Table 4 Summary of included studies characterised using PROFANE domains

Single interventions

Exercise

Buettner 2002 [28]

Does a therapeutic recreation intervention reduce falls in older adults with dementia?

I = 3 month therapeutic recreation program delivered at time of day and location where falls occurred; to increase strength, endurance, flexibility and balance.

25 people with dementia & history of previous falls. Aged 60+ (mean age 83), MMSE <= 23 (M=2.63)

Nursing Home, America

RCT (2 months FU)

C = usual activities

Rosendahl et al. 2008 [43]

Does an exercise program reduce falls in residential care facilities?

I= 3 month individualised weight-bearing exercise intervention

191 people aged 65+ (mean age 85), MMSE 10+ (M=17.8), 52% with dementia

Residential Care, Sweden

Cluster RCT (6 months FU)

C = non-exercise control activity while sitting

Environment/assistive technology

Detweiler et al. 2009 [30] (also Detweiler 2008) [31]

Does a dementia wander garden and medication review reduce number and severity of falls?

I = wander garden and medication review

28 people with dementia aged 74-92 (mean age 81).

Residential Dementia Care unit, America

Uncontrolled before/after study (12 months FU)

Social environment

Detweiler et al. 2005 [29]

Does consistent supervision during day and evening shifts reduce falls in dementipea unit?

I = Supervision focusing on behavioural and environmental factors.

8 older people with dementia aged 74 to 85 (mean age 81)

Dementia Care Home, America

Uncontrolled before/after study (4 month FU)

Shimada et al. 2009 [45]

Does a falls prevention aide using systematic supervision reduce falls?

I = Aide delivered intervention, targeting residents considered to be at high risk of falls

60 people aged 68-105 (mean age 87), 48% Dementia, 5% cognitive impairment , 2% depression

Long-term aged-care facility, Japan.

Uncontrolled before/after study (25 week FU)

Chenowith et al. 2009 [27]

Investigate effectiveness of person-centred care and dementia-care mapping compared with each other and with conventional dementia care

I 1= Person centred care

296 Average age: 83 for dementia care mapping 84 for person centred care 85 usual care

Residential Care sites, Australia

Cluster RCT (4 Month FU)

I 2= Dementia care mapping.

C=Usual care.

Mador et al. 2004 [38]

Does individualized advice on non-pharmacological strategies for hospitalized older patients with confusion and behavioural problems improve levels of agitation and reduce the use of psychotropic medication.

I= Patient assessment, non-pharmacological management plan, on-going support and education for nursing staff. Tailored to patient needs-included addressing patient safety, minimising restraint use, reducing fall risk, communication, behavioural strategies and education.

71 older people with confusion

Acute Hospital, Australia

RCT (FU to discharge)

C= usual care- included review with geriatrican.

Mean age I=82, C=83

Knowledge

Bouwen et al. 2008 [26]

Does a staff-oriented intervention impact on the number of accidental falls in residents with and without cognitive impairment?

I = 6 wk multifaceted intervention involving staff training on falls risk factors, followed by a falls diary and patient questionnaire linking risk with possible interventions.

379 older people with mean age of 83 and MMSE <23 (M=15.72)

Nursing Home, Belgium

Cluster RCT (6 month FU)

C = no staff training, no diary, no questionnaire

Haines et al. 2011 [33]

Evaluative comparison of 2 forms of multimedia patient education intervention alongside usual care for the prevention of falls.

I1 = written and video based intervention materials and 1-to-1 follow-up with a physiotherapist, in addition to usual ward based care (median time spent with patient 25 (20-36) minutes, maximum with one patient 200 minutes).

1206 people aged 60 + (mean age 75), mean SPMSQ = 8.4, 25% cognitive impairment

In-patient, Australia

Cluster RCT (FU to discharge)

I2 = intervention materials provided but without 1-to-1 with physiotherapist, in addition to usual ward based care.

C = usual ward based care

Other

Multisensory stimulation

Klages et al. 2011 [36]

To investigate the influence of multisensory stimulations in a Snoezelen room on the balance of individuals with dementia.

I= 30 mins use of a Snoezelen room twice a week for 6 weeks.

19 older people, mean age 86. MMSE 12 (range 4-22) for IV group, 13 (2-22) for control. Able to walk with minimal assistance and understand simple instructions.

Long term care home, Canada

RCT (FU 6 weeks post intervention)

C= volunteer spending same amount of time 1-to-1 with resident.

Multisensory stimulation

Sakamoto et al. 2012 [44]

Does a lavender olfactory stimulation intervention reduce falls in nursing home residents?

I = 12 month, 24 hour exposure to lavender olfactory stimulation patch on clothes near neck

145 people aged 65+ (mean age 84), mean MMSE = 15. able to transfer independently

Nursing Home, Japan

RCT (360 days FU)

C = same patch and duration as intervention, but no lavender

B: Combination interventions: multiple

Wesson 2013 [51]

To test design and feasibility of a home hazard reduction and balance and strength exercise fall prevention program for people with mild dementia living in the community.

I= Strength & balance training, home hazard reduction, discussion of behaviour and management issues with carers. Carers supervised exercise and responsible for implementation of home safety recommendations.

11 patient and carer dyads.

Community, Australia

RCT (4 month FU)

C= Usual care.

Mean age I= 78.7, C=80.9

Both groups received health promotion brochures on fall prevention and home safety.

Faes et al. 2011 [32]

Is a multifactorial fall prevention program more effective than usual geriatric care?

I = Psychological training for staff & physical training for patients

33 older people, mean age 78, mean MMSE 25, 48% had mild cognitive impairment or dementia

Geriatric outpatient clinic, The Netherlands

RCT (6 month FU)

C= Usual geriatric care

Jensen et al 2002 [34], Jensen et al. 2003 [35]

Does a multi-factorial intervention reduce falls & fall-related injuries, in a high risk population in residential care?

I = 11 week multifactorial intervention including staff education, & resident exercise

40 people aged 65-100 (mean age 83), mean MMSE = 19, 36% with dementia

Residential Care, Sweden

Cluster RCT (34 week FU)

C = Usual care, no staff education.

Mackintosh et al. 2005 [37]

How feasible and effective is a falls-prevention programme for community dwelling people with dementia?

I = Multifactorial including individualised management plan, mobility exercises, foot health, and multidisciplinary referrals.

64 people with dementia aged 53-93 (mean age 80)

Respite Day Centre, Australia

Uncontrolled before/After study (6 month FU)

Neyens et al. 2009 [39] (also Neyens 2006 [40])

Is a multidisciplinary fall prevention intervention effective for psychogeriatric nursing home patients?

I = 12 month multifactorial intervention including assessment & evaluation

518 people with dementia, mean age 82

Psychogeriatric nursing homes, The Netherlands

Cluster RCT (12 months FU)

C= Usual care, staff had no insight in the fall prevention programme.

Rapp et al 2008 [41] (also Becker et al. 2003 [25])

Is a multifactorial fall prevention program effective in pre-specified subgroups of nursing home residents?

I = 12 month intervention including staff training & education and exercise, and environmental assessments for residents

725 people >60 (mean age 86), 46% with cognitive impairment.

Long-term Nursing Homes, Germany

Cluster RCT (12 months FU)

C= No specific fall prevention measures

Ray et al. 1997 [42]

Does a safety intervention prevent falls and associated injury in high-risk nursing home residents?

I = Individual and facility-wide safety and environmental assessment

482 people aged >65 (mean age 83), 49% with cognitive impairment

Nursing Home, America

Cluster RCT (12 months FU)

C = no assessments or activities

Shaw et al. 2003 [46]

Does a multifactorial intervention reduce falls in older patients with cognitive impairment and dementia attending an accident and emergency department?

I = Multidisciplinary risk assessment and intervention

274 people aged 65+ with MMSE <24, 89% with dementia.

Community, UK

RCT (12 months FU)

C = Assessment but no intervention

Stenvall et al. 2007 [47]

Does a post-operative multidisciplinary, multifactorial intervention reduce inpatient falls and fall-related injuries in patients with femoral neck fracture?

I = Post-op care in geriatric ward with special intervention programme (staff education, joined up assessments by OT and dietician)

199 older people aged 70+ (mean 82), 33% with dementia and 33% with depression.

Orthopaedic and geriatric hospital departments, Sweden.

RCT (follow up not clear)

C = Conventional care in orthopaedic ward

Salminen et al. 2009 [43]

Does a multi-factorial fall prevention program reduce falls and which subgroups benefit the most?

I = 12 month intervention based on individual patient risk analysis.

591 people aged 65+, with at least one fall in previous year, and able to walk 10 metres

Community, Finland

RCT (12 months FU)

C = initial counselling and guidance but no follow up over the 12 month period

52 people in I with GDS ≥11, 40 in control group.