Efficacy and Generalizability of Falls Prevention Interventions in Nursing Homes: A Systematic Review and Meta-analysis

JAMDA xxx (2019) 1e12

JAMDA journal homepage: www.jamda.com

Original Study

Efficacy and Generalizability of Falls Prevention Interventions in Nursing Homes: A Systematic Review and Meta-analysis Heidi J. Gulka BSc a, Vaidehi Patel BSc a, Twinkle Arora MA a, Caitlin McArthur MScPT, PhD b, Andrea Iaboni MD, DPhil a, c, * a

Toronto Rehabilitation Institute, University Health Network, Toronto, Ontario, Canada Department of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada c Department of Psychiatry, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada b

a b s t r a c t Keywords: Nursing homes long-term care dementia cognitive impairment fall prevention falls

Objectives: To determine the efficacy of fall intervention programs in nursing homes (NHs) and the generalizability of these interventions to people living with cognitive impairment and dementia. Design: Systematic review and meta-analysis. Setting and Participants: NH residents (n ¼ 30,057) living in NHs defined as residential facilities that provide 24-hours-a-day surveillance, personal care, and some clinical care for persons who are typically aged 65 years with multiple complex chronic health conditions. Methods: Meta-analysis of falls prevention interventions on number of falls, fallers, and recurrent fallers. Results: Thirty-six studies met inclusion criteria for the systematic review. Overall, fall prevention interventions reduced the number of falls [risk ratio (RR) ¼ 0.73, 95% confidence interval (CI) ¼ 0.60-0.88], fallers (RR ¼ 0.80, 95% CI ¼ 0.72-0.89), and recurrent fallers (RR ¼ 0.70, 95% CI ¼ 0.60-0.81). Subanalyses revealed that single interventions have a significant effect on reducing fallers (RR ¼ 0.78, 95% CI ¼ 0.690.89) and recurrent fallers (RR ¼ 0.60, 95% CI ¼ 0.52-0.70), whereas multiple interventions reduce fallers (RR ¼ 0.69, 95% CI ¼ 0.39-0.97) and multifactorial interventions reduce number of falls (RR ¼ 0.65, 95% CI ¼ 0.45-0.94). Conclusions and Implications: Exercise as a single intervention reduced the number of fallers and recurrent fallers by 36% and 41%, respectively, in people living in NHs. Other effective interventions included staff education and multiple and multifactorial interventions. However, more research on exercise including people with cognitive impairment and dementia is needed to improve the generalizability of these interventions to the typical NH resident. Ó 2019 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

Older adults living in nursing homes (NHs) are at a high risk of falling.1 Around 16% of NH residents fall in a given month,2 and publicly reported NH quality indicators show that, on average, 3.4% (range 0%-30%) of NH residents suffer a major injury related to a fall every 3 months.3 Thus, fall prevention is an important target of resident safety and quality improvement in NHs. However, despite investment in falls prevention education and activities, there has been little overall impact on the rate of falls in NHs in the United States and Canada.2

This work was supported by funding from the Walter and Maria Schroeder Institute for Brain Innovation and Recovery. The authors declare no conflicts of interest. * Address correspondence to Andrea Iaboni, MD, DPhil, Toronto Rehabilitation institute, University Health Network, 550 University Avenue, Toronto, Ontario, Canada, M5G 2A2. E-mail address: [email protected] (A. Iaboni). https://doi.org/10.1016/j.jamda.2019.11.012 1525-8610/Ó 2019 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

One possible explanation is that few falls prevention interventions have been developed specifically for the NH population.4 Dementia is highly prevalent in NH residents, with around 60% of NH residents diagnosed with dementia and a further 30% having cognitive impairment without a diagnosis of dementia.5e7 Although older adults with dementia are more likely to experience falls,8,9 they are also often excluded in falls prevention studies, because of concern about their ability to adhere to the intervention.10 This raises questions about the generalizability of falls prevention interventions to the NH population.11 The focus on falls as a measure of safe and quality care may also have unintended consequences, including increasing immobility and other risk factors for falls. For example, staff fear that residents will fall, which can lead to practices such as limiting the physical activity of falls-prone older adults, or in the most extreme case, restraint used to prevent falls.12,13 Although these may achieve a reduction in falls in the immediate short term, they have a negative impact on quality of

2

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

life, contribute to deconditioning,14,15 and ultimately increase the long-term risk of falling.16,17 Falls among older adults in NHs are multifactorial in nature.12,18 Thus, the prevention of falls requires a comprehensive and multidimensional approach. Several previous systematic reviews have examined falls prevention interventions in community, hospital, and NH settings.19e22 Vlaeyen et al20 examined the effectiveness of falls prevention programs in NHs and included 14 studies up until 2013. The review revealed that, overall, falls prevention programs reduced the number of recurrent fallers by 21% but did not reduce the number of falls or fallers. A recent Cochrane Review and meta-analysis reported on falls in both residential care and hospital settings.19 The review found an effect of vitamin D supplementation on reducing the rate of falls, but not fallers, and found no effect of any other type of intervention in the residential care population. This review used a heterogenous sample of residential care facilities by including retirement homes and skilled nursing facilities and did not take into account the generalizability of interventions to those living with dementia. The aim of this systematic review is to update the previous review by Vlaeyen et al20 with a focus on interventions evaluated specifically in

the NH environment and examine the generalizability of these interventions to those living with cognitive impairment and dementia. In addition, we investigated whether balancing measures were included in clinical trials of falls prevention programs to determine the impact of these interventions on important measures of safe and quality care other than falls rate. Methods Search Strategy A systematic literature search for relevant studies was conducted in multiple databases (MEDLINE, EMBASE, CINAHL, SPORTDiscus, Cochrane Central Register of Controlled Trials, and PEDRO; search strategy available in Supplementary Material 1). The search focused on studies published from September 2013 to April 2019, and we included all 14 studies in the previous review encompassing the period from database conception to April 11, 2019. The flow diagram of studies is found in Supplementary Figure 1. One reviewer screened the titles and abstracts of the references to exclude any irrelevant studies. Three independent reviewers

Table 1 Study Characteristics Study (First Author)

Design

Setting

NHs, n

Residents Intervention

Becker30 Chenoweth31

CRCT CRCT

Germany Australia

Colón-Emeric32 Cox33 Dyer34 Fu35 Galik36 García-Gollarte37 Hewitt38 Imaoka39

CRCT CRCT CRCT RCT CRCT CRCT CRCT RCT

United States United Kingdom United Kingdom Hong Kong United States Spain Australia Japan

24 209 20 1 4 37 16 1

Juola40 Kerse41 Kovács42 Lapane43 Law44 McMurdo45 Meyer46 Neyens47 Patterson48 Potter49 Rapp50 Ray51 Sakamoto52 Saravanakumar53

CRCT CRCT RCT CRCT CRCT CRCT CRCT CRCT CRCT RCT CRCT CRCT RCT RCT

Finland New Zealand Hungary United States United Kingdom United Kingdom Germany The Netherlands Northern Ireland Australia Germany United States Japan Australia

NA 41 1 25 118 9 58 12 22 4 6 14 3 1

Schnelle54 Schoenfelder55 Sitjà-Rabert56 Teresi57 Tuunainen58

RCT RCT RCT CRCT RCT

United States United States Spain

van Gaal59 Varela60 Walker61 Ward62 Whitney63 Wylie64 Yokoi65

CRCT RCT CRCT CRCT CRCT RCT CRCT

Netherlands Spain United Kingdom Australia United Kingdom United Kingdom Japan

Finland

6 15

4 2 10 4 1 6 NA 6 88 9 6 5

509 109 98 887 3476 102 30 53 516 113 22 23 23 118 330 43 1711 1762 77 574 249 173 47 365 221 73 11 11 92 9 81 601 18 18 114 17 25 2802 103 23 51

Age, Mean (SD) Control 472 82 907 2753 94 30 50 502 108 23

109 352 43 1492 1955 56 551 269 161 48 360 261 72 11 98 7 78 600 19 127 22 27 2589 88 20 54

Female, n (%)

Duration, mo

Intervention

Control

Intervention

Control

83.5 (7.5) 83 (7.6) 84 (6.4) 80.9 (9.6) NA 87.2 (6.9) 82.4 (3.8) 83.7 (9.9) 84.24 (14.6) 86 (R ¼ 65-100) 82.6 (9.1) 84.6 (7.7) 87.6 (6.5) 82.9 (R ¼ 66-100) 84.4 (7.2) 76.39 (9.6) NA 85 84.9 (6.7) 86 (6) 82.1 (7.7) 82.6 (8.4) 84 (6) NA 82.7 84.2 (7.8) 81.1 (8.0) 84.9 (6.7) 87.3 (8.0) NA 82.3 (7.8) 85.2 (8.9) 84.7 (5.5) 85 (4.2) 78 (9.9) 77.9 (8.8) 84 (14.8) 86y 84.1 (8.4) 86.9 (6.2) 80.2 (7.9)

84.3 (6.9) 85 (6.6)

399 (78.4) 90 (83) 74 (76) 476 (53.7) 2693 (77.9) 81 (79.4) 19 (63.3) 79 (77) 382 (74.0) 71 (62.8) 16 (72.7) 20 (86.9) 18 (78.3) 77 (65.3) 240 (72.7) 36 (83) 1237 (72.3) 1339 (75.9) 60 (77.9) 486 (84.7) 161 (64.6) 125 (72.3) 26 (55) NA 171 (77.4) 14 (19.2) 8 (72.7) 10 (90.9) 73 (79.3) 8 (88.9) 53 (65.4) 443 (73.7) 12 (66.7) 16 (88.9) 70 (61.4) 8 (47.1) 18 (72) 2049 (73.1) 72 (70) 20 (87) 33 (65)

373 (79.0) 60 (73)

12 8

481 (53.0) 2003 (75.4) 72 (76.6) 20 (66.7) 362 (72.1) 73 (68.2) 15 (65.2)

6 12 12 12 6 6 12 6

84 (77.1) 262 (74.4) 34 (79) 1018 (68.2) 1485 (75.9) 48 (85.7) 473 (85.8) 191 (71.0) 119 (73.4) 23 (48) NA 205 (78.5) 13 (18.1) 6 (54.5)

12 12 12 12 10* 12 12 12 12 12 12 12 12 9.5

88 (89.7) 4 (57.1) 54 (69.2) 444 (74) 14 (73.7)

8 6 6 12 36

89 (66.0) 7 (31.8) 17 (63) 1862 (71.9) 60 (68) 15 (75) 30 (56)

26 15 6 17 6 9 12

80.7 (9.1) NA 87.4 (6.9) 82.3 (4.3) 84.5 (10.4) 86 (R ¼ 65-99) 82.5 (10.9)

83.5 (R ¼ 65-102) 84.1 (7.2) 79.29 (12.7) NA 85 83.7 (6.7) 87 (6) 83.3 (7.7) 82.9 (8.4) 84 (8) NA 82.6 84.1 (7.7) 85.4 (9.1) 88.6 (6.7) NA 82.55 (7.1) 85.7 (8.8) 86.1 (7.3) 78 (11.7) 83.6 (7.1) 82 (13.4) 85y 83 (9.2) 85.9 (7.8) 78.5 (5.2)

CRCT, cluster randomized controlled trial; NA, not available; R, range; RCT, randomized controlled trial; SD, standard deviation. *Mean. y Median.

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

3

Table 2 Components of Interventions According to PRoFaNE Taxonomy Study (First Author) Exercise Component

Medication Assessment Type

Frequency

Single intervention Chenoweth31

Cox33

Gait, balance, and functional training (Wii, Nintendo)

Individual

3x/wk, 6 wk

60 min

Galik36

Social: Staff trainingdFunctionFocused Care for the cognitively impaired Social: Staff trainingd education on inappropriate drug prescription

García-Gollarte37

Hewitt38

Gait, balance, and functional training Strength and resistance

Group

2x/wk, 25 wk 60 min

Juola40

Social: Staff trainingdEducation on harmful medication

Kovács42

Gait, balance, and functional training Strength and resistance Endurance

Group

2x/wk, 12 mo

Lapane43 Law44 Meyer46 Patterson48 Potter49 Sakamoto52

Medication review Vitamin D Falls risk Medication review Medication review Olfactory lavender patch

Saravanakumar53 Schoenfelder

55

Sitjà-Rabert56

3D: tai chi Flexibility: yoga Strength and resistance General physical activity Gait, balance, and functional training Strength and resistance Other: WBV

Group

2x/wk, 14 wk 30 min

Group

3x/wk, 3 mo

20 min

Group 3x/wk, 6 wk Individual

30 min

Teresi57

Tuunainen58

Social: Staff trainingdResident-to resident elder mistreatment Gait, balance, and functional training Strength and resistance Flexibility

Group

2x/wk, 3 mo

60 min

van Gaal59

Varela60 Walker61

Other

Social: Staff trainingddementia care mapping and person-centered care Social: Staff trainingdCONNECT for staff interactions Social: Staff trainingdeducation on fracture prevention

Colón-Emeric32

Fu35

Environment

Duration

Social: Staff trainingdSafe or Sorry (adverse events) Endurance

NA

7x/wk, 15 mo 15 min Social: Staff trainingdGuide to Action Care Home falls prevention (continued on next page)

4

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

Table 2 (continued ) Study (First Author) Exercise Component

Medication Assessment Type

Frequency

Ward62

Other: Short stick exercises Multiple interventions Imaoka39 Gait, balance, and functional training Strength and resistance Schnelle54 Strength and resistance General physical activity

Strength and resistance

Group

2x/wk, 6 mo

25 min

Individual

2x/wk, 3 mo

NA

Individual

5x/wk, 8 mo

NA

Individual

3x/wk, 3 mo

NA

2x/wk

75 min

Multifactorial interventions Becker,30 Rapp50 Gait, balance, and Group functional training Strength and resistance General physical activity

Dyer34

Kerse41

McMurdo45

Neyens47

Gait, balance, and functional training Strength and resistance Flexibility General physical activity Gait, balance, and functional training General physical activity Gait, balance, and functional training Strength and resistance Flexibility

3x/wk, 3 mo Group, individual if needed

Vitamin D

Management of urinary incontinence Fluid or nutrition therapy Podiatric surgery or intervention Falls risk Environment

Furnishings and adaptations: dwelling unit including entrances Body-worn protective aids Aids for personal mobility Social: Staff training Social: Staff training Vision Falls risk Podiatry Medication review surgery or Environment intervention

40 min

Individual

Daily or more Short doses

Falls risk

Group

2x/wk, 6 mo

30 min

NA

Individual

NA

NA

Gait, balance, and functional training

2x/wk, 6 mo Group, individual if needed

Vision Furnishings and Falls risk adaptations: Medication review dwelling unit Environment including entrances Furnishings and Falls risk adaptations: Medication review dwelling unit Environment including entrances Aids for personal mobility Body-worn protective aids Social: Staff training Podiatric Furnishings and Falls risk surgery or adaptations: Medication review intervention dwelling unit Environment including entrances Aids for personal mobility Body-worn protective aids Social: Staff training Dementia Furnishings and Falls risk care mapping adaptations: Medication review dwelling unit Environment including entrances Aids for personal mobility Alarm systems Social: Staff training

Ray51

Whitney63

Other

Social: Staff trainingdeducation on falls injury prevention strategies

Yokoi65

Wylie64

Environment

Duration

45 min

Social: Staff training

NA, not available; WBV, whole body vibration.

then completed full-text reviews of all articles meeting initial criteria. Reference lists of studies relevant to the review were searched to find additional relevant studies. Studies were crossreferenced with the Cochrane review19 to ensure all relevant studies were captured. Authors were contacted in cases where clarification on inclusion criteria and/or additional data was required pertaining to the study.

Inclusion Criteria Consistent with the previous meta-analysis, studies were required to be in English language and original randomized or cluster randomized controlled trials, of at least 6 months in duration, including intervention and follow-up, with monitoring of falls or number of fallers over the duration of intervention or post-intervention.20

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

5

Table 3 Results of Interventions Study (First Author)

Falls

Fallers, n/N (%)

Recurrent Fallers, n/N (%)

Intervention

Control

P value or CI

Intervention

Control

P value or CI

Intervention

Control

P value

Becker30 Chenoweth31

1.40/RY NA

2.56/RY d

<.001 d

247/472 (52.3) 25/82 (30.5)

115/472 (24.4) d

.015 d

6.70/RY* NA/3315 2.17/RY 0.54/RY 0.48/R 0.32/R 1.31/RY LE: HR: 0.475 VD: HR: 0.575 C: HR: 0.276 2.25/RY NA 22 total 3.42/RY 2.01/RY 3.02/RY 1.8/R 2.09/RY 1.96/RY 221 total 1.29/RY NA 1.04/RY Tai chi: 1.6/R Yoga: 0.8/R 0.67/RY 42 total 57 total 0.235 M: 42 total MB: 24 total 1.51/RY 1/R 1.9/RY Mean 13/mo 78 total 2.3/R NA

7.11/RY NA/2322 4.02/RY 1.52/RY 1.2/R 0.43/R 2.91/RY HR: 1

.96 .17 .27 <.001 .02 NA CI: 0.17-0.74 .15 .31 .037 <0.001 d NA CI: 0.85-1.01 NA .17 .60 .029 .09 NA CI: 0.42-0.74 d .04 .65

d d 51/94 (54.3) d 25/50 (50.0) 104/372 (28.0) 73/108 (67.6) 9/17 (52.9)

.038 .02 .03 d d .94 d .02 NA NA NA

66/509 (12.9) NA

Colón-Emeric32 Cox33 Dyer34 Fu35 Galik36 García-Gollarte37 Hewitt38 Imaoka39

188/509 (36.9) DM: 22/109 (20.2) PC: 33/98 (33.7) NA NA 56/102 (54.9) NA 15/53 (28.3) 82/344 (23.8) 50/113 (44.2) LE: 7/22 (31.8) VD: 6/17 (35.3) C: 4/19 (21.1) 42/118 (35.6) 162/310 (52.2) 16/32 (50) NA 770/1762 (43.7) 20/52 (38.0) 299/574 (52.1) NA NA 25/45 (55.6) 143/365 (39.2) NA 26/73 (35.6) NA

NA NA 26/102 (25.5) NA NA 2/344 (0.6) 9/113 (8.0) NA

d d 25/94 (26.6) d d 4/372 (1.1) 20/108 (18.5) d

d d .28 d d NA NA d

60/109 (55.0) 146/329 (44.3) 20/30 (66.7) d 833/1955 (42.6) 22/38 (57.9) 291/551 (52.8) d d 31/48 (64.6) 189/360 (52.5) d 36/72 (50.0) d

NA .48 NA d CI: .95-1.25 .10 .88 d d .40 CI: 0.56-0.87 d .08 d

27/118 (22.9) NA 5/32 (15.6) NA NA 13/52 (25.0) NA NA NA NA NA 91/221 (43.8) 14/73 (19.2) NA

41/109 (37.6) d 7/30 (23.3) d d 9/38 (23.7) d d d d d 129/261 (54.1) 23/72 (31.9) d

NA d NA d d 1 d d d d d .03 .08 d

29/98 (29.6) d

d 9/18 (50)

d d NA d NA

d d d d 25/88 (28.4) 12/17 (70.6) 16/54 (29.6)

d d d d CI: 0.58-2.03 NA .019

NA NA 12/159 (7.5) NA M: 6/16 (37.5) MB: 5/14 (35.7) NA NA NA NA NA 7/21 (33.3) NA

d d

d 14/18 (77.8)

NA d NA d NA

d d d d d 9/17 (52.9) d

d d d d d NA d

Juola40 Kerse41 Kovács42 Lapane43 Law44 McMurdo45 Meyer46 Neyens47 Patterson48 Potter49 Rapp50 Ray51 Sakamoto52 Saravanakumar53 Schnelle54 Schoenfelder55 Sitjà-Rabert56 Teresi57 Tuunainen58 van Gaal59 Varela60 Walker61 Ward62 Whitney63 Wylie64 Yokoi65

3.25/RY d 29 total 3.78/RY 2.31/RY 3.85/RY 1.8/R 2.54/RY 1.37/RY 142 total 2.35/RY d 1.40/RY 2.6/R 1.09/RY 12 total 0.18 64 total

.40 NA NA .42 .02

2.02/RY 1.5/R 4/RY Mean 10/mo 41 total 2.7/R d

CI: 0.35-1.16 .47 NA NA CI: 0.67-3.76 .30 d

17/92 (18.5) NA 35/159 (22.0) NA M: 7/16 (43.8) MB: 6/14 (42.9) NA NA NA NA 31/103 (30.1) 10/21 (47.6) 6/51 (11.8)

CI, confidence interval; DM, dementia mapping group; HR, hazard ratio; LE, low exercise group; M, muscle force group; MB, muscle force and balance group; NA, not available; PC, person-centered group; R, per resident; RY, per resident year; VD, vitamin D supplementation group. Bold text indicates results achieving statistical significance at the level of P < .05. *Fall rate for recurrent fallers.

Studies were required to be set in NHs, defined as “residential facilities that provide 24-hour-a-day surveillance, personal care, and limited clinical care for persons who are typically elderly and infirm.”23 For studies that included participants from multiple settings and where the different populations were analyzed separately, only data on NH residents were included. Interventions could be single, multiple, or multifactorial falls prevention programs evaluated for the prevention of falls as a primary or secondary outcome. Outcomes evaluated were number of falls, fallers, or recurrent fallers. Fallers were considered all residents sustaining at least 1 fall during the intervention or follow-up period. In the same way, recurrent fallers were defined as residents sustaining a minimum of 2 or more falls.

Risk of Bias Assessment Each study was assessed for bias using the Cochrane methodological quality assessment scheme24 according to the original review.20 Methodological quality was scored as follows: 0, not fulfilled or not mentioned; 1, partially met; or 2, completely met. Higher scores indicated better quality.

Data Synthesis and Analysis Synthesis of study data to summarize characteristics and effectiveness of the falls prevention programs was completed by V.P. and H.G. Studies were classified based on their intervention characteristics using the Prevention of Falls Network Europe (ProFaNE) taxonomy, as single, multiple, or multifactorial.25 Effects were based on the results reported in the included studies. Relative risk was calculated using crude data for number of falls (rate of falls per person-year in the intervention group to rate of falls per person-year in the control group) and fallers and recurrent fallers (probability of being a faller or recurrent faller in the intervention group to probability of being a faller or recurrent faller in the control group). Subgroup analyses based on intervention (single, multiple, or multifactorial) were also conducted with respect to falls, fallers, and recurrent fallers, and at least 2 studies were required in a group for a subanalysis to be performed. When studies had more than 1 intervention group, the control group results were divided evenly among the intervention arms to prevent the same individual from being included twice. Within single interventions, further subanalyses based on type of intervention (ie, exercise) were conducted.

6

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

Fig. 1. Forest plots: (A) effects of interventions on falls; (B) effects of interventions on fallers; (C) effects of interventions on recurrent fallers. FRA, falls risk assessment; PSRT, progressive strength and resistance training.

Calculations of effect size used inverse variance and a random effects model.26 Publication bias was assessed with a funnel plot using the trim and fill methods. To examine between-study variability, heterogeneity of variance tests were applied to the data. Calculation of I2 was used to assess the proportion of between-study variability to total variability. The calculation of 95% confidence intervals (CIs) and the meta-analyses for fallers and recurrent fallers were conducted using Meta-Essentials, version 1.4.27 The 95% CIs for fall rate were calculated according to the method described by Rothman and Greenland,28 and the Meta-Essentials27 worksheet was modified to incorporate these calculations into a meta-analysis. Because many of the studies measured the degree of cognitive impairment without identifying the proportion who had a diagnosis of dementia, we performed 2 subgroup analyses to determine the effects of cognitive impairment and dementia on fall interventions. Cognitive impairment or dementia were categorized from 0 to 3 based on the description of the study sample,29 with no information provided on cognitive status or unable to determine scored as 0; <40% scored as 1; 40% to 69% scored as 2; and 70% prevalence of cognitive impairment or dementia scored as 3. When the prevalence of cognitive impairment in the study sample was not directly described, approximations based on mean test scores and standard deviations were used to estimate percentage scoring below defined cut-offs for cognitive impairment

using z scores (see Supplementary Table 1 for assessment toolespecific cut-offs applied). Where the prevalence of people with a diagnosis of dementia was not stated, the same strategy was used, using assessments that described severity of dementia. Studies with a score of zero (no information on cognitive status) were excluded from the subanalyses. The results were reported based on the ProFaNE taxonomy and calculated using a weighted subgroup analysis that estimated the effect size of each prevalence category. Studies were reviewed to determine if balancing measures were assessed pre- and post-intervention and what type of balancing measures were present. Balancing measures were categorized as restraints, physical activity, and functional status. Physical activity levels were defined as a measure of the amount of physical activity (ie, kilocalories burned) engaged in by residents. Functional status was defined as measures of physical (ie, timed up and go) and activity of daily living (ie, Barthel Index) function. Results Selected Studies From August 2013 to April 2019, a total of 18 studies met the inclusion criteria, and in addition to the 14 studies published before

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

7

Fig. 1. (continued).

August 2013, 4 more studies were found to meet criteria and added to the review. In total, 36 studies were included in the review (Table 1), with 2 studies30,50 combined as 1 study in the meta-analysis because they reported data on the same cohort of residents. Quality of articles based on the risk of bias assessment ranged from 6 to 17 (Supplementary Table 2).

Study Characteristics As categorized by ProFaNE, there were 25 single interventions,31e33,35e38,40,42e44,46,48,49,52,53,55e62,65 3 multiple interventions,39,54,64 and 8 multifactorial interventions30,34,41,45,47,50,51,63 (Table 2). Thirteen studies used cognitive function as exclusion

8

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

Fig. 1. (continued).

criteria,36,38,41,42,45,47,53e56,60,64,65 whereas 30 studies reported on cognitive status30e32,34,36e45,47e58,60,63e65 (Supplementary Table 3). Twenty-two studies assessed balancing measures,34e36,38,39, 41e43,45,46,49,51,53e56,58,60,61,63e65 with 2 looking at restraint use,46,51 3 36,54,55 reporting on activity level, and 21 reporting on functional status34e36,38,39,41e43,45,49,51,53e56,58,60,61,63e65 (Supplementary Table 4).

falls,30,34,45 more than 3 falls,37 and more than 4 falls.38 Two showed significant decreases of 47% and 19% between intervention and control groups on recurrent fallers.30,51 Combined analysis of 11 studies revealed a significant overall effect of falls prevention interventions on recurrent fallers (RR ¼ 0.70, 95% CI ¼ 0.60-0.81; I2 ¼ 9.56%).30,34,37,38,40,42,45,51,52,58,64 According to the trim-and-fill method, there was publication bias for number of fallers, with an estimate of 3 unpublished studies.

Combined Effect on Falls, Fallers, and Recurrent Fallers Thirty-two studies reported on number of falls (Table 3),30,32e40,42e50,52e64 with 6 single,35,36,38,40,52,58 1 multiple,39 and 3 multifactorial studies30,47,50 seeing 17% to 72% fewer falls in the intervention group compared with the controls. Combined data from 13 studies showed a significant effect of falls prevention programs in reducing number of falls per person-year [risk ratio (RR) ¼ 0.73, 95% CI ¼ 0.60-0.88; I2 ¼ 92.51%] (Figure 1).30,34,38,40,43e45,47,48,50,52,54,59,61 Number of fallers was an outcome measure in 22 studies.30,31,34,36e42,44e46,49,50,52,54,56,58,63e65 Of these studies, 3 singleintervention studies reported 43% to 60% fewer fallers,31,36,65 with 1 reporting 11% more fallers31 seen in the intervention groups compared with the control group. One multifactorial study30,50 reported 29% fewer fallers in the intervention group. The combined effect from 20 studies showed that there was a significant effect of falls prevention programs on number of fallers (RR ¼ 0.80, 95% CI ¼ 0.72-0.89; I2 ¼ 67.40%).30,31,34,36e42,44e46,49,50,52,54,58,63e65 Twelve studies looked at recurrent fallers but used a variety of definitions, including more than 1 fall,40,42,51,52,56,58,64 more than 2

Single With respect to number of falls, single interventions did not have a significant combined effect (RR ¼ 0.78, 95% CI ¼ 0.58-1.05; I2 ¼ 91.11%).38,40,43,44,48,52,59,61 In a subgroup analysis, 3 studies investigating staff education showed a significant combined effect (RR ¼ 0.70, 95% CI ¼ 0.52-0.93; I2 ¼ 0.00%),40,59,61 but the 2 studies implementing medication review as an intervention did not (RR ¼ 1.13, 95% CI ¼ 0.06-20.45; I2 ¼ 94.42%).43,48 Single interventions showed a significant effect of reducing numbers of fallers (RR ¼ 0.78, 95% CI ¼ 0.69-0.89; I2 ¼ 62.15%).31,36e40,42,44,46,49,52,58,65 In a subgroup analysis, exercise interventions reduced the number of fallers (RR ¼ 0.64, 95% CI ¼ 0.550.75; I2 ¼ 0.00%),38,39,42,58,65 but staff education (RR ¼ 0.75, 95% CI ¼ 0.56-1.00; I2 ¼ 24.35%)31,36,37,40 and vitamin D supplementation (RR ¼ 1.02, 95% CI ¼ 0.69-1.52; I2 ¼ 0.00%)39,44 did not. A significant effect of single interventions on recurrent fallers was found (RR ¼ 0.60, 95% CI ¼ 0.52-0.70; I2 ¼ 0.00%).37,38,40,42,52,58 Exercise interventions had a significant combined effect on the number of

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

9

Fig. 2. Forest plot of effects of interventions on falls by proportion of participants with cognitive impairment or diagnosis of dementia.

recurrent fallers (RR ¼ 0.59, 95% CI ¼ 0.38-0.93; I2 ¼ 0%),38,42,58 as did the staff education studies (RR ¼ 0.60, 95% CI ¼ 0.43-0.85; I2 ¼ 0%).37,40 Multiple Three studies administered multiple interventions consisting of an exercise program combined with either vitamin D,39 incontinence care and fluid therapy,54 or podiatry,64 and showed a significant combined effect on reducing number of fallers (RR ¼ 0.62, 95% CI ¼ 0.39-0.97; I2 ¼ 0.00%).39,54,64 Multifactorial Seven studies used multifactorial interventions to decrease fall outcomes.30,34,41,45,47,50,51,63 Multifactorial interventions typically included falls risk assessment,30,34,41,45,47,50,51,63 medication review,34,45,47,51,63 environmental assessment,30,34,45,47,50,51,63 30,34,41,45,47,50,63 exercise, and staff education30,34,41,47,50,51,63 and together saw a significant decrease on number of falls (RR ¼ 0.65, 95% CI ¼ 0.45-0.94; I2 ¼ 88.16%).30,34,45,47,50 On the other hand, the multifactorial studies did not see a reduction in fallers (RR ¼ 0.91, 95% CI ¼ 0.66-1.25; I2 ¼ 84.02%)30,34,41,45,50,63 or recurrent fallers (RR ¼ 0.77, 95% CI ¼ 0.48-1.24; I2 ¼ 66.19%).30,34,45,51

Inclusion of Participants With Cognitive Impairment and Dementia Ten studies used varying degrees of cognitive impairment as an exclusion criteria, and of these, 6 were exercise interventions,38,53,55,56,60,65 2 multiple interventions,54,64 and 2 multifactorial interventions.41,45 Eleven of 25 single interventions could be determined to have a cognitive impairment prevalence of greater than 40%.31,32,36,38,40,42,43,48,49,52,56 Only 13 studies31,37,39,40,42,43,48,53,55,60,62,63,65 reported on the prevalence of people with dementia within the study, with 6 of these studies having a prevalence greater than 40%.31,39,40,42,43,63 Cognitive impairment and dementia classification were used as the basis for subgroup analyses, which combined all types of interventions. The prevalence of cognitive impairment and dementia had no relationship with the effect of the intervention on fallers or recurrent fallers, but did with number of falls (Figure 2). Increasing cognitive impairment had a negative impact on the effect of the interventions. Moving from a prevalence of 40%-69% to 70% cognitive impairment, the relative risk of falls changed significantly from 0.52 (95% CI ¼ 0.41-0.66) to 0.86 (95% CI ¼ 0.67-1.11) (P < .001). Only 3 studies looking at falls included dementia prevalence. Single-study results showed that 1 staff education study (prevalence of dementia 70, RR ¼ 0.69, 95% CI ¼ 0.57-0.84)40 and 1 medication review study (prevalence of dementia 40%-69%, RR ¼ 0.91, 95% CI ¼

10

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

0.82-1.00)43 had a lower incidence of falls in the intervention group, whereas 1 medication review intervention48 saw a higher incidence of falls in the intervention group (prevalence of dementia 40, RR ¼ 1.43, 95% CI ¼ 1.19-1.72). Each of these RRs differs from the others (P < .001) (Figure 2). Balancing Measures The 2 studies that assessed restraint use did not see any significant differences between the intervention and control groups.45,51 Eight studies reported differences in functional status between the intervention and control groups,34e36,38,42,54,64,65 with 1 study reporting on activity of daily living functional status36 and all reporting on physical functional status.34e36,38,42,54,64,65 A single study that found a decrease in physical function was investigating a podiatry intervention,64 and the 7 studies that saw better physical function incorporated exercise and physical activity as falls prevention strategies.34e36,38,42,54,65 Two studies reported significantly greater activity levels in the intervention groups.36,54 All 9 exercise-based intervention programs included balancing measures,35,38,42,53,55,56,58,60,65 whereas 2 of 10 social,36,61 3 of 6 medication,34,43,49 all 3 multiple39,54,64 and 5 of 8 multifactorial studies34,41,45,51,63 reported balancing measures. Discussion In this meta-analysis of studies examining falls prevention interventions in NHs, interventions reduced falls by 27%, fallers by 20%, and recurrent fallers by 30%. This finding updates a previous review focused on falls prevention in NHs, in which a combined effect of interventions was found only on recurrent fallers.20 In this study, both single and multiple interventions reduced fallers, single interventions reduced recurrent fallers, and multifactorial studies reduced the number of falls. Exercise as a single intervention, consisting of various combinations of gait, balance, and functional training as well as strength and resistance training, reduced both fallers and recurrent fallers.36,39,42,48 Exercise was also a component of all multiple interventions that reduced fallers and all but one of the multifactorial interventions that reduced number of falls, demonstrating the critical role of exercise in improving physical and functional performance.66e68 This is the first meta-analysis to identify an effect of exercise on preventing falls in NHs. Previous meta-analyses examining the effect of exercise on falls in older adults in residential care settings have not found that exercise as a single intervention is effective in reducing fall outcomes.19,69,70 There are 2 key differences between this meta-analysis and previous studies: first, we required studies to have at least 6 months of followup, and second, we used strict criteria to focus on high-care residential environments (NHs). The inclusion of brief studies of less than 6 months has an important impact on the effect of the intervention, both due to the insufficient duration and dose of the intervention and due to an inadequate follow-up period to detect a difference between groups in falls events. Including a heterogeneous sample of residential care environments also impacted the effect size of interventions in previous reviews because of the different falls risk factors and lower fall rates of residents in these low-care settings. An additional limitation of the Sherrington et al meta-analysis70 is that it reported only on number of falls and not fallers or recurrent fallers, which are clinically important outcomes demonstrating that the interventions prevented individuals from newly or repeatedly falling. Although we did not perform subanalyses by type of exercise, previous reviews in NH and community populations have consistently identified that gait, balance, and functional training and strength and resistance training69 are the exercise components most strongly associated with a reduction in falls.71 For example, the recent study by Hewitt et al38 provides

a model for a progressive resistance- and balance-focused program with efficacy in preventing falls in selected participants. Several different types of staff education interventions had an effect on reducing recurrent fallers and number of falls. For example, education on the risks of polypharmacy and falls-promoting medication directed to nursing staff40 and NH physicians37 contributed to a decrease in recurrent fallers. These studies were successful in reducing the use of psychotropic medications40 and significantly lowered inappropriate drug use and antipsychotic medication in the intervention group compared with the control group at the end of the study.37 Psychotropic medication use and psychotropic polypharmacy are both important as risk factors for recurrent falling in NHs.72 In contrast, 3 studies43,48,49 looked at the effect of medication review alone on falls43,48 and fallers49 and did not find any benefit. Medication review on its own may therefore not be adequate without the addition of staff education to support deprescribing and to monitor for adverse effects of medications. Other education interventions included in this meta-analysis were based on person-centered care for people with dementia, where falls reduction was a secondary objective of the intervention.31,36 These types of interventions hold promise by shifting focus toward addressing the comfort and needs of the resident as falls-preventive measures. This review raises questions about the generalizability of some interventions to NH residents with cognitive impairment. Although all multiple and multifactorial interventions included participants with cognitive impairment, few exercise interventions did so. The 3 studies included in the dementia subanalysis were focused on staff training40 and medication review43,48dall interventions in which the residents are not directly or actively engaged. There is a clear need for the development of exercise or physical activity interventions specific for people with dementia and their evaluation for the prevention of falls. Promising interventions evaluated thus far in small pilot studies include a person-centered walking program73 and environmental design to promote increased activity.74,75 Practices such as enhanced supervision or hourly rounding also hold promise as interventions designed to better address care needs of people with dementia to prevent falls,76,77 although these have yet to be evaluated in controlled studies in NHs. Most of the included studies in this meta-analysis were designed to evaluate the efficacy of the intervention, rather than to establish real-world effectiveness. This is particularly notable in studies that use restrictive eligibility criteria and where the delivery of the intervention is by trained research staff. Adherence has been shown to be an important factor moderating treatment effectiveness, both for falls prevention78 and for fracture prevention using hip protectors.79 Other contributors to lack of effectiveness in this setting include limited resources, insufficiently vigorous interventions, and poor support from staff because of workload issues.80 Interestingly, 2 studies included in this analysis using more pragmatic designs30,37 had large effect sizes, demonstrating that these types of effectiveness trials are feasible in NHs. An important part of evaluating the effectiveness of an intervention under usual conditions is monitoring for adverse events. When falls reduction is the primary goal, there is the risk of taking a route to achieve this goal that includes restricting the level of activity and restraint use, thereby worsening physical function. Although restraint use has been shown to be ineffective at preventing falls, and the evidence supports restraint reduction or restraint-free environments, restraint use to prevent falls remains prevalent.81e83 Future pragmatic studies of falls prevention in NHs might consider including other safety and quality-of-care outcomes other than falls to monitor for potential unintended consequences of implementing the falls prevention intervention. There are some limitations to our meta-analysis. Falls intervention studies, particularly the single intervention studies, reported rate of falls using many different methods. Future studies should adopt

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

ProFaNE guidelines, which recommends reporting on number of falls, number of fallers/recurrent fallers, fall rate per person-year, and time to first fall.84 Another limitation is in the analysis by cognitive impairment and dementia; because of the variability between studies in describing the cognitive status of the study participants, we had to rely on estimates and could not establish the prevalence of dementia for most studies. Studies represented in this meta-analysis were conducted in 12 different countries, and it is important to note that NHs vary by region in their model of care and access to resources. Conclusions and Implications There is evidence to support the implementation of falls prevention interventions in NHs, although exercise interventions appear to be less effective in people with cognitive impairment and have not been evaluated in people with dementia. It is recommended that future NH falls intervention studies consider pragmatic study designs, follow ProFaNE guidelines for reporting fall outcomes, and include measures to evaluate potential unintended consequences of the intervention. There is also a need for studies of exercise interventions designed and evaluated in people with cognitive impairment. Acknowledgments The authors thank Jessica Babineau and Steven Stewart for their assistance on the project. References 1. Rapp K, Becker C, Cameron ID, et al. Epidemiology of falls in residential aged care: Analysis of more than 70,000 falls from residents of bavarian nursing homes. J Am Med Dir Assoc 2012;13:187.e1e187.e6. 2. Indicator library: Falls in the Last 30 Days in Long-Term Care 2018. Canadian Institute for Health Information. Available at: https://yourhealthsystem.cihi.ca/ hsp/indepth?lang¼en#/indicator/051/2/C5001/. Accessed July 2, 2019. 3. Percentage of long-stay residents experiencing one or more falls with major injury. Nursing Home Compare 2018. Available at: https://data.medicare.gov/ Nursing-Home-Compare/Quality-Measures-Long-Stay/iqd3-nsf3. Accessed July 2, 2019. 4. Simmons SF, Schnelle JF, Sathe NA, et al. Defining safety in the nursing home setting: Implications for future research. J Am Med Dir Assoc 2016;17: 473e481. 5. Seitz D, Purandare N, Conn D. Prevalence of psychiatric disorders among older adults in long-term care homes: A systematic review. Int Psychogeriatr 2010; 22:1025e1039. 6. Bergh S, Holmen J, Saltvedt I, et al. Dementia and neuropsychiatric symptoms in nursing-home patients in Nord-Trøndelag County. Tidsskr Nor Laegeforen 2012;132:1956e1959. 7. Graham JE, Rockwood K, Beattie BL, et al. Prevalence and severity of cognitive impairment with and without dementia in an elderly population. Lancet 1997; 349:1793e1796. 8. Allan LM, Ballard CG, Rowan EN, Kenny RA. Incidence and prediction of falls in dementia: A prospective study in older people. PLoS One 2009;4:e5521. 9. van Doorn C, Gruber-Baldini AL, Zimmerman S, et al. Dementia as a risk factor for falls and fall injuries among nursing home residents. J Am Geriatr Soc 2003; 51:1213e1218. 10. Luk JK, Chan TY, Chan DK. Falls prevention in the elderly: Translating evidence into practice. Hong Kong Med J 2015;21:165e181. 11. Updated American Geriatrics Society/British Geriatrics Society Clinical Practice Guideline for Prevention of Falls in Older Persons and Recommendations 2011. American Geriatrics Society/British Geriatrics Society. Available at: www. geriatricscareonline.org/FullText/CL014/CL014_BOOK003. Accessed March 26, 2019. 12. Iaboni A, Van Ooteghem K, Marcil MN, et al. A palliative approach to falls in advanced dementia. Am J Geriatr Psychiatry 2018;26:407e415. 13. Sze TW, Leng CY, Lin SKS. The effectiveness of physical restraints in reducing falls among adults in acute care hospitals and nursing homes: A systematic review. JBI Libr Syst Rev 2012;10:307e351. 14. Komara FA. The slippery slope: Reducing fall risk in older adults. Prim Care 2005;32:683e697. 15. Takasaki DA. Tying the knot between resident care and quality of life. Balance 1997;1:16, 18. 16. Castle NG, Mor V. Physical restraints in nursing homes: A review of the literature since the Nursing Home Reform Act of 1987. Med Care Res Rev 1998;55: 139e170; discussion 171e176.

11

17. Rubenstein LZ, Josephson KR, Robbins AS. Falls in the nursing home. Ann Intern Med 1994;121:442e451. 18. Cameron ID, Murray GR, Gillespie LD, et al. Interventions for preventing falls in older people in nursing care facilities and hospitals. Cochrane Database Syst Rev; 2010:CD005465. 19. Cameron ID, Dyer SM, Panagoda CE, et al. Interventions for preventing falls in older people in care facilities and hospitals (review). Cochrane Database Syst Rev 2018;12:CD005465. 20. Vlaeyen E, Coussement J, Leysens G, et al. Characteristics and effectiveness of fall prevention programs in nursing homes: A systematic review and metaanalysis of randomized controlled trials. J Am Geriatr Soc 2015;63:211e221. 21. Holte HH, Underland V, Hafstad E. Review of Systematic Reviews on Prevention of Falls in Institutions [Internet]. Oslo, Norway: Knowledge Centre for the Health Services at The Norwegian Institute of Public Health (NIPH), 2015. Available at: http://www.ncbi.nlm.nih.gov/books/NBK390564/. Accessed October 1, 2019. 22. Choi M, Hector M. Effectiveness of intervention programs in preventing falls: A systematic review of recent 10 years and meta-analysis. J Am Med Dir Assoc 2012;13:188.e13e188.e21. 23. Sorenson SV, De Lissovoy G, Kunaprayoon D, et al. A taxonomy and economic consequences of nursing home falls. Drugs Aging 2006;23:251e262. 24. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0. The Cochrane Collaboration, 2011. Available at: 2011. www.handbook.cochrane.org; 2011. 25. ProFaNE Taxonomy 2011. Prevention of Fall Network Europe. Available at: www.profane.eu.org/documents/Falls_Taxonomy.pdf. Accessed March 21, 2019. 26. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to MetaAnalysis. West Sussex, UK: John Wiley & Sons, Ltd; 2009. 27. Suurmond R, van Rhee H, Hak T. Introduction, comparison, and validation of meta-essentials: A free and simple tool for meta-analysis. Res Synth Methods 2017;8:537e553. 28. Rothman KJ, Greenland S. Modern Epidemiology. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2008. 29. Oliver D, Connelly JB, Victor CR, et al. Strategies to prevent falls and fractures in hospitals and care homes and effect of cognitive impairment: Systematic review and meta-analyses. BMJ 2007;334:82. 30. Becker C, Kron M, Lindemann U, et al. Effectiveness of a multifaceted intervention on falls in nursing home residents. J Am Geriatr Soc 2003;51: 306e313. 31. Chenoweth L, King MT, Jeon H, et al. Caring for Aged Dementia Care Resident Study (CADRES) of person-centred care, dementia-care mapping, and usual care in dementia: A cluster-randomised trial. Lancet Neurol 2009;8: 317e325. 32. Colón-Emeric CS, Corazzini K, McConnell ES, et al. Effect of promoting highquality staff interactions on fall prevention in nursing homes: A clusterrandomized trial. JAMA Intern Med 2017;177:1634e1641. 33. Cox H, Puffer S, Morton V, et al. Educating nursing home staff on fracture prevention: A cluster randomised trial. Age Ageing 2008;37:167e172. 34. Dyer CA, Taylor GJ, Reed M, et al. Falls prevention in residential care homes: A randomised controlled trial. Age Ageing 2004;33:596e602. 35. Fu AS, Gao KL, Tung AK, et al. Effectiveness of exergaming training in reducing risk and incidence of falls in frail older adults with a history of falls. Arch Phys Med Rehabil 2015;96:2096e2102. 36. Galik E, Resnick B, Hammersla M, Brightwater J. Optimizing function and physical activity among nursing home residents with dementia: Testing the impact of function-focused care. Gerontologist 2014;54:930e943. 37. García-Gollarte F, Baleriola-Júlvez J, Ferrero-López I, et al. An educational intervention on drug use in nursing homes improves health outcomes resource utilization and reduces inappropriate drug prescription. J Am Med Dir Assoc 2014;15:885e891. 38. Hewitt J, Goodall S, Clemson L, et al. Progressive resistance and balance training for falls prevention in long-term residential aged care: A cluster randomized trial of the sunbeam program. J Am Med Dir Assoc 2018;19:361e369. 39. Imaoka M, Higuchi Y, Todo E, et al. Low-frequency exercise and vitamin D supplementation reduce falls among institutionalized frail elderly. Int J Gerontol 2016;10:202e206. 40. Juola AL, Bjorkman MP, Pylkkanen S, et al. Nurse education to reduce harmful medication use in assisted living facilities: Effects of a randomized controlled trial on falls and cognition. Drugs Aging 2015;32:947e955. 41. Kerse N, Peri K, Robinson E, et al. Does a functional activity programme improve function, quality of life, and falls for residents in long term care? Cluster randomised controlled trial. BMJ 2008;337:a1445. 42. Kovács E, Sztruhár Jónásné I, Karóczi CK, et al. Effects of a multimodal exercise program on balance, functional mobility and fall risk in older adults with cognitive impairment: A randomized controlled single-blind study. Eur J Phys Rehabil Med 2013;49:639e648. 43. Lapane KL, Hughes CM, Daiello LA, et al. Effect of a pharmacist-led multicomponent intervention focusing on the medication monitoring phase to prevent potential adverse drug events in nursing homes. J Am Geriatr Soc 2011;59:1238e1245. 44. Law M, Withers H, Morris J, Anderson F. Vitamin D supplementation and the prevention of fractures and falls: Results of a randomised trial in elderly people in residential accommodation. Age Ageing 2006;35: 482e486.

12

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

45. McMurdo ME, Millar AM, Daly F. A randomized controlled trial of fall prevention strategies in old peoples’ homes. Gerontology 2000;46:83e87. 46. Meyer G, Köpke S, Haastert B, Mülhauser I. Comparison of a fall risk assessment tool with nurse’s judgement alone: A cluster-randomised controlled trial. Age Ageing 2009;38:417e423. 47. Neyens JC, Dijcks BP, Twisk J, et al. A multifactorial intervention for the prevention of falls in psychogeriatric nursing home patients, a randomised controlled trial (RCT). Age Ageing 2009;38:194e199. 48. Patterson SM, Hughes CM, Crealey G, et al. An evaluation of an adapted U.S. model of pharmaceutical care to improve psychoactive prescribing for nursing home residents in Northern Ireland (Fleetwood Northern Ireland Study). J Am Geriatr Soc 2010;58:44e53. 49. Potter K, Flicker L, Page A, Etherton-Beer C. Deprescribing in frail older people: A randomised controlled trial. PLoS One 2016;11:e0149984. 50. Rapp K, Lamb SE, Büchele G, et al. Prevention of falls in nursing homes: Subgroup analyses of a randomized fall prevention trial. J Am Geriatr Soc 2008;56: 1092e1097. 51. Ray WA, Taylor JA, Meador KG, et al. A randomized trial of a consultation service to reduce falls in nursing home. JAMA 1997;278:557e562. 52. Sakamoto Y, Ebihara S, Ebihara T, et al. Fall prevention using olfactory stimulation with lavender odor in elderly nursing home residents: A randomized controlled trial. J Am Geriatr Soc 2012;60:1005e1011. 53. Saravanakumar P, Higgins IJ, van der Riet PJ, et al. The influence of tai chi and yoga on balance and falls in a residential care setting: A randomised controlled trial. Contemp Nurse 2014;48:76e87. 54. Schnelle JF, Kapur K, Alessi C, et al. Does an exercise and incontinence intervention save healthcare costs in a nursing home population? J Am Geriatr Soc 2003;51:161e168. 55. Schoenfelder DP. A fall prevention program for elderly individuals. Exercise in long-term care settings. J Gerontol Nurs 2000;26:43e51. 56. Sitjà-Rabert M, Martínez-Zapata MJ, Fort Vanmeerhaeghe A, et al. Effects of a whole body vibration (WBV) exercise intervention for institutionalized older people: A randomized, multicentre, parallel, clinical trial. J Am Med Dir Assoc 2015;16:125e131. 57. Teresi JA, Ramírez M, Fulmer T, et al. Resident-to-resident mistreatment: Evaluation of a staff training program in the reduction of falls and injuries. J Gerontol Nurs 2018;44:15e23. 58. Tuunainen E, Rasku J, Jäntti P, et al. Postural stability and quality of life after guided and self-training among older adults residing in an institutional setting. Clin Interv Aging 2013;8:1237e1246. 59. van Gaal BGI, Schoonhoven L, Mintjes JAJ, et al. Fewer adverse events as a result of the SAFE or SORRY? Programme in hospitals and nursing homes. Part I: Primary outcome of a cluster randomised trial. Int J Nurs Stud 2011;48: 1040e1048. 60. Varela S, Cancela JM, Seijo-Martinez M, Ayán C. Self-paced cycling improves cognition on institutionalized older adults without known cognitive impairment: A 15-month randomized controlled trial. J Aging Phys Act 2018;26: 614e623. 61. Walker GM, Armstrong S, Gordon AL, et al. The Falls in Care Home Study: A feasibility randomized controlled trial of the use of a risk assessment and decision support tool to prevent falls in care homes. Clin Rehabil 2016;30: 972e983. 62. Ward JA, Harden M, Gibson RE, Byles JE. A cluster randomised controlled trial to prevent injury due to falls in a residential aged care population. Med J Aust 2010;192:319e322. 63. Whitney J, Jackson SHD, Martin FC. Feasibility and efficacy of a multi-factorial intervention to prevent falls in older adults with cognitive impairment living in residential care (ProF-Cog). A feasibility and pilot cluster randomised controlled trial. BMC Geriatr 2017;17:115. 64. Wylie G, Menz HB, McFarlane S, et al. Podiatry intervention versus usual care to prevent falls in care homes: Pilot randomised controlled trial (the PIRFECT study). BMC Geriatr 2017;17:143.

65. Yokoi K, Yoshimasu K, Takemura S, et al. Short stick exercises for fall prevention among older adults: A cluster randomized trial. Disabil Rehabil 2015;37: 1268e1276. 66. Álvarez Barbosa F, Del Pozo-Cruz B, Del Pozo-Cruz J, et al. Factors associated with the risk of falls of nursing home residents aged 80 or older. Rehabil Nurs 2016;41:16e25. 67. Moyer HS, Gale J, Severe S, et al. Outcome measures correlated with falls in nursing home residentsdA pilot study. Physiother Theory Pract 2017;33: 725e732. 68. McGough EL, Logsdon RG, Kelly VE, Teri L. Functional mobility limitations and falls in assisted living residents with dementia: Physical performance assessment and quantitative gait analysis. J Geriatr Phys Ther 2013;36: 78e86. 69. Lee SH, Kim HS. Exercise interventions for preventing falls among older people in care facilities: A meta-analysis. Worldviews Evid Based Nurs 2017;14: 74e80. 70. Sherrington C, Michaleff ZA, Fairhall N, et al. Exercise to prevent falls in older adults: An updated systematic review and meta-analysis. Br J Sports Med 2017; 51:1750e1758. 71. Sherrington C, Fairhall NJ, Wallbank GK, et al. Exercise for preventing falls in older people living in the community. Cochrane Database Syst Rev; 2019:CD012424. 72. Thapa PB, Gideon P, Fought RL, Ray WA. Psychotropic drugs and risk of recurrent falls in ambulatory nursing home residents. Am J Epidemiol 1995; 142:202e211. 73. Chu CH, Puts M, Brooks D, et al. A feasibility study of a multifaceted walking intervention to maintain the functional mobility, activities of daily living, and quality of life of nursing home residents with dementia. Rehabil Nurs; 2019 [Epub ahead of print]. 74. Detweiler MB, Murphy PF, Kim KY, et al. Scheduled medications and falls in dementia patients utilizing a wander garden. Am J Alzheimers Dis Other Demen 2009;24:322e332. 75. Jensen L, Padilla R. Effectiveness of environment-based interventions that address behavior, perception, and falls in people with Alzheimer’s disease and related major neurocognitive disorders: A systematic review. Am J Occup Ther 2017;71:7105180030p1e7105180030p10. 76. Shimada H, Tiedemann A, Lord SR, Suzuki T. The effect of enhanced supervision on fall rates in residential aged care. Am J Phys Med Rehabil 2009;88:823e828. 77. Goldsack J, Bergey M, Mascioli S, Cunningham J. Hourly rounding and patient falls: What factors boost success? Nursing 2015;45:25e30. 78. Nyman SR, Victor CR. Older people’s recruitment, sustained participation, and adherence to falls prevention interventions in institutional settings: A supplement to the Cochrane systematic review. Age Ageing 2011;40:430e436. 79. Korall AM, Feldman F, Scott VJ, et al. Facilitators of and barriers to hip protector acceptance and adherence in long-term care facilities: A systematic review. J Am Med Dir Assoc 2015;16:185e193. 80. Neyens JC, van Haastregt JC, Dijcks BP, et al. Effectiveness and implementation aspects of interventions for preventing falls in elderly people in long-term care facilities: A systematic review of RCTs. J Am Med Dir Assoc 2011;12:410e425. 81. Hofmann H, Schorro E, Haastert B, Meyer G. Use of physical restraints in nursing homes: A multicentre cross-sectional study. BMC Geriatr 2015;15:129. 82. Estévez-Guerra GJ, Fariña-López E, Núñez-González E, et al. The use of physical restraints in long-term care in Spain: A multi-center cross-sectional study. BMC Geriatr 2017;17:29. 83. Lam K, Kwan JSK, Wai Kwan C, et al. Factors associated with the trend of physical and chemical restraint use among long-term care facility residents in Hong Kong: Data from an 11-year observational study. J Am Med Dir Assoc 2017;18:1043e1048. 84. Lamb SE, Jørstad-Stein EC, Hauer K, Becker C. Development of a common outcome data set for fall injury prevention trials: The prevention of falls network Europe consensus. J Am Geriatr Soc 2005;53:1618e1622.

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

Appendix

Iden fica on

Combined search results of all electronic databases: (total n = 1073)

Excluded duplicates (n = 336)

Poten ally relevant ar cles iden fied and screened for retrieval (n = 737)

Ar cles retrieved for full-text assessment of eligibility ar cles assessed for eligibility (n = 149)

Full-text ar cles excluded, with reasons (n = 131) Se ng: not conducted in a nursing home (n = 27) Study Design: not a RCT (n = 25) No relevant outcomes on number of falls, fallers, or recurrent fallers (n = 27) Not English (n = 2) Study objec ve: not designed to prevent falls (n = 20) Protocol/Thesis/Conference abstracts (n = 25) Study Length (n = 2) Fall data not consistently reported (n=3)

Included

Eligibility

Screening

Ar cles excluded on the basis of tle and abstract (total n = 588) Se ng: not conduced in a nursing home (n = 102) Study design: not a RCT (n = 149) Study objec ve: not designed to prevent falls (n = 240) No relevant outcomes on number of falls, fallers, or recurrent fallers (n = 47) Not English (n = 3) Protocol/Thesis/Conference abstracts/Incomplete Studies (n = 47)

Studies included (n = 18)

Addi onal records prior to 2013 (n = 4)

Supplementary Fig. 1. PRISMA flow diagram.

Supplementary Table 1 Cut-off Scores to Define Cognitive Impairment and Dementia Test Abbreviated Mental Test (AMT)

Cut-off 7/10

Addenbrooke’s Cognitive Examination-Revised (ACE-R) Clinical Dementia Rating Scale (CDR) Dementia Global Deterioration Scale (DGDS) Hasegawa Dementia ScaleeRevised (HDS-R) Mini-Examen Cognoscitivo (MEC)

82/100

Mini-Mental State Examination (MMSE)

23/30

1/3 4/6 20/30 23/30

Measure Cognitive impairment Dementia Dementia Dementia Dementia Cognitive impairment Cognitive impairment

12.e1

12.e2

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

Supplementary Table 2 Risk of Bias Assessment Study (First Author)

Fall Incident Clearly Defined and Related to Staff

Inclusion and Blinded Randomization Exclusion Criteria Clearly Defined

Treatment and Control Groups Comparable at Baseline

Intention- Loss to Total Blinded Identical Blinded Blinded Identical Follow- Score Standard Care Treatment Outcome Subjects Ascertainment of to-Treat (0-22) up Analysis Outcomes Programs for Providers Assessors Both Groups

Becker30 Chenoweth31 Colón-Emeric32 Cox33 Dyer34 Fu35 Galik36 García-Gollarte37 Hewitt38 Imaoka39 Juola40 Kerse41 Kovács42 Lapane43 Law44 McMurdo45 Meyer46 Neyens47 Patterson48 Potter49 Rapp50 Ray51 Sakamoto52 Saravanakumar53 Schnelle54 Schoenfelder55 Sitjà-Rabert56 Teresi57 Tuunainen58 van Gaal59 Varela60 Walker61 Ward62 Whitney63 Wylie64 Yokoi65

1 0 0 0 0 2 0 0 1 2 0 2 2 1 0 1 1 0 0 2 1 1 2 1 1 1 0 0 0 0 0 2 0 1 2 2

2 0 2 1 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 2 2 1 2 2 2 2 0 2 0 2 2 0 2 2 2

1 1 1 0 1 2 2 1 1 2 1 2 2 1 2 2 2 2 1 1 2 2 2 2 2 1 2 1 1 1 2 1 1 1 1 2

0 0 0 0 0 2 2 0 0 2 0 0 2 0 0 0 0 0 0 0 0 0 2 2 0 0 0 0 2 0 0 0 0 0 0 0

2 2 2 2 2 1 1 1 2 1 1 2 1 0 2 1 2 2 2 2 2 2 2 2 2 0 2 1 1 1 1 2 2 2 2 2

0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 2 1 0 0 2 2 0 2 0 2 2 2 0 1 2 0 0 1 1 0 2 2 1 2 0 2 2 0 0 2 1 0 1 1 1

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

2 2 2 0 2 2 2 0 2 0 2 2 0 0 2 2 0 2 1 2 0 2 2 2 0 0 2 0 0 2 1 0 2 0 2 2

2 2 0 1 2 2 0 0 2 0 0 0 0 0 0 0 2 0 0 0 0 0 0 2 0 2 2 2 2 0 0 2 0 2 0 0

12 11 10 6 11 17 13 8 14 11 10 14 13 6 11 12 10 10 9 12 9 13 16 16 11 8 14 8 10 6 10 12 7 11 12 13

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

12.e3

Supplementary Table 3 Cognitive Impairment Measures and Prevalence Study (First Author)

30

Cognition-Based Exclusion Criteria

CI Prevalence Score

Dementia Prevalence Score

CI: 273 (57.8%) DGDS 5.3 (1.1)

2 3

0 3

CI: 583 (64.3%) d AMT: 7.4 (2.6) d d DGDS 1-3: 355 (70.7%) 4-5: 104 (20.7%) 6: 39 (7.8%) N/A: 4 (0.8%) CI: 45 (41.7%) HDS-R: 14.6 (7.9)

2 0 2 0 3 0

0 0 0 0 0 1

2 3

0 3

CDR 0: 1 (0.9%) 0.5: 6 (5.6%) 1: 17 (15.7%) 2: 27 (25%) 3: 57 (52.8%) AMT: 7.1 (2.4)

3

3

2

0

MMSE Scores, Mean (SD)

Other Measure of Cognition Scores, n (% of Residents) or Mean (SD)

Intervention

Control

Intervention

Control

CI: 275 (54%) DGDS DM ¼ 5.6 (1.3) PC ¼ 5.6 (0.73) CI: 571 (64.4%) d AMT: 6.2 (3.1) d d DGDS 1-3: 371 (71.9%) 4-5: 98 (19%) 6: 43 (8.3%) N/A: 4 (0.8%) CI: 63 (55.8%) HDS-R: 12.4 (7.9) 12.8 (8.6) 12.7 (9.5) CDR 0: 0 (0%) 0.5: 8 (6.8%) 1:16 (13.6%) 2: 31 (26.3%) 3: 63 (53.4%) AMT: 7.4 (2.3)

Becker Chenoweth31

d

d

d

Colón-Emeric32 Cox33 Dyer34 Fu35 Galik36 García-Gollarte37

d d d d MMSE score >15 d

d d d d 8.7 (4) d

d d d d

Hewitt38 Imaoka39

MMSE score 15 d

d d

d d

Juola40

d

8.8 (R ¼ 0-30)

10 (R ¼ 0-30)

Inability to discuss and remember a goal and participate in programming for the goal MMSE score >23, inability to follow simple instructions d

d

d

20.85 (3.2)

20.93 (3.8)

Dementia mild: 30 (69.5%) moderate: 13 (30.2%)

Dementia mild: 29 (67.4%) moderate: 14 (32.6%)

3

3

d

d

2

d 19.1 (3.9) d 9.3 (7.8)

d 18.0 (5.9) d 7.0 (6.6)

CI moderate: 47.1% severe: 26.2% Dementia: 43.4% AD: 14.6% 22% in EMI facilities d d d

3

0 3 0 3

0 0 0 0

Patterson48

d MMSE score <12 d Nonpsychogeriatric patients d

CI moderate: 47.8% severe: 22.3% Dementia: 35.4% AD: 12.7% 25% in EMI facilities d d d

d

d

3

1

Potter49 Rapp50

d d

Kerse41

Kovács42

Lapane43

Law44 McMurdo45 Meyer46 Neyens47

Ray51 Sakamoto52 Saravanakumar53

Schnelle54 55

Schoenfelder Sitjà-Rabert56

Teresi57 Tuunainen58 van Gaal59 Varela60 Walker61

d Severe cognitive impairment, inability to understand and follow simple instructions Inability to follow a simple 1-step instruction MMSE score <20 Cognitive disorders that could interfere with intervention d

A score 24 on MEC, dementia d

d

15 (10)

13 (8)

d

d

d 15.3 (8.4) 25.7 (2.5) 27 (2.4)

d 14.6 (8.1) 25.2 (2.7)

12 (8)

14 (7)

CI CI mild: 56 (34.8%) mild: 67 (38.7%) moderate to moderate to severe: 105 (65.2%) severe: 106 (61.3%) Dementia (20.6) Dementia (20.6%) AD (13.2%) AD (13.2%) d d >40% reported symptoms of cognitive impairment CI: 48.4% CI: 49.4% <24 MMSE ¼ 82.2% <24 MMSE ¼ 81.9 % d d

3 2

0 0

2 3 1

0 0 1

d

d

3

0

25.79 d

26.71 d

d CI: 40 (49.4%)

d CI: 33 (42.3%)

1 2

1 0

19 (R ¼ 11-29) 19 (R ¼ 2-29) d d

21 (R ¼ 2-29)

CDDS: 8.55 (4.62) d

CDDS: 7.15 (4.51) d

0 0

0 0

d d

d MEC: 28.58 (2.5)

d MEC: 28.54 (2.8)

0 1

0 1

d

d

d

d

0

0

(continued on next page)

12.e4

H.J. Gulka et al. / JAMDA xxx (2019) 1e12

Supplementary Table 3 (continued ) Study (First Author)

Ward62 Whitney63 Wylie64 Yokoi65

Cognition-Based Exclusion Criteria

d d Inability to provide informed consent Dementia

CI Prevalence Score

Dementia Prevalence Score

DSC: 22% ACE-R: 27.2 (27) d

0 3 2

1 3 0

d

1

1

MMSE Scores, Mean (SD)

Other Measure of Cognition Scores, n (% of Residents) or Mean (SD)

Intervention

Control

Intervention

Control

d d 21.2 (5.4)

d d 21.0 (5.3)

DSC: 20% ACE-R: 29.1 (26) d

25.2 (2.0)

25.4 (2.6)

d

ACE-R, Addenbrooke’s Cognitive Examination; AD, Alzheimer’s disease; AMT, Abbreviated Mental Test; CDDS, CARE Dementia Diagnostic Scale; CDR, Clinical Dementia Rating; CI, cognitive impairment; DGDS, Dementia Global Deterioration Scale; DSC, dementia-specific care; HDS-R, Hasegawa Dementia ScaleeRevised; MEC, Mini-Examen Cognoscitivo (Spanish adapted version of MMSE); MMSE, Mini-Mental State Examination; R, range; SD, standard deviation.

Supplementary Table 4 Balancing Measures Study (First Author)

Balancing Type

Measure

Description

Dyer Fu35 Galik36

Functional Functional Activity, functional

Physical Physical Activity, physical, ADL

Hewitt38 Imaoka39 Kerse41 Kovács42 Lapane43 McMurdo45 Meyer46 Potter49 Ray51 Saravanakumar53 Schnelle54 Schoenfelder55 Sitjà-Rabert56 Tuunainen58 Varela60 Walker61 Whitney63 Wylie64 Yokoi65

Functional Functional Functional Functional Functional Functional Restraints Functional Restraints, functional Functional Activity, functional Activity, functional Functional Functional Functional Functional Functional Functional Functional

Physical, ADL Physical, ADL Physical, ADL Physical, ADL Physical Physical Restraints ADL Restraints, ADL Physical Activity, physical Activity, physical Physical Physical Physical, ADL ADL Physical, ADL Physical, ADL Physical

TUG, unsupported stand, 180 turn, POMA PPA PA: Actigraph, kcals, PAS-LTC; F: BI, POMA SPPB, UAB-LSA, SF-36 Grip strength, functional independence TUG, EMS, FICSIT-4, LLFDI POMA, TUG, Katz index Physical function from MDS Functional reach, reaction time, TUG, grip strength, spinal flexion Restraints MBI Restraints, life space, functional status BBS Physical activity level, upper extremity strength, mobility endurance How often residents participate in exercise, ankle strength, 6-m walk speed POMA, TUG, sit to stand Posturography TUG, Katz Index BI MBI, TUG, sit to stand, balance BBS, TUG, BI, ankle strength Sit to stand, TUG, 5-m walk, sit and reach, one-leg balance, bar gripped reaction test, functional reach

34

ADL, Activities of Daily Living scale; BBS, Berg Balance Scale; BI, Barthel Index; EMS, Elderly Mobility Scale; F, functional; FICSIT, Frailty and Injuries: Cooperative Studies of Intervention Techniques; FIM, Functional Independence Measure; LLFDI, Late-Life Function and Disability Instrument; MBI, Modified Barthel Index; MDS, Minimum Data Set; PA, physical activity; PAS-LTC, Physical Activity Survey in Long-term Care; POMA, Performance Oriented Mobility Assessment; PPA, Physiological Profile Assessment; SF-(12/ 36), Short Forme12/36 Health Survey; SPPB, Short Physical Performance Battery; TUG, Timed-up-and-go; UAB-LSA, University of Alabama BirminghameLife Space Assessment.