Study reference

Study characteristics

Patient characteristics²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Exercises

Day, 2015

Type of study: RCT [parallel]

Setting: Disabled community-dwelling people

Country: Australia

Source of funding: Non-commercial

Inclusion criteria:

• Aged 70 years or over
• Community residents
• Preclinally disabled defined by self-reported difficulty with, or modification of, 2 mobility tasks

N total at baseline:

Intervention: 250

Control:253

Important prognostic factors²:

Described for analysed population

Age category, n (%)

I:

70-74 years: 69 (34)

75-79 years: 62 (30)

80-84 years: 50 (25)

≥85 years: 23 (11)

C:

70-74 years: 64 (31)

75-79 years: 64 (31)

80-84 years: 56 (27)

≥85 years: 21 (10)

Sex:

I: 30% M

C: 30% M

Groups comparable at baseline? Yes

Modified Sun style tai-chi and covered agility, mobility, balance, strength, breathing, and relaxation

Classes of 12-16 participants were held twice weekly for 60 minutes per class, for up to 24 weeks.

Flexibility and stretching program, conducted primarily in the seated position

Classes of 12-16 participants were held twice weekly for 60 minutes per class, for up to 24 weeks.

Length of follow-up:

24 weeks

Loss-to-follow-up:

Intervention:

N = 46 (18%)

Reasons: refused fall calendars

Control:

N = 48 (19%)

Reasons: refused fall calendars

Incomplete outcome data:

Intervention:

N = 46 (18%)

Reasons: refused fall calendars

Control:

N = 48 (19%)

Reasons: refused fall calendars

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 100 patient days

Number of fallers, risk ratio

I: 38 (out of 204)

C: 42 (out of 205)

RR 0.91 (95%CI 0.61-1.35)

Rate of falls, rate ratio

I: 56.7 (46.7-66.7)

C: 60.6 (50.8-70.4)

RR 0.94 (95%CI 0.63-1.38)

2. Fractures

Not reported

Not only community dwellers included, but also residents from retirement villages.

Follow-up lasted till 48 weeks; however, as not all participants could participate till 48 weeks due to budget limitations, only data from 24 weeks was considered.

Gleeson, 2015

Type of study: RCT [parallel]

Setting: Participants’ homes

Country: Australia

Source of funding: Non-commercial

Inclusion criteria:

• Aged 50-90
• Live within the Sydney metropolitan area
• Not need an interpreter
• Have had Orientation & Mobility training within the previous five years

Exclusion criteria:

• Not independently mobile
• Did not have conversational English
• Were planning cataract surgery within 12 months

N total at baseline:

Intervention: 60

Control: 60

Important prognostic factors²:

Age ± SD:

I: 74 (11)

C: 74 (11)

Sex:

I: 28% M

C: 30% M

Groups comparable at baseline? Yes

Alexander Technique lesson for 12 weeks in addition to usual care.

Lessons were typically 30 minutes in length. A lesson protocol was developed using everyday activities such as movements between sitting and standing, getting to and from the floor, and walking, climbing stairs and carrying everyday articles.

Usual care provided by Guide Dog (see comment)

Length of follow-up: 12 months

Loss-to-follow-up:

Intervention:

N = 5 (7%)

Reasons: 1 refusal, 4 withdrawals

Control:

N = 4 (7%)

Reasons: 2 deaths and 2 withdrawals

Incomplete outcome data:

Intervention:

N = 5 (7%)

Reasons: 1 refusal, 4 withdrawals

Control:

N = 4 (7%)

Reasons: 2 deaths and 2 withdrawals

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I: 25 (42)

C: 27 (45)

RR 0.93 (95%CI 0.61-1.39)

Number of falls per person years, rate ratio

I1: 1.01

C: 1.49

RR 0.67 (95%CI 0.36-1.26)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 1 (2)

C: 2 (2)

“Participants were recruited from the client database of a community organisation providing support for people with visual impairments in Sydney, Australia (Guide Dogs).”

Falls were collected through calendars over 12 months. Authors stated that the study was not powered to detect an effect on fall rates.

Patil, 2015

Type of study: RCT [parallel]

Setting: Community

Country: Finland

Source of funding: Non-commercial

Inclusion criteria:

• Aged 70 to 80 years
• Fallen at least once during the previous months
• Did not use vitamin D supplements
• No contraindications to exercise

Exclusion criteria:

• Participation in moderate to vigorous exercise more than 2 hours per week.

N total at baseline:

Intervention: 205

Control: 204

Important prognostic factors²:

Age ± SD:

I: 74 (3)

C: 74 (3)

Sex:

I: 0% M

C: 0% M

Groups comparable at baseline? Yes

Exercise – Exercises classes twice a week for 12 months and once a week for the subsequent 12 months and home exercises

Control – Current physical activity

Length of follow-up:

24 months

Loss-to-follow-up:

Intervention:

N = 18 (9%)

Reasons: 5 lost interest; 13 due to health reasons

Control:

N = 21 (10%)

Reasons; 6 lost interest; 10 due to health reasons; 4 died; 1 spouse unwell

Incomplete outcome data:

Intervention:

N = 18 (9%)

Reasons: 5 lost interest; 13 due to health reasons

Control:

N = 21 (10%)

Reasons; 6 lost interest; 10 due to health reasons; 4 died; 1 spouse unwell

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 100 person years

Number of fallers, risk ratio

I: 140 (68)

C: 141 (69)

RR 0.99 (95%CI 0.78-1.25)

Number of falls per person years, rate ratio

I1: 233.8

C: 250.3

RR 1.00 (95%CI 0.79-1.26)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 8 (4)

C: 12 (6)

RR 0.66 (95%CI 0.27-1.61)

Data on falls was collected through monthly fall diaries.

Gianoudis, 2014

Type of study: RCT [parallel]

Setting: Community

Country: Australia

Source of funding: Non-commercial

Inclusion criteria:

• Aged >60 years
• Living independently in the community

Exclusion criteria:

• Aged <60 years
• BMI >40 kg/m²
• History of osteoporosis or a recent (past 6 months) low-trauma fracture
• Participated in resistance training or a structured weight-bearing exercise program more than once a week in the past three months
• Current smokers
• Any medical condition or used medication known to influence bone metabolism or fracture risk
• Had initiated calcium or vitamin D supplementation in the preceding 6 months
• Were expecting to travel for more than 6 weeks throughout the intervention
• For women: used hormone replacement therapy

N total at baseline:

Intervention: 81

Control: 81

Important prognostic factors²:

Age ± SD:

I: 67 (7)

C: 67 (6)

Sex:

I: 26% M

C: 27% M

Groups comparable at baseline? Yes

Osteo-cise: Strong Bones for Life multimodal exercise, osteoporosis education/awareness and behavioural change program

Standard care self-management control group

Length of follow-up: 12 months

Loss-to-follow-up:

Intervention:

N = 5 (6%)

Reasons: 1 death, 1 illness, 2 no time, 1 overseas

Control:

N = 7 (9%)

Reasons: 1 death; 1 illness, 5 lack of interest/other

Incomplete outcome data:

Intervention:

N = 5 (6%)

Reasons: 1 death, 1 illness, 2 no time, 1 overseas

Control:

N = 7 (9%)

Reasons: 1 death; 1 illness, 5 lack of interest/other

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I: 29 (36)

C: 25 (31)

RR 1.15 (95%CI 0.66-1.99)

Number of falls per person years, rate ratio

I: 0.57

C: 0.42

RR 1.22 (95%CI 0.72-2.04)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 1 (1)

C: 0 (0)

All participants were provided with and advised to take one vitamin D supplement and two calcium supplements each day throughout the study.

Data on falls were collected via monthly falls calendar.

Kim, 2014

Type of study: RCT [parallel]

Setting: Community

Country: Japan

Source of funding: Non-commercial

Inclusion criteria:

• Women aged 70 years or older
• Experienced at least one fall incident in the previous year
• No missing fall-related baseline data

Exclusion criteria:

• Severe knee or back pain
• Severe walking disability
• Unstable cardiac conditions

N total at baseline:

Intervention: 52

Control: 53

Important prognostic factors²:

Age ± SD:

I: 77 (4)

C: 77 (4)

Sex:

I: 0% M

C: 0% M

Groups comparable at baseline? Yes

Exercise – Group-based 60-min exercise class focusing on muscle strength and balance training twice a week for 3 months for a total of 24 sessions

Education – 60-min class once a month for three months for a total of three times. The classes focused on undernutrition, cognitive function and oral hygiene.

Length of follow-up:

1 year

Loss-to-follow-up:

Intervention:

N = 1 (2%)

Reasons: 1 hip fracture

Control:

N = 1 (2%)

Reasons: 1 withdrew due to hospitalization

Incomplete outcome data:

Intervention:

N = 1 (2%)

Reasons: 1 hip fracture

Control:

N = 1 (2%)

Reasons: 1 withdrew due to hospitalization

1. Falls

Measured as the number of fallers, n (%)

Number of fallers, risk ratio

I: 10 (20)

C: 21 (41)

RR 0.49 (95%CI 0.25-0.93)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 1 (2)

C: 2 (4)

Of all women aged 70 years or older, only 17% were included.

Fall were collected through diaries at the end of the one year follow-up.

Lliffe, 2014

Type of study: RCT [parallel]

Setting: General practices

Country: United Kingdom

Source of funding: Non-commercial

Inclusion criteria:

• Aged ≥65 years
• Independently mobile
• Physically able to take part in a group exercise class

Exclusion criteria:

• Had experienced three or more falls in the previous year
• Unstable clinical conditions
• Unable to follow instructions about exercise safely
• Receiving palliative care
• Already exercising at, or above, the target level

N total at baseline:

Intervention1: 411

Intervention2: 387

Control: 458

Important prognostic factors²:

Only reported for the entire population together

Age (range): 73 (65-94)

Sex: 38% M

Groups comparable at baseline? Cannot be judged.

1) Home-based Otago Exercise Programme (OEP) – 30 minutes of leg muscle strengthening and balance retraining exercises at home at least three times per week, and a walking plan for up to 30 minutes at a moderate pace two times per week for 24 weeks.

2) Community-based group exercise programme- 1-hour-long postural stability instructor-delivered group exercise class and two 30-minute home exercise sessions per week for 24 weeks.

Usual care – participants were free to participate in any other non-trial-related exercise

Length of follow-up:

18 months (6 month intervention and 12 follow-up)

Loss-to-follow-up:

Intervention1:

N = 158 (38%)

Reasons: 27 at allocation; 13 GP excluded; 3 died; 43 through illness; 15 research burden; 17 other; 40 not known

Intervention2:

N = 147 (38%)

Reasons: 33 at allocation; 32 GP excluded; 3 died; 29 through illness; 12 research burden; 20 other; 18 not known

Control:

N = 188 (41%)

Reasons: 38 at allocation; 24 GP excluded; 4 died; 26 through illness; 30 research burden; 10 other; 56 not known

Incomplete outcome data:

Intervention1:

N = 158 (38%)

Reasons: 27 at allocation; 13 GP excluded; 3 died; 43 through illness; 15 research burden; 17 other; 40 not known

Intervention2:

N = 147 (38%)

Reasons: 33 at allocation; 32 GP excluded; 3 died; 29 through illness; 12 research burden; 20 other; 18 not known

Control:

N = 188 (41%)

Reasons: 38 at allocation; 24 GP excluded; 4 died; 26 through illness; 30 research burden; 10 other; 56 not known

1. Falls

Measured as the rate of falls per 365 patient days

Number of falls per person years, rate ratio

During intervention

I1: 0.80

I2: 0.81

C: 0.87

I1 vs C: RR 0.93 (95%CI 0.64-1.37)

I2 vs C: RR 0.91 (95%CI 0.54-1.52)

Post intervention

I1: 0.54

I2: 0.57

C: 0.71

I1 vs C: RR 0.76 (95%CI 0.53-1.09)

I2 vs C: RR 0.74 (95%CI 0.55-0.99)

2. Fractures

Not reported

Randomisation took place at the general practices level.

Statistical analyses were adjusted for effects of size, size of practice, deprivation of practice, and clustering due to practice

Suttanon, 2013

Type of study: RCT [parallel]

Setting: Community

Country: Australia

Source of funding: Non-commercial

Inclusion criteria:

• Diagnosis of Alzheimer’s disease confirmed from specialist or Memory Clinic assessment
• Alzheimer’s disease symptoms were of mild to moderate severity
• Could walk outdoors with nor more support than a single-pint stick.

Exclusion criteria:

N total at baseline:

Intervention: 19

Control: 21

Important prognostic factors²:

Age ± SD:

I: 83 (5)

C: 80 (6)

Sex:

I: 32% M

C: 43% M

Groups comparable at baseline? Yes

Exercise programme – standing balance and strengthening exercises and a graduated walking programme.

Participants and caregivers were encouraged to regular exercise (five days/week).

Participants in both groups continued with usual care and other activities while participating in their allocated programme.

Control programme – education and information sessions on the topic of dementia and ageing.

Designed to provide the same number of home visits and phone calls as the exercise programme.

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 8 (42%)

Reasons: 3 refused to continue; 3 moved to care facility; 1 hospitalisation; 1 passed away

Control:

N = 3 (14%)

Reasons: 2 refused to continue; 1 moved to care facility

Incomplete outcome data:

Intervention:

N = 8 (42%)

Reasons: 3 refused to continue; 3 moved to care facility; 1 hospitalisation; 1 passed away

Control:

N = 3 (14%)

Reasons: 2 refused to continue; 1 moved to care facility

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 1000 patient days

Number of fallers, risk ratio

I: 9 (47)

C: 7 (33)

RR 1.42 (95%CI: 0.66-3.06)

Number of falls per person years, rate ratio

I: 3.10

C: 2.46

RR 1.00 (95%CI 0.42-2.39), adjusted for baseline measurements

2. Fractures

Not reported

Falls were reported by participants or caregivers at each of the home visits and follow-up phone calls.

Yamada, 2013

Type of study: RCT [parallel]

Setting: Community

Country: Japan

Source of funding: Non-commercial

Inclusion criteria:

• Aged ≥65 years
• Community dwelling frail
• Certified for long-term care insurance service requirement
• No severe cognitive impairment
• Ability to walk independently (or with a cane)
• Willing to participate in group exercise classes for at least six months
• Access to transportation
• No significant hearing or vision impairment
• Had not exercised regularly in the previous 12 months

Exclusion criteria:

• Serious visual impairment
• Severe cardiac, pulmonary, or musculoskeletal disorders
• Comorbidities associated with greater risk of falls
• Use of psychotropic drugs.

N total at baseline:

Intervention: 132

Control: 132

Important prognostic factors²:

Age ± SD:

I: 76 (9)

C: 77 (8)

Sex:

I: 40% M

C: 45% M

Groups comparable at baseline? Yes

Multitarget stepping task combined twice a week for 24 weeks

30-minute physical exercise program (aerobic exercise, mild strength training and flexibility and balance exercises)

Control – indoor walking program twice a week for 24 weeks

30-minute physical exercise program (aerobic exercise, mild strength training and flexibility and balance exercises)

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 20 (15%)

Reasons: not stated

Control:

N = 14 (11%)

Reasons: not stated

Incomplete outcome data:

Intervention:

N = 20 (15%)

Reasons: not stated

Control:

N = 14 (11%)

Reasons: not stated

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I: 13 (12)

C: 39 (33)

RR 0.33 (95%CI: 0.19-0.60)

Number of falls per person years, rate ratio

I: Not reported

C: Not reported

RR 0.35 (95%CI 0.19-0.66)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 3 (3)

C: 13 (11)

RR 0.22 (95%CI: 0.06-0.80)

No participant in either group received any specific instructions regarding activities undertaken outside the program or during the follow-up year.

Data on falls were collected through monthly diaries.

Fractures were ascertained through radiological evidence of fracture.

Clemson, 2012

Type of study: RCT [parallel]

Setting: Residents in metropolitan area

Country: Australia

Source of funding: Non-commercial

Inclusion criteria:

• Men and women aged 70 years or older
• Two or more falls or one injurious fall in the past 12 months (determined by self report)

Exclusion criteria:

• Moderate to severe cognitive problems
• No conversational English
• Inability to ambulate independently
• Neurological condition which severely influenced their gait and mobility
• Resident in a nursing home or hostel
• Any unstable or terminal illness that would preclude the planned exercises

N total at baseline:

Intervention1: 107

Intervention2: 105

Control: 105

Important prognostic factors²:

Age ± SD:

I1: 82 (4)

I2: 84 (4)

C: 83 (4)

Sex:

I1: 45% M

I2: 46% M

C: 45% M

Groups comparable at baseline? Yes

1: LIFE programme

Movements specifically prescribed to improve balance or increase strength embedded within everyday activities

2: Structured exercise

Seven exercises for balance and six for lower limb strength using ankle cuff weights; three times a week

Both interventions programs were taught over five sessions with two booster sessions and two follow-up phone calls over a six month period.

Control (gentle exercise)

12 gentle and flexibility exercises while seated, lying down, or standing while holding on.

Exercises were taught over two sessions with one booster sessions and six follow-up phone calls.

Length of follow-up: 12 months

Loss-to-follow-up:

Intervention1:

N = 8 (7%)

Reasons: 1 deceased; 1 too ill; 1 terminal illness; 2 injury or pain attributed to exercise; 2 partner need; 1 declined.

Intervention2:

N = 9 (9%)

Reasons: 3 deceased; 2 too ill; 1 terminal illness; 2 injury or pain attributed to exercise; 1 declined

Control:

N = 14 (13%)

Reasons: 3 deceased; 5 too ill; 2 terminal illness; 1 partner need; 3 declined

Incomplete outcome data:

Intervention1:

N = 8 (7%)

Reasons: 1 deceased; 1 too ill; 1 terminal illness; 2 injury or pain attributed to exercise; 2 partner need; 1 declined.

Intervention2:

N = 9 (9%)

Reasons: 3 deceased; 2 too ill; 1 terminal illness; 2 injury or pain attributed to exercise; 1 declined

Control:

N = 14 (13%)

Reasons: 3 deceased; 5 too ill; 2 terminal illness; 1 partner need; 3 declined

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I1: 60

I2: 65

C: 71

I1 vs C: RR 0.83 (95%CI 0.67-1.03)

I2 vs C: RR 0.92 (95%CI 0.75-1.12)

Number of falls per person years, rate ratio

I1: 1.66

I2: 1.90

C: 2.28

I1 vs C: RR 0.69 (95%CI 0.48-0.99)

I2 vs C: RR 0.81 (95%CI 0.56-1.17)

2. Fractures

Not reported

Potential limitations discussed by authors:

The control group received less contact time than both interventions, which could have caused a bias, but we saw no difference in the return rates of fall surveillance diaries, so this is unlikely.

The control group had an intervention that could have diluted the effect of the outcomes. Since the control exercises were gentle, flexible, mostly non-weightbearing, and not upgraded by the therapists, their effect on fall reduction or balance would have been marginal, although we did observe some minimal strength improvements.

Freiberger, 2012

Type of study: RCT [parallel]

Setting: Community

Country: Germany

Source of funding: Commercial

Inclusion criteria:

• Aged 70 and older
• Fallen in the past 6 months or fear of falling
• Completed a baseline assessment
• Membership to a health insurance company

Exclusion criteria:

• Unable to ambulate independently
• Had cognitive impairment

N total at baseline:

Intervention1: 63

Intervention2: 64

Intervention3: 73

Control: 80

Important prognostic factors²:

Age ± SD:

I1: 76 (4)

I2: 75 (4)

I3: 75 (4)

C: 76 (4)

Sex:

I1: 52% M

I2: 64% M

I3: 56% M

C: 54% M

Groups comparable at baseline? Yes

1) Strength and balance

2) Fitness (endurance training)

3) Multifaceted (fall risk education)

All lasting 16 weeks and including two 1-hour sessions per week. All interventions included strength and balance exercises.

No intervention

Length of follow-up: 24 months

Loss-to-follow-up:

Intervention1:

N = 14 (22%)

Reasons: Not stated

Intervention2:

N = 16 (25%)

Reasons: Not stated Intervention3:

N= 15 (18%)

Reasons: Not stated

Control:

N = 28 (35%)

Reasons: Not stated

Incomplete outcome data:

Intervention1:

N = 14 (22%)

Reasons: Not stated

Intervention2:

N = 16 (25%)

Reasons: Not stated Intervention3:

N= 15 (18%)

Reasons: Not stated

Control:

N = 28 (35%)

Reasons: Not stated

1. Falls

Measured as the number of falls

Number of falls, rate ratio

I1: 74

I2: 51

I3: 90

C: 82

I1 vs C: RR 0.97 (95%CI 0.58-1.62)

I2 vs C: RR 0.68 (95%CI 0.40-1.16)

I3 vs C: RR 0.94 (95%CI: 0.58-1.53)

2. Fractures

Not reported

Participants were recruited from a health insurance company membership database, which includes approximately 25% of the population.

Data on falls were collected monthly via fall calendars.

Taylor, 2012

Type of study: RCT [parallel]

Setting: Community

Country: New Zealand

Source of funding: Commercial

Inclusion criteria:

• Aged 65 years and older (55 years if Maori or Pacific Islander to account for Ethnic disparities in health)
• Had experienced at least one fall in the previous 12 month or considered to be at risk of falling.

Exclusion criteria:

• Unable to ambulate independently
• Had a chronic medical condition
• Had severe cognitive limitations
• Participated in tai chi within the last year
• Currently par

N total at baseline:

Intervention1: 233

Intervention2: 220

Control: 231

Important prognostic factors²:

Age ± SD:

I1: 75 (7)

I2: 74 (6)

C: 73 (6)

Sex:

I1: 31% M

I2: 25% M

C: 24% M

Groups comparable at baseline? Yes

1) Tai chi once a week

2) Tai chi twice a week

for 20 weeks in a group setting

Low-level exercise program for 20 weeks

Length of follow-up:

intervention for 20 weeks followed by 12 months follow-up

Loss-to-follow-up:

Intervention1:

N = 53 (23%)

Reasons: 42 partially returned, 11 no returns of the calendars

Intervention2:

N = 46 (21%)

Reasons: 36 partially returned, 10 no returns of the calendars

Control:

N = 57 (25%)

Reasons: 40 partially returned, 17 no returns of the calendars

Incomplete outcome data:

Intervention:

N (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I1: 132 (60)

I2: 111 (53)

C: 140 (65)

I1 vs C: RR 0.83 (95%CI 0.67-1.03)

I2 vs C: RR 0.92 (95%CI 0.75-1.12)

Number of falls per person years, rate ratio

I1: 1.55 (1.23-1.97)

I2: 1.16 (0.92-1.48)

C: 1.38 (1.24-1.53)

I1 vs C: RR 1.13 (95%CI 0.87-1.45)

I2 vs C: RR 0.84 (95%CI 0.64-1.11)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 1 (1)

C: 0 (0)

Falls were collected through monthly falls calendars.

Fluid or nutrition therapy

Suominen, 2015

Type of study: RCT [parallel]

Setting: Persons with Alzheimer disease living with a spouse

Country: Finland

Source of funding: Commercial and industry

Inclusion criteria:

• Person with Alzheimer disease living together with a spouse
• Age over 64 years
• Ability to reach the study place by taxi
• Ability to stand on a scale
• Residency in the Helsinki metropolitan area
• Absence of terminal disease, an estimated life expectancy of at least half a year

N total at baseline:

Intervention: 50

Control: 44

Important prognostic factors²:

Reported for analysed population (I: 40 and C: 38 participants)

Age ± SD:

I: 78 (6)

C: 76 (6)

Sex:

I: 53% M

C: 47% M

Groups comparable at baseline? Yes

Tailored nutritional guidance with home visits during one year

Written guide about nutrition in older adults and all community-provided normal care

Length of follow-up: 12 months

Loss-to-follow-up:

Intervention:

N = 10 (20%)

Reasons: 1 moved to another city; 3 long-term care; 3 died; 3 food records not received

Control:

N = 11 (22%)

Reasons: 4 refusals; 1 long-term care; 1 died; 5 food records not received

Incomplete outcome data:

Intervention:

N = 10 (20%)

Reasons: 1 moved to another city; 3 long-term care; 3 died; 3 food records not received

Control:

N = 11 (22%)

Reasons: 4 refusals; 1 long-term care; 1 died; 5 food records not received

1. Falls

Measured as the rate of falls per 365 patient days

Number of falls per person year, rate ratio

I: 0.55 (95%CI: 0.34-0.83)

C: 1.39 (95%CI: 1.04-1.82)

RR 3.74 (95%CI 2.16-6.46)*, adjusted for age, sex and MMSE

2. Fractures

Not reported

*Calculated with the intervention group as reference. This results could because of adjustment for age, sex and MMSE not be included in the meta-analysis.

Authors did not stated how data on falls was collected.

Vitamin D

Uusi-Rasi, 2015

Type of study: RCT [parallel]

Setting: home-dwelling women

Country: Finland

Source of funding: Non-commercial

Inclusion criteria:

• Aged 70 to 80 years
• Fallen at least once during the previous months
• Did not use vitamin D supplements
• No contraindications to exercise

Exclusion criteria:

• Participation in moderate to vigorous exercise more than 2 hours per week.

N total at baseline:

Intervention: 102

Control: 102

Important prognostic factors²:

Age ± SD:

I: 74 (3)

C: 74 (3)

Sex:

I: 0% M

C: 0% M

Groups comparable at baseline? Yes

Vitamin D3 (800 IU/d) for 24 months

Placebo

Length of follow-up:

24 months

Loss-to-follow-up:

Intervention:

N = 14 (14%)

Reasons: 4 lost interest; 7 health reasons; 1 spouse illness; 2 died

Control:

N = 7 (7%)

Reasons: 2 lost interest; 3 health reasons; 2 died

Incomplete outcome data:

Intervention:

N = 14 (14%)

Reasons: 4 lost interest; 7 health reasons; 1 spouse illness; 2 died

Control:

N = 7 (7%)

Reasons: 2 lost interest; 3 health reasons; 2 died

1. Falls

Measured as the rate of falls per 100 person-years

Number of falls per 100 person-years, rate ratio

I: 132.1

C: 118.2

RR 1.08 (95%CI: 0.78-1.52)

2. Fractures

Not reported

NB: same population as Patil, 2015

The number of falls was obtained from prospective fall diaries returned monthly via mail, and details of each registered fall were ascertained by a telephone call.

Environment/ assistive technology

Kamei, 2015

Type of study: RCT [parallel]

Setting: Community

Country: Japan

Source of funding: Non-commercial

Inclusion criteria:

• First time participation in the program
• Allowed by their primary physician to undergo physical exercise
• Living in their own residence
• Aged 65 years or older

Exclusion criteria:

• Low cognitive function
• Dementia
• Poor physical condition such as inability to walk by themselves

N total at baseline:

Intervention: 67

Control: 63

Important prognostic factors²:

Age ± SD:

I: 75 (7)

C: 75 (6)

Sex:

I: 16% M

C: 14% M

Groups comparable at baseline? Yes

Home Hazard modification education program (HHMP) (weekly for 2 hours for a total of four weeks)

Week 1: Education: elder’s fall and safety + exercise program

Week 2: Education: food and nutrition + exercise program

Week 3: Education and practice home hazard self-modification program (HHMP) using mock up and equipment + exercise program

Week 4: Education and practices: foot self-care + exercise program

Control program (weekly for 2 hours for a total of four weeks)

Week 1: Education: elder’s fall and safety + exercise program

Week 2: Education: food and nutrition + exercise program

Week 3: Education: aging and health + exercise program

Week 4: Education and practices: foot self-care + exercise program

Length of follow-up:

52 weeks

Loss-to-follow-up:

Intervention:

N = 11 (16%)

Reasons: Not stated

Control:

N = 9 (14%)

Reasons: Not stated

Incomplete outcome data:

Intervention:

N = 11 (16%)

Reasons: Not stated

Control:

N = 9 (14%)

Reasons: Not stated

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I: 15 (22)

C: 21 (33)

RR 0.67 (95%CI: 0.38-1.18)

Number of falls per person years, hazard ratio

I: Not reported

C: Not reported

RR 0.59 (95%CI: 0.31-1.15)

2. Fractures

Only reported for indoor falls

Falls were monitored prospectively using a daily falls calendar and self-report. Moreover, participants were interviewed twice: once at the 12 week follow-up as a midterm evaluation and then again at 52 weeks for the 1 year evaluation.

Tchalla, 2013

Type of study: Minimized trial [parallel]

Setting: Community

Country: France

Source of funding: Not stated

Inclusion criteria:

• Suffering from Alzheimer disease
• Aged 65 years and older
• Living at home

Exclusion criteria:

• Severe dementia (MMSE <10)
• Already in a falls prevention rehabilitation program

N total at baseline:

Intervention: 49

Control: 47

Important prognostic factors²:

Age ± SD:

I: 87 (7)

C: 85 (6)

Sex:

I: 23% M

C: 23% M

Groups comparable at baseline? Yes

Home-based technologies coupled with teleassistance service (a nightlight path for preventing falls at home)

All participants undertook a fall reduction program following the initial Comprehensive Gerontological Assessment

All participants undertook a fall reduction program following the initial Comprehensive Gerontological Assessment

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 4 (8%)

Reasons: 2 died; 2 admissions to institution

Control:

N = 4 (9%)

Reasons: 1 died; 3 admissions to institution

Incomplete outcome data:

Intervention:

N = 4 (8%)

Reasons: 2 died; 2 admissions to institution

Control:

N = 4 (9%)

Reasons: 1 died; 3 admissions to institution

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, odds ratio

I: 15 (22)

C: 21 (33)

OR 0.37(95%CI: 0.15-0.88) adjusted for six variables (unclear which)

Cumulative incidence of falls

I: 32.7% (95%CI: 21.2-46.6)

C: 63.8% (95%CI: 49.5-76.0)

2. Fractures

Not stated

Because of the calculated OR and cumulative incidence these results could not be pooled with the data from the other trials.

Falls were collected by GPs and caregivers.

Multifactorial interventions

Fairhall, 2014

Type of study: RCT [parallel]

Setting: Community

Country: Australia

Source of funding: Non-commercial

Inclusion criteria:

• 70 years or older
• Frail
• Did not live in a residential aged-care facility
• MMSE >18
• Life expectancy of at least 12 months

N total at baseline:

Intervention: 120

Control: 121

Important prognostic factors²:

Age ± SD:

I: 83 (6)

C: 83 (6)

Sex:

I: 33% M

C: 32% M

Groups comparable at baseline? Yes

Multifactorial, interdisciplinary intervention – Targeting frailty characteristics with an individualised home exercise programme prescribed in 10 home visits from a physiotherapist and interdisciplinary management of medical, psychological and social problems.

Participants also received usual care.

Control – Usual care available to older residents from community services and their general practitioner

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 13 (11%)

Reasons: 12 died; 1 withdrew

Control:

N = 12 (10%)

Reasons: 10 died; 2 withdrew

Incomplete outcome data:

Intervention:

N = 13 (11%)

Reasons: 12 died; 1 withdrew

Control:

N = 12 (10%)

Reasons: 10 died; 2 withdrew

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I: 72 (60)

C: 67 (55)

RR 1.08 (95%CI 0.87-1.35)

Number of falls per person, rate ratio

I: 1.54

C: 1.50

RR 1.12 (95%CI 0.78-1.63)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 13 (11)

C: 12 (10)

RR 1.09 (95%CI: 0.52-2.30)

Authors of the paper were employee at an insurance company.

Data on falls were collected through monthly calendars and follow-up telephone calls.

Moller, 2014

Type of study: RCT [parallel]

Setting: Urban and rural

Country: Sweden

Source of funding: Non-commercial

Inclusion criteria:

• 65 years or older
• Resident in the study area
• Need of help with at least two activities of daily living
• Admitted to hospital at least twice
• Had at least four outpatient contacts during the previous 12 months
• Able to communicate verbally
• No cognitive impairments

N total at baseline:

Intervention: 80

Control: 73

Important prognostic factors²:

Age ± SD:

I: 81 (6)

C: 81 (7)

Sex:

I: 35% M

C: 31% M

Groups comparable at baseline? No, less participants fell in the last 3 months

1 Case management

2 General information (e.g. exercise, nutrition, social activities, healthy system and other)

3 Specific information (participant’s individual needs, medication, and more)

4 Safety and continuity

At least one home visit per month for twelve months

Control

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 24 (30%)

Reasons: unclear

Control:

N = 21 (29%)

Reasons: unclear

Incomplete outcome data:

Intervention:

N = 24 (30%)

Reasons: unclear

Control:

N = 21 (29%)

Reasons: unclear

1. Falls

Measured as the number of fallers, n (%)

Number of fallers, risk ratio

I: 43 (54)

C: 35 (48)

RR 1.12 (95%CI 0.82-1.53)

2. Fractures

Not reported

After the pilot study, the intervention was expanded in 2008 by also employing two physiotherapists. 61 of 80 participants from the intervention group also received visitors from physiotherapists.

Data on falls was collected every three months.

Palvanen, 2014

Type of study: RCT [parallel]

Setting: Fall prevention clinics

Country: Finland

Source of funding: Non-commercial

Inclusion criteria:

• Aged 70 years or older

At least one of the following risk factors for falls and injuries:

• Problems in mobility and everyday function
• 3 or more falls during the last 12 months
• A previous fracture after the age 50
• Osteoprotic fracture (hip fracture) in a close relative (mother or father)
• Osteoporosis
• Low body weight
• Sickness or illness essentially increasing the risk for osteoporosis, falls, and fractures

Exclusion criteria:

• Inability to give informed consent
• Disabilities or illnesses preventing physical activity and training
• Inability to move
• Terminal illness (predicted lifetime less than 12 months)

N total at baseline:

Intervention: 661

Control: 653

Important prognostic factors²:

Age ± SD:

I: 77 (6)

C: 77 (6)

Sex:

I: 14% M

C: 14% M

Groups comparable at baseline? Yes

Multifactorial, individualized 12-month falls prevention program concentrating on strength and balance training, medical review and referrals, medication review, proper nutrition (calcium, vitamin D), and home hazard assessment and modification.

Individually tailored preventive measures judged necessary at the baseline assessment.

Control group – a general injury prevention brochure

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 72 (11%)

Reasons: 35 illness/sickness; 31 refusal to continue; 3 died; 3 other reason

Control:

N = 97 (15%)

Reasons: 54 illness/sickness; 29 refusal to continue; 8 died; 4 moved; 2 other reason

Incomplete outcome data:

Intervention:

N = 72 (11%)

Reasons: 35 illness/sickness; 31 refusal to continue; 3 died; 3 other reason

Control:

N = 97 (15%)

Reasons: 54 illness/sickness; 29 refusal to continue; 8 died; 4 moved; 2 other reason

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 100 person years

Number of fallers, risk ratio

I: 296 (45)

C: 349 (53)

RR 0.84 (95%CI 0.75-0.94)

Number of falls per 100 person years, rate ratio

I: 95

C: 131

RR 0.72 (95%CI 0.61-0.86)

2. Fractures

Measured as the number of participants with a fracture related to a fall, n (%)

I: 33 (5)

C: 42 (6)

RR 0.78 (95%CI: 0.50-1.21)

The Chaos Clinic professionals (who were not blinded to group allocation, as noted above) recorded the number of falls and fall-related injuries in three month intervals, by phone interview at 3 and 9 months, and at the follow-up visit at the Clinic at 6 and 12 months.

Lee, 2013

Type of study: RCT [parallel]

Setting: Community

Country: Taiwan

Source of funding: Non-commercial

Inclusion criteria:

• 65 years or older

Fulfilled 1 of the following criteria:

• Encountered recurrent falls during the previous year
• Had medical history known to carry high risk of falls
• Fell only once during the previous year, coupled with gait or balance problems ascertained by the Timed Up & Go test

Exclusion criteria:

• Unable to provide study information and comply with study procedures
• Any medical condition precluding regular exercise participation

N total at baseline:

Intervention: 313

Control: 303

Important prognostic factors²:

Age ± SD:

I: 75 (7)

C: 76 (7)

Sex:

I: 53% M

C: 56% M

Groups comparable at baseline? Yes

Intervention group – A risk-based multifactorial fall prevention program including exercise intervention, health education, and home hazards evaluation/modification, along with medication review and referral to an ophthalmologist or other specialists.

Control group – health education brochures, along with medication review and medical referrals without direct exercise training and structures health education classes.

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 22 (7%)

Reasons: not stated

Control:

N = 25 (8%)

Reasons: not stated

Incomplete outcome data:

Intervention:

N = 22 (7%)

Reasons: not stated

Control:

N = 25 (8%)

Reasons: not stated

1. Falls

Measured as the number of fallers, n (%) and the rate of falls per 365 patient days

Number of fallers, risk ratio

I: 86 (28)

C: 84 (28)

RR 0.99 (95%CI 0.77-1.28)

Number of falls per person year, rate ratio

I: 0.41

C: 0.44

RR 0.94 (95%CI 0.73-1.21)

2. Fractures

Not reported

The participants in each group were contacted monthly through phone calls or outpatient department visits, whenever available, regarding fall occurrence, fall-related circumstances, and consequences for up to 12 months. For those who were cognitively impaired or unable to express themselves clearly, their primary caregivers were interviewed concurrently to confirm the information.

Perula, 2012

Type of study: RCT [parallel]

Setting: Health centers

Country: Spain

Source of funding: Non-commercial

Inclusion criteria:

• 70 years or older residing in the community
• Ability to walk independently
• Informed consent

Exclusion criteria:

• Institutionalized, immobilized or bedridden
• Terminal disease or severe psychiatric illness present
• Contraindications to physical exercise

N total at baseline:

Intervention: 133

Control: 271

Important prognostic factors²:

Age ± SD:

I: 76 (4)

C: 76 (5)

Sex:

I: 41% M

C: 50% M

Groups comparable at baseline? Yes

Multifactorial – individual advice, information leaflet, physical exercise workshop, and home visits

Control group – a brief piece of advice at the consultation on falls prevention and an information leaflet and received the usual clinical care in their health center.

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N = 12 (9%)

Reasons: 4 voluntary departures; 3 not found in the time of assessment; 2 illnesses unrelated to falls; 2 changes of address and location; 1 died

Control:

N = 13 (5%)

Reasons: 4 discontinued interventions; 2 voluntary departures; 1 not found in the time of assessment; 2 illnesses unrelated to falls; 2 changes of address and location; 2 died

Incomplete outcome data:

Intervention:

N (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

1. Falls

Measured as the rate of falls

Number of falls over a year, rate ratio

I: 17%

C: 24%

RR 0.73 (95%CI 0.48-1.12)

2. Fractures

Not reported

Data on falls was collected through a questionaire

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Exercise

Day, 2015

“Participants were randomized by the study statistician using a computerized random number generator.

Likely. “The allocation list was e-mailed directly to the exercise program administrator who managed exercise class delivery, independent of the data collectors.”

Likely

Likely

Unlikely. “The interviewer was blind to group assignment.”

Unlikely. Trial was registered during enrolment.

Unclear. Similar number of participants refused to return fall calendar, except the reasons for refusal were not clear.

Unclear. Not stated.

Gleeson, 2015

“[…] were then block-randomized (block permutation size 1, 2 and 4) using a computer generated list from […] kept by a separate centre-based investigator who had no contact with the participants.”

Unlikely. List was kept separate from participants.

Likely

Likely

Unlikely. “All outcome assessors remained masked to group allocation for all assessments and participants were asked not to reveal their allocation status.”

Unlikely. Trial protocol prospectively registered and outcomes are reported in paper.

Unlikely. Although twice as many participants withdrew in the intervention group, the numbers are very low (4 versus 2)

Unlikely. “Data were analysed on an intention to treat basis.”

Patil, 2015

“Participants were randomly assigned to an exercise or control group.”

Unclear. Not stated

Likely

Likely

Unclear. Not stated.

Unclear. Trial was registered but this analyse was not stated.

Unlikely. Similar number of dropouts for similar reasons.

Unlikely. Analysed intention to treat

Gianoudis, 2014

“Participants were randomized to one of the two groups, stratified by sex, by an independent staff member not involved in the study using a computer generated random numbers table.”

Unclear. Not stated

Likely

Likely

Unclear. Not stated.

Unclear whether the trial was registered.

Unlikely. Although more participants from the control group, dropped out because of lack of interest, this most likely would not affect the results.

Unlikely. “All data were analysed on an intention-to-treat basis; ie, data from all participants were included in the analyses regardless of compliance.”

Kim, 2014

“[…] based on computer generated random numbers, […]”

Unlikely. “The randomization procedure was blinded, […]”

Likely

Likely

Unclear. Only stated that the investigators evaluating the effects of the exercise treatment were blind to interventions allocations.

Unclear whether the trial was registered.

Unlikely. One participant in each arm was lost to follow-up

Unclear, not stated

Llife, 2014

“[…] using computer-generated random number tables, embedded in a computer for minimisation.” Minimisation on trial centre, practice size and index of multiple deprivation.

Unlikely. “Practices were allocated to intervention or usual care, only after all participants had been recruited. The practices, their patients and the researchers undertaking baseline assessments were all blinded to allocation until this point. “

Likely

Likely

Likely. “Researchers assessing outcomes were not blinded for pragmatic reasons alone; […]”

Unclear. Trial was registered, however, this was done retrospectively.

Unlikely. Used the method described by Gillespie to evaluate this, because there is a large number of lost-to-follow-up.

Unlikely. Data were analysed intention to treat.

Suttanon, 2013

“[…], using a concealed randomization procedure. A random numbers table with group allocation was computer-generated by one of the research team who was not involved in assessments or interventions.“

Unlikely. “These allocations were placed in opaque sealed envelopes. Following, baseline assessment, the next numbered envelope was opened by another researcher who took no part in the assessments, […].”

Likely

Likely

Unclear, not stated

Unlikely. Trial was registered before start of enrolment and outcomes were reported in publication

Likely. More participants were lost to follow-up in the intention group

Unclear. Not clearly stated.

Yamada, 2013

“Randomized controlled trial” According to the trial registration stratified and block randomization were used.

Unclear, not stated

Likely

Likely

Unclear, not stated for the outcome falls

Unlikely. Trial was registered before start of enrolment and outcomes were reported in publication

Unclear, reasons for withdrawal were not reported

Unclear, not reported

Clemson, 2012

“Randomisation done by computer generated random blocks, stratified by sex and fall history and concealed by an independent secure website.”

Unlikely. “The randomisation […] and concealed by using an automated secure website that was operated by an off-site independent service.”

Likely

Likely

Unlikely. “A research assistant blinded to group allocation telephoned participants if they failed to return the calendar to ascertain whether they had fallen. “

Unlikely. Trial was registered before start of enrolment and outcomes were reported in publication

Unlikely. Small number of lost to follow-up with fairly similar reasons.

Unlikely. “We did a ‘complete case’ intention to treat analysis.”

Freiberger, 2012

“[…] were randomly assigned […]”

Unlikely. “All randomizations were concealed. (A third party not involved in the study applied a computerized random-number generator.”

Likely

Likely

Unlikely. “Outcome assessors were blinded to group assignment.”

Unclear. Trial not registered

Unclear. Number of lost to follow-up were different, unclear for what reasons.

Unlikely. “All analyses were performed according to a modified intention-to-treat design. “

Taylor, 2012

“[…], participants were randomly assigned to one of the three arms using a central, Web-based, computer-generated blocked random number system.”

Unlikely. “At the end of the baseline assessment, each participant was given a sealed opaque envelope containing group allocation details and was instructed to open the envelope after leaving the assessment venue and not to discuss the assignment with any of the assessors.”

Likely

Likely

Unlikely. “[…], assessors blinded to group allocation collected data related to any falls over the telephone.”

Likely. Trial was registered with the mention of a secondary outcome health quality of life. Results for this outcome were not reported.

Unclear. Similar number of participants dropped out but the reasons are unclear. It is only stated that the fall calendars were not returned.

Unclear, not stated

Fluid or nutrition therapy

Suominen, 2015

“[…] were randomly allocated to the intervention and control groups according to a computer-generated, blocked randomisation list. The block size was six, […]”

Unlikely. “A person unrelated to the investigation and unfamiliar with the procedure performed the randomisation.”

Likely

Likely

Unclear. Not stated how data on falls were assessed

Unclear. Authors do not state whether the trial was registered.

Unlikely. Although reasons for lost to follow-up were slightly different, these differences could most likely not explain the results on the rate of falls.

Likely. Only participants with follow-up were included.

Vitamin D

Uusi-Rasi, 2015

“[…] were randomly assigned to 1 of 4 groups using a computer-generated list based on simple randomization with random allocation sequence to ensure equal group sizes.”

Unclear. Not stated

Unlikely. Placebo was used.

Unlikely. Placebo was used.

Unlikely. “Assessors were blinded to group assignments.”

Unlikely. Trial was registered. Outcomes are reported.

Likely. Reasons for lost to follow-up different and could have influenced the results.

Unlikely. Stated that data was analysed intention to treat.

Environment/ assistive technology

Kamei, 2015

“Research assistants allocated participants randomly into either the HHMP group or the control group and without the presence of the researchers.”

Unclear. Not stated

Likely

Likely

Unclear. Not stated

Unclear. Not mentioned whether the trial was registered.

Unclear. Number of participants who withdrew was reported and less than 20% in each arm. However, the reasons for withdrawal are not stated.

Unlikely. Stated that analyses were undertaken on the basis of intention to treat.

Tchalla, 2013

“[…] involved dyanamic random allocation using a minimization (criteria: age, sex, fall history in previous 12 months and MMSE scale) […]”

Unclear. Not stated

Likely

Likely

Unclear. Not stated

Unclear. Not mentioned whether the trial was registered.

Unlikely. Few participants were lost to follow-up with similar reasons

Unclear. Not stated.

Multifactorial interventions

Fairhall, 2014

“[…] research staff randomised participants to the intervention or control group, with stratification by degree of frailty […].”

Unclear. “The group allocation schedule was generated and managed by an investigator independent of participant recruitment, using a computer-generated random number schedule with varying block sizes.”

Likely

Likely

Unclear. Not stated

Likely. Trial was registered. Falls was not defined as a primary or secondary outcome.

Unlikely. Few participants were lost to follow-up with similar reasons

Unclear. Not stated.

Moller, 2014

“[…] participants were consecutively randomized by using sealed envelopes containing information about group allocation to the intervention group or a control group.”

Unclear. “[…] participants were consecutively randomized by using sealed envelopes containing information about group allocation to the intervention group or a control group.”

Likely

Likely

Likely. “[The data collection] was not blinded due to the nature of the intervention.”

Likely. Trial was registered after completion of the trial.

Unclear. Similar proportion of participants were lost to follow-up. The flow diagram is not clear about reasons for lost to follow-up.

Unlikely. “[…] analyses according to the intention-to-treat principle using last observation carried forward as the imputation technique […]”

Palvanen, 2014

“Randomisation was stratified by gender, age group and study clinic. Within each of these eight strata, randomly varying block size of 6, 8, or 12 was used to ensure the equality of group sizes. The randomisation schedule for each stratum was generated by a statistician who was not a part of the research team.”

Unlikely. “[…] by sealed opaque envelopes. […] Group allocation remained fully concealed until opening of the envelope.”

Likely. “[…] it was not possible to blind the participants of the Chaos Clinic professionals.”

Likely. “[…] it was not possible to blind the participants of the Chaos Clinic professionals.”

Likely. “The Chaos Clinic professionals (who were not blinded to group allocation, as noted above) recorded the number of falls and fall-related injuries in three month intervals, by phone interview at 3 and 9 months, and at the follow-up visit at the Clinic at 6 and 12 months.”

Unlikely. Trial was registered with primary outcomes stated online reported in the paper.

Likely. More participants in the control group were lost due to illness/sickness. Meaning the intervention group may potentially be ‘healthier’.

Unlikely. “The data was analyzed on an intention-to-treat basis, using the data for all randomized participants. Follow-up time for falls, fallers and fall-induced injuries were calculated from the day of randomization to the end of the study period (12 months) or until participants died or withdrew the study.”

Lee, 2013

“All randomizations were concealed through a computerized random number generator or lottery drawing dependent on the regular practice pattern of each study center.”

Unclear. Not stated

Likely

Likely

Unclear not stated.

Unclear. Not mentioned whether the trial was registered.

Unclear. Reasons for lost to follow-up not stated

Unlikely. Analysed intention to treat.

Perula, 2012

“[…] randomized by cluster at 2 parallel arms in a ratio of 1:2 and stratified by type of health center (urban and rural).”

Unclear. Not stated.

Likely

Likely

Unclear not stated.

Unclear. Not mentioned whether the trial was registered.

Unlikely. Few participants were lost to follow-up for similar reasons

Unlikely. “[…], on an intention-to-treat basis.”