Table 3 Summary of evidence for included systematic reviews

Citation details:

Hayes et al. (2017)

Number of relevant primary studies:

13

AMSTAR 2 rating:

Moderate

Neurological condition:

MS

To evaluate the effectiveness of interventions to reduce falls in people with MS, specifically to compare falls prevention interventions to controls and to compare different types of falls prevention interventions.

N = 839 (range: 12–177)

Female participants range: 59–98%

Mean age = 52 years (range: 36–62 years)

Participants in the majority of RCTs included people with mild to moderate

severity of MS.

Group-based exercise session or individualised HEP × 1 RCT

One-to-one motor and sensory rehabilitation or motor rehabilitation × 1 RCT

Balance treatment × 1 RCT

Group-based exercise circuit class or one-to-one physiotherapy or yoga classes × 1 RCT

FES × 1 RCT

Supervised sensory integration balance training × 1 RCT

Interactive exergames × 1 RCT

Group-based exercise × 1 RCT

Wii Fit Plus balance games × 1 RCT

Progressive HEP × 1 RCT

Exercise or education or combined exercise and education × 1 RCT

Activity Through Movement × 1 RCT

FES and core stability exercises × 1 RCT

Falls rate:

There was no significant effect of exercise compared to control on falls rate (RaR 0.68, 95% CI 0.43 to 1.06).

Moderate

There was no significant effect of FES compared to exercise on falls rate (RaR 0.91, 95% CI 0.78 to 1.06).

Low

Number of fallers:

There was no significant effect of exercise on the number of fallers post-intervention (RR 0.85, 95% CI 0.51 to 1.43).

Moderate

There was no evidence of an effect of education-based interventions on number of fallers (RR 0.83, 95% CI 0.40 to 1.76).

Moderate

There was no evidence of an effect of multicomponent interventions on number of fallers (RR 0.30, 95% CI 0.04 to 2.20).

Moderate

Citation details:

Sosnoff & Sung (2015)

Number of relevant primary studies:

9

AMSTAR 2 rating:

Critically low

Neurological condition:

MS

To review the effects of falls prevention interventions on falls incidence among people with MS and determine characteristics of these programmes that might optimise the reduction of falls.

N = 504 (range: 28–111)

Percentage of female participants: N/R

EDSS median range: 3.0–6.0 (N/R ×3 studies: 1x RCT, 2x NRSIs)

Mean age range: 46–63 years

Exercise-based ×7 studies:

Motor-sensory rehabilitation or motor rehabilitation × 1 RCT

Group physiotherapy or 1-to-1 physiotherapy or yoga × 1 NRSI

Wii balance board system training × 1 RCT

Exergame training on an unstable platform or single-task exercises on the unstable surface × 1 NRSI

Balance exercise targeting core stability, dual tasking and sensory strategies × 1 NRSI

Home-based exercise or education or exercise and education × 1 RCT

Sensory integration rehabilitation × 1 RCT

Technology-based × 2 studies:

FES for 12 weeks and exercise with FES for 12 weeks × 1 RCT

FES × 1 RCT

Total number of falls:

3x studies reported a significant reduction in the number of falls in the exercise-based intervention groups (1x RCT, 2x NRSIs).

Low

2x RCTs reported a reduction in total number of falls for groups receiving FES.

Very low

Number of fallers:

2x RCTs reported that the number of fallers was lower following exercise-based intervention, 1x NRSI reported no difference between groups.

Very low

Number of recurrent fallers:

1x RCT reported a significantly lower number of recurrent fallers in the exercise-based intervention groups compared to the control group.

Not assessed

Mean number of falls:

1x NRSI reported that the exercise-based intervention group had a lower mean number of falls than the control group.

Not assessed

Citation details:

Booth et al. (2014)

Number of relevant primary studies:

1

AMSTAR 2 rating:

Critically low

Neurological condition:

MS

To evaluate whether virtual reality interventions, including interactive gaming systems, are effective at improving balance in adults with impaired balance.

N = 80

Percentage of female participants: N/R

Age: N/R

Nintendo WiiFit balance exercise programme ×1 RCT

Total number of falls:

1x RCT identified that those in the intervention group experienced less falls (n = 10) compared with the control group (n = 14) during the study.

Critically low

Citation details:

Denissen et al. (2019)

Number of relevant primary studies:

14

AMSTAR 2 rating:

Moderate

Neurological condition:

Stroke

To evaluate the effectiveness of interventions aimed at preventing falls in people after stroke.

N = 1358 (range: 34–170, median: 91)

Mean percentage of female participants: 40% (range: 29–65%)

Mean age range: 57 (+/−11) - 79 (+/−8) years

Exercise-based interventions ×8 RCTs:

Treadmill training without body weight support plus overground walking × 1 RCT

Treadmill walking with harness plus conventional stroke rehabilitation × 1 RCT

WEBB programme plus HEP plus advice to increase walking × 1 RCT

Physiotherapy treatment × 1 RCT

Whole-body vibration × 1 RCT

External perturbation training × 1 RCT

Exercise programme challenging dynamic balance and emphasising agility and multisensory approach in between × 1 RCT

Tai Chi or SilverSneakers × 1 RCT

Environment/assistive technology ×3 RCTs:

Predischarge home assessment visit × 1 RCT

Prescription of single lens distance glasses × 1 RCT

Walking training using the I-Walker plus exercises on hand recovery, tone control and improvement of global ability × 1 RCT

Other interventions/ procedures × 1 RCT:

Active repeated tDCS plus physical rehabilitation × 1 RCT

Multifactorial intervention × 1 RCT:

Multifactorial, individually-tailored falls prevention programme plus usual care after discharge × 1 RCT

Multiple intervention × 1 RCT:

HIFE programme plus individualised HEP × 1 RCT

Falls rate:

There was a significant reduction in falls rate for the exercise group (RaR 0.72, 95% CI 0.54 to 0.94).

Moderate

There was no significant reduction in falls rate when comparing a home visit to a predischarge assessment in the hospital setting (RaR 0.85, 95% CI 0.43 to 1.69).

Low

There was no significant reduction in falls rate when single lens distance vision glasses replaced multifocal glasses (RaR 1.08, 95% CI 0.52 to 2.25).

Low

There was no significant reduction in falls rate for the I-walker group compared to the control group (RaR 0.56, 95% CI 0.19 to 1.66).

Low

Number of fallers:

When pooled, there was no significant effect of exercise on number of fallers (RR 1.03, 95% CI 0.90 to 1.19).

Moderate

There was no significant difference in number of fallers between the home visit or hospital assessment groups (RR 1.48, 95% CI 0.71 to 3.09).

Low

There was no significant reduction in number of fallers when single lens distance vision glasses replaced multifocal glasses (RR 0.74, 95% CI 0.47 to 1.18).

Low

There was no significant reduction in number of fallers for the I-walker group compared to the control group (RR 0.44, 95% CI 0.16 to 1.22).

Low

There was a significant reduction in the number of fallers in the active tDCS group compared to the control group (RR 0.30, 95% CI 0.14 to 0.63).

Low

Citation details:

Pollock et al. (2014)

Number of relevant primary studies:

3

AMSTAR 2 rating:

Low

Neurological condition:

Stroke

To determine the effect of interventions that alter the starting posture on ability to STS independently and to determine the effect of rehabilitation interventions on ability to STS independently.

N = 276 (range: 54–156)

Percentage of female participants: N/R

Mean age range (intervention group): 60 (+/−7) - 72 (+/−10.4) years

Mean time since stroke range (intervention group): 21 (+/−8) – 171 days

Repetitive STS training × 1 RCT

Falls prevention programme × 1 RCT

Endurance and resistance exercises × 1 RCT

Number of fallers:

There was no evidence of an effect of intervention on the number of fallers compared to control (OR 0.81, 95% CI 0.35 to 1.87).

Moderate

Citation details:

Batchelor et al. (2010)

Number of relevant primary studies:

10 (based on 7 RCTs)

AMSTAR 2 rating:

Critically low

Neurological condition:

Stroke

To systematically evaluate the effects of any interventions on falls in people after stroke.

N = 723 (range: 39–258)

Percentage of female participants: N/R ×6 RCTs, 46% × 1 RCT

Age = N/R ×5 RCTs, range: 18–90 years × 1 RCT, mean: 74.7 years × 1 RCT

Time since stroke range: < 24 h to > 2 years post-stroke

Group STS practice plus usual care × 1 RCT

Very early mobilisation plus usual care × 1 RCT

Standing symmetry training and STS training plus usual care × 1 RCT

Community physiotherapy sessions × 1 RCT

Fresnel prisms applied to affected hemi-field plus usual rehabilitation × 1 RCT

Sunlight exposure outdoors × 1 RCT

Home rehabilitation × 1 RCT

Falls rate:

These was no significant effect of exercise on falls rate compared to usual care (RaR 1.22, 95% CI 0.76 to 1.98).

Very low

The application of fresnel prisms to the affected hemi-field had no significant effect on falls rate.

Very low

Increased sunlight exposure had no significant effect on falls rate.

Low

Home rehabilitation with multi-disciplinary outreach service had no significant effect on falls rate.

Very low

Number of fallers:

There was no significant effect of exercise on number of fallers compared to usual care (RR 0.77, 95% CI 0.24 to 2.43).

Low

Citation details:

Rutz et al. (2020)

Number of relevant primary studies:

1

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To investigate the evidence for physical interventions for freezing of gait and gait impairments in PD and establish recommendations for clinical practice.

N = 21

Percentage of female participants: N/R

Age: N/R

H&Y range: 2–3

All participants had freezing of gait

HEP with rhythmic auditory cueing and functional walking exercises × 1 RCT

Total number of falls:

Intervention did not significantly reduce falls.

Very low

Citation details:

Owen et al. (2019)

Number of relevant primary studies:

3

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To identify and review falls self-management interventions for people with PD and, where possible, assess their efficacy for improving patient and caregiver outcomes, quality of life and psychological outcomes.

N = 574 (range: 133–231)

Percentage of female participants: N/R

Mean age range: 67.9 (+/− 9.6) - 71.4 (+/− 8.1) years

Majority participants H&Y stage ≤2 (indicating reduced falls risk) × 2 RCTs

Range of participants that had fallen in year preceding intervention: 55–78%

Physiotherapy plus education ×3 RCTs

Falls rate:

Physiotherapy plus falls self-management education significantly reduced falls rate.

Moderate

Number of fallers:

Physiotherapy plus falls self-management education did not have a significant effect on number of fallers.

Low

Number of recurrent fallers:

No significant effect of intervention on number of recurrent fallers.

Not assessed

Citation details:

Rodrigues-Krause et al. (2019)

Number of relevant primary studies:

1

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To review dance as a form of intervention to promote functional and metabolic health in older adults.

N = 33

Percentage of female participants: 39%

Intervention group mean age = 68.4 +/− 7.7 years

Control group mean age = 74.4 +/− 6.5 years

Tango dance × 1 NRSI

Total number of falls:

Tango group had reduced number of falls compared to education group.

Very low

Number of fallers:

There was no significant effect of Tango on number of fallers compared to education.

Very low

Citation details:

Winser et al. (2019)

Number of relevant primary studies:

1

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To identify evidence evaluating the cost-effectiveness of physiotherapy treatment techniques for people with neurological disorders.

N = 231

Percentage of female participants: 41.6%

Mean age: 70.7 years

Group based exercise classes with home visits from physical therapist and provision of standard fall prevention booklet × 1 RCT

Mean number of falls:

Lower mean number of falls in intervention group (4.106 falls) than control group (7.053 falls).

Not assessed

Citation details:

Mak et al. (2017)

Number of relevant primary studies:

8 (based on 6 RCTs)

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To investigate the long-term effects of exercise and physical therapy in people with PD.

N = 790 (range: 23–195)

Percentage of female participants: N/R

Age range: 40–89 years

H&Y stage range: 1–4

Balance training ×4 RCTs:

Progressive strengthening, balance, cueing for FOG and fall prevention advice × 1 RCTs

Mobility and balance training with movement strategies and fall prevention advice or progressive resistance training and fall prevention advice × 1 RCT

Technology-assisted balance and mobility training × 1 RCT Balance and mobility training × 1 RCT

Complementary exercises × 2 RCTs:

Tai Chi × 1 RCT

Tai Chi or progressive strength training × 1 RCT

Falls rate:

Balance training significantly reduced falls rate.

Low

Tai Chi significantly reduced falls rate.

Low

Citation details:

Ramazzina et al. (2017)

Number of relevant primary studies:

5 (based on 4 RCTs)

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To assess the effectiveness of resistance training on muscle strength improvement.

N = 154 (range: 29–51)

Percentage of female participants: N/R

Age: N/R

H&Y range: 1.5–3

Strength training × 4 RCTs:

Using pneumatic resistance equipment × 1 RCT

Using dynamometers and leg-press machines, in addition to rowing exercises, repetitive step on a 6-in. curb, and weighted walking × 1 RCT

Using dynamometers and leg-press machines, in addition to rowing exercises, repetitive step on a 6-in. curb, and weighted walking plus home training × 1 RCT

Hydrotherapy with perturbation-based balance and strength training × 1 RCT

Total number of falls:

1x RCT reported a significant reduction with strength training (hydrotherapy), 3x RCTs reported no significant reduction in number of falls with strength training.

Low

Citation details:

Song et al. (2017)

Number of relevant primary studies:

3

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To investigate the effects of Tai Chi/Qigong on motor and non-motor function, and quality of life in people with PD.

N = 305 (range: 34–195)

Percentage of female participants: 38%

Mean age range: 66–69.5 years

Tai Chi × 2 RCTs

Qigong × 1 NRSI

Total number of falls:

Tai Chi significantly reduced number of falls compared to control (ES − 0.403, 95% CI − 0.677 to − 0.129).

Moderate

1x NRSI reported Qigong reduced number of falls.

Low

Citation details:

Shen et al. (2016)

Number of relevant primary studies:

8

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To examine the effects of exercise on improving balance and gait ability and reducing falls among people with PD over the short-term and long-term.

N = 925 (range: 64–231)

Percentage of female participants: 38%

Age range: 61.6 (+/− 8) - 72.2 (+/− 9.2) years

H&Y range: 1–4

Balance, gait, strength, other exercises × 2 RCTs

Gait × 1 RCT

Balance × 2 RCTs

Balance, strength × 1 RCT

Strength × 1 RCT

Balance and gait × 2 RCTs

Falls rate:

The fall rate showed a significant overall reduction over the short-term with exercise training (RaR 0.485, 95% CI 0.329 to 0.715).

Moderate

The fall rate showed a significant overall reduction over the long-term with exercise training RaR 0.413,95% CI 0.270 to 0.630).

Moderate

Numbers of fallers:

The number of fallers did not decrease significantly over the short-term with exercise training (RR 0.939, 95% CI 0.822 to 1.072).

Moderate

The number of fallers did not decrease significantly over the long-term with exercise training (RR 0.787, 95% CI 0.605 to 1.024).

Moderate

Citation details:

Tomlinson et al. (2014)

Number of relevant primary studies:

3

AMSTAR 2 rating:

Moderate

Neurological condition:

PD

To assess the effectiveness of one physiotherapy intervention compared with a second approach in people with PD.

N = 469 (range: 64–210)

Percentage of female participants: 36%

Mean age range: 67.3–69 years

Mean disease duration range: 6.7–10.4 years

Tai Chi or resistance training × 1 RCT

Movement strategy training and individualised home practice session and weekly structured falls risk education and a single home visit or progressive strength training and individualised HEP and once weekly structured falls risk education and a single home visit × 1 RCT

Balance training × 1 RCT

Total number of falls:

1x RCT found number of falls were significantly reduced during the intervention period in the progressive strength training group, 2x RCTs found no significant effect on number of falls.

Low

Time to first fall:

No significant difference in time to first fall between the progressive strength training and the movement strategy training arms × 1 RCT.

Not assessed

Citation details:

Tomlinson et al. (2012a)

Number of relevant primary studies:

7

AMSTAR 2 rating:

Low

Neurological condition:

PD

To assess the effectiveness of physiotherapy intervention compared with no intervention or placebo in patients with PD.

N = 532 (range: 18–153)

Percentage of female participants: 38% female across 6x RCTs (N/R × 1 RCT, male participants only × 1 RCT)

Mean age range: 63.4–73.7 years (N/R × 1 RCT)

Mean H&Y range: 2–3.14 (N/R × 2 RCTs)

Mean disease duration range: 4.7–9.1 years (N/R × 2 RCTs)

Exercise × 3 RCT

Tai Chi × 2 RCTs

Cueing × 1 RCT

Treadmill training × 1 RCT

Total number of falls:

1x RCT reported a significant reduction in number of falls for the Tai Chi compared with no intervention, 5x RCTs reported no significant effect of exercise-based intervention on number of falls.

Low

There was no significant effect of cueing intervention on number of falls.

Low

Citation details:

Monti et al. (2011)

Number of relevant primary studies:

5

AMSTAR 2 rating:

Critically low

Neurological condition:

PD

To research the effectiveness of physiotherapy intervention on the prevention of falls among people with PD.

N = 456 (range: 18–230)

Percentage of female participants: N/R

Mean age range: 71.8 +/−  6.4 years to 72.5 years (mean age N/R × 3 RCTs)

Age range: 44–91 years (age range N/R × 4 RCTs)

Exercise-based interventions × 4 RCTs:

Exercises, cueing with the integration in the ADL plus received a booklet with advice for the prevention of falls × 1 RCT.

Treadmill walking, exercise to increase ROM and stretching exercises × 1 RCT

Exercises to strengthen the muscles of the legs, to increase the ROM, for the equilibrium, for walking outdoor × 1 RCT

Treadmill walking × 1 RCT

Cueing intervention × 1 RCT

Total number of falls:

4x RCTs report a reduction in the number of fall episodes for the exercise-based intervention groups.

Low

The cueing intervention group had a decrease in the number of falls in the ADL × 1 RCT.

Very low

Citation details:

Winser et al. (2018)

Number of relevant primary studies:

4

AMSTAR 2 rating:

Low

Neurological condition:

Stroke and PD

To determine whether Tai Chi training improves balance and reduces falls incidence when compared to control conditions of either active treatment or no treatment in people with neurological diseases.

PD:

N = 288 (range 17–195)

Percentage of female participants: 36%

Age: 72 +/−  8.5 years (N/R × 2 RCTs)

Stroke:

N = 145

Percentage of female participants: 47%

Age: 69.9 +/−  10 years

PD:

Tai Chi × 3 RCTs

Stroke:

Tai Chi × 1 RCT

Total number of falls (PD):

There was a statistically significant effect of Tai Chi compared with active therapies on total number of falls (OR 0.47, 95% CI 0.29 to 0.77).

Moderate

There was a statistically significant effect of Tai Chi compared with no treatment on total number of falls (OR 0.29, 95% CI 0.11 to 0.79).

Low

Total number of falls (Stroke):

There was a statistically significant effect of Tai Chi compared with active therapies on total number of falls (OR 0.21, 95% CI 0.09 to 0.48).

Low