147/CIA.S43284] [DOI] [PMC free article] [PubMed] [Google Scholar]

Moller 2014 {published data only}

Möller UO, Kristensson J, Midlöv P, Ekdahl C, Jakobsson U. Effects of a one-year home-based case management intervention on falls in older people: a randomized controlled trial. Journal of Aging and Physical Activity 2014;22(4):457-64. [DOI: 10.1123/japa.2013-0101] [DOI] [PubMed] [Google Scholar]

Oddsson 2020 {published data only}

Oddsson L, Bisson T, Cohen H, Jacobs L, Khoshnoodi M, Kung D, et al. The effects of a wearable sensory prosthesis on gait and balance function after 10 weeks of use in persons with peripheral neuropathy and high fall risk – the walk2Wellness trial. Aging Neuroscience 2020;12:592751. [DOI: 10.3389/fnagi.2020.592751] [DOI] [PMC free article] [PubMed] [Google Scholar]

Tannenbaum 2019 {published data only}

Tannenbaum C, Fritel X, Halme A, den Heuvel E, Jutai J, Wagg A. Long-term effect of community-based continence promotion on urinary symptoms, falls and healthy active life expectancy among older women: cluster randomised trial. Age and Ageing 2019;48(4):526-32. [DOI: 10.1093/ageing/afz038] [DOI] [PMC free article] [PubMed] [Google Scholar]

Tiefenbachova 2019 {published data only}

Tiefenbachová P, Zeleníková R. The effect of educational intervention by nurses on home environmental risk factors for falls. Central European Journal of Nursing and Midwifery 2019;10(2):1019-25. [DOI: 10.15452/CEJNM.2019.10.0009] [DOI] [Google Scholar]

Wesson 2013 {published data only}

Wesson J, Clemson L, Brodaty H, Lord S, Taylor M, Gitlin L, et al. A feasibility study and pilot randomised trial of a tailored prevention program to reduce falls in older people with mild dementia. BMC Geriatrics 2013;13:89. [DOI: 10.1186/1471-2318-13-89] [DOI] [PMC free article] [PubMed] [Google Scholar]

Yamashita 2019 {published data only}

Yamashita T, Yamashita K, Rinoie C, Takase Y, Sato M, Yamada K, et al. Improvements in lower-limb muscle strength and foot pressure distribution with foot care in frail elderly adults: a randomized controlled trial from Japan. BMC Geriatrics 2019;18:83. [DOI: 10.1186/s12877-019-1097-z] [DOI] [PMC free article] [PubMed] [Google Scholar]

References to studies awaiting assessment

Stark 2016 {published data only}

Stark S, Keglovits M, Somerville E. A randomized controlled feasibility trial of tailored home modifications to improve activities of daily living. American Journal of Occupational Therapy 2016;70(4 Suppl 1):7011520290p1. [DOI: 10.5014/ajot.2018.021774] [DOI] [Google Scholar]

References to ongoing studies

Hatton 2019 {published data only}

ACTRN12617000543381. Do different shoe insole surfaces affect balance and walking in adults with diabetic peripheral neuropathy? A randomised controlled trial. www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=372682&isReview=true (first received 11 April 2017).
Hatton AL, Gane E, Maharaj J, Burns J, Paton J, Kerr G, et al. Textured shoe insoles to improve balance performance in adults with diabetic peripheral neuropathy: study protocol for a randomised controlled trial. BMJ Open 2019;9:e026240. [DOI: 10.1136/bmjopen-2018-026240] [DOI] [PMC free article] [PubMed] [Google Scholar]

Additional references

Aliberti 2019

Aliberti MJR, Covinsky KE. Home modifications to reduce disability in older adults with functional disability. JAMA Internal Medicine 2019;179(2):211-2. [DOI: 10.1001/jamainternmed.2018.6414] [DOI] [PubMed] [Google Scholar]

Aplin 2015

Aplin T, Jonge D, Gustafsson L. Understanding home modifications impact on clients and their family's experience of home: a qualitative study. Australian Occupational Therapy Journal 2015;62(2):123-31. [DOI: 10.1111/1440-1630.12156] [DOI] [PubMed] [Google Scholar]

Campbell 1990

Campbell AJ, Borrie MJ, Spears GF, Jackson SL, Brown JS, Fitzgerald JL. Circumstances and consequences of falls experienced by a community population 70 years and over during a prospective study. Age and Ageing 1990;19:136-41. [DOI] [PubMed] [Google Scholar]

Canning 2015

Canning CG, Allen NE, Bloem BR, Keus SHJ, Munneke M, Nieuwboer A, et al. Interventions for preventing falls in Parkinson's disease. Cochrane Database of Systematic Reviews 2015, Issue 3. Art. No: CD011574. [DOI: 10.1002/14651858.CD011574] [DOI] [PMC free article] [PubMed] [Google Scholar]

Carnemolla 2019

Carnemolla P, Bridge C. Housing design and community care: how home modifications reduce care needs of older people and people with disability. International Journal of Environmental Research and Public Health 2019;16(11):1951. [DOI: 10.3390/ijerph16111951] [DOI] [PMC free article] [PubMed] [Google Scholar]

Clemson 2003

Clemson L, Manor D, Fitzgerald MH. Behavioral factors contributing to older adults falling in public places. OTJR: Occupation, Participation and Health 2003;23(3):107-17. [Google Scholar]

Clemson 2008

Clemson L, Mackenzie L, Ballinger C, Close JCT, Cumming RG. Environmental interventions to prevent falls in community-dwelling older people: a meta-analysis of randomized trials. Journal of Aging and Health 2008;20(8):954-71. [DOI] [PubMed] [Google Scholar]

Clemson 2021

Clemson L, Pighills A. Environmental interventions to prevent falls at home and in the community. In: Lord S, Sherrington C, Naganathan V, editors(s). Falls in Older People: Risk factors, Strategies for Prevention and Implications for Practice. Cambridge: Cambridge University Press, 2021:360-77. [DOI: 10.1017/9781108594455.024] [DOI] [Google Scholar]

Cochrane Norway 2017

Cochrane Norway. How to write a plain language summary of a Cochrane intervention review. www.ahajournals.org/doi/10.1161/circ.114.suppl_18.II_525-c (accessed prior to 11 January 2023).

Cockayne 2021b

Cockayne S, Pighills A, Adamson J, Fairhurst C, Crossland S, Drummond A. Home environmental assessments and modification delivered by occupational therapists to reduce falls in people aged 65 years and over: the OTIS RCT. Health Technology Assessment 2021;25(46):1-118. [DOI: 10.3310/hta25460] [ISSN: 1366-5278] [DOI] [PMC free article] [PubMed]

Deandrea 2010

Deandrea S, Lucenteforte E, Bravi F, Foschi R, La Vecchia C, Negri E. Risk factors for falls in community-dwelling older people: a systematic review and meta-analysis. Epidemiology 2010;21(5):658-68. [DOI] [PubMed] [Google Scholar]

Díaz‐Arribas 2017

Díaz-Arribas MJ, Fernández-Serrano M, Royuela A, Kovacs F, Gallego-Izquierdo T, Ramos-Sánchez M, et al. Minimal clinically important difference in quality of life for patients with low back pain. Spine 2017;42(24):1908-16. [DOI: 10.1097/BRS.0000000000002298] [DOI] [PubMed] [Google Scholar]

EuroQol Research Foundation

EuroQol Research Foundation. EQ-5D. euroqol.org/ (accessed 2 February 2019).

Fange 2005

Fange A, Iwarsson S. Changes in accessibility and usability in housing: an exploration of the housing adaptation process. Occupational Therapy International 2005;12(1):44-59. [DOI] [PubMed] [Google Scholar]

Ferrari 2012

Ferrari M, Harrison B, Lewis D. The risk factors for impulsivity-related falls among hospitalized older adults. Rehabilitation Nursing 2012;37(3):145-50. [DOI: 10.1002/RNJ.00046] [DOI] [PubMed] [Google Scholar]

Gillespie 2012

Gillespie LD, Robertson MC, Gillespie WJ, Sherrington C, Gates S, Clemson LM, et al. Interventions for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews 2012, Issue 9. Art. No: CD007146. [DOI: 10.1002/14651858.CD007146.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

Gitlin 2003

Gitlin LP. Conducting research on home environments: lessons learned and new directions. Gerontologist 2003;45(5):628-37. [DOI] [PubMed] [Google Scholar]

Gitlin 2009

Gitlin L. Environmental adaptations for older adults and their families in the home and community. In: Soderback I, editors(s). International Handbook of Occupational Therapy Interventions. London: Springer, 2009. [Google Scholar]

Guyatt 2011

Guyatt GH, Oxman AD, Montori V, Vist G, Kunz R, Brozek J, et al. GRADE guidelines: 5. Rating the quality of evidence - publication bias. Journal of Clinical Epidemiology 2011;64(12):1277–82. [DOI: 10.1016/j.jclinepi.2011.01.011] [DOI] [PubMed] [Google Scholar]

Hannan 2010

Hannan MT, Gagnon MM, Aneja J, Jones RN, Cupples LA, Lipsitz LA, et al. Optimizing the tracking of falls in studies of older participants: comparison of quarterly telephone recall with monthly falls calendars in the MOBILIZE Boston Study. American Journal of Epidemiology 2010;171(9):1031-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

He 2016

He W, Goodkind D, Kowal P, US Census Bureau. An Aging World: 2015. Washington, DC: International Population Reports. US Government Publishing Office, 2016. [Google Scholar]

Herdman 2011

Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Quality of Life Research 2011;20(10):1727-36. [DOI] [PMC free article] [PubMed] [Google Scholar]

Higgins 2011

Higgins JPT, Green S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from training.cochrane.org/handbook/archive/v5.1/.

Hill 2018

Hill KD, Suttanon P, Lin SI, Tsang WWN, Ashari A, Hamid TAA, et al. What works in falls prevention in Asia: a systematic review and meta-analysis of randomized controlled trials. BMC Geriatrics 2018;18(1):3. [DOI: 10.1186/s12877-017-0683-1] [DOI] [PMC free article] [PubMed] [Google Scholar]

Hopewell 2018

Hopewell S, Adedire O, Copsey BJ, Boniface GJ, Sherrington C, Clemson L, et al. Mutlifactorial and multiple component interventions for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews 2018, Issue 7. Art. No: CD012221. [DOI: 10.1002/14651858.CD012221.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]

Horowitz 2013

Horowitz BP, Nochajski SM, Schweitzer JA. Occupational therapy community practice and home assessments: use of the home safety self-assessment tool (HSSAT) to support aging in place. Occupational Therapy in Health Care 2013;27(3):216-27. [DOI] [PubMed] [Google Scholar]

IBM Corp 2019 [Computer program]

IBM SPSS Statistics for Windows. Version 24.0. Armonk, NY: IBM Corp, 2019.

Iliffe 2015

Iliffe S, Kendrick D, Morris R, Griffin M, Haworth D, Carpenter H, et al. Promoting physical activity in older people in general practice: ProAct65. British Journal of General Practice 2015;65:e731–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

Iwarsson 2009

Iwarsson S, Horstmann V, Carlsson G, Oswald F, Wahl HW. Person-environment fit predicts falls in older adults better than the consideration of environmental hazards only. Clinical Rehabilitation 2009;23(6):558-67. [DOI] [PubMed] [Google Scholar]

Karlsson 2013

Karlsson MK, Vonschewelov T, Karlsson C, Cöster M, Rosengen BE. Prevention of falls in the elderly: a review. Scandinavian Journal of Public Health 2013;41(5):442-54. [DOI: 10.1177/1403494813483215] [DOI] [PubMed] [Google Scholar]

Keglovits 2020

Keglovits M, Clemson L, Hu YL, Nguyen A, Neff AJ, Mandelbaum C, et al. A scoping review of fall hazards in the homes of older adults and development of a framework for assessment and intervention. Australian Occupational Therapy Journal 2020;67(5):470-8. [DOI: 10.1111/1440-1630.12682] [DOI] [PubMed] [Google Scholar]

Kellog International Work Group 1987

Kellog International Work Group. The prevention of falls in later life. A report of the Kellogg International Work Group on the Prevention of Falls by the Elderly. Danish Medical Bulletin 1987;34(Suppl 4):1-24. [PubMed] [Google Scholar]

Lamb 2005

Lamb SE, Jørstad-Stein EC, Hauer K, Becker C, Prevention of Falls Network Europe and Outcomes Census Group. Development of a common outcome data set for fall injury prevention trials: the Prevention of Falls Network Europe consensus. Journal of the American Geriatrics Society 2005;53(9):1618-22. [DOI] [PubMed] [Google Scholar]

Lamb 2007

Lamb SE, Hauer K, Becker C. Manual for the fall prevention classification system. www.profane.eu.org/documents/Falls_Taxonomy.pdf (accessed prior to 18 December 2018).

Law 1997

Law M, Cooper BA, Strong S, Stewart D, Rigby PJ, Letts LT. The person environment occupation model: a transactive approach to occupational therapy. Canadian Journal of Occupational Therapy 1997;63:9-23. [DOI] [PubMed] [Google Scholar]

Leavy 2015

Leavy B, Byberg L, Michaëlsson K, Melhus H, Åberg AC. The fall descriptions and health characteristics of older adults with hip fracture: a mixed methods study. BMC Geriatrics 2015;15(1):40. [DOI: 10.1186/s12877-015-0036-x] [DOI] [PMC free article] [PubMed] [Google Scholar]

Lee 2021

Lee S. Falls associated with indoor and outdoor environmental hazards among community-dwelling older adults between men and women. BMC Geriatrics 2021;21(1):547. [DOI: 10.1186/s12877-021-02499-x] [DOI] [PMC free article] [PubMed] [Google Scholar]

Lefebvre 2019

Lefebvre C, Glanville J, Briscoe S, Littlewood A, Marshall C, Metzendorf MI, et al. Chapter 4: Searching for and selecting studies. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions version 6.0 (updated July 2019). Cochrane, 2019. Available from training.cochrane.org/handbook/archive/v6.

Lord 2006

Lord SR, Menz HB, Sherrington C. Home environment risk factors for falls in older people and the efficacy of home modifications. Age and Ageing 2006;35(Suppl 2):ii55-9. [DOI] [PubMed] [Google Scholar]

Lusardi 2017

Lusardi MM, Fritz S, Middleton A, Allison L, Wingood M, Phillips E, et al. Determining risk of falls in community dwelling older adults: a systematic review and meta-analysis using posttest probability. Journal of Geriatric Physical Therapy 2017;40(1):36. [DOI] [PMC free article] [PubMed] [Google Scholar]

Makino 2018

Makino K, Makizako H, Doi T, Tsutsumimoto K, Hotta R, Nakakubo S, et al. Impact of fear of falling and fall history on disability incidence among older adults: prospective cohort study. International Journal of Geriatric Psychiatry 2018;33(4):658-62. [DOI] [PubMed] [Google Scholar]

Microsoft Excel [Computer program]

Microsoft Excel. Microsoft Corporation, 2018.

Moreland 2020

Moreland B, Kakara R, Henry A. Trends in nonfatal falls and fall-related injuries among adults aged ≥65 years - United States, 2012-2018. Morbidity and Mortality Weekly Report 2020;69(27):875–81. [DOI: 10.15585/mmwr.mm6927a5] [DOI] [PMC free article] [PubMed]

Nahemow 2000

Nahemow L. The ecological theory of aging: Powell Lawton's legacy. In: Rubinstein RL, Moss M, Kieban M, Lawton MP, editors(s). The Many Dimensions of Aging. New York: Springer, 2000:22-40. [Google Scholar]

Peterson 2008

Peterson EW, Clemson L. Understanding the role of occupational therapy in fall prevention for community dwelling older adults. OT Practice 2008;13(3):CEA0208. [Google Scholar]

Pighills 2016

Pighills A, Ballinger C, Pickering R, Chan S. A critical review of the effectiveness of environmental assessment and modification in the prevention of falls amongst community dwelling older people. British Journal of Occupational Therapy 2016;79(3):133–43. [Google Scholar]

Pighills 2019

Pighills A, Drummond A, Crossland S, Torgerson DJ. What type of environmental assessment and modification prevents falls in community dwelling older people? BMJ 2019;364:1880. [DOI: 10.1136/bmj.l880] [DOI] [PubMed] [Google Scholar]

Pighills 2021

Pighills A, Clemson L. Environmental risk factors for falls. In: Lord S, Sherrington C, Naganathan V, editors(s). Falls in Older People: Risk Factors, Strategies for Prevention and Implications for Practice. Cambridge: Cambridge University Press, 2021:202-10. [DOI: 10.1017/9781108594455.013] [DOI] [Google Scholar]

Prabhakaran 2020

Prabhakaran K, Gogna S, Pee S, Samson DJ, Con J, Latifi R. Falling again? Falls in geriatric adults - risk factors and outcomes associated with recidivism. Journal of Surgical Research 2020;247:66-76. [DOI: 10.1016/j.jss.2019.10.041] [DOI] [PubMed] [Google Scholar]

Review Manager 2020 [Computer program]

Review Manager 5 (RevMan 5). Version 5.4. Copenhagen: The Cochrane Collaboration, 2020.

Robertson 2005

Robertson MC, Campbell AJ, Herbison P. Statistical analysis of efficacy in falls prevention trials. Journals of Gerontology Series A: Biological Sciences & Medical Sciences 2005;60(4):530-4. [DOI] [PubMed]

Rubenstein 2006

Rubenstein L. Falls in older people: epidemiology, risk factors and strategies for prevention. Age and Ageing 2006;35-S2:37-41. [DOI] [PubMed] [Google Scholar]

Salkeld 2000

Salkeld G, Cumming RG, O'Neill R, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. [DOI] [PubMed] [Google Scholar]

Sanford 2004

Sanford JA, Jones M, Daviou P, Grogg K, Butterfield T. Using telerehabilitation to identify home modification needs. Assistive Technology 2004;16(1):43-53. [DOI] [PubMed] [Google Scholar]

Santesso 2014

Santesso N, Carrasco‐Labra A, Brignardello‐Petersen R. Hip protectors for preventing hip fractures in older people. Cochrane Database of Systematic Reviews 2014, Issue 3. Art. No: CD001255. [DOI: 10.1002/14651858.CD001255.pub5] [DOI] [PMC free article] [PubMed] [Google Scholar]

Schünemann 2017

Schünemann HJ, Oxman AD, Higgins JPT, Vist GE, Glasziou P, Guyatt GH on behalf of the Cochrane GRADEing Methods Group (formerly Applicability and Recommendations Methods Group) and the Cochrane Statistical Methods Group. Chapter 11: Completing ‘Summary of findings’ tables and grading the confidence in or quality of the evidence. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS, editor(s). Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017). Cochrane, 2017. Available from training.cochrane.org/handbook/archive/v5.2.

Schünemann 2022

Schünemann HJ, Vist GE, Higgins JPT, Santesso N, Deeks JJ, Glasziou P, et al. Chapter 15: Interpreting results and drawing conclusions. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from training.cochrane.org/handbook.

Shi 2014

Shi J, Zhou BY, Tao YK, Yu PL, Zhang CF, Qin ZH, et al. Incidence and associated factors for single and recurrent falls among the elderly in an urban community of Beijing. Biomedical & Environmental Sciences 2014;27(12):939-49. [DOI] [PubMed] [Google Scholar]

Smeeth 2002

Smeeth L, Ng ES. Intraclass correlation coefficients for cluster randomized trials in primary care: data from the MRC Trial of the Assessment and Management of Older People in the Community. Controlled Clinical Trials 2002;23(4):409-21. [DOI] [PubMed] [Google Scholar]

Stark 2009

Stark S, Landsbaum A, Palmer JL, Somerville EK, Morris JC, Harvey A, et al. Client-centred home modifications improve daily activity performance of older adults . Canadian Journal of Occupational Therapy 2009;76(1 Suppl):135-45. [DOI: 10.1177/000841740907600s09] [DOI] [PMC free article] [PubMed] [Google Scholar]

Stevens 2014

Stevens JA, Mahoney JE, Ehrenreich H. Circumstances and outcomes of falls among high risk community-dwelling older adults. Injury Epidemiology 2014;1(5):1-9. [DOI: 10.1186/2197-1714-1-5] [DOI] [PMC free article] [PubMed] [Google Scholar]

Verheyden 2013

Verheyden GSAF, Weerdesteyn V, Pickering RM, Kunkel D, Lennon S, Geurts ACH, et al. Interventions for preventing falls in people after stroke. Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No: CD008728. [DOI: 10.1002/14651858.CD008728.pub2] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

Verma 2016

Verma SK, Willetts JL, Corns HL, Marucci-Wellman HR, Lombardi DA, Courtney TK. Falls and fall-related injuries among community-dwelling adults in the United States. PLOS ONE 2016;11(3):e0150939. [DOI: 10.1371/journal.pone.0150939] [DOI] [PMC free article] [PubMed] [Google Scholar]

Walters 2005

Walters SJ, Brazier JE. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Quality of Life Research 2005;14(6):1523-32. [PMID: 10.1007/s11136-004-7713-0] [DOI] [PubMed] [Google Scholar]

Wellecke 2022

Wellecke C, D'Cruz K, Winkler D, Douglas J, Goodwin J, Davis E, et al. Accessible design features and home modifications to improve physical housing accessibility: a mixed-methods survey of occupational therapists. Disability and Health Journal 2022;101281:1-8. [DOI: 10.1016/j.dhjo.2022.101281] [DOI] [PubMed] [Google Scholar]

WHO 2004

World Health Organization. Assistive devices and technologies. www.who.int/disabilities/technology/en/ (accessed 20 November 2017).

Williams 2015

Williams JS, Kowal P, Hestekin H, O'Driscoll T, Peltzer K, Yawson A, et al. Prevalence, risk factors and disability associated with fall-related injury in older adults in low- and middle-income countries: results from the WHO Study on global AGEing and adult health (SAGE). BMC Medicine 2015;13:147. [DOI: 10.1186/s12916-015-0390-8] [DOI] [PMC free article] [PubMed] [Google Scholar]

Zecevic 2006

Zecevic AA, Salmoni AW, Speechley M, Vandervoort AA. Defining a fall and reasons for falling: comparisons among the views of seniors, health care providers, and the research literature. Gerontologist 2006;46(3):367-76. [DOI] [PubMed] [Google Scholar]

References to other published versions of this review

Clemson 2019

Clemson L, Stark S, Pighills A, Torgerson DJ, Sherrington C, Lamb S. Environmental interventions for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews 2019, Issue 2. Art. No: CD013258. [DOI: 10.1002/14651858.CD013258] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0001] Campbell AJ, Robertson MC, La Grow SJ, Kerse NM, Sanderson GF, Jacobs RJ, et al. Randomised controlled trial of prevention of falls in people aged > or =75 with severe visual impairment: the VIP trial. BMJ 2005;331(817):1-8. [DOI: 10.1136/bmj.38601.447731.55] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0002] ISRCTN15342873. VIP falls prevention trial. Interventions to prevent falls and injury in elderly people with impaired vision. www.isrctn.com/ISRCTN15342873 (first received 2 March 2005). [DOI: 10.1186/ISRCTN15342873] [DOI]

[CD013258-bib-0003] Jacobs R, Campbell AJ, Robertson MC. Randomized controlled trial of falls prevention in people 75 years and older with severe visual impairment [abstract]. American Academy of Optometry Meeting, 2005 Dec 8-12; San Diego (CA) 2005. [CENTRAL: CN-00634854]

[CD013258-bib-0004] Kiata L, Kerse NM, Hughes WE, Hayman KJ, Robertson MC, La Grow SJ, et al. Agreement and compliance with advice on removing mats or rugs by older people with visual impairments. Journal of Visual Impairment & Blindness 2008;102(3):167-72. [Google Scholar]

[CD013258-bib-0005] La Grow SJ, Robertson MC, Campbell AJ, Clarke GA, Kerse NM. Reducing hazard related falls in people 75 years and older with significant visual impairment: how did a successful program work? Injury Prevention 2006;12(5):296-301. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0006] Chu MM, Fong KN, Lit AC, Rainer TH, Cheng SW, Au FL, et al. An occupational therapy fall reduction home visit program for community-dwelling older adults in Hong Kong after an emergency department visit for a fall. Journal of the American Geriatrics Society 2017;37(1):5-13. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0007] Cockayne S, Pighills A, Adamson J, Fairhurst C, Crossland S, Drummond A, et al. Home environmental assessments and modification delivered by occupational therapists to reduce falls in people aged 65 years and over: the OTIS RCT. Health Technology Assessment 2021;25(46):1-117. [DOI: 10.3310/hta2546] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0008] Cockayne S, Pighills A, Adamson J, Fairhurst C, Drummond A, Hewitt C, et al. Can occupational therapist-led home environmental assessment prevent falls in older people? A modified cohort randomised controlled trial protocol. BMJ Open 2018;8:e022488. [DOI: 10.1136/ bmjopen-2018-022488] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0009] Cockayne S, Pighills A, Fairhurst C, Adamson J, Crossland S, Drummond A, et al. Home hazard assessment and environmental modification to prevent falls in older people: the OTIS trial [version 1; peer review: 1 approved, 1 approved with reservations]. F1000Research 2021;10(500):1-18. [DOI: 10.12688/f1000research.52313.1] [DOI] [Google Scholar]

[CD013258-bib-0010] ISRCTN22202133. Occupational therapist home assessment and modification for prevention of falls. www.isrctn.com/ISRCTN22202133 (first received 20 June 2016). [DOI: 10.1186/ISRCTN22202133] [DOI]

[CD013258-bib-0011] Cumming RG, Thomas M, Szonyi G, Frampton G, Salkeld G, Clemson L. Adherence to occupational therapist recommendations for home modifications for falls prevention. American Journal of Occupational Therapy 2001;55(6):641-8. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0012] Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: a randomized trial of falls prevention. Journal of the American Geriatrics Society 1999;47(12):1397-402. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0013] Salkeld G, Cumming RG, O'Neill E, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0014] Cumming RG, Ivers R, Clemson L, Cullen J, Hayes MF, Tanzer M, et al. Improving vision to prevent falls in frail older people: a randomized trial. Journal of the American Geriatrics Society 2007;55(2):175-81. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0015] Day L, Fildes B, Gordon I, Fitzharris M, Flamer H, Lord S. Randomised factorial trial of falls prevention among older people living in their own homes. BMJ 2002;325(7356):128-31. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0016] Fitzharris MP, Day L, Lord SR, Gordon I, Fildes B. The Whitehorse NoFalls trial: effects on fall rates and injurious fall rates. Age & Ageing 2010;39(6):728-33. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0017] Haran MJ, Cameron ID, Ivers RQ, Simpson JM, Lee BB, Tanzer M, et al. Effect on falls of providing single lens distance vision glasses to multifocal glasses wearers: VISIBLE randomised controlled trial. BMJ 2010;340:c2265. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0018] Haran MJ, Lord SR, Cameron ID, Ivers RQ, Simpson JM, Lee BB, et al. Preventing falls in older multifocal glasses wearers by providing single-lens distance glasses: the protocol for the VISIBLE randomised controlled trial. BMC Geriatrics 2009;9:10. [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0019] NCT00350389. VISIBLE study (Visual Intervention Strategy Incorporating Bifocal & Long-Distance Eyewear) (VISIBLE) [Randomised controlled falls prevention trial of plain distance glasses in elderly multifocal wearers]. clinicaltrials.gov/show/NCT00350389 (first received 10 July 2006).

[CD013258-bib-0020] Kamei T, Fumiko K, Yuko Y, Yukako I, Rumi K, Tomoko S, et al. Effectiveness of a home hazard modification program for reducing falls in urban community-dwelling older adults: a randomized controlled trial. Japan Journal of Nursing Science 2015;12(3):184-97. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0021] Keall MD, Howden-Chapman P, Baker MG, Kamalesh V, Cunningham M, Cunningham C, et al. Formulating a programme of repairs to structural home injury hazards in New Zealand. Accident Analysis & Prevention 2013;57:124-30. [DOI: 10.1016/j.aap.2013.04.011] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0022] Keall MD, Pierse N, Howden-Chapman P, Cunningham C, Cunningham M, Guria J, et al. Home modifications to reduce injuries from falls in the home injury prevention intervention (HIPI) study: a cluster-randomised controlled trial. Lancet 2015;385(9964):231-8. [DOI: 10.1016/S0140-6736(14)61006-0] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0023] Lannin NA, Clemson L, McCluskey A, Lin CW, Cameron ID, Barras S. Feasibility and results of a randomised pilot-study of pre-discharge occupational therapy home visits. BMC Health Services Research 2007;7:42. [DOI: 10.1186/1472-6963-7-42] [MEDLINE: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0024] Lin MR, Wolf SL, Hwang HF, Gong SY, Chen CY. A randomized, controlled trial of fall prevention programs and quality of life in older fallers. Journal of the American Geriatrics Society 2007;55(4):499-506. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0025] Lockwood KJ, Harding KE, Boyd JN, Taylor NF. Predischarge home visits after hip fracture: a randomized controlled trial. Clinical Rehabilitation 2019;33(4):681-92. [DOI: 10.1177/0269215518823256] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0026] McKiernan FE. A simple gait-stabilizing device reduces outdoor falls and nonserious injurious falls in fall-prone older people during the winter. Journal of the American Geriatrics Society 2005;53(6):943-7. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0027] Nikolaus T, Bach M. Preventing falls in community-dwelling frail older people using a home intervention team (HIT): results from the randomized falls-HIT trial. Journal of the American Geriatrics Society 2003;51(3):300-5. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0028] Nikolaus T, Specht-Leible N, Bach M, Wittmann-Jennewein C, Oster P, Schlierf G. Effectiveness of hospital-based geriatric evaluation and management and home intervention team (GEM-HIT). Rationale and design of a 5-year randomized trial. Zeitschrift fur Gerontologie und Geriatrie 1995;28(1):47-53. [MEDLINE: ] [PubMed] [Google Scholar]

[CD013258-bib-0029] Pardessus V, Puisieux F, Di P, Gaudefroy C, Thevenon A, Dewailly P. Benefits of home visits for falls and autonomy in the elderly: a randomized trial study. American Journal of Physical Medicine and Rehabilitation 2002;81(4):247-52. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0030] Perry SD, Radtke A, McIlroy WE, Fernie GR, Maki BE. Efficacy and effectiveness of a balance-enhancing insole. Journals of Gerontology Series A: Biological Sciences & Medical Sciences 2008;63(6):595-602. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0031] ISRCTN07575807. Reducing home hazards to prevent falls in older people [Environmental assessment and modification to prevent falls in older people: a randomised controlled single centre trial]. controlled-trials.com/ISRCTN07575807 (first received 10 May 2009).

[CD013258-bib-0032] Pighills A, Torgerson DJ, Sheldon T. Publicity does not increase recruitment to falls prevention trials: the results of two quasi-randomized trials. Journal of Clinical Epidemiology 2009;62(12):1332-5. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0033] Pighills AC, Torgerson DJ, Sheldon TA, Drummond AE, Bland JM. Environmental assessment and modification to prevent falls in older people. Journal of the American Geriatrics Society 2011;59(1):26-33. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0034] Stark S, Keglovits M, Somerville E, Hu Y-L, Barker A, Sykora S, et al. Home hazard removal to reduce falls among community-dwelling older adults. A randomized clinical trial. JAMA Network Open 2021;4(8):e2122044. [DOI: 10.1001/jamanetworkopen.2021.22044] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0035] Stark S, Somerville E, Keglovits M, Conte J, Li M, Hu Y-L, et al. Protocol for the home hazards removal program (HARP) study: a pragmatic, randomized clinical trial and implementation study. BMC Geriatrics 2017;17(90):1-9. [DOI: 10.1186/s12877-017-0478-4] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0036] Stevens M, Holman CD, Bennett N, Klerk N. Preventing falls in older people: outcome evaluation of a randomized controlled trial. Journal of the American Geriatrics Society 2001;49(11):1448-55. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0037] Stevens M, Holman CD, Bennett N. Preventing falls in older people: impact of an intervention to reduce environmental hazards in the home. Journal of the American Geriatrics Society 2001;49(11):1442-7. [MEDLINE: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0038] Tchalla AE, Lachal F, Cardinaud N, Saulnier I, Rialle V, Preux PM, et al. Preventing and managing indoor falls with home-based technologies in mild and moderate Alzheimer's disease patients: pilot study in a community dwelling. Dementia & Geriatric Cognitive Disorders 2013;36(3/4):251-61. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0039] Thomas KS, Akobundu U, Dosa D. More than a meal? A randomized control trial comparing the effects of home-delivered meals programs on participants' feelings of loneliness. Journals of Gerontology: Social Sciences 2016;71(6):1049-58. [DOI: 10.1093/geronb/gbv111] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0040] Thomas KS, Rarikh RB, Zullo AR, Dosa D. Home-delivered meals and risk of self-reported falls: results from a randomized trial. Journal of Applied Gerontology 2018;37(1):41-57. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0041] NCT02819011. OHI - randomized control trial to evaluate efficacy, acceptability, and perception of benefit of an innovative custom AFO. clinicaltrials.gov/ct2/show/NCT02819011 (first received 30 June 2016).

[CD013258-bib-0042] Wang C, Goel R, Zhang Q, Lepow B, Najafi B. Daily use of bilateral custom-made ankle‐foot orthoses for fall prevention in older adults: a randomized controlled trial. Journal of the American Geriatrics Society 2019;67(6):1656-61. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0043] Casteel C, Bruening R, Carson M, Berard-Reed K, Ashida S. Evaluation of a falls and fire safety program for community‑dwelling older adults. Journal of Community Health 2021;45:717-27. [DOI: 10.1007/s10900-019-00786-8] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0044] Deems-Dluhy S, Hoppe-Ludwig S, Mummidisetty CK, Semik P, Heinemann AW, Jayaraman A. Microprocessor controlled knee ankle foot orthosis (KAFO) vs stance control vs locked KAFO: a randomized controlled trial. Archives of Physical Medicine and Rehabilitation 2021;102:233-44. [DOI: 10.1016/j.apmr.2020.08.013] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0045] Ferrer A, Formiga F, Sanz H, Vries OJ, Badia T, Pujol R, et al. Multifactorial assessment and targeted intervention to reduce falls among the oldest-old: a randomized controlled trial. Clinical Interventions in Aging 2014;9:PMC3940644. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0046] Greenberg MR, Goodheart V, Jacoby JL, Barraco RD, Crowley LM, Day R, et al. Emergency Department Stopping Elderly Accidents, Deaths and Injuries (ED STEADI) program. Journal of Emergency Medicine 2020;59(1):1-11. [DOI: 10.1016/j.jemermed.2020.04.019] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0047] Hill A, McPhail SM, Haines TP, Morris ME, Etherton-Beer C, Shorr R. Reducing falls after hospital discharge using patient education. Journal of the American Geriatrics Society 2019;67(S1):S50. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0048] Hill AM, McPhail SM, Haines TP, Morris ME, Etherton-Beer C, Shorr R, et al. Falls after hospital discharge: a randomized clinical trial of individualized multimodal falls prevention education. Journals of Gerontology Series A: Biological Sciences & Medical Sciences 2019;74(9):1511-7. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0049] Johansson E, Jonsson H, Dahlberg R, Patomella AH. The efficacy of a multifactorial falls-prevention programme, implemented in primary health care. British Journal of Occupational Therapy 2018;81(8):474-81. [DOI: 10.1177/0308022618756303] [DOI] [Google Scholar]

[CD013258-bib-0050] Lee H-C, Chang K-C, Tsauo J-Y, Huang Y-C, Lin SI. Effects of a multifactorial fall prevention program on fall incidence and physical function in community-dwelling older adults with risk of falls. Archives Physical Medicine & Rehabilitation 2013;94(4):606-15. [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0051] Liu M, Xue Q-L, Gitlin L, Wolff J, Guralnik J, Leff B, et al. Disability prevention program improves life-space and falls efficacy: a randomized controlled trial. American Geriatrics Society 2021;69:85-90. [DOI: 10.1111/jgs.16808] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0052] Luck T, Motzek T, Luppa M, Matschinger H, Fleischer S, Sesselmann Y, et al. Effectiveness of preventive home visits in reducing the risk of falls in old age: a randomized controlled trial. Clinical Interventions in Aging 2013;8:697-702. [DOI: 10.2147/CIA.S43284] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0053] Möller UO, Kristensson J, Midlöv P, Ekdahl C, Jakobsson U. Effects of a one-year home-based case management intervention on falls in older people: a randomized controlled trial. Journal of Aging and Physical Activity 2014;22(4):457-64. [DOI: 10.1123/japa.2013-0101] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0054] Oddsson L, Bisson T, Cohen H, Jacobs L, Khoshnoodi M, Kung D, et al. The effects of a wearable sensory prosthesis on gait and balance function after 10 weeks of use in persons with peripheral neuropathy and high fall risk – the walk2Wellness trial. Aging Neuroscience 2020;12:592751. [DOI: 10.3389/fnagi.2020.592751] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0055] Tannenbaum C, Fritel X, Halme A, den Heuvel E, Jutai J, Wagg A. Long-term effect of community-based continence promotion on urinary symptoms, falls and healthy active life expectancy among older women: cluster randomised trial. Age and Ageing 2019;48(4):526-32. [DOI: 10.1093/ageing/afz038] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0056] Tiefenbachová P, Zeleníková R. The effect of educational intervention by nurses on home environmental risk factors for falls. Central European Journal of Nursing and Midwifery 2019;10(2):1019-25. [DOI: 10.15452/CEJNM.2019.10.0009] [DOI] [Google Scholar]

[CD013258-bib-0057] Wesson J, Clemson L, Brodaty H, Lord S, Taylor M, Gitlin L, et al. A feasibility study and pilot randomised trial of a tailored prevention program to reduce falls in older people with mild dementia. BMC Geriatrics 2013;13:89. [DOI: 10.1186/1471-2318-13-89] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0058] Yamashita T, Yamashita K, Rinoie C, Takase Y, Sato M, Yamada K, et al. Improvements in lower-limb muscle strength and foot pressure distribution with foot care in frail elderly adults: a randomized controlled trial from Japan. BMC Geriatrics 2019;18:83. [DOI: 10.1186/s12877-019-1097-z] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0059] Stark S, Keglovits M, Somerville E. A randomized controlled feasibility trial of tailored home modifications to improve activities of daily living. American Journal of Occupational Therapy 2016;70(4 Suppl 1):7011520290p1. [DOI: 10.5014/ajot.2018.021774] [DOI] [Google Scholar]

[CD013258-bib-0060] ACTRN12617000543381. Do different shoe insole surfaces affect balance and walking in adults with diabetic peripheral neuropathy? A randomised controlled trial. www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=372682&isReview=true (first received 11 April 2017).

[CD013258-bib-0061] Hatton AL, Gane E, Maharaj J, Burns J, Paton J, Kerr G, et al. Textured shoe insoles to improve balance performance in adults with diabetic peripheral neuropathy: study protocol for a randomised controlled trial. BMJ Open 2019;9:e026240. [DOI: 10.1136/bmjopen-2018-026240] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0062] Aliberti MJR, Covinsky KE. Home modifications to reduce disability in older adults with functional disability. JAMA Internal Medicine 2019;179(2):211-2. [DOI: 10.1001/jamainternmed.2018.6414] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0063] Aplin T, Jonge D, Gustafsson L. Understanding home modifications impact on clients and their family's experience of home: a qualitative study. Australian Occupational Therapy Journal 2015;62(2):123-31. [DOI: 10.1111/1440-1630.12156] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0064] Campbell AJ, Borrie MJ, Spears GF, Jackson SL, Brown JS, Fitzgerald JL. Circumstances and consequences of falls experienced by a community population 70 years and over during a prospective study. Age and Ageing 1990;19:136-41. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0065] Canning CG, Allen NE, Bloem BR, Keus SHJ, Munneke M, Nieuwboer A, et al. Interventions for preventing falls in Parkinson's disease. Cochrane Database of Systematic Reviews 2015, Issue 3. Art. No: CD011574. [DOI: 10.1002/14651858.CD011574] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0066] Carnemolla P, Bridge C. Housing design and community care: how home modifications reduce care needs of older people and people with disability. International Journal of Environmental Research and Public Health 2019;16(11):1951. [DOI: 10.3390/ijerph16111951] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0067] Clemson L, Manor D, Fitzgerald MH. Behavioral factors contributing to older adults falling in public places. OTJR: Occupation, Participation and Health 2003;23(3):107-17. [Google Scholar]

[CD013258-bib-0068] Clemson L, Mackenzie L, Ballinger C, Close JCT, Cumming RG. Environmental interventions to prevent falls in community-dwelling older people: a meta-analysis of randomized trials. Journal of Aging and Health 2008;20(8):954-71. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0069] Clemson L, Pighills A. Environmental interventions to prevent falls at home and in the community. In: Lord S, Sherrington C, Naganathan V, editors(s). Falls in Older People: Risk factors, Strategies for Prevention and Implications for Practice. Cambridge: Cambridge University Press, 2021:360-77. [DOI: 10.1017/9781108594455.024] [DOI] [Google Scholar]

[CD013258-bib-0070] Cochrane Norway. How to write a plain language summary of a Cochrane intervention review. www.ahajournals.org/doi/10.1161/circ.114.suppl_18.II_525-c (accessed prior to 11 January 2023).

[CD013258-bib-0071] Cockayne S, Pighills A, Adamson J, Fairhurst C, Crossland S, Drummond A. Home environmental assessments and modification delivered by occupational therapists to reduce falls in people aged 65 years and over: the OTIS RCT. Health Technology Assessment 2021;25(46):1-118. [DOI: 10.3310/hta25460] [ISSN: 1366-5278] [DOI] [PMC free article] [PubMed]

[CD013258-bib-0072] Deandrea S, Lucenteforte E, Bravi F, Foschi R, La Vecchia C, Negri E. Risk factors for falls in community-dwelling older people: a systematic review and meta-analysis. Epidemiology 2010;21(5):658-68. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0073] Díaz-Arribas MJ, Fernández-Serrano M, Royuela A, Kovacs F, Gallego-Izquierdo T, Ramos-Sánchez M, et al. Minimal clinically important difference in quality of life for patients with low back pain. Spine 2017;42(24):1908-16. [DOI: 10.1097/BRS.0000000000002298] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0074] EuroQol Research Foundation. EQ-5D. euroqol.org/ (accessed 2 February 2019).

[CD013258-bib-0075] Fange A, Iwarsson S. Changes in accessibility and usability in housing: an exploration of the housing adaptation process. Occupational Therapy International 2005;12(1):44-59. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0076] Ferrari M, Harrison B, Lewis D. The risk factors for impulsivity-related falls among hospitalized older adults. Rehabilitation Nursing 2012;37(3):145-50. [DOI: 10.1002/RNJ.00046] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0077] Gillespie LD, Robertson MC, Gillespie WJ, Sherrington C, Gates S, Clemson LM, et al. Interventions for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews 2012, Issue 9. Art. No: CD007146. [DOI: 10.1002/14651858.CD007146.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0078] Gitlin LP. Conducting research on home environments: lessons learned and new directions. Gerontologist 2003;45(5):628-37. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0079] Gitlin L. Environmental adaptations for older adults and their families in the home and community. In: Soderback I, editors(s). International Handbook of Occupational Therapy Interventions. London: Springer, 2009. [Google Scholar]

[CD013258-bib-0080] Guyatt GH, Oxman AD, Montori V, Vist G, Kunz R, Brozek J, et al. GRADE guidelines: 5. Rating the quality of evidence - publication bias. Journal of Clinical Epidemiology 2011;64(12):1277–82. [DOI: 10.1016/j.jclinepi.2011.01.011] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0081] Hannan MT, Gagnon MM, Aneja J, Jones RN, Cupples LA, Lipsitz LA, et al. Optimizing the tracking of falls in studies of older participants: comparison of quarterly telephone recall with monthly falls calendars in the MOBILIZE Boston Study. American Journal of Epidemiology 2010;171(9):1031-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0082] He W, Goodkind D, Kowal P, US Census Bureau. An Aging World: 2015. Washington, DC: International Population Reports. US Government Publishing Office, 2016. [Google Scholar]

[CD013258-bib-0083] Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Quality of Life Research 2011;20(10):1727-36. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0084] Higgins JPT, Green S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from training.cochrane.org/handbook/archive/v5.1/.

[CD013258-bib-0085] Hill KD, Suttanon P, Lin SI, Tsang WWN, Ashari A, Hamid TAA, et al. What works in falls prevention in Asia: a systematic review and meta-analysis of randomized controlled trials. BMC Geriatrics 2018;18(1):3. [DOI: 10.1186/s12877-017-0683-1] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0086] Hopewell S, Adedire O, Copsey BJ, Boniface GJ, Sherrington C, Clemson L, et al. Mutlifactorial and multiple component interventions for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews 2018, Issue 7. Art. No: CD012221. [DOI: 10.1002/14651858.CD012221.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0087] Horowitz BP, Nochajski SM, Schweitzer JA. Occupational therapy community practice and home assessments: use of the home safety self-assessment tool (HSSAT) to support aging in place. Occupational Therapy in Health Care 2013;27(3):216-27. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0088] IBM SPSS Statistics for Windows. Version 24.0. Armonk, NY: IBM Corp, 2019.

[CD013258-bib-0089] Iliffe S, Kendrick D, Morris R, Griffin M, Haworth D, Carpenter H, et al. Promoting physical activity in older people in general practice: ProAct65. British Journal of General Practice 2015;65:e731–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0090] Iwarsson S, Horstmann V, Carlsson G, Oswald F, Wahl HW. Person-environment fit predicts falls in older adults better than the consideration of environmental hazards only. Clinical Rehabilitation 2009;23(6):558-67. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0091] Karlsson MK, Vonschewelov T, Karlsson C, Cöster M, Rosengen BE. Prevention of falls in the elderly: a review. Scandinavian Journal of Public Health 2013;41(5):442-54. [DOI: 10.1177/1403494813483215] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0092] Keglovits M, Clemson L, Hu YL, Nguyen A, Neff AJ, Mandelbaum C, et al. A scoping review of fall hazards in the homes of older adults and development of a framework for assessment and intervention. Australian Occupational Therapy Journal 2020;67(5):470-8. [DOI: 10.1111/1440-1630.12682] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0093] Kellog International Work Group. The prevention of falls in later life. A report of the Kellogg International Work Group on the Prevention of Falls by the Elderly. Danish Medical Bulletin 1987;34(Suppl 4):1-24. [PubMed] [Google Scholar]

[CD013258-bib-0094] Lamb SE, Jørstad-Stein EC, Hauer K, Becker C, Prevention of Falls Network Europe and Outcomes Census Group. Development of a common outcome data set for fall injury prevention trials: the Prevention of Falls Network Europe consensus. Journal of the American Geriatrics Society 2005;53(9):1618-22. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0095] Lamb SE, Hauer K, Becker C. Manual for the fall prevention classification system. www.profane.eu.org/documents/Falls_Taxonomy.pdf (accessed prior to 18 December 2018).

[CD013258-bib-0096] Law M, Cooper BA, Strong S, Stewart D, Rigby PJ, Letts LT. The person environment occupation model: a transactive approach to occupational therapy. Canadian Journal of Occupational Therapy 1997;63:9-23. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0097] Leavy B, Byberg L, Michaëlsson K, Melhus H, Åberg AC. The fall descriptions and health characteristics of older adults with hip fracture: a mixed methods study. BMC Geriatrics 2015;15(1):40. [DOI: 10.1186/s12877-015-0036-x] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0098] Lee S. Falls associated with indoor and outdoor environmental hazards among community-dwelling older adults between men and women. BMC Geriatrics 2021;21(1):547. [DOI: 10.1186/s12877-021-02499-x] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0099] Lefebvre C, Glanville J, Briscoe S, Littlewood A, Marshall C, Metzendorf MI, et al. Chapter 4: Searching for and selecting studies. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions version 6.0 (updated July 2019). Cochrane, 2019. Available from training.cochrane.org/handbook/archive/v6.

[CD013258-bib-0100] Lord SR, Menz HB, Sherrington C. Home environment risk factors for falls in older people and the efficacy of home modifications. Age and Ageing 2006;35(Suppl 2):ii55-9. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0101] Lusardi MM, Fritz S, Middleton A, Allison L, Wingood M, Phillips E, et al. Determining risk of falls in community dwelling older adults: a systematic review and meta-analysis using posttest probability. Journal of Geriatric Physical Therapy 2017;40(1):36. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0102] Makino K, Makizako H, Doi T, Tsutsumimoto K, Hotta R, Nakakubo S, et al. Impact of fear of falling and fall history on disability incidence among older adults: prospective cohort study. International Journal of Geriatric Psychiatry 2018;33(4):658-62. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0103] Microsoft Excel. Microsoft Corporation, 2018.

[CD013258-bib-0104] Moreland B, Kakara R, Henry A. Trends in nonfatal falls and fall-related injuries among adults aged ≥65 years - United States, 2012-2018. Morbidity and Mortality Weekly Report 2020;69(27):875–81. [DOI: 10.15585/mmwr.mm6927a5] [DOI] [PMC free article] [PubMed]

[CD013258-bib-0105] Nahemow L. The ecological theory of aging: Powell Lawton's legacy. In: Rubinstein RL, Moss M, Kieban M, Lawton MP, editors(s). The Many Dimensions of Aging. New York: Springer, 2000:22-40. [Google Scholar]

[CD013258-bib-0106] Peterson EW, Clemson L. Understanding the role of occupational therapy in fall prevention for community dwelling older adults. OT Practice 2008;13(3):CEA0208. [Google Scholar]

[CD013258-bib-0107] Pighills A, Ballinger C, Pickering R, Chan S. A critical review of the effectiveness of environmental assessment and modification in the prevention of falls amongst community dwelling older people. British Journal of Occupational Therapy 2016;79(3):133–43. [Google Scholar]

[CD013258-bib-0108] Pighills A, Drummond A, Crossland S, Torgerson DJ. What type of environmental assessment and modification prevents falls in community dwelling older people? BMJ 2019;364:1880. [DOI: 10.1136/bmj.l880] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0109] Pighills A, Clemson L. Environmental risk factors for falls. In: Lord S, Sherrington C, Naganathan V, editors(s). Falls in Older People: Risk Factors, Strategies for Prevention and Implications for Practice. Cambridge: Cambridge University Press, 2021:202-10. [DOI: 10.1017/9781108594455.013] [DOI] [Google Scholar]

[CD013258-bib-0110] Prabhakaran K, Gogna S, Pee S, Samson DJ, Con J, Latifi R. Falling again? Falls in geriatric adults - risk factors and outcomes associated with recidivism. Journal of Surgical Research 2020;247:66-76. [DOI: 10.1016/j.jss.2019.10.041] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0111] Review Manager 5 (RevMan 5). Version 5.4. Copenhagen: The Cochrane Collaboration, 2020.

[CD013258-bib-0112] Robertson MC, Campbell AJ, Herbison P. Statistical analysis of efficacy in falls prevention trials. Journals of Gerontology Series A: Biological Sciences & Medical Sciences 2005;60(4):530-4. [DOI] [PubMed]

[CD013258-bib-0113] Rubenstein L. Falls in older people: epidemiology, risk factors and strategies for prevention. Age and Ageing 2006;35-S2:37-41. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0114] Salkeld G, Cumming RG, O'Neill R, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0115] Sanford JA, Jones M, Daviou P, Grogg K, Butterfield T. Using telerehabilitation to identify home modification needs. Assistive Technology 2004;16(1):43-53. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0116] Santesso N, Carrasco‐Labra A, Brignardello‐Petersen R. Hip protectors for preventing hip fractures in older people. Cochrane Database of Systematic Reviews 2014, Issue 3. Art. No: CD001255. [DOI: 10.1002/14651858.CD001255.pub5] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0117] Schünemann HJ, Oxman AD, Higgins JPT, Vist GE, Glasziou P, Guyatt GH on behalf of the Cochrane GRADEing Methods Group (formerly Applicability and Recommendations Methods Group) and the Cochrane Statistical Methods Group. Chapter 11: Completing ‘Summary of findings’ tables and grading the confidence in or quality of the evidence. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS, editor(s). Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017). Cochrane, 2017. Available from training.cochrane.org/handbook/archive/v5.2.

[CD013258-bib-0118] Schünemann HJ, Vist GE, Higgins JPT, Santesso N, Deeks JJ, Glasziou P, et al. Chapter 15: Interpreting results and drawing conclusions. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from training.cochrane.org/handbook.

[CD013258-bib-0119] Shi J, Zhou BY, Tao YK, Yu PL, Zhang CF, Qin ZH, et al. Incidence and associated factors for single and recurrent falls among the elderly in an urban community of Beijing. Biomedical & Environmental Sciences 2014;27(12):939-49. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0120] Smeeth L, Ng ES. Intraclass correlation coefficients for cluster randomized trials in primary care: data from the MRC Trial of the Assessment and Management of Older People in the Community. Controlled Clinical Trials 2002;23(4):409-21. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0121] Stark S, Landsbaum A, Palmer JL, Somerville EK, Morris JC, Harvey A, et al. Client-centred home modifications improve daily activity performance of older adults . Canadian Journal of Occupational Therapy 2009;76(1 Suppl):135-45. [DOI: 10.1177/000841740907600s09] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0122] Stevens JA, Mahoney JE, Ehrenreich H. Circumstances and outcomes of falls among high risk community-dwelling older adults. Injury Epidemiology 2014;1(5):1-9. [DOI: 10.1186/2197-1714-1-5] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0123] Verheyden GSAF, Weerdesteyn V, Pickering RM, Kunkel D, Lennon S, Geurts ACH, et al. Interventions for preventing falls in people after stroke. Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No: CD008728. [DOI: 10.1002/14651858.CD008728.pub2] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0124] Verma SK, Willetts JL, Corns HL, Marucci-Wellman HR, Lombardi DA, Courtney TK. Falls and fall-related injuries among community-dwelling adults in the United States. PLOS ONE 2016;11(3):e0150939. [DOI: 10.1371/journal.pone.0150939] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0125] Walters SJ, Brazier JE. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Quality of Life Research 2005;14(6):1523-32. [PMID: 10.1007/s11136-004-7713-0] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0126] Wellecke C, D'Cruz K, Winkler D, Douglas J, Goodwin J, Davis E, et al. Accessible design features and home modifications to improve physical housing accessibility: a mixed-methods survey of occupational therapists. Disability and Health Journal 2022;101281:1-8. [DOI: 10.1016/j.dhjo.2022.101281] [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0127] World Health Organization. Assistive devices and technologies. www.who.int/disabilities/technology/en/ (accessed 20 November 2017).

[CD013258-bib-0128] Williams JS, Kowal P, Hestekin H, O'Driscoll T, Peltzer K, Yawson A, et al. Prevalence, risk factors and disability associated with fall-related injury in older adults in low- and middle-income countries: results from the WHO Study on global AGEing and adult health (SAGE). BMC Medicine 2015;13:147. [DOI: 10.1186/s12916-015-0390-8] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013258-bib-0129] Zecevic AA, Salmoni AW, Speechley M, Vandervoort AA. Defining a fall and reasons for falling: comparisons among the views of seniors, health care providers, and the research literature. Gerontologist 2006;46(3):367-76. [DOI] [PubMed] [Google Scholar]

[CD013258-bib-0130] Clemson L, Stark S, Pighills A, Torgerson DJ, Sherrington C, Lamb S. Environmental interventions for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews 2019, Issue 2. Art. No: CD013258. [DOI: 10.1002/14651858.CD013258] [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Environmental interventions for preventing falls in older people living in the community

Lindy Clemson

Susan Stark

Alison C Pighills

Nicola J Fairhall

Sarah E Lamb

Jinnat Ali

Catherine Sherrington

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Summary of findings 1. Home fall‐hazard reduction compared with control (e.g. usual care) for preventing falls in older people living in the community .

Summary of findings 2. Assistive technology for vision improvement interventions compared with usual care or attention control for preventing falls in older people living in the community.

Summary of findings 3. Education interventions compared with control (e.g. usual care) for preventing falls in older people living in the community .

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Timing of outcome measurement

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Risk of bias in included trials

Meta‐analysis decisions

Summary of findings and assessment of the certainty of the evidence

Results

Description of studies

Results of the search

1.

Included studies

1. Study design, length of follow‐up, setting, and trial size.

2. Key characteristics of participants and intervention scope.

Home fall‐hazard interventions

Study design

Study size

Study setting

Participants

Interventions

Outcomes

Assistive technology interventions

Study design

Study size

Study setting

Participants

Interventions

Outcomes

Education interventions

Study design, size, and setting

Intervention and outcomes

Excluded studies