Intervention Review

You have free access to this content

Physical rehabilitation for older people in long-term care

  1. Tom Crocker1,
  2. Anne Forster1,2,*,
  3. John Young1,2,
  4. Lesley Brown1,
  5. Seline Ozer1,
  6. Jane Smith1,
  7. John Green1,
  8. Jo Hardy1,
  9. Eileen Burns3,
  10. Elizabeth Glidewell4,
  11. Darren C Greenwood5

Editorial Group: Cochrane Stroke Group

Published Online: 28 FEB 2013

DOI: 10.1002/14651858.CD004294.pub3


How to Cite

Crocker T, Forster A, Young J, Brown L, Ozer S, Smith J, Green J, Hardy J, Burns E, Glidewell E, Greenwood DC. Physical rehabilitation for older people in long-term care. Cochrane Database of Systematic Reviews 2013, Issue 2. Art. No.: CD004294. DOI: 10.1002/14651858.CD004294.pub3.

Author Information

  1. 1

    Bradford Institute for Health Research, Bradford Teaching Hospitals NHS Trust, Academic Unit of Elderly Care and Rehabilitation, Bradford, UK

  2. 2

    Leeds Institute of Health Sciences, University of Leeds, Academic Unit of Elderly Care and Rehabilitation, Bradford, UK

  3. 3

    Leeds Teaching Hospitals NHS Trust, Department of Elderly Care, Leeds, UK

  4. 4

    Leeds Institute of Health Sciences, University of Leeds, Academic Unit of Primary Care, Leeds, UK

  5. 5

    University of Leeds, Centre for Epidemiology and Biostatistics, Leeds, UK

*Anne Forster, Academic Unit of Elderly Care and Rehabilitation, Bradford Institute for Health Research, Bradford Teaching Hospitals NHS Trust, Temple Bank House, Bradford Royal Infirmary, Bradford, BD9 6RJ, UK. a.forster@leeds.ac.uk.

Publication History

  1. Publication Status: New search for studies and content updated (conclusions changed)
  2. Published Online: 28 FEB 2013

SEARCH

 
Characteristics of included studies [ordered by study ID]

MethodsDesign: RCT
Duration: 14 weeks
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: none


ParticipantsCountry: USA
Setting: community nursing home
Randomised: 29
% women = 90
Age: mean = 88.3 ± 5.7 years; range = not reported

Consent: assent accepted

Inclusion criteria: urinary incontinence

Exclusion criteria: cognitive = comatose, severe physical aggression; medical = life expectancy < 3 months, length of stay < 3 months

% Eligible within home: 49.6
% Eligible that participate: 45.3

Intervention: N = 15; % women = 92.9; age (mean) = 88.6 years ± 10.4
Control: N = 14; % women = 92.9; age (mean) = 88.3 years ± 5.7


InterventionsStudy aim or objective: to test whether an intervention combining increased daytime physical activity with improvement in the night-time environment improves sleep and decreases agitation in nursing-home residents

Intervention group: FIT programme, individualised intervention, session duration = n/a, number of sessions per week = maximum of 20
Exercise features: upper limb and lower limb exercises, walking/wheelchair propulsion delivered by research personnel twice hourly up to a maximum of 4 sessions per day, 5 days a week for 14 weeks
Nonexercise features: night-time program commenced in 14th week for 5 nights, reduction of noise, reducing sleep-disruptive nursing care practices, night-time incontinence care

Control group: usual care for 14 weeks, then 1 week of night-time programme


OutcomesPhysical function in ADL: 10-minute walk/wheel (time), 10-minute walk/wheel (average distance)
Agitation: agitation (daytime behavioural observation)
Physical activity: in-bed time (% of daytime behavioural observations)
Energy expenditure: physical activity (kCal/hr)
Sleeping: sleep episodes (maximum duration), sleep episodes (average duration), sleep (night-time) %, sleeping (daytime behavioural observation, % of observations asleep)
Environment (physical): night-time noise (> 60 dB in 2-minute period), night-time light


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Randomization occurred after baseline assessment"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskQuote: "Randomization occurred after baseline assessment"
Insufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskStudy was performed in a single home with an observable intervention

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskQuote: "...performed by independent evaluators..."
Details of blinding or likelihood of it being broken not reported

Incomplete outcome data (attrition bias)
All outcomes
High risk25 participants lost prior to baseline assessment because of time taken for parent study. 4 further participants lost to follow up, but unclear if already randomised and if so from which group

Selective reporting (reporting bias)Unclear riskProtocol unavailable. Some baseline assessments could have been outcome measures (e.g. OSAI)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: randomised, assessor-blind trial, matched pairs
Duration: 18 sessions over 2 months, 3-month follow-up
Method of randomisation: drawing lots - matched according to age, sex, ambulatory status, medical history, length in home; each pair randomly allocated to control or intervention by physiotherapist not involved in the exercise programme by drawing lots
Concealment of allocation: assessors blinded to allocation of participants
Intra and interrater reliability established before study
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: 31 consented; 13 dropped out because of lack of interest, medical problems, or personal reasons


ParticipantsCountry: Hong Kong
Setting: 3 private old-age homes
Randomised: 31
% women = 78
Age: mean = approximately 80 years; range = not reported
Consent: fully-informed consent

Inclusion criteria: able to understand and follow verbal instructions, ambulate independently (with or without aids), tolerate standing, and walking for at least 5 minutes

Exclusion criteria: medical= acute musculoskeletal pain, neurological signs and symptoms not under medication control, unstable medical conditions, complaint of dizziness and blurred vision leading to difficulty walking, medical conditions contraindicative to physical activity

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention: N = baseline not reported; % women = not reported; age (mean) = 79.1 years ± 8.41

Control: N = baseline not reported; % women = not reported; age (mean) = 81.0 years ± 7.45


InterventionsStudy aim or objective: to examine the effects of a short-term mobility programme on the balance and mobility of elderly residents in private old-age homes in Hong Kong
Number of experimental groups: 2
Groups: 2 exercise programmes
Group intervention delivery
Session duration: 45 minutes
Number of sessions per week: 3 (18 sessions over 2 months conducted in the home by qualified physiotherapist or 2 students)

Intervention: M programme (N = 10) = lower limb strengthening and balance training based on the overloading principle for strengthening and specificity for challenging balance in the upright position

Control: C programme (N = 8) = general light exercises performed while sitting without progression

Training session adherence: 98%
Mobility: 17.2 ± 1.4
Control: 18 ± 1.07 of 18 sessions


OutcomesPhysical function in ADL: TUG test (seconds) (Podsiadlo 1991), 4 metre walk (time seconds)
Balance: BBS


NotesFunding: not reported
Pilot study


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Subjects of each matched pair were randomly allocated...by drawing of lots"

Allocation concealment (selection bias)Unclear riskQuote: "...by drawing of lots..."

Unclear if concealed

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNot mentioned. Both participant groups received exercise interventions

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "They were blinded to the allocation of subjects to the exercise programmes"

Incomplete outcome data (attrition bias)
All outcomes
High risk31 participants consented; 13 dropped out (unclear numbers per group). Analysed as treated

Selective reporting (reporting bias)Unclear riskThe three measures were appropriate. However, may not have been exclusive. Protocol not available

Other biasLow riskNo other apparent risks of bias


MethodsDesign: randomised controlled semi-cross-over trial
Duration: 12 months (control group joined exercise group at 6 months)
Method of randomisation: after baseline assessments, randomisation was determined by computer-generated algorithm stratified by place of residence (nursing home or assisted-living)
Concealment of allocation: unclear - all participants had been promised they would receive the intervention eventually
Outcome assessor blinding: yes - all performance tests carried out by occupational therapists and physiotherapists blinded to allocation
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: 2 participants non-compliant with their assignment - 1 switched to exercise immediately and another assigned to exercise refused to attend -; some analyses included them; 13% missing measurements


ParticipantsCountry: USA
Setting: 50-bed long-term care facility comprising 32 nursing-home beds and 18 assisted-living beds
Randomised: 21
21 met criteria, and 20 consented (5 from nursing home; 15 from assisted-living)
% women = 75
Age: mean = 88 years; range = 75 to 99 years

Consent: not specified

Inclusion criteria: age > 65, residence at facility > 3 months, ability to ambulate alone (included with assistive devices or carer)

Exclusion criteria: cognitive = inability to follow 2-step command; medical = acute unstable illness (e.g. pneumonia),chronic illness (e.g. uncompensated congestive heart failure); functional = assaultive behaviour pattern, unwillingness to discontinue current physical therapy

% Eligible within home: 42
% Eligible that participate: 95.2

Intervention: N = 11; % women = 82; age (mean) = 88 years, range = 75 to 96 years

Control: N = 9; % women = 67; age (mean) = 88 years, range = 78 to 99 years


InterventionsStudy aim or objective: to determine whether a strength and flexibility programme in frail long-term care facility residents would result in improved function

Number of experimental groups: 2
Group intervention delivery
Session duration: 60 minutes
Number of sessions per week: 3
Seated: yes
Attendance records kept

Intervention: conducted by exercise physiologist in the lounge, exercises done in seated position (frailty), warm-up; upper body strengthening; lower-body strengthening; cool down, soft ankle and wrist weights (2 to 4 lbs), Thera-Bands® (resistance 2.5 to 9 lbs), weighted hand-sized balls, beach balls for kicking and throwing, weekly evaluations of progress

Control: art therapist or social worker; sessions of drawing, painting, puzzles, or cards; encouraged to continue normal activities; discouraged from joining exercise regime during the intervention


OutcomesPhysical function in ADL: Physical Performance Test, TUG test (seconds) (Podsiadlo 1991)
Balance: BBS
Cognition: MMSE


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomisation was determined by a computer-generated algorithm (permuted blocks) stratified by place of residence within the LTC facility"

Allocation concealment (selection bias)Low riskQuote: "Assignment to the study group was done by opening sealed envelopes with the random numbers supplied in sequence by the study coordinator"

Blinding of participants and personnel (performance bias)
All outcomes
High riskParticipants were not blinded

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "All the performance tests were administered by physical and occupational therapists who were blinded to the group assignments. The MMSE was administered by two trained medical students and research nurses also blinded to group membership"
FIM not administered after baseline because could not be blinded

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk13% of total repeated measures after baseline missing because of death or acute illness, but not reported which groups. ITT analysis performed

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasHigh riskQuote: "Two patients were non-compliant with their assignment; one switched to exercise immediately...results...included these patients as they were assigned...When they were eliminated from analysis, the results were slightly more positive in favour of the exercise intervention"

Evidence of contamination reported


MethodsDesign: RCT
Duration: 6-week intervention
Follow up: none
Method of randomisation: done by lottery at the same time; stratification was applied for gender, ADL dependence, and age
Concealment of allocation: yes - participants in the control group thought they were also receiving the vibration treatment
Outcome assessor blinding: yes - all functional assessments done by assessors blind to allocation
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: 3 (13%)


ParticipantsCountry: Belgium
Setting: nursing home with capacity for 102 beds
Randomised: 24
% women = 63
Age (mean) = 77.5 years ± 11.0; range = not given
Consent: informed consent

Inclusion criteria: dependent in no more than 2 of 6 ADL categories (Katz Scale)

Exclusion criteria: cognitive = cognitive dysfunction interfering with test and training procedures; medical = presence of infectious disease, insulin-dependent diabetes mellitus, endogenous osteosynthetic material, knee or hip prosthesis, pacemaker, epilepsy, musculoskeletal disorders

% Eligible within home: 33.7
% Eligible that participate: 72.7

Intervention: N = 13; men:women ratio = 5:8; age (mean) = 76.6 years ± 11.8

Control: N = 11; men:women ratio = 4:7; age (mean) = 78.6 years ± 10.4


InterventionsStudy aim or objective: to investigate the feasibility of whole body vibration in the institutionalised elderly and its impact on functional capacity and muscle performance
Number of experimental groups: 2
Individualised intervention delivery
Session duration: not reported
Number of sessions per week: 3
Seated: no
Both groups attended "two-weekly seated gymnastic sessions together with other residents of the nursing home" organised by independent physical therapists unaware of the participant's group

Targeted social interaction

Intervention: used Power Plate vibration platform, sessions 3 times a week with at least 1 day of rest between; 6 static exercises targeting lower limb muscles, exercise volume, and intensity gradually increased

Control: the same exercise regimen on the same vibration platform but machine switched off and sound produced by tape recorder


OutcomesPhysical function in ADL: TUG test (seconds) (Podsiadlo 1991)
Physical function (other): Tinetti Mobility Scale (gait and balance) (Tinetti 1986), Tinetti Test (gait)
Muscle power (anaerobic): leg extension (60 cm/second) - work, leg extension (40 cm/second) - maximal explosivity, leg extension (40 cm/second) - maximal force, leg extension (40 cm/second) - work, leg extension (60 cm/second) - maximal force, leg extension (60 cm/second) - maximal explosivity, hand grip strength (maximal), leg extension (40 cm/second) - maximal power, leg extension (60 cm/second) - maximal power
Balance: Tinetti Test - Body Balance
Flexibility: 'Sit-and-reach' test, Back Scratch test (Rikli 1999)
Adverse events (other): occurrence of complications
Feasibility and acceptability: attendance


NotesFunding: not reported, were given the loan of the vibration platform by 'Power Plate Belgium'


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomisation was done for all 24 participants together at the same moment by lottery"
Cards representing participants (stratified) assigned to intervention or control by means of lottery. Starting sequence determined by tossing a coin

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskBlinding of participants attempted - reproduced sound of vibration platform to convince control group that it was working, but participants may have felt that it was not working

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Functional performance assessment was done by a physical therapist who was unaware of group assignment of the participants"

Incomplete outcome data (attrition bias)
All outcomes
High riskThree dropouts in whole body vibration group compared with no dropouts in control group - two dropouts likely to be related to whole body vibration program

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 8 weeks
Follow up: none
Method of randomisation: random numbers table
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: none


ParticipantsCountry: USA
Setting: integrated health services of Dallas multi-level assisted-living facility
Randomised: 16
% women = 87
Age: 25% older than 90 years, mean = approximately 82 years; range = 69 to 96
Consent: fully-informed consent

Inclusion criteria: residential status, > 65 years old, ambulatory (± assistive device)

Exclusion criteria: history of heart attack/stroke within previous 6 months, unstable angina, any condition that the physician felt might be worsened by exercise

% Eligible within home: not reported
% Eligible that participate: not reported

Training group 1: N = 8; mean age = 84 years ± 9.6; range = 71 to 96 years; 6 women, 2 men

Training group 2: N = 8; mean age = 80 years ± 6.6; range = 69 to 90 years; all women


InterventionsStudy aim or objective: to evaluate the effect of a 8-week progressive functional fitness strength programme using dumbbells and ankle weights on strength, functional capability, balance, and selected psychological variables in residents in an assisted-living facility
Number of experimental groups: 2
Group intervention delivery
Session duration: 30 minutes
Number of sessions per week: 3
Seated: unclear

Intervention: both groups followed the same exercise routine, only the weights varied (see Notes), the exercise routine comprised 5 upper and 5 lower body strengthening exercises targeting the major muscle groups, using different weights of dumbbells as resistance, cadence exercises wearing ankle weights were also performed, a gerontologist specialising in exercise training for older adults led the exercise sessions, which included all of the participants in 1 large group (see Notes)

Training group 1: the dumbbell and ankle weights and number of exercise repetitions were gradually increased over the course of the study

(Control) Training group 2: Used 1 lb dumbbells throughout the study, and cadence exercises were performed wearing ankle straps without the addition of weights


OutcomesFalls: number of falls
Fear of falling: fear of falling (single item 4-point scale) (Tinetti 1990)
Physical function in ADL: steps required to walk 6 metres, Stair Climb test (1 flight of 7 steps), ADL score (Brill 1998), six-metre walk (time)
Physical function (other): sit-to-stand (fastest time to stand up)
Muscle power (anaerobic): hand grip strength, upper body strength (chest press), lower body strength (leg extension)
Balance: balance (Brill 1998)
Mood related: trait anxiety (State-Trait Anxiety Inventory), depression (Beck Depression Inventory)
Pain: ADL level of pain (Brill 1998)


NotesFunding: University of North Texas Research and Professional Development Grant
An issue arose when the participants refused to serve as controls; all wanted to participate in the strength-training programme. In addition to this, the facility would only permit use of 1 time period and 1 room. The investigators resolved these issues by combining the 2 treatment groups together. Cross-over was prevented by the exercise leader handing out the appropriate weights to each participant


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Subjects were entered into training group 1 or training group 2 through random assignment by a random number table"

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskThe two treatment groups trained together - participants may have noticed that different groups were receiving different weighted dumbbells and wearing different ankle weights

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo report of blinding of assessors

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskInterviewer not described

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo report of missing data. No reports of attrition and exclusions

Selective reporting (reporting bias)High riskPrespecified measures 'balance' and 'number of falls' not reported at 8 weeks

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering accounted for
Details: exploratory cluster RCT
Duration: 5 weeks
Follow up: yes


ParticipantsCharacterisation: residents with self-care dependency needs
Country: UK
Setting: nursing and residential homes
Randomised: 56
% women = 71%
Age details: mean (SD) for control = 82 (9.98); mean (SD) for exercise = 87 (6.99)
Statistically significant difference between groups
Inclusion criteria:

Nursing and residential homes: +5 beds with residents > 65 years with self-care dependency needs
Residents: expected to survive for more than 9 months, had reduced mobility indicated by a Barthel Activities of Daily Living Index score of 17, less than or equal to 16
Exclusion criteria: none stated
ADL status details: mean (SD) BI Score for control = 11.0 (4.19); mean (SD) BI Score for exercise = 11.1 (4.20)
Cognitive status details: MMSE, n (%): < 21: 39 (70%); 21 to 23: 9 (16%); > 24: 8 (14%)
Significant comorbidities: at least 1 confirmed stroke, n (%): 13 (23%)
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to investigate the feasibility, acceptability, and potential efficacy of group exercise for residents in care homes

2 groups:

Intervention: exercise group (N = 28)
Format: group, delivered by 2 physiotherapists in lounge area
Session length: 40 to 60 minutes, twice weekly
Interactive group exercise class including:
- warm-up and cool-down period
- flexibility: range of movement and stretching
- sitting balance: postures that progressively reduce the base of support and dynamic movements, such as reaching and throwing that perturb the body's centre of gravity
- Posture: education and practice of good posture during exercises
- Co-ordination: reaching targets and dual tasking
- Strengthening of the clinically major muscle groups
- Cardiovascular, e.g. marching on the spot (in sitting or standing)
Feedback to participants: group cohesion, peer reinforcement, social support

Control: control group (N = 28)
Usual care (no provision of regular physiotherapy or exercise training)


OutcomesPhysical function in ADL: RMI (Collen 1991)
Mood related: depression subscore, (HADS-D) (Zigmond 1983)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomization was performed...using computer-generated random numbers"

Allocation concealment (selection bias)Low riskQuote: "Randomization was performed by an independent principal statistician from Birmingham Clinical Trials Unit"

Blinding of participants and personnel (performance bias)
All outcomes
High riskPersonnel designed and delivered the intervention to exercise, but control received usual care, so personnel knew. No report of blinding of participants - although, intervention and control groups were in separate homes, so may not have been aware of which group they were in

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "Assessments were conducted...by one of two research staff... masked to group allocation"

Incomplete outcome data (attrition bias)
All outcomes
Low riskSame number of losses to follow up in each group

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: Cluster RCT
Duration: 9 weeks; 2-week baseline, 5-week intervention for experimental group 1 followed by 2-week intervention for experimental group 2
Follow up: none
Method of randomisation: coin toss
Concealment of allocation: unclear
Outcome assessor blinding: unclear - pre-test and post-test data collected by the same person for consistency
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: none reported for phase 1; of group B, 21 did not enter the second phase


ParticipantsCountry: USA
Setting: 2 rural public nursing homes; home A had 98 beds; home B had 100
Randomised: 66
% women = 82
Age: mean = approximately 82 years; range = 60 to 96 years
Consent: assent accepted

Inclusion criteria: their current health status did not preclude participation, aged = 60, could speak and understand English, cognitively comprehend and answer questions, communicate verbally or in writing, were willing to participate in indoor gardening activities for 6 weeks

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported

Experimental group 1: N = 33, men = 6
Experimental group 2: N = 33, men = 6
Experimental group 3: N = 12, men = 6


InterventionsStudy aim or objective: the effects of indoor gardening on socialisation, activities of daily living, and perceptions of loneliness
Number of experimental groups: 3
Group intervention delivery
Session duration: 20 minutes
Number of sessions per week: 2
Seated: yes

Phase 1: residents in home A comprised experimental group 1 and participated in an indoor gardening project once a week for 5 weeks; home B, the control group, received 20-minute visits over the same 5-week period

Phase 2: residents of home B became experimental group 2 and participated in indoor gardening twice a week for 2 weeks

Intervention: decorating flowerpots and planting bulbs of their choice, choosing and transplanting colourful flowering plants, discussing proper care of plants, viewing video on gardening, arranging plants in a hanging basket, arranging fresh cut flowers and greenery

Control: 20-minute visits during the 5-week intervention period to control for social interaction and changes due to the presence of experimenters; control group then invited to participate in the gardening (phase 2)


OutcomesPhysical function in ADL: MDS: transfer item (Brown 2004), MDS: eating item (Brown 2004), MDS: locomotion item (Brown 2004), MDS: grooming item (Brown 2004), MDS: dressing item (Brown 2004), MDS: bathing item (Brown 2004), MDS: Physical Functioning scale (6 items, Brown 2004)
Mood related: UCLA Loneliness Scale (version 3)
Social support: Revised Social Provisions Scale


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "A coin toss was used to determine the assignment of each nursing home"

Allocation concealment (selection bias)High riskCoin toss - allocation could have been foreseen by researchers

Blinding of participants and personnel (performance bias)
All outcomes
High riskUnblinded participants and personnel

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskOutcome assessor unreported

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo report of loss to follow up or otherwise

Selective reporting (reporting bias)High riskOnly some sub-scales reported for some comparisons

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: none
Method of randomisation: not specified
Concealment of allocation: not specified
Outcome assessor blinding: not specified
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: 18 (46%)


ParticipantsCountry: Denmark
Setting: nursing homes
Randomised: 39
% women = 99
Age: mean = approximately 89 years; range = 85 to 95 years
Consent: not specified

Inclusion criteria: see exclusion criteria

Exclusion criteria: cognitive = moderate/severe cognitive impairment; medical = acute illness, hypertension, severe cardiovascular disease, severe impairment of motor function, neurological disorder

% Eligible within home: not reported
% Eligible that participate: 53.8

Intervention: N = 10; men:women ratio = 1:9; age = 88.6 years (86 to 95 years)

Control: N = 11; men:women ratio = 1:10; age = 90.6 years (86 to 95 years)


InterventionsStudy aim or objective: to test the hypothesis that physical exercise induces an anti-inflammatory response that is associated with reduced chronic activation of the tumour necrosis factor (TNF)-alpha system in frail elders and that the increase in muscle strength after resistance training is limited by systemic low-grade inflammation

Number of experimental groups: 2
Unclear whether intervention delivery was group or individual
Session duration: 45 minutes
Number of sessions per week: 3
Seated: yes

Exercise features: training protocol from Harridge 1999. 3 exercise sessions a week for 12 weeks, low repetitions with high weight resistance, seated upright in training chair. 3 sets of 8 knee extensions

Non-exercise features: subgroup of participants gave blood samples for examining inflammatory marker

Control: occupational therapist supervised social activities twice a week for 12 weeks; no physical training


OutcomesMuscle power (anaerobic): knee flexor muscle strength, knee extensor muscle strength
Physiology: plasma levels of cytokines


NotesFunding: Danish Medical Research Council

NOVO Foundation


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Participants were randomly assigned..."
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo report of blinding of participants, but social control intervention with physical outcome measures, so intervention would have been obvious

Incomplete outcome data (attrition bias)
All outcomes
High riskExcluded results of the two men from the analysis

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 6 weeks
Follow up: none
Method of randomisation: not specified
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: no, significant group differences, P < 0.05

Treatment group significantly older than control group (P = 0.03)
Treatment group had significantly better TUG scores at baseline (P = 0.04)

Losses to follow up: 6 (14%)


ParticipantsCountry: Belgium
Setting: nursing home in Liège, Belgium

Randomised: 42
% women = 73
Age: mean = 81.9 ± 6.9 years; range= 63 to 98 years

Consent: not specified

Inclusion criteria: ambulatory, no major cognitive disorders that would effect their ability to complete questionnaires

Exclusion criteria: medical = people with a high risk or thromboembolism, history of hip or knee replacement

% Eligible within home: not reported
% Eligible that participated: 87.5

Intervention (vibration therapy plus physiotherapy): N = 22; % women = 81; age: mean: 83.6 years ± 4.8 years

Control (physiotherapy alone): N = 20; % women = 65; age(mean) = 78.9 years ± 6.9 years


InterventionsStudy aim or objective: to investigate the effects of whole body vibration in the elderly
Number of experimental groups: 2
Individual intervention delivery
Session duration: 10 minutes
Number of sessions per week: 3
Seated: no

Groups: randomised to receive vibration intervention plus a standard physical training regimen or physical training alone

Exercise features:

Intervention – controlled whole body vibration: at each session stood on vertical vibrating platform for 4 series of 1 minute of vibration alternating with 90 seconds of rest, vibration set at 10 Hz for the first and third series with peak to peak amplitude of 3 mm; for second and fourth series, vibration set at 26 Hz with peak to peak 7 mm, blood pressure and pulse were taken before the first series, immediately after the second and fourth series, and 2 minutes after the fourth series in each session

Physical therapy: standard exercise programme, gait and balance exercises, training in transfer skill, strengthening exercises with resistive mobilisation of lower limbs, 3 times weekly for 10 minutes during the 6-week study, provided by only 1 physical therapist


OutcomesPhysical function in ADL: TUG test (seconds) (Podsiadlo 1991)
Physical function (other): Tinetti Mobility Scale (gait and balance) (Tinetti 1986), Tinetti Test (gait), SF-36 physical function
Balance: Tinetti Test - Body Balance
Perceived health status: SF-36 vitality, SF-36 social function, SF-36 Role-physical, SF-36 Role-emotional, SF-36 mental health, SF-36 health change, SF-36 general health
Pain: SF-36 pain


NotesFunding: not reported; no commercial party had any financial interests


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo evidence of blinding. The potential influence of the additional treatment in the intervention group, and outcome expectations of the intervention provider could have influenced participant response

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo information provided

Incomplete outcome data (attrition bias)
All outcomes
High riskLost to follow up from intervention (27%); none lost from control. ITT analysis undertaken utilising last available data

Selective reporting (reporting bias)High riskSome results were reported as ITT analysis, others as per-protocol (possible selective reporting)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 30 weeks
Method of randomisation: name draw
Concealment of allocation: unclear
Outcome assessor blinding: yes - evaluators blind to participant assignment
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 14 (21%); 12 deaths - 2 became unstable on their medications, does not report which group(s) they were from


ParticipantsCountry: USA
Setting: nursing home
Randomised: N = 66
% women = 88
Age: mean = 86.2 years; range = 54 to 100 years
MMSE score: range = 0 to 19; mean score = 7.5

Consent: assent accepted

Inclusion criteria: diagnosis of dementia, family consent, stable on medications, resident in the home for 3 months

Exclusion criteria: medical: use of tacrine, a drug used in the treatment of Alzheimer's disease (centrally acting anticholinesterase)

% Eligible within home: not reported
% Eligible that participate: not reported


InterventionsStudy aim or objective: to assess the impact of a highly structured interdisciplinary programme of sensorimotor activities on the function and behaviour of nursing-home residents with dementia

Number of experimental groups: 2
Both small groups and personalised interventions
Session duration: n/a
Number of sessions per week: n/a
Seated: unclear

Intervention: first 10-week period: Intervention provided by certified therapeutic recreation specialists in collaboration with the unit managers; design based on level of functioning, personal care schedule, and interests; small group activities among people of similar functioning; co-ordinated schedule of care established for the treatment group including all aspects of care and therapeutic programming; staff were encouraged to walk with residents, interact socially, and promote functional independence during activities; intervention participants received therapeutic programming and diversional stimulatory activities throughout the day and evening; every aspect of the day considered programming and outcome-based – all sensory motor activities, no matter how mundane (e.g. hand washing, waking to meals); cooking, herb gardening, group cognitive therapy, fitness sessions, various sensory (water, relaxation activities); second 10-week period: home staff took over 50% of the programming; third 10-week period: nursing-home staff took over all aspects of the programming

Control: usual care: same schedule of regular nursing-home activities and standard nursing care


OutcomesPhysical function in ADL: timed walk over 50 feet
Physical function (other): Timed Manual Performance (TMP) test
Muscle power (anaerobic): grip strength (lbs)
Mood related: GDS
Agitation: Cohen-Mansfield Agitation Inventory
Cognition: MMSE
Flexibility: 'Sit-and-Reach' test (Modified Wells)
Psychosocial and physical functioning: Multidimensional Observation Scale for Elderly Subjects (MOSES)


NotesFunding: National Alzheimer's Association
Evaluators were unit managers; participants came from the same home, and as a result, were probably aware of group allocation during the 30-week period


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "...by name draw without replacement"
Comment: N = 33 in each group. Assume drawn names were allocated alternately to each group

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo blinding of participants

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskAll evaluators blind to group assignment

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskAll evaluators blind to group assignment

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk12 participants died, and 2 were not stable on medications - data from these 14 eliminated from final data analysis

Selective reporting (reporting bias)High riskMOSES outcomes not reported

Other biasUnclear riskRisk of contamination: randomisation of individual residents, but intervention involved some staff training, so possibility of confounding


MethodsDesign: RCT
Duration: 12 weeks
Follow up: no


ParticipantsCharacterisation: visually-impaired elderly
Country: Hong Kong
Setting: care and attention homes
Randomised: 50
% women = 100%
Age details (mean (SD)): experimental group = 83 (4.7) years; control group = 84.4 (6.5) years
Inclusion criteria: visually-impaired people (people with no light perception or with visual acuity of 6/120 or worse on the better eye with corrective device), aged 65 years or older. All participants were independently mobile
Exclusion criteria: those who suffered from painful conditions that affected their mobility or balance, neurological disorders, musculoskeletal problems, cardiovascular disease, unstable blood pressure associated with posture or mental conditions, which limited them from following the exercise instructions
ADL status details: There were no significant differences between the training and control groups in age (P = 0.377) and pre-training scores of the BBS, CST, and TUG
Cognitive status details: excluded neurological disorders that limited following instructions
Significant comorbidities: not reported
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to examine the effects of an exercise programme, which focused on improvement of the functional balance of visually-impaired elderly people
2 groups:

Intervention: exercise training (experimental) (N = 27)
Format: group; delivered by: designed and conducted by 2 physiotherapists
Session length: 45 minutes, 3 times weekly
Protocol included:
(1) warm up - range of motion and stretching exercises for the upper limbs in a sitting position, lower limb warm-up exercise including quadriceps and calf stretching, and ankle circumduction in a standing position
(2) lower limbs strengthening exercises - chair stand exercise (sets of 5 repetitions progressing to 10 repetitions), quadriceps strengthening in a sitting position, strengthening of hip extensors and abductors in a standing position, with cuff weights in 3 sets of 10 repetitions with progressive weights (e.g. 10 repetitions with 3 lbs, then 5 lbs, followed by 7 lbs, based on the capacity of the individual participant)
(3) balance exercises - supervised stool-stepping exercise, tandem standing, and single-leg standing
(4) cool down exercises - general stretching and mobility exercise
Balance exercise was progressed based on the participant's needs and according to the ability of the participant

Control: control (N = 23)
General exercise (upper limb and lower limb mobilization exercises using shoulder pulley and floor bike/static bike) per week


OutcomesPhysical function in ADL: TUG (Timed Up and Go) test (seconds) (Podsiadlo 1991)
Physical function (other): sit-to-stand (up and down 5 times)
Balance: BBS

Falls, risk and fear of falling: falls (any episodes for participant)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on random sequence generation

Allocation concealment (selection bias)Low riskQuote: "Participants were randomly assigned...by drawing from a sealed opaque envelope that contained the number that determined the allocation"

Blinding of participants and personnel (performance bias)
All outcomes
High riskBoth control and experimental exercise programmes were designed and conducted by two physiotherapists

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Assessment of the functional status of the subjects was conducted by a third physiotherapist who was blinded to the grouping of the subjects"

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo losses to follow up in either group

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 6 months
Follow up: no


ParticipantsCharacterisation: older adults
Country: Netherlands
Setting: long-term care facilities, i.e. homes for the aged with services ranging from independent living to skilled nursing
Randomised: 224
% women = 80%
Age details: mean (SD) = 81.7(5.4); range = 64 to 94 years
Inclusion criteria:

  1. aged 65 or older
  2. living in a nursing home or residential care facility
  3. able to walk six metres or more (with or without a walking aid)
  4. able to comprehend the study procedures
  5. no medical contraindication for study participation
  6. no rapidly progressive or terminal illness
  7. not moving away from the home within the 6-month intervention period


(5 and 6 were evaluated by their general practitioner)
Exclusion criteria: ADL status details: number of participants with ADL disability: 27
Number of participants using a walking aid indoors: 23
Cognitive status details: not reported
Significant comorbidities: not reported
Assessed: not reported
Excluded: 17 on eligibility criteria, 5 other


InterventionsStudy aim or objective: to examine the effect of different training protocols on quality of life, vitality, and depression of older adults living in long-term care facilities
4 groups:

Intervention: combined training (N = 56)
Format: group; delivered by: trained physical therapist and assistant
Session length: 40 to 60 minutes, twice weekly
A complete strength training session (2 x 12 reps of 5 exercises, once weekly) and a complete functional training session (once weekly)

Strength: resistance increased until 2 sets of 8 to 12 reps possible. Progressively increased when 2 x 12 reps achieved at 2 sessions. Warm-up of 10 to 20 reps at minimal resistance. The 5 exercises were leg press; latissimus pull down; biceps curl and triceps press on TechnoGym equipment; and heel raises with dumbbells (1 to 5 kg each), or ankle or wrist weights (1 and 2 kg per pair), or some combination of these. For the heel raises, the number of repetitions were increased if the participants could lift the maximum weight (2 × 5 kg dumbbells + 2 × 2 kg ankle weights)
Functional: warm-up of walking and movement to music, followed by 30 to 35 minutes of skills training in game-like and co-operative activities, such as throwing and catching a ball while standing up and sitting down on a chair, musical chairs, and team pursuit races

Control 1 (of 3): functional-skills training (N = 60)
Program designed to improve muscle strength, speed, endurance, co-ordination and flexibility to improve functional performance of common daily activities. An emphasis was placed on skills training, meaning that the specific activities required for independence in daily activities were practiced. Classes started with 5 to 10 minutes of warm-up activities: walking (whenever possible), exercise-to-music routines, becoming familiar with the equipment, followed by 30 to 35 minutes of skills training in game-like and co-operative activities, such as throwing and catching a ball while standing up and sitting down on a chair, musical chairs, and team pursuit races. The cool-down period (5 to 10 minutes) consisted of stretching and relaxation activities (e.g. finger and wrist rolls, shoulder rolls, reaching, leg stretches). Exercises adjusted to individual mobility level. The intensity was gradually increased: the number of repetitions increased, exercises were performed more often standing up straight, and the use of wrist and ankle weights (1 and 2 kg per pair) was stimulated

Control 2 (of 3): strength training (N = 57)
Program designed to improve muscle strength of major muscle groups of both upper and lower body, important for functional performance on common daily activities. 5 exercises: leg press, latissimus pull down, biceps curl and triceps press on TechnoGym equipment, and heel raises with dumbbells (1 to 5 kg each); ankle or wrist weights, or both (1 and 2 kg per pair). In the first 2 weeks, participants were familiarised with the equipment and the technique of the exercises by exercising with minimal resistance. The following weeks, resistance increased until 2 sets of 8 to 12 repetitions were possible. Resistance was to be increased after the participant could complete 2 sets of 12 repetitions for 2 consecutive sessions. As a warm-up, each exercise was first performed 10 to 20 repetitions with minimal resistance. Progression was monitored with exercise logs filled out by the supervising physical therapist and assistant. Sessions closed with stretching exercises

Control 3 (of 3): control (educational) (N = 51)
An 'educational' program designed to provide attention and social interaction (i.e. group discussions about topics of interest to older people, such as history of the 20th century, music, relaxation, etc)


OutcomesPhysical function in ADL: walking speed over 8 metres, disability in 17 ADLs (Chin A Paw 2006), putting on and off a coat, picking up a pen from the floor while standing
Physical function (other): eye-hand co-ordination (block-transfer test for measuring manual dexterity), reaction time, sit-to-stand (up and down 5 times)
Muscle power (anaerobic): ankle dorsiflexor strength, elbow extensor strength
Mood related: GDS
Flexibility: 'Sit-and-Reach' test
Physical activity: physical activity (accelerometer data), LASA Physical Activity Questionnaire (Stel 2004)
Quality of life: Dementia Quality of Life Instrument (Brod 1999)
Perceived health status: Vitality Plus Scale, VPS (Myers 1999)
Continence: laxatives (number of participants using), constipation (number of participants with)
Anthropometry: body fat (%), BMI


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "The random allocation sequence was generated by computer"

Allocation concealment (selection bias)Low riskTwo independent students assigned participants to their group - implies allocation was concealed

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo mention of blinding of participants, but there were four different groups so participants may not have been aware which ones were experimental and which was the control

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Data were collected at baseline and after 6 months intervention by three trained research assistants who were blinded to group assignment"

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "Data were collected at baseline and after 6 months intervention by three trained research assistants who were blinded to group assignment"

Incomplete outcome data (attrition bias)
All outcomes
Low riskQuote: "Dropout of participants was not significantly different among the four groups".

Similar reasons for dropout between groups

Selective reporting (reporting bias)Unclear riskStated that they were going to report body composition measurements - only referred to in text in associated paper (Chin A Paw 2006)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: Cluster RCT
Duration: 12 weeks
Follow up: none
Method of randomisation: randomly assigned to either the experimental or control group by coin tossing, cluster allocation of the 2 homes
Concealment of allocation: no
Outcome assessor blinding: not possible because of the non-random participant assignment (homes were allocated)
Group comparability at entry: no, significant differences, P < 0.05

Strength of ankle dorsiflexors (control group stronger)

Balance (control group had better balance both with eyes open and eyes closed)

Mobility (control group more mobile)

Losses to follow up: 9 (13.2%)


ParticipantsCountry: South Korea
Setting: unclear - based on the number of residents, location, and facilities; 2 facilities with similar characteristics were selected from list provided by Korean Council on social welfare
Randomised: 68
% women = 75
Age: mean = 77.86 years; range = 61 to 91 years
61% fall in the previous year

Consent: assent accepted

Inclusion criteria: ambulatory adults aged over 60 with at least 1 of the following: impaired gait (score of < 10 on gait sub-scale (maximum score of 12) of the Performance Orientated Assessment of Mobility (POAM)), impaired balance (score of < 14 on POAM balance sub-scale (maximum 16), history of falling in the previous year, postural hypotension (drop in systolic blood pressure of 20 mmHg from lying to standing, use of 4 or more prescription medications that may affect balance

Exclusion criteria: cognitive = severe dementia (score < 20 on Folstein MMSE); medical = inability to complete 12 weeks of exercise because of physical illness; functional = current involvement in any type of regular exercise

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention: N = 29; % women = 79; age (mean) = 76.96 years ± 7.7 years

Control: N = 30; % women = 70; age (mean) = 78.73 years ± 6.9 years


InterventionsStudy aim or objective: to determine changes in physical fitness (knee and ankle muscle strength, balance, flexibility, and mobility), fall avoidance efficacy, and fall episodes of institutionalised adults after participating in a 12-week Sun-style Tai Chi exercise programme

Number of experimental groups: 2
Group intervention delivery
Session duration: 35 minutes
Number of sessions per week: 3
Seated: unclear

Exercise features: Sun style Tai Chi, 10 minutes warming up, 20 minutes of 12 Tai Chi movements, 5 minutes of cooling down, done to music for soothing effect

Control: maintained routine activities; did not participate in any regular exercise classes


OutcomesFalls: falls (any episodes for participant)
Fear of falling: Falls Efficacy Scale (Tinetti 1990) (10 item fear of falling)
Physical function in ADL: 6-metre walk (time)
Muscle power (anaerobic): ankle plantar/flexor strength, ankle dorsiflexor strength, knee extensor muscle strength, knee flexor muscle strength
Balance: balance: time standing on 1 leg (eyes open), balance: time standing on 1 leg (eyes closed)
Flexibility: flexibility (touch toes) (Choi 2005)


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskSelected two facilities (based on number of residents, location, and facilities) and randomly assigned them to either the experimental or control group by coin tossing

Allocation concealment (selection bias)High riskCoin toss - allocation could have been foreseen by researchers

Blinding of participants and personnel (performance bias)
All outcomes
High riskQuote: "Participants were aware of their group assignment"

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskQuote: "Professional team rather than the research team measured physical fitness"
No report of blinding of assessors, although hinted at. Because of cluster assignment and non-blinding of participants, all allocations could have been revealed by 1 participant

Incomplete outcome data (attrition bias)
All outcomes
Low risk'As treated' analysis done - measures not reported for all participants initially included, but only two departed from allocation

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 6 months
Follow up: no


ParticipantsCharacterisation: institutionalised elderly people with mixed dementia
Country: Brazil
Setting: residential care
Randomised: 54
% women = 69%
Age details (mean (SD)): 74.3 (1.4)
Inclusion criteria: (a) a primary diagnosis of dementia, based on ICD-1011 Classification of Mental and Behavioral Disorders, and confirmed by the participant's performance on the MMSE and on Katz Activities Daily Living Scale; (b) no other neurological diagnosis or neuropsychiatric conditions associated to cognitive impairment; (c) medically fit in order to participate in therapy; (d) no prescriptions of antidepressant medications with central anticholinergic or sedation actions; (e) no drug-related impairment of cognition or balance; (f ) residing in a long-term psychiatric institution
Exclusion criteria: included above
ADL status details: Katz ADL means (SD), N =
5.0 (0.3), N = 17
4.6 (0.5), N = 17
4.5 (0.5), N = 20
Cognitive status details: MMSE means (SD), N =
18.7 (1.7), N = 17
12.7 (2.1), N = 17
14.6 (1.2), N = 20
Significant comorbidities: not reported
Assessed: not reported
Excluded: 8 on criteria, 1 other


InterventionsStudy aim or objective: to analyse the effects of multidisciplinary or physiotherapeutic rehabilitation interventions on the cognition and balance of institutionalised elderly people with mixed dementia
3 groups:

Intervention: Group 1 (physiotherapy + occupational therapy + physical exercise) (N = 17)
Format: group; delivered by: physiotherapy, occupational therapy, and physical exercise professionals
Session length: 120 minutes, 5 times per week
An interdisciplinary programme that consisted of physiotherapy, occupational therapy, and physical exercise
Physiotherapy: individualised kinesiotherapeutic exercises that stimulated strength, balance, and cognition using bars, Bobath balls, elastic ribbons, and proprioceptive stimulation plates
Occupational therapy: arts and crafts activities (pictures, paintings, drawings, and embroidering)
Physical exercise: walking and upper and lower limb exercises

Control 1 (of 2): Group 2 (physiotherapy) (N = 17)
Physiotherapy sessions involving the same kinesiotherapeutic exercises as detailed for group 1

Control 2 (of 2): Group 3 (control) (N = 20)
Control without motor intervention


OutcomesPhysical function in ADL: TUG test (steps) (Christofoletti 2008), TUG test (seconds) (Podsiadlo 1991)
Balance: BBS
Cognition: Brief Cognitive Screening Battery (BCSB: includes Identification/nomination, incidental memory, immediate memory, learning memory, delayed memory, clock drawing test, verbal fluency test, recognition), MMSE


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "A sealed envelope with an identification number was assigned to each subject, each one filled with a slip giving the group. When a patient was registered and given a number, the appropriate envelope was opened"

Allocation concealment (selection bias)Low riskQuote: "A sealed envelope with an identification number was assigned to each subject, each one filled with a slip giving the group. When a patient was registered and given a number, the appropriate envelope was opened"

Blinding of participants and personnel (performance bias)
All outcomes
High riskQuote: "As a common bias presented on most rehabilitation trials, it was not possible to 'blind' the subjects regarding the treatments"

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Assessors were 'masked' with respect to the data collected and to those patients that were included or not in this trial"

Incomplete outcome data (attrition bias)
All outcomes
High riskBecause of small initial group sizes (mean = 18), the loss of 5 from each intervention group compared with the loss of 3 from the control group may present potential bias

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: 4 weeks
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: not described
Group comparability at entry: unclear
Losses to follow up: 2 participants (8.7%) failed to complete the study; both in the social activity group


ParticipantsCountry: USA
Setting: 4 long-stay psychiatric wards, of which 3 were secure
Randomised: 23
% women = 52
Age: mean = 69 years; range = 50 to 77 years

Consent: assent accepted

Inclusion criteria: capable of communicating and following simple instructions

Exclusion criteria: hypertension, debilitating arthritic impairment, requiring cardiac medication

% Eligible within home: not reported
% Eligible that participate: not reported

Activity group: N = 10; 5 women

Social group: N = 6; 4 women

Control group: N = 7; 3 women


InterventionsStudy aim or objective: hypothesised that 12-week physical activity programme would (1) increase total daily activity level, (2) upgrade participant self care, and (3) increase activity tolerance levels

Number of experimental groups: 3
Group intervention delivery
Session duration: 60 minutes
Number of sessions per week: 5
Seated: no

Activity group: stretching and postural exercise, modified weight and circuit training, dancing and walking; led by a therapist trained in physical education instruction and an assistant, for 1 hour, 5 sessions per week for 12 weeks

Social group: recreational activities involving no physical exertion, e.g. board games, arts and crafts; led by a recreational therapist and an assistant, for 1 hour, 5 sessions per week for 12 weeks

Control group: usual care


OutcomesPhysical function in ADL: self-care personal neatness evaluation (NOSIE-30)
Balance: balance (beam stand), balance (one foot stand), balance (toe stand)
Endurance (physical other): heart rate
Physical activity: Total Daily Activity Level Assessment (TDAL)


NotesFunding:

National Institute of Mental Health (NIMH)

U.S. Department of Health, Education, and Welfare (DHEW) Hospital improvement grant

Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) grant


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Subjects were assigned randomly to one of three groups"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskRCT with obvious control

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo information reported

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskNo information reported

Incomplete outcome data (attrition bias)
All outcomes
Low riskQuote: "Only two subjects failed to complete the 12-week experimental period, both in the social activity group"

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 16 weeks
Follow up: none
Method of randomisation: randomly assigned to 1 of 3 groups using a table of random numbers
Concealment of allocation: yes
Outcome assessor blinding: yes, maintained
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 12 (8.7%), 5 lost from talk-only group and 7 lost from control group
Reasons stated: death and surgery; no separate data reported


ParticipantsCountry: Canada
Setting: 3 long-term care facilities
Randomised: 86
% women = 53
Age: mean = 82 years ± 8 years; range = not reported
Consent: assent accepted

Inclusion criteria: diagnosis of Alzheimer's disease, MMSE less than 20, MMSE item 8 score of less than 3, ability to walk 5 metres with or without walking aid or supervision

Exclusion criteria: medical: cardiac conditions precluding ambulation

% Eligible within home: not reported
% Eligible that participate: 81.5

Walk and talk group: N = 30; mean age = 83.23 years (SD 8.34); 16 women

Talk-only group: N = 30; mean age = 81.68 years (SD 7.36); 15 women

Control group: N = 26; mean age = 79.78 years (SD 8.30); 8 women


InterventionsStudy aim or objective: to investigate the effects of a walking/talking programme on communication, ambulation, and level of function on people with Alzheimer's disease

Number of experimental groups: 3
Group intervention delivery
Session duration: 30 minutes
Number of sessions per week: 5
Seated: no

Exercise features: walking

Non-exercise features: talking

Walk-and-talk group: to walk and talk as much as possible with rest as necessary (guided conversation), 30-minute sessions, 5 sessions per week for 16 weeks, led by a research assistant

Talk-only group: guided conversation only, 30 minutes, 5 sessions per week for 16 weeks, led by a research assistant

Control group: usual care


OutcomesPhysical function in ADL: 2-minute walk test (Cooper 1968; Cooper 1970)
Communication: Functional Assessment of Communication Skills for Adults (Frattali 2003)
Psychosocial and physical functioning: London Psychogeriatric Rating Scale (LPRS) (Hersch 1978)


NotesFunding: Alzheimer's Society of Canada


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Residents were assigned within each site to one of three groups using a table of random numbers"

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskResearch assistants were blind to the group membership of the residents when completing the measures, but not to the study design
No report of blinding of participants, but usual care so intervention would have been obvious

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskDoes not specifically report blinded assessment for this measurement. However, it does report 1 measure where they were not used, so assume blinded RAs used for measurement

 

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
High riskThe nurses caring for the residents completed the LPRS. They were not blind to group membership

Incomplete outcome data (attrition bias)
All outcomes
High riskLosses from each group significantly different P = 0.01 (talk-only = 5, control = 7, and talk-and-walk = 0)

Selective reporting (reporting bias)High riskNot all observed ADL measures (i.e. LPRS in this study) were reported

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: not described
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 3 (6%)


ParticipantsCountry: Canada
Setting: sheltered apartments, rest homes, and nursing homes
Randomised: 50
% women = 100
Age: mean = 82.2 years; range = 71 to 91 years

Consent: fully-informed consent

Inclusion criteria: ability to ambulate independently without walking aid, eyesight sufficiently good to read large new print, hearing sufficiently good to hear instructions in normal speaking voice, ability to understand instructions, and ability to participate in exercise programme

Exclusion criteria: see inclusion criteria

% Eligible within home: 69
% Eligible that participate: 43

Intervention: N = 25
Control: N = 25


InterventionsStudy aim or objective: to test the hypothesis that increase in postural sway is due to nervous system deterioration, and as a consequence, no improvement is possible – irreversible loss of function

Number of experimental groups: 2
Group intervention delivery
Session duration: 25 minutes
Number of sessions per week: 3
Seated: no

Exercise features: exercise class delivered by physiotherapists, activities conducted aim to improve breathing, single and double limb balance, co-ordination, flexibility, antigravity strength, trunk and ankle strength, and promote general relaxation

Control: usual care


OutcomesBalance: postural sway (measured on a force platform; lateral and anteroposterior sway measured with eyes open and eyes closed)


NotesFunding: Canadian Geriatrics Society


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Subjects were randomized into an exercise or control group using the Rand Corporation random tables"

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo mention of blinding of participants, but usual care in same setting, so intervention would have been obvious

Incomplete outcome data (attrition bias)
All outcomes
Low risk2 participants dropped out of the exercise group - may be related to intervention, but less than 10% of group

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Details: participants were matched in pairs of 2 and then randomised because participants were expected to train with a regular partner (participants were allowed to choose their preferred training partner)
Duration: 12 weeks
Follow up: no


ParticipantsCharacterisation: senior hostel residents
Country: Switzerland
Setting: senior residents hostel
Randomised: 32
% women = 64%
Age details: mean (SD)
Strength group: 86.5 (4.9)
Strength and balance group: 85.39 (5.4)
Inclusion criteria: residential status, age over 70 years, signed informed consent statement, ability to walk 6 metres
Exclusion criteria: severe cognitive impairment, rapidly progressive or terminal illness, acute illness or unstable chronic illness, myocardial infarction, fracture of a lower extremity within 6 months prior to study beginning, insulin-dependent diabetes mellitus, undergoing resistance training at the time of enrolment, musculoskeletal or cardiovascular abnormalities (revealed by muscle strength test)

ADL status details: not reported
Cognitive status details: MMSE, mean (SD)
Strength group: 26.5 (3.8)
Strength and balance group: 27.3 (2.2)
Significant comorbidities: not reported
Assessed: not reported
Excluded: 6 on eligibility criteria, 0 other


InterventionsStudy aim or objective: to evaluate the additional effect of functional exercises on balance and lower extremity function among hostel-dwelling elderly people partaking in strength training
2 groups

Intervention: strength and balance group (N = 16)
Format: group, delivered by: physical therapist and exercise trainer
Session length: 45 minutes strength exercises twice a week plus half-hour balance training on same day as second weekly strength training. Twice weekly strength exercises as reported in strength group, plus an additional half-hour of balance training on the same day as the second weekly strength training. Classes started with 5 to 10 minutes of warm-up activities, followed by 20 to 25 minutes of skills training in game activities, such as throwing and catching a ball while standing on a foam surface. All exercise adjusted to individual mobility level. The intensity was gradually increased and previously formulated recommendations were applied with progressive difficulty as tolerated
Control: Strength group (N = 16)
A regimen of progressive resistance training of the hip and knee extensors, the hip abductors, and the foot plantar flexors. The total training period was divided into 2 phases of 6 weeks each. In phase 1, participants performed progressive resistance. The first set performed with no weight or with 50% of maximal exercise weight and 15 repetitions. In the 2nd set, 100% of the maximal exercise weight was taken with the aim of 8 to 12 repetitions. In the 2nd training phase, participants trained in high-intensity progressive resistance training


OutcomesPhysical function (other): functional test for physical performance (Guralnik 1994), Tinetti Mobility Scale (gait and balance) (Tinetti 1986)
Muscle power (anaerobic): isometric knee extensor muscle force (maximal)
Balance: dynamic postural stability


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskStated used random number table

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskStated that all exercise sessions were undertaken by a single exercise trainer. As a result he/she would have been aware of group status

Incomplete outcome data (attrition bias)
All outcomes
Low riskSimilar numbers lost for similar reasons

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT, repeated measures cross-over design
Duration: n/a, one-off intervention
Follow up: n/a
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: no
Group comparability at entry: n/a
Losses to follow up: none


ParticipantsCountry: USA
Setting: nursing home and retirement home
Randomised: 28
% women: not reported
Age: mean = 84.86 years ± 6.08 years; range = 76 to 98 years

Consent: unspecified

Inclusion criteria: score of 25 to 40 points on the Paracheck Geriatric rating scale

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported

Group 1: N = 10
Group 2: N = 8
Group 3: N = 10


InterventionsStudy aim or objective: (1) materials-based intervention would elicit more repetitions and greater distance of movement than imagery-based occupation and rote exercise; (2) imagery-based occupation would elicit more repetitions and greater distance of movement during physical activity than rote exercise

Number of experimental groups: 3
Single, individual session
Session duration: not reported
Number of sessions per week: n/a
Seated: yes

Exercise features: materials-based occupation involved kicking a balloon; imagery-based occupation involved kicking an imaginary balloon; rote exercise involved being asked to kick your foot as in a demonstration; participants were asked to kick with the same foot as many times as possible before becoming tired

Group 1: materials-based occupation, followed by imagery-based occupation, followed by rote exercise
Group 2: imagery-based occupation, followed by rote exercise, followed by materials-based occupation
Group 3: rote exercise, followed by materials-based occupation, followed by imagery-based occupation


OutcomesPhysical function (other): vertical kicking speed, vertical kicking distance
Endurance (physical other): kicking repetitions


NotesFunding: not reported
Replication and extension of Lang 1992


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Subjects were randomly assigned to three groups"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskCross-over design - all participants received all interventions - it is likely that participants would have been able to tell which was the intervention under experiment and which was the control

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskCross-over design, but intervention unlikely to produce a carry-over effect


MethodsDesign: RCT
Duration: 8 weeks
Follow up: no


ParticipantsCharacterisation: ambulatory independent
Country: Turkey
Setting: nursing home
Randomised: 42
% women = 68%
Age details: median years (range):
UHE group = 79 (21)
SGE group = 81 (19)
Drop-outs = 79 (20)
Inclusion criteria: > 65 years and being non-obese (i.e. BMI B/30), increased risk of falling.
Exclusion criteria: psychological disease
Score of > 52 on the BBS or if they had attended an exercise programme regularly in the 2 months prior to the study, not recently suffered a stroke or recovery from an acute illness, no unstable or uncontrolled medical conditions (e.g. diabetes mellitus, hypertension), no resting angina, no recurrent heart failure or arrhythmias, no uncontrolled seizure disorder, no progressive neurological disease, no blindness or deafness, and no severe osteoporosis with 2 or more fractures
ADL status details: assistive device: cane inside (N = 1); cane outside (N = 3); 2 unilateral crutches, 2 canes outside (N = 4)
Cognitive status details: not reported.
Significant comorbidities: number of chronic diseases: 0 to 3 (N = 27), 4 to 6 (n = 15)
Assessed: 535
Excluded: overall = 493, 272 on criteria, 112 declined, 109 other


InterventionsStudy aim or objective: to compare the effectiveness of unsupervised home and supervised group exercise on parameters related to risk of falling among older adults
2 groups

Intervention: Supervised group exercise (SGE) group (N = 21)
Format: group; delivered by: physiotherapist in exercise class
Session length: 45 minutes, 3 times per week
Exercise programme based on balance training, strengthening and stretching of the lower limbs, increasing flexibility, posture exercises, and functional activities
Sessions consisted of warm-up, posture exercises, balance and lower limb co-ordination exercises, sit-to-stands, walking training, stretching, strengthening, and cool down (details in Table 1)
Body weight was used for strengthening of the lower limbs. Balance exercises progressed from being undertaken with eyes open to eyes closed and from undertaken with a support to without a support. For each exercise, the duration and number of repetitions increased as the participants' tolerance increased and as time progressed

Control: Unsupervised home exercise (UHE) group (N = 21)
Same exercise programme as SGE group
The physiotherapist demonstrated the exercises to the participants and then observed them practising the exercises once to ensure they understood how to do them properly. Participants were given a written exercise programme with a weekly chart to help them keep a record of what they had accomplished. On this chart the participants marked the exercises they completed on the various days. Furthermore, they were asked to come to meet the physiotherapist at the end of the second and fourth weeks in order to ensure the exercise programme was being performed properly and to discuss any problems encountered. At these meetings, the participants were also instructed on how to make the appropriate progressions to higher level exercises


OutcomesPhysical function in ADL: TUG test (seconds) (Podsiadlo 1991)
Muscle power (anaerobic): leg strength (leg dynamometer)
Balance: balance: time standing on 1 leg (eyes open), balance: time standing on 1 leg (eyes closed), balance: tandem stance (up to 30 seconds, eyes open), balance: tandem stance (up to 30 seconds, eyes closed), BBS
Flexibility: 'Sit-and-Reach' test
Falls, risk and fear of falling: fear of falling (VAS) (Wolf 2001a)
Proprioception: lower limb matching task (position sense of knee joint)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "...code number of each subject was written on a card and placed in an envelope, and the envelopes were then put in two groups by a person with no knowledge of the codes"
No report of shuffling of envelopes; therefore, unclear risk of bias

Allocation concealment (selection bias)Low riskQuote: "...code number of each subject was written on a card and placed in an envelope, and the envelopes were then put into two groups by a person with no knowledge of the codes"

Blinding of participants and personnel (performance bias)
All outcomes
High riskParticipants were in the same home; 1 group supervised, and 1 unsupervised so knowledge of allocation likely

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "The physiotherapist, who carried out all measurements, both at baseline and after the exercises, was also unaware of the group the subjects were in"

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskSlight imbalance in number of participants lost to follow up between groups (more in UHE group than in SGE group), although similar reasons across groups

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Details: to achieve statistically-similar groups, participants of the same sex, similar age, and comparable MMSE scores and degree of muscle strength were sorted in pairs. Then each pair was divided randomly; 1 participant assigned to training group and other to control group
Duration: 10 weeks
Follow up: no


ParticipantsCharacterisation: frail, long-term care facility residents, aged 75 years or older
Country: Austria
Setting: geriatric long-term care facility
Randomised: 42
% women = 77%
Age details: mean (SD) = 86.8 (5.8); range = 77 to 98
Inclusion criteria: > 75 years, physical ability to take part in strength and balance training (assessed as the ability to walk 5 metres or more, with or without the assistance of walking aids)
Exclusion criteria: high degree of dementia (MMSE score < 10), severe acute diseases
ADL status details: not reported
Cognitive status details: MMSE score, mean (SD): 20.9 (5.2)
Range: 11 to 28
Significant comorbidities: most participants were multi-morbid people with several diagnosed illnesses: osteoporosis: 52%, dementia: 45%, depression: 42%, history of stroke: 39%, coronary heart disease: 29%
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to examine the effects of structured strength and balance training in frail, elderly long-term care residents
2 groups

Intervention: Training group (N = 21)
Format: group; delivered by: sports scientist
Session length: 50 minutes 3 times weekly
Structured strength and balance training: warm up (10 minutes), strength training (25 minutes), performed with elastic resistance bands and soft weights, 1 set per muscle group, 10 to 15 repetitions

Balance training (10 minutes): exercise balls, balance discs and blocks (20 cm high) used
Cool down (5 minutes)

Control: Control group (N = 21)
At baseline, according to their skills and deficits, some participants received rehabilitative physical, occupational, psychotherapy, speech therapy, or both (possible contamination)


OutcomesPhysical function in ADL: BI (0 to 100 scale), FIM
Physical function (other): Tinetti Mobility Scale (gait and balance) (Tinetti 1986)
Muscle power (anaerobic): muscle function (upper and lower extremity)
Mood related: GDS
Cognition: MMSE
Anthropometry: lean body mass (kg), lean body mass (%), BMI


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMatched pairs "divided randomly"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo mention of blinding of participants, but RCT in same home and usual care as alternative

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskNeither the psychologist nor the physiotherapist, who tested muscle function, were informed by the study organisers to which group participants were assigned. Unclear who assessed BI and FIM and if assessors were blinded to allocation

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskSame number of losses to follow-up in each group, but unclear if reasons for losses to follow-up were different between groups

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 20 weeks
Follow up: 52 weeks
Method of randomisation: participating homes allocated to 1 of 2 exercise interventions using sealed envelopes. Participants in those homes were then randomly assigned to exercise programme or control using computer-generated random numbers. Maximum size of exercise group at each home is 12 and control at least 5
Concealment of allocation: yes
Outcome assessor blinding: not specified
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: recruited 278 participants; 40 (14.4%) dropped out immediately after randomisation (equally distributed across both groups), 6 excluded from fall analyses because no reliable data, 30 excluded from physical function and disability analyses because they did not come to the post-intervention assessment, 4 perceived their health to be too poor, 4 lost interest, 1 suffered fracture, 5 hospitalised > 2 weeks; 4 died, and 6 were ill


ParticipantsCountry: the Netherlands
Setting: 15 homes for the elderly
Randomised: 278
% women = 79
Age: mean = 84.9 years; range = 63 to 98 years
Consent: fully-informed

Inclusion criteria: see exclusion criteria

Exclusion criteria: cognitive: impaired cognition to the extent that they could not process information provided during the testing and exercising; medical = GP judged whether there was a medical contraindication to exercising; functional = unable to walk more than 6 metres independently (aids allowed)

% Eligible within home: not reported
% Eligible that participate: not reported

Functional walking: N = 130 (7 residences); 80 allocated intervention, 50 allocated control

In balance: N = 148 (8 residences); 94 allocated intervention, 54 allocated control


InterventionsStudy aim or objective: to determine the effects of moderate intensity group-exercise programme on falls, functional performance, and disability in older adults, and to investigate the effect of frailty on outcome

Number of experimental groups: 4
Group intervention delivery
Session duration: 90 minutes
Number of sessions per week: 2
Seated: no

All participants (including control) required to report levels of physical activity to monitor and control contamination from the intervention

Interventions: 2 exercise programmes, both with evidence that they were effective in preventing falls, 1 session per week for 4 weeks followed by bi-weekly sessions for 16 weeks, 90-minute sessions including 30-minute social element intended to increase motivation; all groups had their own instructor and assistant

Functional walking: 10 exercises: balance, mobility, and transfer training
In balance: (derived from Tai Chi) elements of Tai Chi most beneficial to elderly people
Control: there was a control group for each exercise type; details not reported


OutcomesRisk of falling: time to first fall
Falls: falls (any episodes for participant), fall rate (falls per person years), number of falls
Physical function in ADL: GARS (Groningen Activity Restriction Scale)
Physical function (other): Physical Performance Score (PPS) (Faber 2006), Performance Orientated Mobility Assessment (POMA)


NotesFunding: not reported; no commercial party had any financial interests


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "The participating homes were randomly assigned to one of the two exercise intervention programs, using sealed envelopes. Participants in each of the homes were then randomly distributed across an intervention and a control group, using computer-generated random numbers. The maximum size of the exercise group in each home was set at 12, with the provision that the control group should contain at least 5 participants"
Unclear whether size of participant groups was constrained using a truly random process

Allocation concealment (selection bias)Unclear riskQuote: "The participating homes were randomly assigned to one of the two exercise intervention programs, using sealed envelopes"
Unclear if second allocation of participants was concealed

Blinding of participants and personnel (performance bias)
All outcomes
High riskQuote: "The outcome of the randomization was notified to the participants in a letter after baseline assessment"

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskInsufficient information - no information included about who undertook the outcome measures

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskInsufficient information - no information included about who undertook the outcome measures

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskITT analysis stated. However, baseline figures do not match number of participants initially randomised. 6 participants excluded from analyses of fall data because there was no reliable data available; 30 participants excluded from analyses of physical function and disability data because of missing post-intervention assessment (24 of whom had dropped out of the study). Reasons not given per group and unclear impact on outcomes

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 10 weeks
Follow up: none
Method of randomisation: unclear
Concealment of allocation: unclear
Outcome assessor blinding: partial - measurements of muscle function were made by a single observer who was aware of group assignments, but not involved in training. CT evaluation of the mid-thigh was conducted by a single investigator in a blinded fashion; no details of blinding were described for the other outcomes
Group comparability at entry: no, significant differences, P < 0.05

Baseline difference in strength: exercise + nutrition participants significantly weaker than exercise-alone participants

Losses to follow up: 3 lost from exercise-only group (1 lack of interest, 1 musculoskeletal pain, 1 pneumonia), 2 from supplement-only group (1 death, 1 lack of interest), and 1 lost from control group because of death


ParticipantsCountry: USA
Setting: elderly long-term care facility
Randomised: 100
% women = 63
Age: 38% older than 90 years; mean = 87.1 years ± 0.6 years; range = not reported

Consent: not specified

Inclusion criteria: aged over 70 years, residential status, ability to walk 6 metres

Exclusion criteria: cognitive = severe cognitive impairment; medical = rapidly progressive or terminal illness, acute illness, unstable chronic illness, myocardial infarction, fracture of a lower extremity within 6 months before the study, insulin-dependent diabetes mellitus, if they were on a weight loss diet or undergoing resistance training at the time of enrolment, if test of muscle strength revealed a musculoskeletal or cardiovascular abnormality

% Eligible within home: 26.7
% Eligible that participate: 28.7

Exercise-only group: N = 25; mean age = 86.2 years ± 1.0 mean SE; range = 72 to 95 years; 64% women

Supplement-only group: N = 24; mean age = 85.7 years ± 1.2 mean SE; range = 75 to 97 years; 71% women

Exercise and supplement group: N = 25; mean age = 87.2 years ± 1.2 mean SE; range = 76 to 98 years; 64% women

Control group: N = 26; mean age = 89.2 years ± 0.8 mean SE; range = 78 to 98 years; 54% women


InterventionsStudy aim or objective: hypothesis: physical frailty is partially mediated by skeletal-muscle disuse and marginal nutritional intake, and should therefore be reduced by interventions designed to reverse those deficits

Number of experimental groups: 4
Individualised intervention
Session duration: 45 minutes
Number of sessions per week: 3
Seated: no

A therapeutic recreation specialist delivered the exercise components

Exercise-only group: high-intensity progressive resistance training of the hip and knee extensors, commencing at 80% of 1 repetition max and progressing as able

Supplement-only group: 240 ml Exceed micronutrient supplement drink daily, representing 360 kilocalories, delivered in an unmarked container

Exercise and supplement group: comprised both interventions

Control group: 240 ml of a minimally nutritive liquid delivered in the same way, plus 3 activities of the participants' choice offered by the same service, but excluding resistance training


OutcomesPhysical function in ADL: gait velocity over a 6.1 metre course (Fiatarone 1994)
Physical function (other): Stair Climbing Power (Bassey 1992)
Muscle power (anaerobic): leg press (Fiatarone 1994), hip extensor muscle strength, knee extensor muscle strength
Physical activity: physical activity (GMM activity monitor)
Anthropometry: thigh-muscle area (Fiatarone 1994), weight
Physiology: whole body potassium (Cohn 1980)
Energy consumption: energy intake (kCal/day)


NotesFunding:

National institute of Ageing

Agricultural Research Service

Public Health Service of Hebrew Rehabilitation Centre, Massachusetts

Brookdale Foundation


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskStudy performed in one home. As a result, it would have been difficult to blind participants and would have been obvious which groups they were assigned to

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNot clear who performed mobility outcome assessments

Incomplete outcome data (attrition bias)
All outcomes
High riskOutcome data missing for some participants because of technical problems or illness at time of testing

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 12 weeks
Follow up: none
Method of randomisation: by residential home
Concealment of allocation: unclear
Outcome assessor blinding: no
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 5 (25%)
Intervention group: N = 4 (2 due to illness, 1 loss of interest, 1 out with relatives on day of classes)
Control group: N = 1 (refused test)


ParticipantsCountry: UK
Setting: 2 residential homes
Randomised: 20
% women = 95 (only data from women analysed)
Age: mean = approximately 88 years; range = not reported

Consent: fully-informed

Inclusion criteria: > 70 years, mobile, able to participate in test battery, no medical conditions that would interfere with safety regarding training program

Exclusion criteria: 6 participants were excluded, but no reasons given

% Eligible within home: 76.9
% Eligible that participate: not reported

Intervention group: N = 10; mean age = 88 ± 5 years; all women

Control group: N = 10; mean age = 87 years ± 4 years; 9 women


InterventionsStudy aim or objective: study question: is it possible to improve functional ability in older people by getting them to practise the functional tasks themselves?

Number of experimental groups: 3
Group intervention delivery
Session duration: unclear
Number of sessions per week: 2
Seated: unclear

Exercise features: circuit of 8 functional exercises for 30 seconds initially progressing to a maximum of 1 minute, then increasing difficulty of task

Control: reminiscence and recreational sessions; gentle, seated range of movement exercises (trunk and upper limbs only)

Personnel delivering interventions not specified


OutcomesPhysical function in ADL: stair ascent/descent, walking distance in 15 seconds
Physical function (other): sit-to-stand (fastest time to stand up)
Anthropometry: weight


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskThis was a cluster randomised trial. No evidence as to how randomisation was determined

Allocation concealment (selection bias)Unclear riskNo evidence to demonstrate if individuals were recruited into the trial before/after the homes had been randomised to intervention or control

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo evidence of blinding

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo evidence of blinding

Incomplete outcome data (attrition bias)
All outcomes
High riskMissing outcome data (5 dropouts; 4 from intervention)

Selective reporting (reporting bias)Unclear riskIn the results, confidence intervals are only reported for some results (not clear why)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 10 weeks
Follow up: none
Method of randomisation: randomly assigned using a 1:2 ratio in a lottery format to control group or exercise group
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 5 (17%), dropped out because of health reasons (exercise = 2, control = 3)


ParticipantsCountry: Canada
Setting: long-term care facility
Randomised: 30
% women = 80
Age: mean = approximately 83 years; range = 75 to 94 years

Consent: assent accepted

No participants were currently involved in any physical exercise programme or had any recent exercise history

Inclusion criteria: ability to follow directions; ability to walk across a room (with or without assistive device); no recent history of cardiovascular, cerebral vascular, respiratory, systemic, muscular, or uncontrolled metabolic disease

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported

Control: N = 10; 75 to 87 years
Exercise: N = 20; 76 to 94 years


InterventionsStudy aim or objective: examine the effect of an onsite and simple progressive lower body training programme designed to improve muscle power on functional abilities in frail older adults

Number of experimental groups: 2
Group intervention delivery
Session duration: 20 to 60 minutes
Number of sessions per week: 3
Seated: If necessary

Exercise features: 10-minute warm-up and stretch, strengthening components utilised seated and standing components focusing in lower-body muscle groups, Thera-Bands® gradually introduced to increase resistance, number of exercise repetitions were gradually increased and a speed element introduced, 10-minute cool-down, personnel delivering interventions not specified

Control: no active or placebo intervention, asked to perform no more or no less activity than normal on a daily basis


OutcomesPhysical function in ADL: six-metre walk (time), TUG test (modified to 8 feet) (Bassey 1992; Rikli 1999)
Physical function (other): sit-to-stand (average number in 30 seconds)
Muscle power (anaerobic): knee extensor muscle strength


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Subjects were randomly assigned in a lottery format"

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskRCT with usual care control, so obvious group assignment

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo report of blinding

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskDifferential loss to follow up but seems unlikely to be related to intervention

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 4 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: unclear
Losses to follow up: none stated


ParticipantsCountry: USA
Setting: long-term residents of an intermediate care facility
Randomised: 19
% women = 16
Age: mean = not reported; median = 73 years; range = 62 to 95 years

Consent: assent accepted

Inclusion criteria: > 65 years old, some deficits in self care – requiring assistance with dressing, grooming, and feeding

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: 47.5


InterventionsStudy aim or objective: to test the assumption that elderly individuals participating in a range of motion exercise programme will show more of an increase in self care in hygiene and eating than those who do not
Number of experimental groups: 2
Group intervention sessions
Session duration: 30 minutes
Number of sessions per week: 2
Seated: If necessary

Exercise features: upper limb and lower limb range of movement exercises; personnel delivering the intervention were not described

Control: movies only


OutcomesPhysical function in ADL: Performance Test of Activities of Daily Living (PADL)


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo mention of blinding of participants, but control group watched movies, and outcome measure was physical so intervention would have been obvious

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskNo details provided

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo information about losses to follow up

Selective reporting (reporting bias)High riskResults for pre and post PADL not reported

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering accounted for
Duration: 6 months
Follow up: yes


ParticipantsCharacterisation: elderly care-home residents
Country: New Zealand
Setting: low-level dependency residential care homes
Randomised: 682
% women = 74%
Age details: mean (SD) = 84.3 (7.2)
Inclusion criteria: residents aged 65 years and over, able to engage in a conversation about a goal, remember the goal, and participate in a programme to achieve the goal (a proxy for cognitive state)
Exclusion criteria: residents unable to communicate to complete the study measures, anxiety as their main diagnosis, were acutely unwell, or in a terminal state
ADL status details: Late Life Function and Disability Instrument (LLFDI) total disability score, mean (SD) = 37.7 (6.3)
Cognitive status details: Abbreviated Mental Test Score (AMTS), mean (SD) = 7.2 (2.4)
Significant comorbidities: total number of diagnoses, mean (SD) = 4.9 (2.2)
No (%) Depression (GDS > 4): 215/614 (35%)
Assessed: 1584
Excluded: overall = 902; 762 on criteria, 140 declined, 0 other


InterventionsStudy aim or objective: to assess the effectiveness of an activity programme in improving function, quality of life, and falls in older people in residential care
2 groups

Intervention: activity group (N = 330)
Format: individual, delivered by gerontology nurses and healthcare assistants
Session length: not reported. Daily
Trained nurses delivered the promoting independence in residential care (PIRC) intervention. The resident set a goal to promote physical activity. The nurse then designed an individualised programme of physical activities based on ADLs (daily or several times a day). The gerontology nurse trained healthcare assistants in implementing the plan and provided ongoing support. Exercise activities were undertaken several times daily

Control: social group (N = 352)
Received usual care and were offered 2 social visits by a social science researcher to control for the attention received by the resident from the gerontology nurse visits


OutcomesPhysical function in ADL: Late Life Function and Disability Instument (LLFDI) (Sayers 2004), TUG test (seconds) (Podsiadlo 1991), Elderly Mobility Scale (EMS) (Smith 1994)
Balance: FICSIT-4 balance test
Mood related: GDS
Quality of life: Life Satisfaction Index (Neugarten 1961)
Perceived health status: EuroQoL (Brooks 1996)
Falls, risk and fear of falling: fear of falling (modified) (Hill 1996), falls (any episodes for participant)
Acute health events: adverse effects, hospitalisations


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomised homes to the intervention or control group using computer generated random numbers

Allocation concealment (selection bias)Low riskQuote: "...a biostatistician not involved in recruitment randomized homes to the intervention or control group..."

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskCluster design so potential for blinding, but not specifically reported

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "...research nurses blinded to the group allocation of the homes used standardised methods to assess outcomes..."

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "...research nurses blinded to the group allocation of the homes used standardised methods to assess outcomes..."

Incomplete outcome data (attrition bias)
All outcomes
Low riskLarge losses, but balanced and similar reasons

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 8 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: unclear
Losses to follow up: none


ParticipantsCountry: USA
Setting: assisted-living home for the aged
Randomised: 24
% women = 75
Age: mean = approximately 85 years; range = 72 to 101 years

Consent: fully-informed consent

Inclusion criteria: permission of participants' physicians was sought; no participants had any acute illness

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention group: N = 12; mean age = 87 year; range = 72 to 101 years; 9 women

Control group: N = 12; mean age = 82 years; range = 74 to 100 years; 9 women


InterventionsStudy aim or objective: the programme addressed physical activity and psychosocial needs, such as learned helplessness and sadness, without placing additional strain on the hectic schedules of staff

Number of experimental groups: 2
Group session delivery
Session duration: 30 minutes
Number of sessions per week: 3
Seated: yes

Exercise features:
Sit-and-get-fit group: performed seated range of movement exercises, included measures to promote psychosocial well-being; programme was delivered by a paraprofessional

Control: participated in usual home activities, with opportunity to participate in sit and get fit programme after the study period


OutcomesFlexibility: ankle dorsiflexion and plantarflexion (range of motion), shoulder abduction, hip flexion and extension, elbow flexion and extension, shoulder anterior flexion, knee flexion (range of movement)
Feasibility and acceptability: subjective experience data


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo information provided, but RCT with obvious intervention/control

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo attrition

Selective reporting (reporting bias)High riskPre and post data not reported in an appropriate format (only reported as percentage change; no absolute figures)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: n/a (one-off intervention)
Follow up: n/a
Method of randomisation: unclear - according to a counter-balanced design
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: n/a
Losses to follow up: none


ParticipantsCountry: USA
Setting: 2 nursing homes
Randomised: 15
% women = not reported
Age: mean = 76.3 ± 9.95 years; range = 56 to 93 years

Consent: not specified

Inclusion criteria: Parachek score > 25

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported


InterventionsStudy aim or objective: tested the hypothesis that materials-based occupation elicits a greater number of repetitions during physical activity in elderly persons than rote exercise

Number of experimental groups: 3
Delivered to groups of participants
Session duration: n/a
Number of sessions per week: n/a

Exercise features: materials-based occupation (kicking balloon); imagery-based occupation (kicking imaginary balloon); rote exercise (kicking foot as demonstrated)

Group 1: materials-based occupation, then imagery-based occupation, then rote exercise
Group 2: imagery-based occupation, then rote exercise, then materials-based occupation
Group 3: rote exercise, then materials-based occupation, then imagery-based occupation

Participants were instructed to kick as many times as possible and stop when too tired to continue

All interventions were supervised by a research assistant, conducted in one-off sessions, with 3 days in between each 1


OutcomesEndurance (physical other): kicking repetitions


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Subjects were randomly assigned..."
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNot addressed; all three groups received each intervention in different orders

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete outcome data

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskCross-over design, but intervention unlikely to produce a carry-over effect


MethodsDesign: RCT
Duration: 4 months
Follow up: none
Method of randomisation: table of random numbers; participants were stratified into 2 levels of mobility, based on their scores and the Timed Up and Go test. Within each mobility category at each site, residents were randomly assigned to either the functional fitness for long-term care (FFLTC) or range of motion (ROM) condition
Concealment of allocation: yes - maintained
Outcome assessor blinding: yes - maintained
Group comparability at entry: unclear
Losses to follow up: N = 28 (29%): FFLTC group N = 19, ROM group N = 9
Similar reasons for both groups: too busy N = 15, medical reasons N = 8, unable to follow exercise N = 3, moved away N = 2


ParticipantsCountry: Canada
Setting: long-term care institution
Randomised: 96
% women = 84
Age: mean = 80 years ± 0.9 years; range = not reported

Consent: assent accepted

Inclusion criteria: able to stand with minimal assistance, ability to follow simple instructions

Exclusion criteria: medical = recent cardiovascular event, vestibular disorder, uncontrolled hypertension, uncontrolled epilepsy, fracture within 4 months, total blindness/deafness, surgery planned for within the next 4 months; functional = holidays planned for within the next 4 months, recent admission (less than 3 months)

% Eligible within home: not reported
% Eligible that participate: not reported

FFLTC (functional fitness for long-term care) group: N = 55; mean age = 79.7 years; 29 women

ROM (range of motion) group: N = 41; mean age = 80.4 years; 30 women


InterventionsStudy aim or objective: this study compared traditional range of motion to a 'functional fitness for long-term care' programme designed to improve strength, balance, flexibility, gait, functional capacity, and strength

Number of experimental groups: 2
Group intervention delivery
Session duration: 45 minutes
Number of sessions per week: 3
Seated: not specified

Exercise features:
FFLTC group: comprised walking, strengthening and balance exercises, tailored to meet each of the 2 groups (high mobility/low mobility), conducted by recreation staff

ROM group: comprised seated exercise to improve range of movement (fingers, hands, arms, knees, ankles), relaxation, vocal exercise, and word/memory games

Groups were of mixed ability, supervised by recreation staff


OutcomesPhysical function in ADL: gait speed over 7 metres (self-selected normal pace), gait speed over 7 metres (fast pace), TUG test (seconds) (Podsiadlo 1991), FIM
Physical function (other): stair climbing power (Bassey 1992)
Muscle power (anaerobic): lower extremity strength, total hip strength, upper extremity strength, isometric strength (elbow flexion; shoulder abduction; knee extension; hip ab/adduction, flexion/extension), hand grip strength, isotonic strength of knee extensors (Connelly 1995)
Balance: BBS
Flexibility: 'Sit-and-reach' test (modified) (Lazowski 1997), upper body flexibility (Leighton flexometer)


NotesFunding: grants from the Canadian Fitness and Lifestyle Research Institute, The Walter J. Blackburn Family Foundation, The Richard Ivey Foundation, and the Ontario Ministry of Health Long-term Care


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskStated used a random number table

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskBlinding attempted, but potential that blinding was broken as participants were from the same setting
However, ROM and FFLTC groups, so questionable whether it was important

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskTUG: stated research assistant blind to study condition

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskFIM: stated research assistant blind to study condition

Incomplete outcome data (attrition bias)
All outcomes
High riskMuch higher attrition rate in FFLTC than ROM, 19/55 (35%) versus 9/41 (22%), and large numbers

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering not accounted for
Details: a non-equivalent pre-test-post-test control-group design
Duration: 26 weeks
Follow up: no


ParticipantsCharacterisation: nursing-home residents
Country: Hong Kong
Setting: nursing homes
Randomised: 175
% women = 68%
Age details: mean (SD) = 82.7 (7.1); range = 66 to 101 years

Inclusion criteria: (a) Chinese, (b) > 65 years of age, (c) intact cognitive function (Abbreviated Mental Test score > 6/10), (d) able to walk independently, (e) able to communicate in Cantonese
Exclusion criteria: (a) experiencing acute symptoms of medical problems, (b) having a pre-existing psychological disorder, (c) having previous Tai Chi training
ADL status details: not reported
Cognitive status details: not reported
Significant comorbidities: mean (SD) number of comorbidities = 2.9 (3.1)
Assessed: not reported

Excluded: not reported


InterventionsStudy aim or objective: to examine the effect of Tai Chi on health-related quality of life in nursing-home residents
2 groups

Intervention: Tai Chi group
Format: group, delivered by: Tai Chi instructor
Session length: 8 to 10 minutes to perform entire sequence, 3 times weekly
Chen-style Tai Chi was taught. This short-form incorporated essential elements of Tai Chi and was relatively gentle
During the sessions, the instructor demonstrated the Tai Chi movements, and participants imitated the motions and postures. New Tai Chi movements were introduced each session, and participants learned the sequence at the end of 26 weeks

Control: Control group
Usual daily activities


OutcomesBalance: balance (Single Limb Stand Timed test)
Flexibility: 'Sit-and-reach' test
Perceived health status: SF-12 health-related quality of life (mental component), SF-12 health-related quality of life (physical component)
Energy expenditure: Physical Activity Questionnaire (PAQ) (Liu 2001)
Other: satisfaction with Nursing Home Instrument Chinese version (Lee 2006)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskAuthor reports homes as randomised, but no information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskStated that the instructor was blind to outcome measures. However, participants could not have been

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskFigures not provided for loss to follow up (hospitalisation, death, or move to other home) for intervention or control group (presented as combined data only)

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 54 weeks
Follow up: none
Method of randomisation: participants were stratified by every 5 years of age and then randomly assigned by computer
Concealment of allocation: unclear
Outcome assessor blinding: no, but objective assessments were performed (assessors were unaware of pre-test scores when collecting post-test data)
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 6 (7%)


ParticipantsCountry: USA
Setting: 2 nursing homes
Randomised: 88
% women = 80.7
Age: mean = 84.1 years ± 6.9 years; range = 65 to 98 years

Consent: informed consent from participant or relative and from their doctor

Inclusion criteria: nursing home long-term care resident, physically capable of safe bilateral lower extremity weight-bearing with supervision or minimal assistance, cognitively able to follow simple directions

Exclusion criteria: cognitive = unable or generally unwilling to follow simple directions; medical = inability or medical restriction to bear weight on both lower extremities, less than 65 years of age; functional = participating in skilled rehabilitation (physical therapy or occupational therapy) immediately prior to study

% Eligible within home: unclear
% Eligible that participate: unclear, possibly N = 88 of 294 residents (30%)

Group breakdown: not given, 29 to 30 participants in each group


InterventionsStudy aim or objective: investigation of the effect of a standing exercise programme on the number of falls and the severity of intrinsic fall risk factors (functional losses of strength, balance and endurance, depression and number of infections)

Number of experimental groups: 3
one-to-one sessions with each participant being supervised by a care assistant 'buddy'
Session duration: 20 minutes
Number of sessions per week: 5 - both interventions were conducted daily Monday to Friday, for 54 weeks
Seated: no

Intervention group: (group 1): comprised exercises in standing and walking activities, triggered when energetic music was played over intercom
Control group (group 2): consisted of talking and listening to music only
Control group (group 3): listened to the music alone


OutcomesFalls: number of falls
Physical function in ADL: 30-feet walk (time s), TUG test (seconds) (Podsiadlo 1991)
Balance: Duncan Functional Reach (DFR) test
Mood related: GDS
Acute health events: infections (average number per month)


NotesFunding: unclear
Unpublished PhD thesis


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Assignment to specific activities will be randomized by computer"
Comment: assume this means a computer random-number generator was used

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskBlinding of participants would not have been possible - participants would have been aware of which group they were in

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskUnclear if assessor was blinded

Incomplete outcome data (attrition bias)
All outcomes
Low riskAuthor states that the 6 participants who were unable to complete the post-tests were "dispersed among the groups, with reasons for dropping out unrelated to the intervention"

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Details: stratified by care level (determined on base of total time required for assistance with ADLs, instrumental ADLs, functional training, and medical services)
Duration: 3 months
Follow up: no


ParticipantsCharacterisation: frail elderly women
Country: Japan
Setting: special nursing homes for the elderly
Randomised: 149
% women = 100%
Age details: mean (SD):
Exercise group: 84.85 (7.3)
Control group: 86.25 (6.59)
Inclusion criteria: women and stable health
Exclusion criteria: not reported
ADL status details: length of time of care per day (number of participants):
30 to 50 minutes: 9
50 to 70 minutes: 19
70 to 90 minutes: 36
90 to 110 minutes: 33
110 minutes or longer: 48
Cognitive status details: not reported
Significant comorbidities: not reported
Assessed: 226
Excluded: Overall = 77


InterventionsStudy aim or objective: to evaluate the effects of exercise therapy using the Takizawa Program
2 groups

Intervention: Exercise group (N = 74)
Format: group, delivered by nurses and care workers as physical exercise instructors
Session length: not reported, 3 times weekly
Takizawa program: following exercises performed while sitting or standing:
1. upper limb range of motion exercise using a movable pulley
2. trunk flexion and twisting exercise with abdominal breathing
3. ankle plantar-dorsal flexion exercise using the instrument 'PATA'
4. knee flexion-extension exercise using the instrument 'KORO'
5. shoulder and elbow flexion-extension exercise
6. knee extension exercise
7. hip flexion exercise


OutcomesPhysical function in ADL: FIM
Flexibility: range of motion (shoulder flexion; knee extension; ankle dorsal flexion/plantar flexion)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo report of blinding of participants, but usual care so intervention would have been obvious

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "The evaluators of FIM in this study were the care workers who provided daily care to the patients, and were blind to the patient’s assignments to the Ex or Co group"

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskSlightly more losses to follow up in exercise group than in control group, although same reason provided across groups

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 7 months
Follow up: none
Method of randomisation: sealed envelopes supplied in sequence by the study co-ordinator and prepared from a computer-generated random numbers table
Concealment of allocation: yes - maintained
Outcome assessor blinding: no
Group comparability at entry: unclear
Losses to follow up: N = 8 (16.3%): intervention group: N = 5 (3 deaths, 2 loss of interest), control group: N = 3 (3 deaths)


ParticipantsCountry: Scotland, UK
Setting: 4 residential homes
Randomised: 49
% women = 80 (of completers)
Age: mean = 81 years; range = 64 to 91 years

Consent: assent accepted

Inclusion criteria: see exclusion criteria

Exclusion criteria: cognitive: severe communication difficulties

Residential homes all had identical entrance criteria, namely that residents should be able to toilet, dress, and walk independently

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention group: N = 20; mean age = 82.3 years (SD 6.9); 12 women

Control group: N = 29; mean age = 79.3 years (SD 6.2); 21 women


InterventionsStudy aim or objective: to evaluate whether participation in regular exercise was acceptable to residents of old people's homes, and whether it produced significant improvements in balance, flexibility, strength, or functional capacity compared with a control group who participated in reminiscence sessions

Number of experimental groups: 2
Group intervention delivery
Session duration: 45 minutes
Number of sessions per week: 2
Seated: yes
Personnel delivering the interventions were not described

Exercise features: full upper limb and lower limb range of movement; seated exercises to music, intended to promote strengthening; exercise groups lasted for 45 minutes, and were conducted twice weekly for 7 months

Control: music and reminiscence therapy designed to prompt social interaction


OutcomesPhysical function in ADL: BI (0 to 20)
Physical function (other): sit-to-stand (fastest time to stand up)
Muscle power (anaerobic): hand grip strength
Balance: postural sway (using Wright's ataxiameter) (Wright 1971 )
Mood related: GDS
Cognition: MMSE
Flexibility: spinal flexion (American Academy of Orthopaedic Surgeons 1965), knee extension (range of movement), knee flexion (range of movement)
Quality of life: Life Satisfaction Index (Neugarten 1961)
Anthropometry: BMI


NotesFunding:

The Mathew Trust

The ICL Discretionary Trust
All of the residential homes exhibited identical admission criteria


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomisation...prepared from a computer-generated random numbers table"

Allocation concealment (selection bias)Low riskQuote: "Randomisation was by opening sealed envelopes supplied in sequence by the study co-ordinator"

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo report of blinding of participants - although exercise groups and reminiscence groups were in different homes, so participants possibly weren't aware if they were intervention or control

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
High riskAll measurements made by the same observer that provided the interventions

Incomplete outcome data (attrition bias)
All outcomes
High riskMore losses to follow up in exercise group (25%) than in control group (10%), and losses from exercise group include lack of interest

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 6 months
Follow up: none
Method of randomisation: sealed envelopes supplied in sequence by the study co-ordinator, prepared from a computer-generated random number table
Concealment of allocation: yes - maintained
Outcome assessor blinding: partial
Group comparability at entry: yes, no significant differences, P > 0.05
Losses to follow up: N = 10 (15.4%): intervention group: N = 4 (3 deaths, 1 fractured neck of femur); control group N = 6 (4 deaths, 2 hospital admissions)


ParticipantsCountry: Scotland, UK
Setting: 4 residential homes
Randomised: 65
% women = 83%
Age: mean = 83 years; range = 67 to 98 years

Consent: assent accepted

Inclusion criteria: volunteers were not excluded on the basis of any medical condition

Exclusion criteria: cognitive = severe communication difficulties

% Eligible within home: not reported
% Eligible that participate: not reported
Intervention group: N = 36; mean age = 83.7 years (6.6); 29 women

Control group: N = 29; mean age = 82.0 years (9.6); 25 women

Residential homes all had identical entrance criteria, namely that residents should be able to toilet, dress, and walk independently


InterventionsStudy aim or objective: (1) what are the mechanisms of improvement seen in McMurdo 1993?, (2) in the institutionalised elderly, does participation in regular seated exercise strengthen the quadriceps muscles?, (3) is participation in such exercise associated with improved psychomotor or cognitive function?

Number of experimental groups: 2
Group intervention sessions
Session duration: 45 minutes
Number of sessions per week: 2
Seated: yes

Exercise features: performed seated exercise to music, number of repetitions and gravity-resisted exercises were increased during the course of the study, group format, supervised by research physiotherapist

Control group: reminiscence therapy designed to prompt social interaction and group discussion, 45 minutes in duration, conducted twice weekly for 6 months, facilitated by research physiotherapist


OutcomesPhysical function in ADL: step test
Physical function (other): reaction Time
Muscle power (anaerobic): quadriceps muscle strength
Cognition: MMSE


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomization...prepared from a computer-generated random table"

Allocation concealment (selection bias)Low riskQuote: "Randomization was by opening sealed envelopes supplied in sequence by the study coordinator"

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskQuote: "Pre-study measurements were undertaken by the research physiotherapist at the same time of day in both groups. Post-study measurements were undertaken by an independent, blinded observer who was not otherwise involved in the project"
No information about blinding of participants and personnel involved in recording outcome measures

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Done by an independent, blinded observer who was not otherwise involved in the project"

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing outcome data relatively balanced across the groups (all participants accounted for)

Selective reporting (reporting bias)High riskOnly Quadriceps, MMSE and reaction time data reported

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 4 to 8 weeks
Follow up: 1 year
Method of randomisation: groups were determined by computer algorithm stratified by site of care (GEM or nursing home); assignments were in sealed envelopes that were opened after the pre-test was completed (see notes)
Concealment of allocation: yes - maintained
Outcome assessor blinding: yes
Group comparability at entry: no significant differences, P < 0.05

Significantly more training participants had their primary medical problem disability from cerebrovascular accident

Losses to follow up: N = 20 (25.6%) (at initial post-test): training group N = 13 (6 discharged home, 4 due to illness, 2 deaths, 1 withdrew because of shoulder strain); control group N = 7 (4 discharged home, 1 because of illness, 2 deaths)


ParticipantsCountry: USA
Setting: 1 Veteran's Affairs nursing home and geriatric evaluation and management unit, and 1 community nursing home
Randomised: 78
% women = 12
Age: mean = 75 years; range = 60 to 97 years

Consent: assent accepted

Inclusion criteria: > 60 years, impaired functional status (requiring assistance with 1 or more physical activities of daily living) with potential for improvement, able to follow simple commands, wheelchair participants must be able to transfer with modest assistance at most, Veteran's Affairs participants must have an expected length of stay > 4 weeks

Exclusion criteria: severe dementia, uncontrolled hypertension, unstable angina, medical condition that would interfere with safety of training protocol, stroke in previous 3 weeks, pacemaker, chronic atrial fibrillation

% Eligible within home: not reported
% Eligible that participate: not reported

Training group: N = 39; mean age = 74.1 years; range = 60 to 90 years; 2 women (7.7%)

Control group: N = 39; mean age = 76.9 years; range = 60 to 97 years; 7 women (21.9%)


InterventionsStudy aim or objective: to establish whether moderate-intensity endurance training results in short-term improvements in strength, endurance, and function

Number of experimental groups: 2
Group intervention delivery
Session duration: 30
Number of sessions per week: 2
Seated: no

All participants participated in training for 4 to 8 weeks, and a minimum of 10 resistance sessions in the 4-week period were ensured

The interventions were delivered in-group format and led by a physiotherapist and an aide

Exercise features:
Training group: progressive resistance and endurance training regime, comprised a minimum of 10 minutes of endurance training, up to a maximum of 30 minutes before resistance was increased, endurance training took place on a Tuesday and Thursday, resistance training occurred on Monday, Wednesday, and Friday, and comprised 15 repetitions for each muscle group

Control group: usual care


OutcomesPhysical function in ADL: walking speed over 20-foot, self-selected (Friedman 1988), Performance Test of Activities of Daily Living (PADL), Instrumental Activities of Daily Living (IADL, 7 items)
Muscle power (anaerobic): isokinetic eccentric strength (knee extension/flexion; shoulder extension/flexion; elbow extension/flexion; ankle dorsi/plantarflexion), isokinetic concentric strength (knee extension/flexion; shoulder extension/flexion; elbow extension/flexion; ankle dorsi/plantarflexion), isometric strength (knee extension/flexion; shoulder extension/flexion; elbow extension/flexion; ankle dorsi/plantarflexion)
Mood related: GDS
Endurance (physical other): heart rate after 6 minutes of endurance testing (using either an upper extremity ergometer, a stationary cycle, or a recumbent stepper)
Acute health events: hospitalisations


NotesFunding: Veterans affairs health services research and development service

Transportation logistics precluded more than 2 participants from the community nursing home from training at any given time. As a result, the randomisation scheme for these participants was done using a flip coin every time 2 participants had completed the pre-test. Though the assessor was blinded to the participants' group assignment, the assessor was not questioned to ascertain the success of blinding


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "The randomization scheme for the VA participants was determined by a computer algorithm stratified by site of care"

Allocation concealment (selection bias)Low riskQuote: "Assignments to the study groups were concealed in sealed envelopes that were opened after the pretest was completed"

Blinding of participants and personnel (performance bias)
All outcomes
High riskRCT in same setting with usual care, so intervention would be obvious

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskRegarding strength tests: "The assessor was blinded to the subjects' group assignments (training or control)"
No information on blinding of mobility tests

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskQuote: "Whenever possible, the subject's primary nurse was interviewed to provide information for the PADL and IADL scale."

No mention of blinding

Incomplete outcome data (attrition bias)
All outcomes
High riskAlmost double the attrition rate in the training group. 25% overall

Selective reporting (reporting bias)High riskQuote "When post-hoc we stratified the subjects into most dysfunctional score..." (re: ADL)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 8 weeks
Follow up: 1 week
Method of randomisation: randomised by residence - no further details given
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: no, significant differences, P < 0.05

Participants in the strength training programme had significantly greater strength and ADL scores
Losses to follow up: none


ParticipantsCountry: USA
Setting: senior citizen or residential nursing home
Randomised: 58
% women = 83%
Age: mean = 82.67 years ± 7.72 years; range = 71 to 101 years

Consent: not specified

Inclusion criteria: sedentary for at least 6 months prior to commencing programme

Exclusion criteria: none stated

% Eligible within home: not reported
% Eligible that participate: not reported

Experimental group: N = 29
Control group: N = 29


InterventionsStudy aim or objective: (1) effects of upper body high-intensity training on muscular strength, ADLs, and subjective well-being; (2) whether changes in strength were related to subsequent changes in subjective well-being and ADLs

Number of experimental groups: 2
Group intervention delivery
Session duration: 30 minutes
Number of sessions per week: 3
Seated: yes

Exercise features:
Experimental group: progressive resistance exercise regime targeting 5 upper body muscle groups, led by an exercise specialist

Control group: comprised fluid movements that incorporated non-stress exercise and mild stretching activities

Groups were led by the same exercise specialist


OutcomesPhysical function in ADL: Instrumental Activities of Daily Living Scale (modified) (Mihalko 1996)
Muscle power (anaerobic): upper body strength (for each of pectorals, shoulders, back, biceps, triceps), upper body power (sum of strength scores for 5 muscle groups: pectorals, shoulders, back, biceps, triceps)
Mood related:Positive and Negative Affect Schedule (PANAS) (Watson 1988)
Quality of life: Satisfaction with Life Scale (SWLS) (Diener 1985)
Feasibility and acceptability: Subjective Exercise Experiences Scale (SEES) (McAuley 1994)


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskStated "subjects were assigned by residence"

Allocation concealment (selection bias)Unclear riskInsufficient information provided.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo information provided

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo information provided

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskNo information provided

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskThe number of participants who completed intervention/control not reported

Selective reporting (reporting bias)High riskSubjective Exercise Experiences Scale not presented. Comparative statistics for IADL not presented

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 10 months
Follow up: 40 weeks
Method of randomisation: 2 nursing homes were randomly assigned to be control sites, 4 as experimental sites. Homes were matched into sets of triplets, from which sites were randomised to the 3 study conditions. Randomisation procedure not detailed
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes
Losses to follow up: 76 (16.2%): Fit for your life group N = 18 (12 deaths, 6 refusals), Self care for seniors group N = 27 (deaths), Control group N = 31 (28 deaths, 3 refused)


ParticipantsCountry: USA
Setting: 6 nursing homes
Randomised: 468
% women = 79
Age: mean = 84.7 years; range = not reported

Consent: assent accepted

Inclusion criteria: see exclusion criteria

Exclusion criteria: cognitive = severe cognitive disability (Cognitive Performance Scale < 5); medical = unstable cardiac condition (excluded from exercise component only), terminal prognosis, length of stay < 90 days, health complications that prohibited contact

% Eligible within home: 55.1
% Eligible that participate: not reported

Fit for your life group: N = 142
Self-care for seniors group: N = 171
Control group: N = 155


InterventionsStudy aim or objective: to evaluate how weight training or nursing-based rehabilitation programmes in nursing homes impact on resident performance of ADLs and objective tests of physical performance

Number of experimental groups: 3
Individualised intervention delivery
Session duration: 20 minutes
Number of sessions per week: 3
Seated: unclear

Fit for your life group: progressive resistance training of major muscle groups related to function and mobility; led by staff, family, and volunteers; walking for 1 to 5 minutes initially, up to a maximum of 20 minutes continuous walking; resistance training comprised 2 sets of 8 repetitions, with progressively heavier weights; resistance training was conducted 3 times per week, non-consecutive days, with walking on alternate days, for a minimum of 4 months over a 10-month study period

Self-care for seniors group: Nursing rehabilitation intervention tailored to individual, with aim of maintaining function or preventing decline

Control group: usual care


OutcomesPhysical function in ADL: number of feet walked in 6 minutes - scale score (Morris 1999), MDS: Locomotion (on and off unit items) (Morris 1999), MDS: Late loss ADL (transfer, toilet use, bed mobility, and eating) (Morris 1999), MDS:early loss ADL (dressing and personal hygiene) (Morris 1999), MDS: ADL summary (8 items) (Morris 1999)
Physical function (other): sit-to-stand (scale based on time required to stand up 5 times in a row) (Morris 1999)
Balance: balance (time able to stand normally in 5-feet positions) (Morris 1999)
Mood related: GDS


NotesFunding: Grant from National Institute of Health, National Institute on Ageing


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskFacilities were "randomly designated" to be control or experimental sites
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNot mentioned (although, different homes were assigned to different interventions, so participants would not have been aware of which group they were in)

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNot clear who performed observed ADL outcome assessments

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskADL summary: "Assessments were completed by trained research staff, all of whom were blinded to the intervention status of the study subjects"

Incomplete outcome data (attrition bias)
All outcomes
High riskPerformance - large number of residents were unable to start to initiate these activities

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 1 year
Follow up: none
Method of randomisation: performed by calling a central number, randomisation was blocked into groups of 4, and stratified by nursing-homes site
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes
Losses to follow up: N = 14 (7.2%): Intervention group N = 5 (deaths); control group: N = 9 (7 deaths)


ParticipantsCountry: USA
Setting: 1 academic nursing home and 8 community nursing homes
Randomised: 194
% women = 71
Age: mean = approximately 80 years; range = not reported

Consent: not specified

Inclusion criteria: > 60 years of age, residence in nursing home > 3 months, dependent in 2 or more activities of daily living

Exclusion criteria: cognitive = severe dementia, inability to follow 2-step command; medical = terminal illness/acute medical condition; functional = assaultive behaviour, receiving physiotherapy currently, or within last 2 months

% Eligible within home: 7.3
% Eligible that participate: 77

Intervention group: N = 97; mean age = 79.7 years (8.5); 70% women

Control group: N = 97; mean age = 81.4 years (7.9); 71 % women


InterventionsStudy aim or objective: to assess whether a physical therapy programme tailored to long-stay residents' disabilities improved their physical function and long-term health

Number of experimental groups: 2
Individualised intervention delivery
Session duration: 30 minutes
Number of sessions per week: 3
Seated: unclear

Exercise features:
Intervention group: physical therapy tailored to the individual, incremental programme, used algorithm for treatment priorities, a specific number of repetitions were performed for each exercise category, sessions conducted on an individual basis, in either Spanish or English, by a physiotherapist

Control group: friendly visits, reading to participants in language of their choice, activities avoided exercise and psychosocial interventions, personnel not described


OutcomesFalls: number of falls
Physical function in ADL: Physical Disability Index (Mobility Score) (Gerety 1993), Katz ADL Scale (Katz 1963)
Physical function (other): Physical Disability Index (Summary Score) (Gerety 1993)
Muscle power (anaerobic): Physical Disability Index (Strength Score) (Gerety 1993)
Balance: Physical Disability Index (Balance Score) (Gerety 1993)
Mood related: GDS (short version) (Sheikh 1986)
Cognition: MMSE
Flexibility: Physical Disability Index (Range of Motion Score) (Gerety 1993)
Perceived health status: Sickness Impact Profile (Total Score) (Bergner 1981), Sickness Impact Profile (Psychosocial Score) (Bergner 1981), Sickness Impact Profile (Physical Score) (Bergner 1981)
Medications: medications (number)
Acute health events: adverse effects, emergency department and physician visits, hospitalisations
Feasibility and acceptability: compliance
Cost: cost


NotesFunding: grants from the National Institute on Ageing and Veterans Affairs Health Services Research and Development
San Antonio nursing-home policy routinely prohibits independent bathing, which results in de facto classification of all residents as dependent in at least 1 ADL


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on random sequence generation procedure

Allocation concealment (selection bias)Low riskQuote: "Randomization was performed after baseline assessments by calling a central number" - central allocation

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo report of blinding of participants. RCT in same setting, but friendly visits could have blinded intervention/control

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Unclear riskUnclear if Katz ADL scale was performed by a blinded assessor

Incomplete outcome data (attrition bias)
All outcomes
Low riskMore losses to follow up in control group (N = 9) than in intervention group (N = 5) but all for same reason (death). Total = 194 participants

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 1 year
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: unclear
Losses to follow up: N = 4 (26.7%)


ParticipantsCountry: USA
Setting: nursing home
Randomised: 15
% women: not reported
Age: mean = not reported; range = 64 to 97

Consent: fully-informed

Inclusion criteria: see exclusion criteria

Exclusion criteria: cognitive = mental impairment (unable to understand programme description); medical = serious cardiac disease (CCF, angina), other active illness, significant

% Eligible within home: 10
% Eligible that participate: 100

Intervention group: N = 8; age range = 66 to 97 years

Control group: N = 7; age range = 64 to 87 years


InterventionsStudy aim or objective: to examine the effectiveness of an upper extremity and lower extremity exercise programme on endurance

Number of experimental groups: 2
Individual session delivery
Session duration: 15 minutes
Number of sessions per week: 3
Seated: unclear

Exercise features:
Intervention group: upper and lower body endurance programme based on target heart rates; personnel not described

Control group: usual care


OutcomesEndurance (physical other): heart rate after 2 minutes of exercise, duration of exercise


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo mention of blinding of participants, but usual care so intervention would have been obvious

Incomplete outcome data (attrition bias)
All outcomes
Low riskModerate loss to follow up, but balanced across groups with similar reasons

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: randomised controlled cross-over trial
4 phases: (1) participant screening and enrolment; (2) baseline assessment and randomisation by computer-generated random numbers into immediate intervention (group 1) or delayed intervention (group 2); (3) immediate intervention phase for which group 2 acted as group 1's control; (4) delayed intervention phase - group 2 receive the intervention and group 1 cross over to no intervention to assess the durability of the intervention's effects
Duration: 16 weeks - after 8 weeks the groups switched
Method of randomisation: computer-generated random numbers
Outcome assessor blinding: assessor masked to treatment group, but at some treatment sites, the group assignment became apparent in a small number of participants
Group comparability at entry: yes
Losses to follow up: (N = 29, 27%) 61 of 107 allocated, and 178 eligible completed all assessments


ParticipantsCountry: USA
Setting: 4 Veterans Affairs medical-centre nursing homes in the South-East of the USA (Atlanta, Georgia; Durham, North California; Salisbury, North Carolina; West Palm Beach, Florida) selected because of their proximity to the researcher's home institution in Atlanta, willingness to participate, and size of potential population
All residents at each facility screened
Randomised: 107
% women = 10
Age: mean = approximately 78 years; range = not reported
75% white
More than three-quarters of participants had at least 1 psychiatric diagnosis

Consent: informed consent from participant where capable, if not, from facility staff or from a responsible party - assent

Inclusion criteria: long-stay resident (at least 30 days and not initially admitted for short-term care), able to state their name, or in the presence of aphasia, capable of reliably pointing to 2 objects, required assistance by 2 or fewer people for transfer from bed to chair, incontinent of urine or stool, or would be without assistance from staff, not severely behaviourally disturbed, not known to be terminally ill, life expectancy of at least 6 months, not receiving active physical therapy, aged 60 and older

Exclusion criteria: see inclusion criteria

% Eligible within home: 44
% Eligible that participate: 60

Intervention: N = 52; % women = 53; age: mean = 77.8 years ± 7.6 years

Control: N = 55; % women = 50; age: mean = 78.8 years ± 6.3 years

N = 528 assessed for eligibility
N = 350 did not meet criteria
Unable to obtain consent for: N = 21
Attrition before assessment complete: N = 50
N = 107 randomised
Allocated to intervention (group 1): N = 52
Allocated to delayed intervention (group 2): N = 55


InterventionsStudy aim or objective: to test the effects of a rehabilitative intervention directed at continence, mobility, endurance, and strength (FIT) in older people living in nursing homes

Number of experimental groups: 2
Individualised intervention delivery
Session duration: n/a
Number of sessions per week: n/a
Seated: yes

Functional incidental training intervention: trained research aides provided opportunities to participate in FIT for 4 to 5 participants every 2 hours between 8am and 4pm Monday to Friday for 8 weeks, so each participant could participate in 4 FIT sessions a day; the intervention included prompted voiding and functionally-orientated endurance and strengthening exercises; individualised exercise programmes created from baseline data and modified every 2 weeks; goal for 3 sessions of FIT to involve endurance exercise (sit-to-stands, walking or wheelchair mobility to a goal time) and 4 sessions to involve strengthening exercises (bicep curls, straight arm exercises, knee extensions, and hip abductions and flexions); daily adherence recorded; supervisors conducted periodic process observations and provided additional training and reinforcement on the protocol where needed to endure quality and consistency

Control: usual care


OutcomesPhysical function in ADL: transfer time (seconds; chair to chair and back), FIM (toileting score), FIM (locomotion score), Walk or Wheel Total Time (seconds) during a 10 minute trial, Walk or Wheel (time over 6 metres), Walk or Wheel (total distance in up to 10 minutes (feet)
Physical function (other): sit-to-stand (time for first, seconds), sit-to-stand (maximum number in 30 seconds), sit-to-stand (average number in 30 seconds)
Muscle power (anaerobic): lower body strength (right hip flexion), upper body strength (right biceps curl)
Continence: appropriate toileting ratio (urine), appropriate toileting ratio (stool), fecal incontinence frequency, urinary incontinence frequency


NotesFunding: grant from the Department of Veterans Affairs Rehabilitation Services Research Service
Baseline significant difference between the intervention and control group in number of sit-to-stand exercises, with immediate intervention group able to do more


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomisation by computer-generated random numbers"

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskCross-over design, so participants would have been aware of each intervention

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
High riskProject manager (outcome assessor) was masked, but allocation revealed in some cases

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
High riskProject manager (outcome assessor) was masked, but allocation revealed in some cases

Incomplete outcome data (attrition bias)
All outcomes
High risk17/35 (49%) in immediate intervention lost versus 12/43 (28%) from control

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskThis was a cross-over trial of an intervention likely to have long-term effects. Therefore, high risk of carry-over effects. However, only the first period used in this review; therefore, low risk of bias


MethodsDesign: cluster RCT
Clustering accounted for
Details: matched wings allocated by coin toss by independent researcher
Duration: 6 months
Follow up: no


ParticipantsCharacterisation: ambulatory, need minimal assistance with transferring, independent in eating, but dependent in instrumental ADLs
Country: New Zealand
Setting: residential care homes
Randomised: 149
% women = 85%
Age details (mean (SD)):
Control: 84.7 (6.7)
Intervention: 86.8 (5.5)
Overall mean 85.7
Overall SD 6.2
Calculated from N = 149
Inclusion criteria: none
(All other residents regardless of cognitive or physical ability were eligible for participation and were invited to take part by research staff)
Exclusion criteria: under the age of 65 years, admission for respite or terminal care, quadriplegia
ADL status details: BI Score (mean, SD):
Control: 17.5 (2.7)
Intervention: 17.8 (2.2)
Cognitive status details: Abbreviated Mental Test Score (mean, SD):
Control: 7.0 (2.8)
Significant comorbidities: not reported
Assessed: 208


InterventionsStudy aim or objective: to determine whether a repetitive ADL activity programme improves health status, life satisfaction, and mobility for older people living in residential care
2 groups

Intervention: intervention group (N = 73)
Format: individual, delivered by gerontology nurse; healthcare assistant within the care home and the community
Session length: not reported, daily
Goal-setting physical activity programme
(1) goal setting phase - goal encompassing physical functionality set with the resident (e.g. gardening, attending a community-based senior citizens' club)
(2) gerontological nursing assessment
(3) development of Promoting Independent Plan (PIP) - using info from assessment, a prescriptive activity programme was developed and tailored to meet the identified goal. Activities designed to increase strength, balance, and endurance through increased ADLs. Repetitive activities included bed mobility, sitting to standing, and transfers. Longer walking routes
(5) healthcare assistant training - individualised programme explained to resident and healthcare assistant. Care plan displayed in resident's room and included in medical record. Healthcare assistant responsible on a daily basis for ensuring that residents carried out their activity programme
The residents' individualised plan was reviewed monthly at a staff meeting attended by the research staff, healthcare assistants, and nurse manager. Goals were modified or reset by the resident, if requested, following the review meeting

Control: control group (N = 76)
Usual care, waiting list for goal setting physical activity intervention


OutcomesPhysical function in ADL: TUG test (seconds) (Podsiadlo 1991), Elderly Mobility Scale (EMS) (Smith 1994)
Physical function (other): SF-36 Physical Function
Quality of life: Life Satisfaction Index (Neugarten 1961)
Perceived health status: SF-36 Mental Health
Falls, risk and fear of falling: falls (any episodes for participant)
Acute health events: adverse effects


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskToss of a coin used for allocation (between two wings of a home)

Allocation concealment (selection bias)Low riskCoin toss performed by independent researcher - suggests allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskQuote: "Care delivery in each wing was administered independently with no crossover of staff or residents during the study period"
No report of blinding of participants, although intervention and control groups were in separate wings, so may not have been aware of which group they were in

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "A research nurse blinded to the allocation then collected baseline, 3- and 6-month outcome measures"

Incomplete outcome data (attrition bias)
All outcomes
Low riskSlight imbalance in number of participants lost to follow up between groups (more in control than in intervention), although reasons similar between groups

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasHigh riskContamination: "Observations of research staff indicated that in two homes the control group residents were observed participating in activities with intervention group residents in the lounge or during walking group outings"
Quote: "The mobility measures used may not have been sensitive enough to show small but important changes"


MethodsDesign: RCT
Duration: 15 weeks
Follow up: none
Method of randomisation: randomised cross-over design, using a random numbers table; participants were allocated to either group 1 or group 2 until 1 group contained 12 participants; the remaining participants were then allocated to the other group
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: n/a
Losses to follow up: N = 8 (33%) Group 1: N = 4 (1 discharged home because of improvement with treatment, 2 deaths, 1 hip fracture during control phase); Group 2: N = 4 (2 general deteriorations, 1 circulation problems, 1 death)


ParticipantsCountry: UK
Setting: long-stay psychiatric hospital participants
Randomised: 24
% women = 67
Age: mean = not reported; range = 61 to 91 years

Consent: not specified

Inclusion criteria: diagnosis of dementia, resident of facility, requiring assistance of 1 to 2 persons for transfers, weight-bearing not precluded by hip/knee contractures, < 18 mobility score, unable to stand/mobilise independently, medically fit to participate

Exclusion criteria: medical = signs of severe osteoarthritis, cardiovascular disease, alcoholism, neurological pathology

% Eligible within home: not reported
% Eligible that participate: not reported

Group 1: N = 12
Group 2: N = 12


InterventionsStudy aim or objective: Does provision of physiotherapy input improve or maintain mobility skills in elderly people with dementing illness?

Number of experimental groups: 2
Individual sessions
Session duration: 30 minutes
Number of sessions per week: 3
Seated: yes

Group 1: physiotherapy followed by no intervention
Group 2: no intervention followed by physiotherapy

Physiotherapy comprised movement, music, body awareness, and individual functional mobility training; sessions were conducted by a physiotherapist in individual format, 3 times per week for 12 weeks, followed by 3 weeks of videoing


OutcomesPhysical function (other): Southampton Assessment of Mobility (Pomeroy 1990)
Cognition: CAPE (Clifton Assessment Procedures for the Elderly) information/orientation score (Pattie 1981)


NotesFunding: Research into Ageing grant
Pilot study


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskUsed a random numbers table

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskCross-over design - all participants experienced the intervention and a control phase. The control phase consisted of no intervention. Therefore, it would have been apparent to the participants what the intervention under study was

Incomplete outcome data (attrition bias)
All outcomes
High riskMissing data for participants who did not complete all phases - "these were equally distributed between groups 1 and 2". However, 8/24 (33%) lost to follow up

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskThis was a cross-over trial of an intervention with the potential for carry-over effects. However, only the first period used in this review so low risk of bias


MethodsDesign: RCT
Duration: 2 years
Follow up: none
Method of randomisation: not described, but participants were stratified a priori by severity of their condition using the resident classification system (RCS)
Concealment of allocation: unclear
Outcome assessor blinding: yes - maintained
Group comparability at entry: unclear
Losses to follow up: N = 52 (45%) (29 deaths/discharges, 3 unable to complete tests, 20 insufficient test results)
Individual group data not described


ParticipantsCountry: Canada
Setting: nursing Home
Randomised: 115
% women = approximately 77
Age: mean = approximately 85 years (for original participants); range = 62 to 101 years

Consent: fully-informed consent

Inclusion and exclusion criteria: not reported

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention group: N = 58; age range = 62 to 97 years; women:men ratio = 3.5:1

Control group: N = 57; age range = 63 to 101; women:men ratio = 3.1:1


InterventionsStudy aim or objective: to determine whether there is a difference in functional status among residents receiving 1 full-time physiotherapist and occupational therapist per 50 beds (enhanced) or per 200 beds (control)

Number of experimental groups: 2
Individualised intervention delivery
Session duration: not reported
Number of sessions per week: not reported
Seated: n/a

Intervention group: enhanced therapy (physiotherapy/occupational therapy), i.e. increased hours of service on a 1.0 FTE/50 bed ratio, therapy tailored to individual, content/frequency not described

Control group: usual treatment comprising minimal therapy input on a 1.0 FTE/200 bed ratio; no further details given


OutcomesPhysical function in ADL: FIM, Functional Assessment Measure, Clinical Outcomes Variables Scale
Cost: cost


NotesFunding: not reported; states that no commercial parties had any interest
Physical and occupational therapists and their assistants operated conjointly on the programmes. No differentiation between these 2 disciplines was made in this study. Treatment was offered in a restorative, consultative, monitoring, low/high maintenance programme format, as suited each participant's needs


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)High riskNo information provided on initial sequence generation. Intervention was implemented over two years, with 29 new participants recruited throughout to replace participants who died or were discharged. The researchers had no control over who died, was discharged, or recruited to the groups, and made the assumption that this was a random process
Inadequate as there is a non-random component

Allocation concealment (selection bias)High riskNo information provided on concealment during initial allocation. The replacement process may have been visible to staff, so allocation could not have been concealed

Blinding of participants and personnel (performance bias)
All outcomes
High riskStated that the testers and the staff were all blind to which group residents had been assigned. However, potential contact between residents and the occupational therapy and physiotherapy staff with other staff members means unlikely that blinding was achieved

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskStated that the testers and the staff were all blind to which group residents had been assigned

Incomplete outcome data (attrition bias)
All outcomes
High riskLarge numbers lost including due to non-testing, not clear which groups they came from

Selective reporting (reporting bias)High riskPre-intervention outcome scores were not provided (only 6 months, 12 months, 18 months, and 24 months)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering accounted for
Duration: 12 months
Follow up: no


ParticipantsCharacterisation: care-home residents
Country: USA
Setting: 12 nursing homes (8 'for profit' and 4 'not for profit')
Randomised: 487
% women = 80%
Age details: mean (SD) = 83.8 years (8.2); range = 65 to 102 years
Inclusion criteria: residents were eligible to participate if they were aged 65 and older, had a MMSE score of 11 or greater (level II eligibility - criteria applied post-consent), had a life expectancy of longer than 6 months, and were not receiving skilled rehabilitation services
ADL status details: not reported
Cognitive status details: mean MMSE score: 20.4 +/- 5.3
Significant comorbidities: not reported
Assessed: 2058
Excluded: overall = 1571, 1266 on eligibility criteria, 305 declined, 0 other


InterventionsStudy aim or objective: to test the effectiveness of a restorative care (Res-Care) intervention on function, muscle strength, contractures, and quality of life of nursing-home residents, with secondary aims focused on strengthening self-efficacy and outcome expectations
2 groups

Intervention: Res-Care Intervention group (N = 256)
Format: not reported, delivered by nursing assistants
Session length: not reported
Res-Care was a 2-tiered self-efficacy-based intervention focused on motivating nursing assistants and residents to engage in functional and physical activities. Restorative care is a philosophy of care that focuses on restoring and maintaining, residents to their highest possible functional and physical status, given comorbidities. Examples include using verbal cues during bathing, transfers, mobility, so the resident performs the tasks, rather than the nursing assistant. Homes in the Res-Care intervention group were also provided with a research restorative care nurse as a prompt to nursing assistants and residents to engage in functional physical activities and to develop short- and long-term goals for residents

Control: Control group (N = 231)
A single in-service program on managing difficult behaviours. The control site did not receive any information about restorative care, and they were not provided with a restorative care nurse


OutcomesPhysical function in ADL: BI (0 to 100 scale)
Physical function (other): Tinetti Mobility Scale (gait and balance) (Tinetti 1986), self-efficacy for functional ability (Resnick 1999; Resnick 2003)
Muscle power (anaerobic): hand grip strength
Flexibility: muscle contractures (upper extremities (fingers, wrist, elbow, and shoulders) and lower extremities (hip, knee, and ankle)
Quality of life: Dementia Quality of Life Instrument (Brod 1999)
Expectations of effects of exercise: Outcome Expectation for Functional Ability Scale (Resnick 1999; Resnick 2003)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskStaff involved in delivering intervention were not blind to participants in study/not in study

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "A team of evaluators who were blinded to randomization and unfamiliar with the details of the Res-Care intervention measured all outcomes"

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "A team of evaluators who were blinded to randomization and unfamiliar with the details of the Res-Care intervention measured all outcomes"

Incomplete outcome data (attrition bias)
All outcomes
Low riskLarge numbers lost to follow up, but balanced with similar reasons

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: n/a (one-off intervention)
Follow up: n/a
Method of randomisation: in accordance with a counterbalanced design; no further details given
Concealment of allocation: unclear
Outcome assessor blinding: no
Group comparability at entry: n/a
Losses to follow up: none


ParticipantsCountry: USA
Setting: nursing home, residential retirement home, foster care home
Randomised: 30
% women = 100
Age: mean = 80.9 years ± 9.2 years; range = 62 to 96 years

Consent: not specified

Inclusion criteria: Parachek score > 25

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported


InterventionsStudy aim or objective: to examine the effects of verbally-elicited imagery in the encouragement of exercise in elderly women

Number of experimental groups: 4
Group intervention delivery
Session duration: not reported
Number of sessions per week: n/a
Seated: yes

Exercise features:
Order 1: control condition followed by imaging
Order 2: imaging followed by control condition

Imaging: added-purpose activity, e.g. reach down as if you are picking up something from the floor

Control: rote exercise activity, e.g. reach down to the floor with both hands

2 exercises were performed as above – a reaching-up exercise and a reaching-down exercise. Interventions were one-offs, supervised by a researcher


OutcomesEndurance (physical other): duration of exercise, frequency of repetition


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "...subjects randomly assigned to different orders in accordance with a counterbalanced design"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo information given. Participants received both interventions

Incomplete outcome data (attrition bias)
All outcomes
Low riskThree lost from group 1; zero from group 2. Reasons unlikely to be related to intervention-conflicting appointments and refused to participate (cross-over)

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskCross-over design, but intervention unlikely to produce a carry-over effect


MethodsDesign: RCT
Details: multi-centre
Duration: 12 months
Follow up: no


ParticipantsCharacterisation: ambulatory participant with Alzheimer's disease
Country: France
Setting: nursing homes
Randomised: 134
% women = 75%
Age details: mean (SD) = 83 years (7.4); range = 62 to 103 years
Inclusion criteria: meet the National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association criteria for probable or possible Alzheimer's disease
Resident in the nursing home for at least 2 months
Able to transfer from a chair and walk at least 6 metres without human assistance
Exclusion criteria: evidence of vascular dementia or Parkinson's disease, planned transfer from the nursing home for surgery in the year to come, a cardiac condition that might deteriorate during exercise, diagnosis of a terminal illness with a life expectancy of less than 6 months
ADL status details: Katz ADL mean (SD) = 3.1 (1.3)
Cognitive status details: MMSE 8.8 (6.6)
Significant comorbidities: comorbidities, N (%) exercise group and routine-care group
1 comorbidity: 9 (13.4) + 16 (23.9) = 25 (18.7%)
2 comorbidities: 21 (31.3) + 15 (22.4) = 36 (26.9%)
3 or more comorbidities: 26 + (38.9) 12 (17.9) = 38 (28.4%)
Assessed: 429
Excluded: overall = 295, 187 on eligibility criteria, 65 declined, 43 other


InterventionsStudy aim or objective: to investigate the effectiveness of an exercise program in improving ability to perform ADLs, physical performance, and nutritional status and decreasing behavioral disturbance and depression in participants with Alzheimer's disease
2 groups

Intervention: Physical exercise program (N = 67)
Format: group, delivered by occupational therapist in care home
Session length: 60 minutes, twice weekly
Personalised walk, strength, balance, and flexibility training to music: A walking route was established around each home passing all participants' rooms. Participants were paired and encouraged to walk fast enough to become somewhat breathless. Around the trail were stations for strength, balance, and flexibility training. Strength training included squats, heel raises and leg lifts; balance training involved 1 or 2 leg balance exercises and a small step test with cones and hoops; flexibility training involved repetition of demonstrated exercises

Control: control (N = 67)
Usual care


OutcomesPhysical function in ADL: Katz ADL Scale (Katz 1963), Get Up and Gotest (Mathias 1986), six-metre walk (speed m/s)
Balance: one-leg balance test (Vellas 1997)
Mood related: depression (Montgomery-Assberg Depression Rating Scale; MADRS)
Anthropometry: weight
Nutrition: Mini-Nutritional Assessment (MNA)
Psychiatric status: Neuropsychiatric Inventory (NPI) (Cummings 1994)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Staff not involved in intervention or assessment performed separate randomization at each site by lottery draw"

Allocation concealment (selection bias)Unclear riskQuote: "Staff not involved in intervention or assessment performed separate randomization at each site by lottery draw"

However, unclear if drawing was concealed, e.g. open hat

Blinding of participants and personnel (performance bias)
All outcomes
High riskQuote: "...single-blind study..."
Residents assigned to exercise or usual care so assignment obvious

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskA single geriatrician who was blinded to the intervention assignment measured outcomes at baseline, 6 months, and 12 months on different days from the intervention

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskA single geriatrician who was blinded to the intervention assignment measured outcomes at baseline, 6 months, and 12 months on different days from the intervention

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing outcome data balanced across groups (13 control, 11 exercise) with similar reasons

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: approximately 3 months (13 weeks) - 29 occasions
Follow up: 12 weeks
Method of randomisation: after inclusion of participants and baseline assessments, 34 clusters of 3 to 9 participants living on the same floor, wing, or unit were randomly assigned to exercise or control activity
Randomisation was stratified in order to have both groups in each facility; within each cluster, the nutrition intervention was randomised individually
Randomisation using lots in sealed envelopes
Concealment of allocation: yes
Outcome assessor blinding: yes, checked at 3 months - if they correctly guessed the participant's group, they were replaced - the case for 11% of participants; checked again at 6 months (1%)
Group comparability at entry: no, significant differences, P < 0.05 for perception of health and prescriptions for proton pump inhibitors
Losses to follow up: 28 (15%)


ParticipantsCountry: Sweden (Umeå) - frail older people - activity and nutrition study (FOPANU study)
Setting: 9 residential care facilities
Randomised: 191 (of 487 screened)
% women = 73
Age: mean = 84.7 years ± 6.5 years; range = 65 to 100 years

Consent: assent accepted

Inclusion criteria: aged 65 years, dependent on assistance from a person in 1 or more ADL according to the Katz index, able to stand up from a chair with arm rests with help from no more than 1 person, MMSE score of 10 or higher, approval from the resident's physician

Exclusion criteria: see inclusion criteria

% Eligible within home: 39
% Eligible that participate: not reported

Exercise + protein drink: 85.0 years ± 6.7 years; women = 78%

Exercise + placebo drink: 85.5 years ± 5.5 years; women = 69%

Control + protein drink: 82.9 years ± 6.4 years; women = 70%

Control + placebo drink: 85.6 years ± 7.0 years; women = 74%


InterventionsStudy aim or objective: to determine whether a high-intensity functional exercise programme improves balance, gait ability, and lower limb strength in activities of daily living, and if an intake of protein-enriched energy supplement immediately after the exercise increases the effects of the training

Number of experimental groups: 4
Group intervention delivery
Session duration: no longer than 45 minutes
Number of sessions per week: 5
Seated: unclear
Groups: both an exercise intervention compared with control activity and a nutrition intervention compared with a placebo in a 2 x 2 factorial model
Both exercise and control held within the facility; similar distance from where participants stayed; where a participant did not attend the session, individual activity was offered where possible

Exercise features: groups of 3 to 9 participants, supervised by 2 physiotherapists
Exercise intervention: based on the high-intensity functional exercise programme (HIFE programme of Littbrand 2006), functional exercises consisting of everyday tasks challenging leg strength, postural stability, and gait ability; exercises selected for each participant according to their deficits; all performed in weight-bearing positions; encouraged to exercise at high intensity and to increase load and difficulty progressively, considering changes in function and health status; tasks followed up after 3 months by asking staff about compliance during the previous 2 weeks

Control: developed by occupational therapists and involved activities while sitting watching films, reading, singing, and conversing; groups of 3 to 9 participants, supervised by 1 occupational therapist; based on themes - the old country shop, famous persons, games from the past; designed to be stimulating, even to people with cognitive impairment

Nonexercise features:
Nutrition intervention: protein enriched energy supplement, placebo-drink control packaged in the same way as the intervention drink and had similar flavours


OutcomesFalls: falls (any episodes for participant), fall rate (falls per person years)
Physical function in ADL: Modified Chair Stand (Guralnik 1994), BI (0 to 20), gait speed (self paced) 2.4 metres, gait speed (maximum) 2.4 metres
Muscle power (anaerobic): lower-limb strength (1 repetition-maximum)
Balance: BBS
Adverse events (other): adverse event rate (within intervention)
Feasibility and acceptability: performance of high-intensity strength training and balance exercises, attendance


NotesFunding: grants from the City Council of Västerbotten, the Vårdal foundation, the Magnus Bergvalls Foundation, the Äldrecentrum Västerbotten, the Umeå University Foundation for Medical research, the Gun and Bertil Stohne Foundation, Erik and Anne-Marie Detlof’s foundation, the Loo and Hans Ostermans Foundation, the Borgerskapet in Unmeå Research foundation, the Swedish Research Council and the Swedish Council for Working Life and Social Research and Norrmejerier


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskStated that researchers not involved in the study performed the randomisation using lots in sealed non-transparent envelopes

Allocation concealment (selection bias)Low riskStated that researchers not involved in the study performed the randomisation using lots in sealed non-transparent envelopes

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskEven though the clusters were in separate flats, they were within the same facilities. As a result, there may have been contact between the exercise intervention group and the control group

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskTrained physiotherapists blind to group allocation undertook assessment for mobility/balance outcome measures

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskStated ITT, but results appear to be 'as treated' (N at 3 months and 6 months is different to N randomised)
N lost = 14 in both groups, but twice as many died in the exercise group

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Duration: 12 weeks
Follow up: 12 weeks
Method of randomisation: carried out independently by a statistician, homes grouped into 4 strata, using combinations of type (residential, nursing, both), funding source (private or local authority), and setting (urban or rural). Within each stratum, pairs of homes were allocated randomly, using computer-generated random numbers
Concealment of allocation: yes
Outcome assessor blinding: yes
Group comparability at entry: yes
Losses to follow up: N = 13 (11%)


ParticipantsCountry: UK
Setting: 12 care homes, (approached the managers of 14 homes: 1 refused, 1 home used as a pre-pilot) 12 entered into the study in 3 groups of 4 to control therapists' workload
Randomised: 118
% women = 82
Age: mean: approximately 87 years; range = 44 to 102 years

Consent: not specified

Inclusion criteria: residents with stroke, staff asked to screen people with the BI, information on stroke history and cognitive status for the purpose of consent

Exclusion criteria: medical = acute illness, terminally ill

% Eligible within home: 46
% Eligible that participate: 61.8

Intervention: 88.6 years ± 6.5 years (62 to 102 years); 83% women
Control: 86.3 years ± 8.8 years (44 to 99 years); 82% women

Intervention: 6 homes; 63 residents, (3 months: 59 assessed, 3 died; before occupational therapy: 1 died during/after treatment; 6 months: 53 assessed, 6 died) lost 10 to follow up.
Control: 6 homes; 55 residents (3 months: 46 assessed, 9 died; 6 months: 35 assessed, 11 died) lost 20 to follow up


InterventionsStudy aim or objective: evaluation of occupational therapy intervention to improve self-care independence for residents with stroke related disability living in care homes

Number of experimental groups: 2
Individualised intervention delivery
Session duration: n/a
Number of sessions per week: n/a
Seated: n/a

Intervention: provided by experienced occupational therapist delivered to the individual, targeted at improving independence in personal activities of daily living, frequency and duration dependent on resident's and therapist's agreed goals, took place over a 3-month period, intervention group given interview of 1 hour to establish functional ability and agree goals

Control: usual care


OutcomesPhysical function in ADL: RMI (Collen 1991), BI (0 to 20)
Death and physical deterioration (chronic): poor global outcome (deterioration in BI or death)


NotesThe Stroke Association, Health Foundation, Department of Health Research Capacity Development Program
Pilot study


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Homes were allocated randomly, using computer-generated random numbers"

Allocation concealment (selection bias)Low riskQuote: "Randomisation was carried out independently by a statistician"

Blinding of participants and personnel (performance bias)
All outcomes
High riskQuote: "Allocation was revealed only to the occupational therapist, not to the assessors"
Control was usual care, so participants could not be blinded to allocation

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "Allocation was revealed only to the occupational therapist, not to the assessors"
Quote: "Assessments were completed...by 1 of 4 research staff masked to the trial allocation"

Incomplete outcome data (attrition bias)
All outcomes
Low riskImbalance in numbers lost to follow up between intervention and control group, but this was largely due to many deaths in the control group, and unlikely to be related to intervention

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering not accounted for.
Details: phase II, exploratory randomised controlled trial with cluster randomisation (at the level of care home)
Duration: 4 weeks
Follow up: yes


ParticipantsCharacterisation: care-home residents with a range of functional, cognitive, and continence impairments
Country: UK
Setting: 6 care homes (2 provided nursing)
Randomised: 34
% women = 88%
Age details: mean (SD) = 86 (9); range = 76 to 101 years
Inclusion criteria: staff used their knowledge of residents' functional, cognitive, and continence status to select 10 residents with a range of functional, cognitive, and continence impairments
Exclusion criteria: none stated
ADL status details: BI ADL, mean (SD): Intervention: 10.7 (5.3); Control: 9.8 (4.4)
Cognitive status details: Short Orientation-Memory-Concentration test < 22: 23 (70%)
Significant comorbidities: incontinent of urine: 21 (64%)
Assessed: 211
Excluded: overall = 177, 0 on criteria, 26 declined, 151 other


InterventionsStudy aim or objective: to assess feasibility, acceptability and potential efficacy of group exercise and staff education intervention to promote continence in older people residing in care homes. To establish measures and information to inform a larger trial.
2 groups

Intervention: Intervention (N = 17)
Format: group, delivered by physiotherapist
Session length: 60 minutes, twice weekly
Mobility training protocol - participants were encouraged to walk or wheel to class. They then practised the task-related training of functional ADLs (e.g. practising standing up from a chair) and strength, balance, endurance, and flexibility exercises. Music played during the class, and exercises were fun, making use of balloons and balls
Pre-training and post-training prompted voiding and fluid intake were included to promote continence
Residents set their own pace but were encouraged to improve on previously achieved goals. Residents’ progress was reviewed and their views were gathered
The staff education component comprised separate 2-hour workshops on continence care and mobility care. The continence training was delivered by specialist nurses from the local Primary Care Trust Continence Team and the mobility training by a qualified physiotherapist and occupational therapist

Control: control (N = 17)
Standard care which involved very little expert care


OutcomesPhysical function in ADL: RMI (Collen 1991)
Continence: continence (urodynamic questionnaire) (Matharu 2005)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskAn independent statistician randomly allocated three care homes to each study group, using computer-generated random numbers

Allocation concealment (selection bias)Low riskAn independent statistician randomly allocated three care homes to each study group, using computer-generated random numbers

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo report of blinding of participants; those in control group received usual care, so would have been obvious that they weren't receiving an intervention

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "Study outcomes were assessed...by an assessor who was masked to allocation"

Incomplete outcome data (attrition bias)
All outcomes
Low riskImbalance in number of participants lost to follow up between groups (more in control than in intervention - unlikely to be related to the intervention)

Selective reporting (reporting bias)Unclear riskProtocol unavailable. Mean mobility scores (RMI) reported without providing SD or CI

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering accounted for
Duration: 3 months
Follow up: yes


ParticipantsCharacterisation: care home residents with mobility limitations and limitations in ADLs
Country: UK
Setting: care homes providing care for physical disability and older people with more than 5 beds
Randomised: 249
% women = 74%
Age details: mean = 85, SD = 9
Inclusion criteria: 5 ≤ BI score ≤ 16
Exclusion criteria: admitted to hospital with acute illness
Admitted to the care home for end-of-life care
ADL status details: none stated
Cognitive status details: MMSE scores:
< 21: 168 (67%)
21 to 23: 21 (8%)
> 24: 60 (24%)
Significant comorbidities: 56% with arthritis
23% with cardiovascular problems
40% dementia
36% diabetes
22% with at least 1 confirmed stroke (conservative, also reported as 24% and 46%)
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to compare effectiveness of physiotherapy and occupational therapy with standard care in care-home residents who have mobility limitations and are dependent in performing ADLs
2 groups

Intervention: intervention (physiotherapy + occupational therapy) (N = 128)
Format: individual, delivered by 2 qualified physiotherapists, occupational therapists, staff training providers unstated
Session length: not reported, but therapy customised to the individual, not reported

Physiotherapy and occupational therapy intervention:
Physiotherapy aimed at enhancing mobility and ability to perform ADLs independently through practising functional tasks and therapy for components (e.g. flexibility, balance). Customised to the individual
Occupational therapy aimed at increasing independence in ADL through routine assessment, treatment, and reassessment. In addition, staff were trained in promoting independence and use of therapeutic aids

Control: control (N = 121)
Standard care as before the trial. Occupational therapy not routinely used by any of the homes and physiotherapy only accessed by GP referral


OutcomesPhysical function in ADL: TUG test (seconds) (Podsiadlo 1991), RMI (Collen 1991), BI (0 to 20)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Randomisation was performed by an independent principal statistician who used a computer-generated randomisation list"

Allocation concealment (selection bias)Low riskQuote: "Randomisation was performed by an independent principal statistician who used a computer-generated randomisation list"

Blinding of participants and personnel (performance bias)
All outcomes
High riskQuote: "Treatment arm was revealed to the treating therapists only, thereby ensuring that allocation was concealed from the independent assessors responsible for all subsequent assessments"
Control group received usual care

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Treatment arm was revealed to the treating therapists only, thereby ensuring that allocation was concealed from the independent assessors responsible for all subsequent assessments"

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "Treatment arm was revealed to the treating therapists only, thereby ensuring that allocation was concealed from the independent assessors responsible for all subsequent assessments"

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups
Note: If TUG had been analysed and reported it probably would have led to an assessment of high risk of bias due to incomplete outcome data (see selective reporting)

Selective reporting (reporting bias)High riskHADS-D was described as an outcome measure, but not reported post-intervention and its exclusion not discussed. TUG was not analysed or reported because so few participants were able to complete it

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Details: blocked randomisation (by gender)
Single-blind
Duration: 12 weeks
Follow up: no


ParticipantsCharacterisation: people with Alzheimer's disease
Country: Spain
Setting: residential nursing home
Randomised: 16
% women = 38%
Age details: mean (SD):
Training group: 76 (4)
Control group: 73 (4)
Inclusion criteria: diagnosed by a trained geriatrician; with Alzheimer's disease of low-medium grade, i.e. score ranging between 18 to 23 in the Spanish; validated version for the general geriatric population of the MMSE; to have lived in the nursing home for at least 4 months; free of neurological (other than Alzheimer's disease), vision, muscle, or cardiorespiratory disorders
Exclusion criteria:
ADL status details: not reported
Cognitive status details: mean (SD) MMSE score:
Training group: 20.1 (2.3)
Control group: 19.9 (1.7)
Significant comorbidities: Alzheimer's disease
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to determine the effects of a 12-week training program for Spanish participants with Alzheimer's disease on their (1) overall functional capacity (muscle strength and flexibility, agility and balance while moving, and endurance fitness), and (2) ability to perform ADLs
2 groups

Intervention: training group (N = 8)
Format: group, delivered by exercise scientist in a room inside the nursing home
Session length: approximately 75 minutes, 3 times weekly
36 programmed training sessions
Each session started and ended with a 15-minute warm-up and 15-minute cool-down period, respectively, consisting of walking without reaching breathlessness (on an inside walking trail) and "gentle" stretching exercises for all major muscle groups. The core portion of the training session was divided into joint mobility, resistance, and co-ordination exercises. Joint mobility exercises focused on shoulder, wrist, hip, knee, and ankle joints. Resistance training included 9 exercises with elastic medium-resistance bands (3 sets of 15 repetitions each) engaging some of the major muscle groups: chest, biceps, triceps, shoulder, knee extensors, abductor and adductor muscles, and calf muscles. All exercises were performed through the full range of motion normally associated with correct technique for each exercise. Stretching exercises of muscles were performed at the end of each set of resistance exercises. Co-ordination exercises were performed with foam balls of gradually decreasing size over the program, e.g. bouncing a ball with both hands, tossing and catching a ball, etc. Music (from the participants' youth years) accompanied each session

Control: Control group (N = 8)
Routine nursing/medical care
Did not perform any type of programmed physical activity, except those necessary for daily living, i.e. normal ambulation inside the nursing home


OutcomesPhysical function in ADL: TUG test (modified to 8 feet) (Bassey 1992; Rikli 1999), Katz ADL Scale (Katz 1963), BI (0 to 100 scale)
Physical function (other): sit-to-stand (average number in 30 seconds), Tinetti Mobility Scale (gait and balance) (Tinetti 1986), 2-minute step in place
Muscle power (anaerobic): arm curl
Flexibility: Back Scratch test (Rikli 1999), 'Sit-and-reach' test


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskUnclear if participants were blinded, but they all came from the same nursing home, and the control group did not receive any intervention so would probably have been obvious which group they were in

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskReported that the study was single-blind, and the exercise scientist performing the evaluations was different to the 1 delivering the intervention, implying that the assessor was blind to allocation

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskReported that the study was single-blind, and the exercise scientist performing the evaluations was different to the one delivering the intervention, implying that the assessor was blind to allocation

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo losses to follow up reported

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: none
Method of randomisation: not specified
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes
Losses to follow up: N = 2 (16.7%)


ParticipantsCountry: USA
Setting: VA Medical Centre nursing unit
Randomised: 14 (12 individuals) (reports for I = 8, C = 6, but 2 participants were lost from the initial intervention group; these were later replaced by 4 control participants; therefore, outcome data are only reported for different 10 individuals, with 4 completing both intervention and control conditions)
% women = 0
Age: mean = 73 ± 4 years; range = not reported

Consent: fully-informed

Inclusion criteria: > 60 years, independently mobile without aid, evidence of gait/balance difficulties (Tinetti Score 30 or less), lower limb weakness (quadriceps/hamstrings < 5), isokinetic quadriceps and hamstrings < 80% of age-predicted normal

Exclusion criteria: cognitive = moderate-severe dementia (MMSE < 22); medical = asymmetrical focal neurological deficit, lower limb amputation, lower limb discrepancy > 1 inch, significant systematic disease, e.g. cancer; functional = refusal of consent

% Eligible within home: 12
% Eligible that participate: 52.2

Intervention group: N = 6; mean age = 73.38 years ± 4.04 years; all male

Control group: N = 6; mean age = 73.83 years ± 4.74; all male


InterventionsStudy aim or objective: to determine whether a moderate to high intensity strengthening and aerobic exercise programme can improve the strength, exercise capacity, gait, and balance of deconditioned nursing-home residents

Number of experimental groups: 2
Group intervention delivery
Session duration: 60 minutes
Number of sessions per week: 3
Seated: unclear

Intervention group: progressive-resistance lower-limb weight training and aerobic conditioning, group format: 20 minutes of aerobic exercise, 10 repetitions per lower limb exercise, conducted 3 times per week for 12 weeks; personnel not described

Control group: usual care with maintenance physiotherapy when indicated


OutcomesPhysical function in ADL: gait speed/velocity (left and right each)
Physical function (other): Tinetti Test - gait, stride length, steps per minute (step cadence), gait duration (step time, seconds)
Muscle power (anaerobic): isokinetic strength (quadriceps, hamstrings, knee flexion and extension; each and combined, left, right, and combined; Sauvage et al)
Balance: stance time (% of total time), Tinetti Test modified sub-scale (6 items combining strength and balance) (Sauvage 1992), stance time (seconds), Tinetti Test - Body Balance, balance (Sauvage 1992) (eyes open and eyes closed; average distance from centre of pressure and total distance travelled by centre of pressure (mm))
Endurance (physical other): heart rate, VO2 max, knee resistance repetitions (number of reps completed at 180 degrees/second before strength declined to < 50% of peak torque) (Sauvage 1992)
Acute health events: adverse effects, hospitalisations
Feasibility and acceptability: attendance


NotesFunding: Department of Veterans Affairs Medical Research Service and Rehabilitation Research and Development Service
Because of resource limitations, it was not possible for all outcome measures to be assessed by blinded raters. However, raters who did not know the residents' group assignment did blinded ratings of FIT assessment performance during approximately 10% of the post-intervention assessments in order to help minimise potential bias of un-blinded raters


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)High riskQuote: "...were then randomized." No details reported. However, randomisation not specified at all for subsequent reduction to 12 participants or for the 4 control participants crossed over

Allocation concealment (selection bias)High riskQuote: "...were then randomized." No details reported. However, randomisation not specified at all for subsequent reduction to 12 participants or for the 4 control participants crossed over

Blinding of participants and personnel (performance bias)
All outcomes
High riskNot reported, but control was usual care; therefore, obvious

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo report of blinding of outcome assessment

Incomplete outcome data (attrition bias)
All outcomes
High risk2 of 6 original exercise-group participants (33%) lost to follow up. No control group participants lost (1st phase). Unknown number lost from original allocation when group reduced to 6

Selective reporting (reporting bias)High riskOriginal protocol not available, but outcomes reported in such a way that they could not be used in a meta-analysis (mixing of related and independent participants)

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 8 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: partial (see notes)
Group comparability at entry: yes
Losses to follow up: intervention group N = 18 (19.1%)


ParticipantsCountry: USA
Setting: 5 proprietary nursing facilities
Randomised: 94
% women = 78
Age: mean = 85.1 years ± 8.2 years; range = not reported

Consent: assent acceptable

Inclusion criteria: incontinent of urine, passing basic cognitive screen

Exclusion criteria: cognitive = severe cognitive impairment that precluded participation; medical = indwelling catheters; functional = unable to weight bear, unable to propel wheelchairs because of irreversible physical limitations, e.g. paralysis

% Eligible within home: 75
% Eligible that participate: 34.6

Intervention group: N = 36

Control group: N = 40


InterventionsStudy aim or objective: to determine if an exercise intervention (FIT) results in improvements in mobility, endurance, and physical activity when compared with prompted voiding among cognitively and mobility impaired residents

Number of experimental groups: 2
Individualised intervention delivery
Session duration:
Number of sessions per week: 20
All components were conducted after each of 4 prompted voiding episodes per day, 5 days per week, and progressed over the 8-week study period
Seated: unclear

Intervention group: prompted voiding and FIT exercise intervention, comprised incontinence care, and social interaction, 1 to 2 stands, 1 transfer, walking/wheeling exercises, and sit-to-stand

Intervention group received approximately 2 times greater input than the control group, delivered by research staff on an individual basis over 8 weeks

Control group: prompted voiding only, comprised incontinence care and social interaction, 1 to 2 stands, and 1 transfer, conducted every 2 hours, 4 times per day, 5 days a week for the 8-week period, delivered by research staff on an individual basis


OutcomesPhysical function in ADL: wheelchair mobility endurance (wheel as long as he/she could) (Schnelle 1995), walking endurance (walk as long as he/she could (minutes)) (Schnelle 1995), wheelchair speed over maximal wheeling distance, wheelchair speed over 6 metres, walking speed over maximal walking distance, six-metre walk (speed m/s)
Physical function (other): sit-to-stand (average number in 30 seconds)
Agitation: agitation (daytime behavioural observation)
Physical activity: behavioural observations (standing/walking)
Energy expenditure: physical activity (kCal/hour)
Feasibility and acceptability: adherence (percentage of sessions in which all, part, or none of the exercise goal was achieved), attendance


NotesNational Institute on Aging (NIA) Pepper Centre grant
Ratings of FIT assessment performance during approximately 10% of the post-intervention assessments in order to help minimise potential bias of un-blinded raters


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskAuthors state that the residents were randomised to receive either prompted voiding or prompted voiding plus FIT, but do not specify how random sequence was generated

Allocation concealment (selection bias)Unclear riskAuthors state that the residents were randomised to receive either prompted voiding or prompted voiding plus FIT, but do not specify if allocation was concealed

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskBlinding of participants and personnel not reported. Same setting, but two interventions so could have been blinded

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
High riskQuote: "It was not possible for all outcome measures to be assessed by blinded raters"
Did manage to perform blinded ratings of approximately 10% of assessments

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskUnclear if there was any missing outcome data - number of residents on which outcome data were assessed was not reported

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 9 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: partial (50% of assessments were blind)
Group comparability at entry: yes
Losses to follow up: 26 (26.8%): intervention group N = 12 (8 deaths/transfers, 4 refused to comply); control group N = 14


ParticipantsCountry: USA
Setting: 5 proprietary nursing facilities
Randomised: 97
% women = not reported
Age: mean = 84 years; range = not reported

Consent: not specified

Inclusion criteria: > 65 years of age, medical order for physical restraint or visual documentation of restraint use by research staff, basic cognitive and behavioural responsiveness

Exclusion criteria: medical: paralysis, contracture, foot drop, severe arthritic pain

% Eligible within home: 94
% Eligible that participate: 80.7

Intervention group: N = 47
Control group: N = 50


InterventionsStudy aim or objective: to evaluate an exercise protocol designed to improve strength and mobility and decrease injury risk factors in physically restrained nursing-home residents

Number of experimental groups: 2
Individual intervention delivery
Session duration: not reported
Number of sessions per week: 3
Seated: if necessary

Intervention group: exercise safety intervention protocol; comprised mobility exercise, safety practice, rowing endurance, and strengthening exercises; targeted pre-set goals and progressed by 10% each week; conducted on an individual basis by a research staff member, 3 times per week for 9 weeks

Control group: usual care


OutcomesRisk of falling: Safety Assessment for the Frail Elderly (SAFE)Transition Score (Schnelle 1994), SAFE Walk Score (Schnelle 1994), SAFE Total Score (Schnelle 1994), SAFE Judgement Score (Schnelle 1994)
Physical function in ADL: walk or wheel time per day (estimated from observations), walking endurance (walk as long as he/she could (seconds), Schnelle 1995), wheelchair speed over 6 metres, six-metre walk (speed m/s), wheelchair mobility endurance (wheel as long as he/she could, Schnelle 1995)
Physical function (other): sit-to-stand (average number in 30 seconds), sit-to-stand (time for first, seconds)
Muscle power (anaerobic): row force (force produced during rowing, Schnelle 1996), hand grip strength
Flexibility: row range of motion (Schnelle 1996)
Endurance (physical other): row time (row as long as possible) (Schnelle 1996)
Other: restraint use
Feasibility and acceptability: attendance, adherence (percentage of sessions in which all, part, or none of the exercise goal was achieved)


NotesNIA Pepper Centre grant
Blinded evaluation on all mobility assessments was accomplished in more than 50% of the observations in all homes but the first site; no significant inter-site difference in outcome data was identified as a result of this


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskAuthors state that residents were randomised into 1 of the 2 groups, but do not indicate how random sequence was generated

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo report of blinding of participants, but usual care so intervention would have been obvious

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
High riskQuote: "Blinding was accomplished on only 50% of the observations"

Incomplete outcome data (attrition bias)
All outcomes
High riskMissing data - standing and walking outcomes could only be measured in ambulatory residents; SAFE assessment not available for the first nursing home

Selective reporting (reporting bias)Unclear riskFigures not reported for all prespecified outcome measures, but commented upon in results section

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 8 months
Follow up: none
Method of randomisation: computerised randomisation program
Concealment of allocation: unclear
Outcome assessor blinding: partial - 1 of 2 observers was blind
Group comparability at entry: yes
Losses to follow up: at 8 weeks N = 18 (22%)


ParticipantsCountry: USA
Setting: 4 nursing homes (long-stay beds)
Randomised: 190
N = 330 met inclusion criteria
N = 257 gave informed consent
N = 190 baseline assessments completed
% women = approximately 84
Age: mean = 87 years ± 8 years; range = not reported

Consent: assent accepted

Inclusion criteria: incontinent of urine (free of a catheter), able to follow a one-step instruction

Exclusion criteria: medical: residents of post-acute skilled care units, terminal illness, catheterised

% Eligible within home: 73
% Eligible that participate: 57.6

Intervention group: N = 9; age range = 71 to 95 years; 8 women

Control group: N = 7; age range = 65 to 70 years, 76 to 95 years; 4 women


InterventionsStudy aim or objective: to examine clinical outcomes and describe the staffing requirements of an incontinence and exercise intervention

Number of experimental groups: 2
Individualised intervention delivery
Session duration: not reported
Number of sessions per week: every 2 hours up to a maximum of 4 episodes per day, 5 days per week for 32 weeks
Seated: where necessary

Intervention group: prompted voiding, walking/wheeling, sit-stands, supervised by research staff, once daily upper limb resistance training

Control group: usual care


OutcomesPhysical function in ADL: walked maximum 10 minutes, wheeled (metres average) 10 minutes (Schnelle 2002), wheeled maximum 10 minutes, standing test (level of assistance) (Schnelle 2002), walked (metres average) 10 minutes (Schnelle 2002), 10-minute walk/wheel (average distance), walked and wheeled (metres maximum) 10 minutes (Schnelle 2002)
Physical function (other): sit-to-stand (maximum number in 30 seconds), sit-to-stand (average number in 30 seconds)
Muscle power (anaerobic): arm curl (maximum lift in pounds), arm raise (maximum lift in pounds)
Physical activity: behavioural observations (standing/walking)
Continence: appropriate toileting ratio (stool), urinary incontinence frequency, appropriate toileting ratio (urine), bowel movement frequency (incontinent and continent), fecal incontinence frequency
Anthropometry: weight
Energy expenditure: energy expenditure by motion sensor
Nutrition: food and fluid intake during meals (average total per cent)
Pain: Geriatric Pain Measure (modified 13 items), pain (number of pain reports per metre of mobility), MDS: documentation of pain (Section J, item 2a: frequency)
Acute health events: acute healthcare conditions (episodes) (Schnelle 2003)
Cost: healthcare cost (Schnelle 2003)


Notes"Supported by Grant AG13013 from the National Institutes of Health: Mobility and Incontinence Management Effects on Sickness, and Grant AG10415 from the National Institute on Aging: UCLA Claude D. Pepper Older Americans Independence Center"


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Residents were randomized into intervention and control groups using a computerized randomization program completed after baseline assessments"

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskLikely to have been incomplete blinding. Whilst attempts were made to blind staff involved in the trial "blinding observers to group assignment whenever possible", participants would probably have been aware of treatment allocation

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
High riskStated: "...blinding observers to group assignment whenever possible", suggesting this was not accomplished all of the time

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskLoss to follow up moderate, but individual groups not reported (*Primarily* because of death or prolonged illness)

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: 12 weeks
Method of randomisation: participants were matched in pairs, according to their Risk Assessment for Fall Scale II, and then randomly assigned within each pair to the intervention or control group
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: yes
Losses to follow up: N = 0 at 12 weeks, at follow up: intervention group N = 2 (illness/death); control group: none


ParticipantsCountry: USA
Setting: 2 long-term care facilities
Randomised: 16
% women = 75
Age: mean = 82.8 years; range = 65 to 95 years

Consent: fully-informed consent

Inclusion criteria: > 65 years of age, independently mobile (+/- aid), able to speak/understand English, MMSE score > 20

Exclusion criteria: medical = unstable physical condition, terminal illness; functional = abusive behaviour

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention group: N = 9; age range = 71 to 95 years; 8 women

Control group: N = 7; age range = 65 to 95 years; 4 women


InterventionsStudy aim or objective: to investigate the role of exercise in preventing falls, specifically assessing the effectiveness of an ankle strengthening and walking programme to improve balance, ankle strength, walking speed, and falls efficacy and to decrease falls and fear of falling

Number of experimental groups: 2
Unclear whether intervention delivery is group or individualised
Session duration: 20 minutes
Number of sessions per week: 3
Seated: no

Intervention group: ankle strengthening programme (heel raises), walking programme (increasing speed/distance), intervention delivered by a researcher

Control group: usual care


OutcomesRisk of falling: Fall Risk Assessment (RAFS II) (Ross 1991)
Falls: number of falls
Fear of falling: fear of falling (single item 4-point scale) (Tinetti 1990), Falls Efficacy Scale (modified) (Schoenfelder 2000) (10 item fear of falling)
Physical function in ADL: six-metre walk (time)
Muscle power (anaerobic): ankle strength (number of heel raises in 30 seconds)
Balance: balance (tandem stance, up to 10 seconds), balance (semi-tandem stance, up to 10 seconds), balance (parallel stance, up to 10 seconds)
Cognition: MMSE
Physical activity: behavioural observations (standing/walking)


NotesFunding: Gerontological Nursing Interventions Research Center grant


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Subjects were matched in pairs and assigned randomly within each pair"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo report of blinding of participants, but would probably have been obvious to the participants which group they were in. Control was usual care in same setting

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo report of blinding of outcome measurement

Incomplete outcome data (attrition bias)
All outcomes
High riskMissing outcome data for 2 (of 9) participants in the intervention group who could not complete 6-month measures

Selective reporting (reporting bias)High riskSome outcome measures not reported
No results reported for the statistical analysis procedure reported to have been used in Data Analysis section, simply a comment that "statistical significance was not reached"

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT, matched-pairs design
Duration: 3 months
Follow up: 3 months
Method of randomisation: matched in pairs by Risk Assessment for Falls Scale II scores (RAFS II)
Concealment of allocation: yes - where participants were roommates or spouses, they were assigned to the same group to lessen the possibility of contamination
Outcome assessor blinding: yes - no contact with participants other than assessments
Group comparability at entry: yes
Losses to follow up: at 3 months N = 15 (18.5%)

Intervention:
Baseline N = 42
3 months N = 33
6 months N = 30 (-12)

Control:
Baseline N = 39
3 months N = 33
6 months N = 28 (-11)


ParticipantsCountry: USA
Setting: 10 private urban nursing homes in Eastern Iowa, ranging from 68 beds to 178 beds
Randomised: 81
% women = 77
Age: mean = 84.1 years; range = 64 to 100 years

Consent: fully-informed

Inclusion criteria: = 65 years, able to ambulate independently or with an assistive device (so they could take part in an ankle strengthening and walking programme), could speak English, did not have an unstable physical condition, did not have evidence of an end-stage terminal illness, no history of acting out or abusive behaviour, had score of 20 or above on MMSE, doctor's consent sought

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention: N = 42, women N = 30
Control: N = 39, women N = 32


InterventionsStudy aim or objective: to test a 3-month ankle strengthening and walking programme designed to improve or maintain fall related outcomes

Number of experimental groups: 2
Individualised programme delivery
Session duration: 15 to 20 minutes
Number of sessions per week: 3
Seated: no

Intervention: 3-month ankle strengthening and walking programme, 3 times weekly, 15 to 20 minutes, programme tailored to individual ability

Control: attention placebo to control for effects of attention and motivation, visited weekly by same research team member who conducted the exercise programme, devoted 30 minutes to an activity such as book reading or 'friendly visiting'


OutcomesFear of falling: fear of falling (single item 4-point scale) (Tinetti 1990), Falls Efficacy Scale (modified) (Schoenfelder 2000) (10 item fear of falling)
Risk of falling: Fall Risk Assessment (RAFS II) (Ross 1991)
Physical function in ADL: six-metre walk (time)
Muscle power (anaerobic): ankle plantar/flexor strength
Balance: balance (tandem stance, up to 10 seconds), balance (semi-tandem stance, up to 10 seconds), balance (parallel stance, up to 10 seconds)
Cognition: MMSE


NotesFunding: NIH grant


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Subjects...randomly assigned within each pair to intervention or control group"

No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskBlinding attempted, but may have been broken

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "For all assessments conducted at 3 and 6 months, examiners doing the assessments had no contact with the participants other than the assessments once group assignments were made"

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskQuote: "For all assessments conducted at 3 and 6 months, examiners doing the assessments had no contact with the participants other than the assessments once group assignments were made"

Incomplete outcome data (attrition bias)
All outcomes
Low riskSimilar numbers and reasons for loss to follow up between groups.

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 4 weeks
Follow up: 8 weeks
Method of randomisation: blocks due to 2 sites, randomised unequally into exercise group and control group in anticipation of greater drop-out from exercise group, done by drawing lots
Concealment of allocation: unclear
Outcome assessor blinding: not specified
Group comparability at entry: yes
Losses to follow up: 1 participant at 4 weeks due to illness, 3 participants at 18 weeks (total 18%)


ParticipantsCountry: Finland
Setting: 2 care homes for older people with 79 inhabitants (72 women, 7 men)
Randomised: 28
% women = 100
Age: mean = approximately 81 years; range = not reported

Consent: fully-informed consent

Inclusion criteria: = 70, able to stand without a walking aid, able to see visual feedback from a computer screen, able to follow instructions for testing and training

Exclusion criteria: see inclusion criteria

% Eligible within home: 41% volunteered
% Eligible that participate: 88% of volunteers able to participate

Exercise group: N = 20; 80.7 years ± 6.1 years
Control group: N = 8; 82.9 years ± 4.2 years


InterventionsStudy aim or objective: to investigate the effects of a 4-week visual feedback-based balance training on the postural control of frail elderly women living in residential care
Number of experimental groups: 2
Individual session delivery
Session duration: 20 to 30 minutes
Number of sessions per week: 3
Seated: no

Exercise features:
20- to 30-minute individualised dynamic balance exercise sessions on a force platform balance measurement and training device (Good Balance), 3 times a week for 4 weeks
Goal: teach participants to control the movement of the centre of pressure during dynamic weight shifting, leaning and stepping tasks, and to manage these tasks in different stances, with higher spatial and temporal demands

Control: not specified

Groups: both groups told to continue their normal daily routines and not to change their physical activity


OutcomesBalance: Dynamic Balance Test (3 tests), Standing Balance Test (6 tests), BBS


NotesFunding:

Ministry of Education

Juhno Vainio Foundation in Finland


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskStated randomisation by the drawing of lots

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskAs intervention and control group participants were from the same 2 homes, unlikely they could have been blinded to group allocation

Incomplete outcome data (attrition bias)
All outcomes
Low risk25% (2/8) lost from control. 15% (3/20) lost from exercise. Limited difference unlikely to relate to intervention or lack of it

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskStated participants were randomised unequally to exercise group in anticipation of greater dropouts. Heavily imbalanced groups not judged to cause systematic risk of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: unclear
Losses to follow up: none


ParticipantsCountry: USA
Setting: 2 ambulatory geriatric mental wards at Woodville State Hospital
Randomised: 17
% women = 0
Age: mean = 69 years; range = not reported

Consent: assent accepted

Inclusion criteria: ambulatory, medical screening prior to inclusion

Exclusion criteria: Medical = cardiovascular abnormality – electrocardiograms carried out prior to inclusion

% Eligible within home: not reported
% Eligible that participate: not reported

Experimental group: N = 9; mean age = 71.5 years
Control group: N = 8; mean age = 65.2 years


InterventionsStudy aim or objective: to investigate the effects of physical training on institutionalised old men

Number of experimental groups: 2
Unclear if group or individual intervention delivery
Session duration: 9 minutes +
Number of sessions per week: 5
Seated: no

Experimental group: performed treadmill walking with speed and gradient adjustment to maintain heart rate at 70% of age-adjusted maximum; sessions lasted 9 minutes for the first 3 weeks, and increased by 3 minutes every subsequent 3 weeks; sessions were conducted daily, Monday to Friday for 12 weeks; persons delivering the intervention were not described

Control group: usual care


OutcomesEndurance (physical other): heart rate
Physiology: systolic blood pressure, diastolic blood pressure


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Twenty-five male geriatric mental patients were selected... Patients adjudged eligible for participation were randomly placed in either an experimental or control group"
No information provided about random sequence generation

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo report of blinding of participants but usual care so intervention would have been obvious

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo reporting of losses to follow up

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 12 weeks
Follow up: 12 weeks
Method of randomisation: lottery method
Concealment of allocation: unclear
Outcome assessor blinding: unclear
Group comparability at entry: no
Losses to follow up: 75 complete data sets from 120 volunteers (63%) 37% dropout


ParticipantsCountry: Australia
Setting: 6 aged-care facilities
Randomised: 120
% women = 75
Age: mean 80.5 years; range = not reported

Consent: assent accepted

Inclusion criteria: mild to moderate dementia, assessments made by local Aged Care Assessment Team, level determined by MMSE < 23, resident in an aged care facility, legally and cognitively capable of providing informed consent to participate, able to respond appropriately to the majority of verbal requests, physically capable of undertaking some form of gentle but regular exercise, efforts to ensure participants were joining of their own free choice (frequent questioning where there were memory problems)

Exclusion criteria: cognitive = severe dementia, MMSE of 0 to 9

% Eligible within home: not reported
% Eligible that participate: not reported

Group 1: N = 30; women = 23; mean age = 81 years
Group 2: N = 21; women = 10; mean age = 81.5 years
Group 3: N = 24; women = 23; mean age = 79 years


InterventionsStudy aim or objective: to measure the effects of exercise on cognitive symptoms related to dementia and disability levels

Number of experimental groups: 3
Group intervention delivery
Session duration: 30 minutes
Number of sessions per week: 3
Seated: if necessary

Groups:
(1) control group, no intervention
(2) control group, social visit from researcher; interactive group discussion on health-related issues, but no exercise; visits of equivalent duration to the exercise
(3) 30-minute group exercise programme 3 x week for 12 weeks

Intervention: based on joint and large muscle group movement with an intention to create gentle aerobic exertion, designed to include those in wheelchairs or with impaired movement, generation-appropriate music, data only analysed where participant attended = 75% of sessions


OutcomesCognition: Clock-drawing Tool (Shulman 1993)
Psychosocial and physical functioning: Revised Elderly Persons Disability Scale (Fleming 1993)


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskStated randomised by lottery method

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo blinding reported for participants or personnel. Outcome measurements could have been influenced by lack of blinding

Incomplete outcome data (attrition bias)
All outcomes
High riskData only provided for 75 participants who completed. Ill health, death, or lack of interest resulted in substantial dropouts (128 agreed to participate), but numbers from each groups not reported

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering not accounted for
Duration: 16 weeks
Follow up: no


ParticipantsCharacterisation: elderly women of care facilities
Country: South Korea
Setting: long-term assisted-living facilities
Randomised: 40
% women = 100%
Age details: mean (SD) = 75.8 (5.6)
Inclusion criteria: older than 65 years
Able to walk alone
Not participated in regular exercise programs within the previous 6 months
No cognitive impairment (cognitive score > 23 on MMSE Korean version)
Exclusion criteria: stroke or cardiovascular event within past 6 months
Unstable chronic or terminal illness (e.g. diabetes mellitus, hepatic cancer, liver cirrhosis)
Severe cognitive impairment (MMSE Korean version < 24) or major depression (GDS > 20)
ADL status details: not reported
Cognitive status details: not reported
Significant comorbidities: 75% had chronic conditions including hypertension, arthritis or diabetes mellitus
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to compare the effects of a 16 week group exercise program on the physical function and mental health of older elderly women ( ≥ 75 years) compared with younger elderly women (<75 years)
2 groups

Intervention: exercise group (N = 20)
Format: group, delivered by: physical therapist supervised, and 2 research assistants led the functional exercise component
Session length: functional exercise: 40 minutes; health education: 30 minutes every 2 weeks, 3 times weekly
The program comprised of functional exercise and health education. Exercise was of low to moderate intensity. Exercise consisted of 10 minutes of warm up, 10 minutes of muscle strengthening, 20 minutes of exercise performed with music, and 10-minute cool down. Health education was based on social cognitive theory and explained how to acquire and maintain behaviour changes, benefits and barriers to change

Control: control group (N = 20)
Usual care. Asked not to initiate any exercise or education program during the 16-week period


OutcomesPhysical function (other): sit-to-stand (average number in 30 seconds)
Balance: static balance (Vellas 1997)
Mood related: GDS, self-esteem (Rosenberg Self Esteem Scale) (Rosenberg 1965)
Flexibility: 'Sit-and-reach' test


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskThere appears to have been random sequence generation. However, the process is unclear

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskCluster, but one group usual care

Incomplete outcome data (attrition bias)
All outcomes
Low riskThree lost from control, none from exercise; reasons not related to intervention

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: cluster RCT
Clustering not accounted for.
Details: quasi-experimental design with controlled group
Duration: 12 weeks
Follow up: no


ParticipantsCharacterisation: elderly residents
Country: Thailand
Setting: residential care facilities
Randomised: 70
% women = 58%
Age details: number of participants:
60 to 69 years: 18 (36%)
70 to 79 years: 24 (48%)
80 years and over: 8 (16%)
Inclusion criteria: over 60 years old; communicate in Thai language; able to practice Tai Chi; stayed in the residential facilities for at least 1 month; normal orientation to place, time, and person; no illnesses limiting movements, uncontrolled epilepsy and diabetes mellitus; not engaged in Tai Chi or other exercises except stretching exercise; in the past month, had at least 1 of the following common sleep problems, which occurred more than 2 times a week for at least 1 week: (1) shallow sleep, (2) sleep less than 5 hours a night, (3)awake more than twice at night, (4) take more than 30 minutes to fall asleep, (5) can not go back to sleep when awake at night, or (6) wake up too early and not refreshed
Exclusion criteria: during the study period, the participants were excluded from the study if they possessed any of the following exclusion criteria: participated in Tai Chi practice less than 3 times a week; felt discomfort or had symptoms such as dizziness, palpitation, dyspnoea, nausea, vomiting, fatigue, or severe muscle and joint pain; their prescription had been changed recently; developed an illness or had an injury that could interfere with sleep; significant change in daily physical activity
ADL status details: not reported
Cognitive status details: not reported
Significant comorbidities: not reported
Assessed: not reported
Excluded: not reported


InterventionsStudy aim or objective: to investigate the effects of low-intensity and short-term Tai Chi practice on sleep quality, general well-being, and physical performance
2 groups

Intervention: experimental group (N = 38)
Format: group, delivered by researcher (trained in Tai Chi)
Session length: 22 minutes, at least 3 times weekly
Tai Chi training program requiring the participants in the group to practice Tai Chi exercise at least 3 times a week, 22 minutes each time for 12 weeks starting from the beginning of the 2nd week through the end of the 13th week of the study. This requirement followed a recommended exercise for the elderly. The participants could join the session on any 3 days of their choice

Control: Control group (N = 32)
The participants in the controlled group were asked to continue their usual activities throughout 14 weeks of the study


OutcomesPhysical function (other): 2-minute step in place
Flexibility: 'Sit-and-reach' test
Quality of life: General Well-Being Scale
Anthropometry: lung capacity
Sleeping: Pittsburg Sleep Quality Index (PSQI)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskTwo care facilities were selected. Assignment to experiment or control was done by simple drawing

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo report of blinding of participants, although control and experimental groups were in different homes, so participants possibly were not aware if they were intervention or control

Incomplete outcome data (attrition bias)
All outcomes
High riskImbalance in number of participants lost to follow up between groups (more in intervention than in control)

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 20 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: maintained
Group comparability at entry: yes
Losses to follow up: 9 (12.5%) (4 before pre-testing completed, 1 transfer, 3 hospitalised); no individual group data provided


ParticipantsCountry: USA
Setting: nursing home
Randomised: 72
% women = 75
Age: mean = 84 years ± 8.5 years; range = 59 to 102 years

Consent: not specified

Inclusion criteria: diagnosis of dementia (on MMSE), 6 or more errors out of 10 on Short Portable Mental Status Questionnaire, ability to stand with assistance of 2 people

Exclusion criteria: medical: evidence of stroke, head injury, major psychiatric problem, mental retardation

% Eligible within home: 80
% Eligible that participate: 37.5


InterventionsStudy aim or objective: to compare the effects of skill training, a traditional stimulation approach, and regular care on the ability to perform basic activities of daily living of nursing-home residents with dementia

Number of experimental groups: 3
Group intervention delivery
Session duration: 2.5 hours a day
Number of sessions per week: 5
Seated: unclear

Skill training group: focused on re-gaining function in basic ADL through repeated practice, with graded assistance

Stimulation group: recreation-orientated activities, group discussion, music, and relaxation

Control group: usual care

The interventions were delivered in group format by a clinical specialist in gerontological nursing, assisted by a rehabilitation aide, for 2.5 hours per day, 5 days a week, for 20 weeks


OutcomesPhysical function in ADL: Performance Test of Activities of Daily Living (PADL), Physical Self-Maintenance Scale (Lawton 1969),
Feasibility and acceptability: Goal Attainment (Brody 1971)


NotesFunding: The Robert Wood Johnson Foundation grant


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskStated "randomly assigned", but no information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo evidence that participants or people delivering the intervention were blinded

Blinding of outcome assessment (detection bias)
Reported ADL (e.g. Barthel, FIM)
Low riskData collectors were blind to group assignment

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskOnly 3 were lost to illness/hospitalisation following pre-testing (4 overall). However, it could not be determined if they from intervention or control group

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 16 weeks
Follow up: none
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: yes - maintained
Group comparability at entry: no
Losses to follow up: N = 6 (8.5%): Walking group N = 3, conversation group N = 2, walk and talk group N = 1, no causes specified


ParticipantsCountry: USA
Setting: nursing home
Randomised: 71
% women = 84
Age: mean = 87 years; range = 70 to 105 years

Consent: assent accepted

Inclusion criteria: Alzheimer's disease, MMSE < 23, ability to stand, ability to mobilise with assistance +/- aid

Exclusion criteria: medical = vascular dementia, stroke, Parkinson's disease, major depression, schizophrenia

% Eligible within home: not reported
% Eligible that participate: not reported

Walking group: N = 26; mean age = 87.4 years (SD 5.87)

Conversation group: N = 24; mean age = 89.6 years (SD 6.53)

Walk and talk group: N = 21
Mean age: 84.3 years (SD 7.53)


InterventionsStudy aim or objective: to examine the effect of a combination of exercise and conversation with walking-only exercise and conversation-only treatments on the functional mobility of frail nursing-home residents with Alzheimer's disease

Number of experimental groups: 3
Individual intervention delivery
Session duration: 30 minutes
Number of sessions per week: 3
Seated: no

Walking group: self-paced with rests as required, and physical assistance, a device or both as required, no conversation initiated, but researcher responded to communication

Conversation group: Holland's approach (aphasia) and facilitation for people with Alzheimer’s, used in natural conversation

Walk and talk group: both interventions simultaneously within a 30-minute session


OutcomesPhysical function in ADL: modified 6-minute walk (Tappen 1997)


NotesFunding: National Institute for Nursing Research grant


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Random assignment to treatment group"
No information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo information regarding blinding of participants. Control involved an intervention (talking), but the only measure was physical

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Low riskQuote: "Raters were blinded to treatment group assignment"

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskReasonably balanced losses, but reasons per group not given

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: 8 weeks (see notes)
Follow up: 1 year
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: yes - maintained
Group comparability at entry: no
Losses to follow up: none


ParticipantsCountry: USA
Setting: long-term care facility
Randomised: 13
% women = 92
Age: mean = approximately 86 years; range = 73 to 95 years

Consent: fully-informed consent

Inclusion criteria: able to walk approximately 250 feet independently with or without aid, Tinetti Score 14 to 24

Exclusion criteria: cognitive = dementia; medical = serious illness requiring medical intervention in previous 6 months

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention group: N = 6; mean age = 89 years (SD 5.1); range = 81 to 95 years

Control group: N = 7; mean age = 82.2 years (SD 4.6); range = 73 to 88 years


InterventionsStudy aim or objective: to determine the effects of 2 exercise programmes on balance in elderly ambulatory people

Number of experimental groups: 2
Group intervention delivery
Session duration: not reported
Number of sessions per week: 2
Seated: no

Exercise group I (control group): traditional exercise incorporating balance and strengthening exercises; no specific equipment used; included weight transference, walking, and lower limb strengthening

Exercise group 2 (intervention group): traditional exercise as above, plus Swiss ball exercises to improve dynamic balance and strengthening component

Both interventions delivered in group format by a physiotherapy student


OutcomesFalls: falls (any episodes for participant)
Physical function in ADL: ambulatory (capable of walking approximately 250 feet)
Physical function (other): Tinetti Test - gait, use of assistive devices (walker/cane), Tinetti Mobility Scale (gait and balance) (Tinetti 1986)
Balance: Tinetti Test - Body Balance


NotesFunding: grants from Kentucky Physical Therapy Association and the University of Louisville Graduate Research fund
Tinetti scores were assessed pre- and post-test (8 weeks) for all 13 participants; 8 of the total 13 participants were able to be followed for 1 year post study, although were assessed only in terms of mobility and assistive devices used


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNo information provided; both groups received physical interventions

Blinding of outcome assessment (detection bias)
Observed ADL (e.g. TUG test, mobility)
Unclear riskNo report of blinding of outcome assessment

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk13 participants completed the intervention phase, but unclear if there were 13 at the onset

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias


MethodsDesign: RCT
Duration: n/a (one-off intervention)
Follow up: n/a
Method of randomisation: not described
Concealment of allocation: unclear
Outcome assessor blinding: no
Group comparability at entry: n/a
Losses to Follow up: none


ParticipantsCountry: USA
Setting: 2 nursing homes
Randomised: 30
% women = 100
Age: mean = 81.5 years ± 7.2 years; range = 70 to 92 years

Consent: not specified

Inclusion criteria: used the first 30 participants scoring > 25 on Parachek Geriatric Rating Scale, residential status

Exclusion criteria: see inclusion criteria

% Eligible within home: not reported
% Eligible that participate: not reported

Intervention:
Group A: N = 15
Group B: N = 15


InterventionsStudy aim or objective: hypothesised that participants engaged in the added-purpose, occupationally-embedded exercise would engage in more repetitions, and exercise for a longer duration and with fewer stops than the participants engaged in rote exercise

Number of experimental groups: 2
Individual session delivery
Session duration: 30 minutes
Number of sessions per week: 2 to 3
Seated: unclear

Added-purpose, occupationally-embedded exercise condition designed, through materials and instructions, to elicit a rotary arm exercise with the added purpose of stirring cookie dough

Compared with an occupational form designed to elicit the rotary arm exercise with no added purpose


OutcomesEndurance (physical other): duration of exercise
Feasibility and acceptability: frequency of discontinuities of exercise, frequency of rotations (repetitions of stirring)


NotesFunding: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo information provided on randomisation procedure

Allocation concealment (selection bias)Unclear riskInsufficient information provided

Blinding of participants and personnel (performance bias)
All outcomes
High riskPersonnel unblinded

Incomplete outcome data (attrition bias)
All outcomes
Low riskAppears there were no losses to follow up

Selective reporting (reporting bias)Unclear riskProtocol unavailable

Other biasLow riskNo other apparent risks of bias

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Alessi 1995bNot primarily physical outcomes of interest (sleep-orientated)

Alexander 2001Review team consensus agreed that the study accommodation (congregate housing) was not synonymous with a long-term care environment

Backman 1986Participants were self-caring within a long-term care environment; interventions were of a psychological, rather than physical, nature

Beck 2009Focus of study was compliance, rather than improving physical condition

Becker 2003Multi-faceted intervention included staff and resident education, advice on environmental adaptations, and hip protectors

Binder 1995Comparison of the same exercise intervention; 1 group was provided with vitamin supplement

Blair 1996Focus was behavioural management, rather than physical activity

Brill 1999bAimed to prevent admission to long-term care

Carmeli 2000Participants were not randomly allocated to study conditions

Collier 2007The intervention was not primarily physical rehabilitation

de Carvalho Bastone 2004Participants were assigned to exercise group or comparative group by personal choice (those who did, or did not, want to attend exercise sessions)

Dyer 2004Multi-factorial falls prevention programme including medication review, podiatry, and optometry

Eggermont 2009The intervention was not aimed at improving physical condition

Evans 1995A review paper that described an apparently relevant study, although insufficient information was provided, and no reference cited. The author was contacted for further information; however, we received no reply

Fisher 1991Not a RCT

Fitzimmons 2001Evaluated the effect of exercise on depression, rather than physical outcomes

Fox 2000Review team consensus deemed passive interventions to reduce contractures to be beyond the scope of this review

Friedman 1991Aimed to improve communication, rather than physical performance measures

Goldberg 1980No physical outcomes evaluated

Hagen 2003Participants not randomised

Hara 2007The participants included visitors to the centre

Hopman-Rock 1999Interventions targeted cognitive, rather than physical, functioning

Ikezoe 2005Non-random allocation of participants

Jensen 2002Multi-faceted intervention to address falls prevention

Jensen 2004Multi-faceted intervention to address falls prevention

Judge 1993Communication with the authors identified very few long-term care residents for whom no separate data were available

Kapasi 2007This was an abstract only for a poster and included no detailed results

Kelly 1983Interventions were not aimed primarily at improving physical condition

Kerse 2004Falls risk management programme

Koc 2008Poster with insufficient data to assess inclusion and included no detailed results

Krishnamurthy 2007The intervention was not intended to address their physical condition

Light 1984Review team consensus deemed passive interventions to reduce contractures to be beyond the scope of this review

MacRae 1996Non-random allocation to the study conditions

McMurdo 2000Multi-faceted intervention to address falls risk

Moye 1996Review team consensus excluded this paper on the basis that no objective physical outcomes measures were used

Mozley 2007While the intervention was occupational therapy, the aim was to reduce depression, rather than affect physical condition

Nowalk 2001Primary focus on reduction of falls

O'Hagan 1994Non-random allocation to the study conditions

Ray 1997Multi-facted falls prevention programme

Remsburg 1999Non-random allocation to the study conditions

Rydwik 2004Non-random allocation of participants

Sato 2007Participants were visitors, rather than residents, of care homes

Sherrington 1997Only a small proportion of participants were residing in institutional care; the majority were independently living in the community. The authors were contacted for separate data, but none were available

Shimada 2003Participants were from out-patient facilities and nursing homes

Shumway-Cook 1997Participants were community-dwelling

Stasi 2004Not a physical rehabilitation intervention

Steffen 1995Not a RCT

Stones 1993Focused on memory, rather than physical outcomes of interest

Tan 2004Non-random allocation of participants

Tseng 2006Focus on contractures

van Heugten 2000Particpants recruited from a variety of settings; no separate data available for long-term care residents

Wolf 2001Included independent living and residential-care participants

Yip 2004Non-random allocation of participants

 
Characteristics of studies awaiting assessment [ordered by study ID]

Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


MethodsRCT

ParticipantsN = 36

Frail older nursing-home residents

InterventionsLow-intensity exercise programme

OutcomesDisability, strength, functional capacity, balance, agility, and walking speed
Performance in ADLs significantly improved

NotesTranslation needed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed. Unable to get.


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed. Reports associated with Resnick 2009.


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed. Reports associated with Rosendahl 2006.


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesTranslation needed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesNot yet assessed


Methods-

Participants-

Interventions-

Outcomes-

NotesUnable to get. Request sent to author mzak1@onet.edu - bounced

 
Characteristics of ongoing studies [ordered by study ID]

Trial name or titleOlder people's exercise intervention in residential and nursing accommodation

Methods-

ParticipantsResidents in nursing home, over 65 years, able to participate in baseline assessment, able to transfer

InterventionsPhysical activation and group-based exercise programme

OutcomesImpact on depression, EQ-5D, mobility, falls, cognitive function, pain, medication use, hospital admissions

Starting date1 January 2008

Contact informationMartin Underwood

m.underwood@qmul.ac.uk

Notes-


Trial name or titleThe effect of exercise on muscle, function and cost in VA nursing home residents

Methods-

ParticipantsResidents in VA nursing home, 65 years or older, able to follow one-step commands

InterventionsLow-intensity exercise

OutcomesMuscle mass, physical function, cost

Starting date2002

Contact information-

NotesHSRP20071249


Trial name or titlePhysical and daily activity for residents in a nursing home setting - A Nordic multi-centre study

Methods-

ParticipantsResidents expected to stay in nursing homes > 3 months

InterventionsIndividually-tailored enhanced activities of daily living training

OutcomesPhysical function, well-being, amount of activity, falls

Starting dateAugust 2005

Contact informationKerstin Frandin

kerstin.frandin@neurotec.ki.se

Notes-


Trial name or titleHealth enhancing strength training in nonagenarians (STRONG)

MethodsRCT

ParticipantsSixty residents of a geriatric nursing home (age range = 90 to 102 years)

InterventionsMuscle strengthening and aerobic exercises over 6 months

OutcomesSF-12, Tinetti mobility scale, BI, 1RM leg press, hand grip strength, 8-metre walk test, 4-step stairs test, BMI, physical activity (Actigraph), MMSE, falls, and other adverse events

Starting dateMarch 2009

Contact informationAlejandro Lucia alejandro.lucia@uem.es

Notes-

 
Comparison 1. Rehabilitation versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Barthel Index7857Mean Difference (Random, 95% CI)6.38 [1.63, 11.12]

 2 Functional Independence Measure (FIM)4303Mean Difference (Random, 95% CI)4.98 [-1.55, 11.51]

 3 Rivermead Mobility Index (RMI)3323Mean Difference (Random, 95% CI)0.69 [0.04, 1.33]

 4 Timed Up and Go (TUG) Test7885Mean Difference (Random, 95% CI)-4.59 [-9.19, 0.01]

 5 Walking speed9590Mean Difference (Random, 95% CI)0.03 [-0.01, 0.07]

 6 Death253721Risk Ratio (M-H, Random, 95% CI)0.95 [0.80, 1.13]

 7 Barthel Index (by risk of bias)7Mean Difference (Random, 95% CI)Subtotals only

    7.1 lower risk of bias
2275Mean Difference (Random, 95% CI)3.38 [-2.10, 8.86]

    7.2 higher risk of bias
5582Mean Difference (Random, 95% CI)8.25 [1.15, 15.34]

 8 Barthel Index (by duration of intervention)7Mean Difference (Random, 95% CI)Subtotals only

    8.1 shorter (< 3 months intervention)
246Mean Difference (Random, 95% CI)17.55 [6.97, 28.13]

    8.2 longer (3+ months intervention)
5811Mean Difference (Random, 95% CI)3.08 [-0.03, 6.19]

 9 Barthel Index (by mode of delivery)7Mean Difference (Random, 95% CI)Subtotals only

    9.1 group
4256Mean Difference (Random, 95% CI)10.99 [1.51, 20.48]

    9.2 individual
2275Mean Difference (Random, 95% CI)3.38 [-2.10, 8.86]

    9.3 not reported
1326Mean Difference (Random, 95% CI)2.19 [-4.35, 8.73]

 10 Barthel Index (by baseline Barthel Index score)6Mean Difference (Random, 95% CI)Subtotals only

    10.1 better (baseline Barthel Index score > median)
3511Mean Difference (Random, 95% CI)7.94 [-1.77, 17.64]

    10.2 worse (baseline Barthel Index score < median)
3305Mean Difference (Random, 95% CI)3.97 [-0.83, 8.78]

 11 Barthel Index (by age)7Mean Difference (Random, 95% CI)Subtotals only

    11.1 younger (mean age < 85 years)
4552Mean Difference (Random, 95% CI)8.02 [-0.25, 16.30]

    11.2 older (mean age 85+ years)
3305Mean Difference (Random, 95% CI)3.97 [-0.83, 8.78]

 12 Barthel Index (by gender)7Mean Difference (Random, 95% CI)Subtotals only

    12.1 < 80% female
4402Mean Difference (Random, 95% CI)7.93 [0.18, 15.69]

    12.2 80%+ female
3455Mean Difference (Random, 95% CI)4.29 [-1.25, 9.83]

 13 Functional Independence Measure (by risk of bias)4Mean Difference (Random, 95% CI)Subtotals only

   13.1 lower risk of bias
00Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]

    13.2 higher risk of bias
4303Mean Difference (Random, 95% CI)4.98 [-1.55, 11.51]

 14 Functional Independence Measure (by duration of intervention)4Mean Difference (Random, 95% CI)Subtotals only

    14.1 shorter (< 3 months intervention)
130Mean Difference (Random, 95% CI)2.0 [-10.26, 14.26]

    14.2 longer (3+ months intervention)
3273Mean Difference (Random, 95% CI)5.85 [-2.22, 13.93]

 15 Functional Independence Measure (by mode of delivery)4Mean Difference (Random, 95% CI)Subtotals only

    15.1 group
3240Mean Difference (Random, 95% CI)3.90 [-3.08, 10.88]

    15.2 individual
163Mean Difference (Random, 95% CI)11.76 [-2.66, 26.18]

 16 Functional Independence Measure (by baseline FIM score)3Mean Difference (Random, 95% CI)Subtotals only

    16.1 better (baseline FIM score > median)
295Mean Difference (Random, 95% CI)7.77 [1.39, 14.14]

    16.2 worse (baseline FIM score < median)
1145Mean Difference (Random, 95% CI)0.3 [-1.73, 2.33]

 17 Functional Independence Measure (by age)4Mean Difference (Random, 95% CI)Subtotals only

    17.1 younger (mean age < 85 years)
2128Mean Difference (Random, 95% CI)9.91 [4.41, 15.42]

    17.2 older (mean age 85+ years)
2175Mean Difference (Random, 95% CI)0.35 [-1.65, 2.34]

 18 Functional Independence Measure (by gender)4Mean Difference (Random, 95% CI)Subtotals only

    18.1 < 80% female
293Mean Difference (Random, 95% CI)6.11 [-3.33, 15.55]

    18.2 80%+ female
2210Mean Difference (Random, 95% CI)4.51 [-4.56, 13.58]

 19 Rivermead Mobility Index (by risk of bias)3Mean Difference (Random, 95% CI)Subtotals only

    19.1 lower risk of bias
3323Mean Difference (Random, 95% CI)0.69 [0.04, 1.33]

   19.2 higher risk of bias
00Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]

 20 Rivermead Mobility Index (by duration of intervention)3Mean Difference (Random, 95% CI)Subtotals only

    20.1 shorter (< 3 months intervention)
149Mean Difference (Random, 95% CI)0.6 [-1.48, 2.68]

    20.2 longer (3+ months intervention)
2274Mean Difference (Random, 95% CI)0.69 [0.02, 1.37]

 21 Rivermead Mobility Index (by mode of delivery)3Mean Difference (Random, 95% CI)Subtotals only

    21.1 group
149Mean Difference (Random, 95% CI)0.6 [-1.48, 2.68]

    21.2 individual
2274Mean Difference (Random, 95% CI)0.69 [0.02, 1.37]

 22 Rivermead Mobility Index (by baseline RMI score)3Mean Difference (Random, 95% CI)Subtotals only

    22.1 better (baseline RMI score > median)
2235Mean Difference (Random, 95% CI)0.70 [0.01, 1.39]

    22.2 worse (baseline RMI score < median)
188Mean Difference (Random, 95% CI)0.6 [-1.17, 2.37]

 23 Rivermead Mobility Index (by age)3Mean Difference (Random, 95% CI)Subtotals only

    23.1 younger (mean age < 85 years)
149Mean Difference (Random, 95% CI)0.6 [-1.48, 2.68]

    23.2 older (mean age 85+ years)
2274Mean Difference (Random, 95% CI)0.69 [0.02, 1.37]

 24 Rivermead Mobility Index (by gender)3Mean Difference (Random, 95% CI)Subtotals only

    24.1 < 80% female
2235Mean Difference (Random, 95% CI)0.70 [0.01, 1.39]

    24.2 80%+ female
188Mean Difference (Random, 95% CI)0.6 [-1.17, 2.37]

 25 TUG Test (by risk of bias)7Mean Difference (Random, 95% CI)Subtotals only

    25.1 lower risk of bias
1556Mean Difference (Random, 95% CI)0.6 [-5.36, 6.56]

    25.2 higher risk of bias
6329Mean Difference (Random, 95% CI)-5.92 [-11.29, -0.54]

 26 TUG Test (by duration of intervention)7Mean Difference (Random, 95% CI)Subtotals only

    26.1 shorter (< 6 months intervention)
4185Mean Difference (Random, 95% CI)-7.34 [-13.93, -0.75]

    26.2 longer (6+ months intervention)
3700Mean Difference (Random, 95% CI)0.13 [-4.28, 4.53]

 27 TUG Test (by mode of delivery)7Mean Difference (Random, 95% CI)Subtotals only

    27.1 group
4154Mean Difference (Random, 95% CI)-4.98 [-10.74, 0.77]

    27.2 individual
3731Mean Difference (Random, 95% CI)-4.56 [-14.02, 4.90]

 28 TUG Test (by baseline TUG score)7Mean Difference (Random, 95% CI)Subtotals only

    28.1 better (baseline TUG score < median)
4185Mean Difference (Random, 95% CI)-7.34 [-13.93, -0.75]

    28.2 worse (baseline TUG score > median)
3700Mean Difference (Random, 95% CI)0.13 [-4.28, 4.53]

 29 TUG Test (by age)7Mean Difference (Random, 95% CI)Subtotals only

    29.1 younger (mean age < 85 years)
5741Mean Difference (Random, 95% CI)-5.39 [-10.77, -0.00]

    29.2 older (mean age 85+ years)
2144Mean Difference (Random, 95% CI)-5.40 [-25.75, 14.96]

 30 TUG Test (by gender)7Mean Difference (Random, 95% CI)Subtotals only

    30.1 < 80% female
3594Mean Difference (Random, 95% CI)0.17 [-3.90, 4.24]

    30.2 80%+ female
4291Mean Difference (Random, 95% CI)-7.55 [-14.28, -0.82]

 31 Walking speed (by risk of bias)9Mean Difference (Random, 95% CI)Subtotals only

    31.1 lower risk of bias
175Mean Difference (Random, 95% CI)-0.10 [-0.21, 0.01]

    31.2 higher risk of bias
8515Mean Difference (Random, 95% CI)0.04 [0.01, 0.07]

 32 Walking speed (by duration of intervention)9Mean Difference (Random, 95% CI)Subtotals only

    32.1 shorter (< 3 months intervention)
359Mean Difference (Random, 95% CI)0.24 [-0.74, 1.22]

    32.2 longer (3+ months intervention)
6531Mean Difference (Random, 95% CI)0.02 [-0.03, 0.08]

 33 Walking speed (by mode of delivery)9Mean Difference (Random, 95% CI)Subtotals only

    33.1 group
7475Mean Difference (Random, 95% CI)0.03 [-0.02, 0.07]

    33.2 individual
148Mean Difference (Random, 95% CI)0.26 [-0.32, 0.83]

    33.3 not reported
167Mean Difference (Random, 95% CI)-0.03 [-0.19, 0.13]

 34 Walking speed (by baseline walking speed)9Mean Difference (Random, 95% CI)Subtotals only

    34.1 better (baseline walking speed > median)
5198Mean Difference (Random, 95% CI)-0.00 [-0.15, 0.14]

    34.2 worse (baseline walking speed < median)
4392Mean Difference (Random, 95% CI)0.04 [0.01, 0.07]

 35 Walking speed (by age)9Mean Difference (Random, 95% CI)Subtotals only

    35.1 younger (mean age < 85 years)
9590Mean Difference (Random, 95% CI)0.03 [-0.01, 0.07]

   35.2 older (mean age 85+ years)
00Mean Difference (Random, 95% CI)0.0 [0.0, 0.0]

 36 Walking speed (by gender)9Mean Difference (Random, 95% CI)Subtotals only

    36.1 < 80% female
5437Mean Difference (Random, 95% CI)0.01 [-0.04, 0.07]

    36.2 80%+ female
4153Mean Difference (Random, 95% CI)0.13 [-0.02, 0.28]

 37 Walking speed (by distance walked)9Mean Difference (Random, 95% CI)Subtotals only

    37.1 less far (< 6 metres)
2185Mean Difference (Random, 95% CI)0.04 [0.01, 0.07]

    37.2 further (6+ metres)
7405Mean Difference (Random, 95% CI)0.01 [-0.06, 0.09]

 38 Death (by risk of bias)25Risk Ratio (M-H, Random, 95% CI)Subtotals only

    38.1 lower risk of bias
61366Risk Ratio (M-H, Random, 95% CI)1.05 [0.76, 1.46]

    38.2 higher risk of bias
192355Risk Ratio (M-H, Random, 95% CI)0.88 [0.71, 1.10]

 39 Death (by duration of intervention)25Risk Ratio (M-H, Random, 95% CI)Subtotals only

    39.1 shorter intervention (< 3 months)
10663Risk Ratio (M-H, Random, 95% CI)0.64 [0.18, 2.29]

    39.2 longer intervention (3+ months)
153058Risk Ratio (M-H, Random, 95% CI)0.95 [0.80, 1.14]

 40 Death (by mode of delivery)25Risk Ratio (M-H, Random, 95% CI)Subtotals only

    40.1 group
121007Risk Ratio (M-H, Random, 95% CI)0.82 [0.46, 1.49]

    40.2 individual
92172Risk Ratio (M-H, Random, 95% CI)0.91 [0.70, 1.19]

    40.3 group and individual
124Risk Ratio (M-H, Random, 95% CI)5.0 [0.27, 94.34]

    40.4 not reported
3518Risk Ratio (M-H, Random, 95% CI)1.00 [0.73, 1.36]

 41 Death (by age)25Risk Ratio (M-H, Random, 95% CI)Subtotals only

    41.1 younger (mean age < 85 years)
163001Risk Ratio (M-H, Random, 95% CI)0.97 [0.81, 1.17]

    41.2 older (mean age 85+ years)
9720Risk Ratio (M-H, Random, 95% CI)0.79 [0.49, 1.27]

 42 Death (by gender)25Risk Ratio (M-H, Random, 95% CI)Subtotals only

    42.1 < 80% female
122366Risk Ratio (M-H, Random, 95% CI)0.98 [0.77, 1.25]

    42.2 80%+ female
121340Risk Ratio (M-H, Random, 95% CI)0.91 [0.71, 1.18]

    42.3 not reported
115Risk Ratio (M-H, Random, 95% CI)0.88 [0.16, 4.68]

 43 Sensitivity analysis: Barthel Index (fixed-effect)7857Mean Difference (Fixed, 95% CI)4.54 [1.59, 7.49]

 44 Sensitivity analysis: Barthel Index (cluster trials)7Mean Difference (Random, 95% CI)Subtotals only

    44.1 cluster (adjusted)
5811Mean Difference (Random, 95% CI)3.08 [-0.03, 6.19]

    44.2 individual
246Mean Difference (Random, 95% CI)17.55 [6.97, 28.13]

 45 Sensitivity analysis: Functional Independence Measure (fixed-effect)4303Mean Difference (Fixed, 95% CI)1.46 [-0.42, 3.34]

 46 Sensitivity analysis: Rivermead Mobility Index (fixed-effect)3323Mean Difference (Fixed, 95% CI)0.69 [0.04, 1.33]

 47 Sensitivity analysis: TUG Test (fixed-effect)7885Mean Difference (Fixed, 95% CI)-3.66 [-5.86, -1.45]

 48 Sensitivity anlaysis: TUG Test (cluster trials)7Mean Difference (Random, 95% CI)Subtotals only

    48.1 cluster (adjusted)
2Mean Difference (Random, 95% CI)0.51 [-3.93, 4.95]

    48.2 individual
5Mean Difference (Random, 95% CI)-7.85 [-14.34, -1.37]

 49 Sensitivity analysis: TUG Test (re-including Christofoletti 2008)8914Mean Difference (Random, 95% CI)-8.41 [-15.53, -1.29]

 50 Sensitivity analysis: Walking speed (fixed-effect)9590Mean Difference (Fixed, 95% CI)0.03 [0.00, 0.06]

 51 Sensitivity analysis: Walking speed (cluster trials)9Mean Difference (Random, 95% CI)Subtotals only

    51.1 cluster (unadjusted)
1Mean Difference (Random, 95% CI)0.04 [0.01, 0.07]

    51.2 individual
8Mean Difference (Random, 95% CI)0.01 [-0.05, 0.08]

 52 Sensitivity analysis: Death (random-effects: odds ratio)253721Odds Ratio (M-H, Random, 95% CI)0.93 [0.75, 1.15]

 53 Sensitivity analysis: Death (random-effects: risk difference)253721Risk Difference (M-H, Random, 95% CI)-0.01 [-0.02, 0.01]

 54 Sensitivity analysis: Death (fixed-effect)253721Risk Ratio (M-H, Random, 95% CI)0.95 [0.80, 1.13]

 55 Sensitivity analysis: Death (fixed-effect: Peto odds ratio)253721Peto Odds Ratio (Peto, Fixed, 95% CI)0.93 [0.75, 1.14]

 56 Sensitivity analysis: Death (cluster trials)25Risk Ratio (M-H, Random, 95% CI)Subtotals only

    56.1 cluster (unadjusted)
132644Risk Ratio (M-H, Random, 95% CI)0.95 [0.79, 1.15]

    56.2 individual
121077Risk Ratio (M-H, Random, 95% CI)0.93 [0.60, 1.44]

 57 Sensitivity analysis: Death (including Brittle 2009)263777Risk Ratio (M-H, Random, 95% CI)0.94 [0.79, 1.11]

 
Comparison 2. Rehabilitation (experimental) versus rehabilitation (control)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 TUG Test257Mean Difference (Random, 95% CI)-7.95 [-19.22, 3.31]

 2 Death4118Risk Ratio (M-H, Random, 95% CI)2.67 [0.12, 60.93]

 3 Sensitivity analysis: TUG Test (fixed-effect)257Mean Difference (Fixed, 95% CI)-7.19 [-10.92, -3.46]