<p>Physical activity programs for older people in the community receiving home care services: systematic review and meta-analysis</p>

(1)

R E V I E W

Physical activity programs for older people in the

community receiving home care services:

systematic review and meta-analysis

This article was published in the following Dove Press journal: Clinical Interventions in Aging

Elissa Burton1

Kaela Farrier1

Rose Galvin2

Shanthi Johnson3

N Frances Horgan4

Austin Warters5

Keith D Hill1

1_{School of Physiotherapy and Exercise} Science, Curtin University, Perth, Western Australia, Australia;2_{School of} Allied Health, Faculty of Education and Health Sciences, Health Research Institute, University of Limerick, Limerick, Ireland;3_{School of Public} Health, University of Alberta, Edmonton, AB, Canada;4_{School of Physiotherapy,} Royal College of Surgeons in Ireland, Dublin, Ireland;5_{Health Service} Executive, Dublin North City and County Healthcare Organisation, Dublin, Ireland

Abstract:The proportion of older adults is increasing around the world and most wish to live in their home until they die. To achieve this, many will require services in the home to remain living independently. To maintain function (ie, strength, balance, and endurance), physical activity needs to be undertaken on a regular basis, and is essential as a person ages. Unfortunately, as people age there is a tendency to reduce activity levels, which often leads to loss of function and frailty, and the need for home care services. This updated systematic review includes a mix of study methodologies and meta-analysis, and investigated the effectiveness of physical activity/exercise interventions for older adults receiving home care services. Eighteen studies including ten randomized controlled trials meeting the

selec-tion criteria were identiﬁed. Many of the studies were multi-factorial interventions with the

majority reporting aims beyond solely trying to improve the physical function of home care clients. The meta-analysis showed limited evidence for effectiveness of physical activity for older adults receiving home care services. Future exercise/physical activity studies working with home care populations should consider focusing solely on physical improvements, and need to include a process evaluation of the intervention to gain a better understanding of the

association between adherence to the exercise program and other factors inﬂuencing

effectiveness.

Keywords:exercise, physical function, community care, reablement, seniors

Introduction

Populations throughout the world are increasing in age, with greater proportions of

older people than previously observed.1Most older people have a desire to live in

their home until they die2and to do this, they need to be able to complete activities

of daily living (ADLs) such as showering, dressing, eating, and toileting

indepen-dently. For some older adults these tasks become increasingly difﬁcult due to health

issues, frailty or disability, and they require assistance to continue living in their

home.3This assistance is often provided by home or community care services and

is predominantly funded by the government in high income countries such as

Australia, Canada, and a number of European countries.4–6

Home care services can be delivered over the short or longer term.7Some

short-term reablement services (ie, goal-oriented, person-centered, often aimed at reduced long-term services) include physical activity or exercise programs, whereas the longer term services such as personal care (ie, showering), domestic assistance (ie, cleaning), gardening, transport, and social support usually do not. In order to assist

Correspondence: Elissa Burton School of Physiotherapy and Exercise Science, Curtin University, GPO Box U1987, Perth, Western Australia 6845, Australia

Tel +6 189 266 4926 Fax +6 189 266 3699 Email [email protected]

Clinical Interventions in Aging

Dove

press

open access to scientiﬁc and medical research

Open Access Full Text Article

Clinical Interventions in Aging downloaded from https://www.dovepress.com/ by 118.70.13.36 on 20-Aug-2020

(2)

an older person to continue living in their home and complete their ADLs, they need to maintain strength,

balance, and endurance.8,9

Improving or maintaining strength, balance, and endur-ance requires a desire to be active, the knowledge of what

to do, and the opportunity to be active.10,11 The World

Health Organization and many individual countries (ie, United States, Canada, Great Britain, Australia) have pro-duced National Physical Activity Guidelines for older

adults.12–15 They predominantly recommend 30

minutes of moderate intensity endurance physical activity every day such as walking or swimming, in addition to

strength and balance training twice a week.12–15 The

strength and balance components are essential for

main-taining independence and completing ADLs.16 Strength

and balance training also leads to many health beneﬁts

such as increased strength and bone density, reduction in

sarcopenia, frailty, and chronic illness9,16–18 and has the

strongest evidence for preventing falls for older adults

living in the community.19 Unfortunately, these exercise

modalities are often overlooked when promoting physical activity recommendations to older adults, particularly

those receiving home care services.20

Many older home care clients walk to be active,

how-ever few participate in strength and balance programs.21It

can be difﬁcult for home care clients to leave their home,

and typically, shopping and medical appointments are prioritized, particularly when assistance with transport is required. To support older adults receiving home care services to, at a minimum, maintain their strength and balance, and at best, improve it over time, promoting

physical activity within the home is required.

Organizations delivering home care services have the ideal opportunity to do this, as many home care workers deliver services from once a week to multiple times each

week. A previous systematic review22evaluated the

effec-tiveness of physical activity (exercise) programs for home care clients, however, at the time only eight articles were

identiﬁed. The evidence (ie, outcomes and assessments)

varied widely and was limited, as such, a meta-analysis could not be undertaken. Therefore, as the focus on older adults receiving home care services continues to increase, it is essential to identify whether further studies have been undertaken. This will help guide home and community care organizations to provide evidence-based care for their older clients and assist policy-makers to understand

the beneﬁts of improved physical function for this older

population whom they heavily fund (eg, Irish government

spends €408 million on home support each year which

equates to 17 million visits).6 This current systematic

review looks at the recent evidence and where possible combines the randomized controlled trials (RCTs) (n=2) from the previous systematic review with more recently published studies, to determine the effectiveness of

physi-cal activity/exercise programs delivered speciﬁcally to

older adults receiving home care services.

Material and methods

Eligibility criteria

The review is limited to studies that met the following eligibility criteria:

● population: aged 65 years and older (at least 50% of

sample) and receiving home care services during the intervention. Home care services may include (but not limited to) personal care (showering), domestic assistance, transport, shopping or social care. These services are delivered in an older persons home by home or community care organizations (ie, munici-palities etc) and can be short-term or long-term ongoing services.

● Intervention: has to have a physical activity or

exer-cise aspect to the intervention.

● Comparison: for RCTs the comparison group will be

usual care (ie, which may include a current physical activity or exercise program used) or a non-active control group.

● Outcomes: the physical activity or exercise

interven-tion has to have been assessed using at least one physical performance outcome (ie, mobility, endur-ance, strength, balance).

● Setting: community-dwelling people only.

Any quantitative study could be included, eg, pilot or feasibility, pre- and post-test, retrospective, controlled trials or RCTs. Only peer-reviewed journal articles in English were included. No unpublished data (ie, reports), books, conference proceedings, theses, or poster abstracts were included. Exclusion criteria were: those living in residential care, and samples with >50% with a diagnosis of dementia or neurodegenerative disorder because they often require greater supervision and/or carer input to complete the physical activity/exercise intervention and therefore may not be representative of a typical home care service.

(3)

Information sources

Five databases were searched between October 2012 and August 2018: Medline (Proquest); CINAHL; PubMed; PsycInfo; and SportDiscus. The previous systematic review

search included articles between January 1982 to

September 2012. Reference lists from these papers were also scanned.

Search strategy

The search included the same mix of keywords as used in the previous systematic review. However, this updated

review also included “reablement” as a search term. An

example of the search strategy is presented in Table 1.

Study selection

Study selection was conducted in three stages. Stage one was the initial screening of the titles and scanning the abstracts against the eligibility criteria to identify poten-tial articles (completed by KF). Stage two included screening the full articles by two authors (EB and KF) to identify whether they met the eligibility criteria. In stage three, disagreements were resolved through discus-sion between EB and KF to achieve consensus, after referring to the eligibility criteria and protocol. The PRISMA checklist was used to ensure the results were

reported systematically.23

Data collection process

Each study in the review was evaluated using

a standardized extraction form, which included study design; purpose; intervention; study characteristics includ-ing sample size, sex proportion, participant age,

interven-tion type; and length of follow-up (Tables 2 and3).

Study quality

Study quality for the RCTs was assessed by two

indepen-dent researchers (KF and RG) using the Cochrane’s Risk

of Bias Tool.24 Categories assessed by the tool include

sequence generation, allocation concealment, blinding of participants, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other

sources of bias.24 RCT studies were assessed as having

“low”,“medium”or“high”risk of bias. Discussions were

undertaken between the reviewers and where required, a third researcher (EB), to form consensus.

Study quality was reviewed by two researchers (KF, EB) for all other included papers, using the National Institute for Health and Care Excellence (NICE)

guide-lines appraisal checklist.25The internal and external

valid-ity of the included papers were assessed by addressing key aspects of study design such as participant characteristics, allocation and intervention details, outcomes assessed, and

methods of analyses.25Each study was awarded an overall

study quality grading (ie, ++ all or most of the checklist

criteria have been satisﬁed, where they have not, it was

deemed unlikely to alter the conclusions; + some of the

checklist criteria have been satisﬁed, where they have not,

it is unlikely these will alter the conclusions; and–few or

no checklist criteria have been satisﬁed and are likely or

very likely to alter the conclusions).25

Data analysis

Physical performance outcome measures from the RCTs

included in this current review and the previous review22

were collated to determine whether there were two or

more studies that utilized the same outcome measures.26

For those studies that did, a meta-analysis was undertaken. Authors were contacted if data were not available within the published article. Four performance outcomes were utilized in the meta-analyses and all were measured using continuous data: Timed Up and Go (TUG),

sit-to-stand ﬁve times, grip strength, and walking speed. The

mean difference (MD) and 95% CIs were calculated. Review Manager (RevMan) version 5.3 was used to con-duct the analyses and generate forest plots with a random

Table 1Search strategy

1 Community care

2 Community health care

3 Home care

4 Community nursing

5 Home and community care

6 Home support

7 Community rehabilitations

8 Restorative care

9 Reablement

10 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9

11 Physical activity

12 Exercis*

13 11 or 12

14 Ageing

15 Aging

16 Aged

17 Older

18 Elderly

19 14 or 15 or 16 or 17 or 18

20 10 and 13 and 19

(4)

T able 2 Summar y of included randomized contr olled trial (RCT) studies Refer ence (coun-tr y) study design Study purpose Sample siz e; % female; a g e (y ears) (SD) comparator P opulation Outcomes and measur e-ment; follo w-up Inter v ention effect Bur ton et al, 28 (Australia) Pragmatic randomized con-tr olled trial Determine whether LiFE w ould be undertak en mor e often and result in gr eater functional gains than the current structur ed ex er -cise pr ogram deliv er ed as part of a restorativ e home car e ser vice LiFE =40; 75%; 80.2 years (6.4) Structur ed ex er cise =40; 90%; 79.6 years (6.2) Ye s Living in P erth suburbs (W estern Australia) re ferr ed for a restorativ e home car e ser vice Functional re ach, chair sit-to-stand (1 and 5), TUG, T andem walk, FES, ABC Scale, Vitality Plus Scale, The Late Life Function and Disability instruments 8 w eeks LiFE gr oup signi fi cantly impr ove d in 95% (19 of 20) of outcome measur es, compared with the structur ed ex er cise gr oup , which signi fi cantly impr ov ed in 70% (14 of 20). Bur ton et al, 29 (Australia) Pragmatic randomized con-tr olled trial Examine ov er the longer term, the effectiv eness and maintenance of a (modi fi ed) LiFE compared to the curr ent, structur ed ex er cise pr ogram used in a restorativ e home care ser vice LiFE =40; 75%; 80.2 years (6.4) Structur ed ex er cise =40; 90%; 79.6 years (6.2) Ye s Living in P erth suburbs (W estern Australia) re ferr ed for a restorativ e home car e ser vice Functional re ach, chair sit-to-stand (1 and 5), TUG, T andem walk, FES, ABC Scale, Vitality Plus Scale, The Late Life Function and Disability instruments, 6 month fall histor y 8 w eeks and 6 months A signi fi cant main effect of time was found for all measur es. LiFE show ed signi fi cantly gr eater impr ov ement than the structur ed ex er cise gr oup for tandem walk, ABC score, and Vitality Plus scale Danilovich et al, 30 (USA) Randomized contr olled trial T o test the effects of SFL ex er cise pr ogram on the ph ysical perfor -mance and self-rated health of older adults enr olled in the Community Car e Pr ogram, a Medicaid waiv er pr ogram 42; 83%; 74.8 years Inter vention =24; 83%; 74.1 years Contr ol =18; 83%; 75.6 ye ars Ye s Community-dwelling older adults receiving HC A ser vices thr ough the Community Car e Systems, Inc. in Illinois Grip and quadriceps str ength using a hand-held dynamometer , TUG, 10 ft gait speed, Self-Ef fi cacy for Ex er cise, pain, and PR OMIS-global health 12 w eeks Effect sizes w er e moderate for grip str ength, pain, and PR OMIS-global health. SFL participants sig-ni fi cantly impr ov ed median quad-riceps and TUG (median only) scor es compared to contr ols. No adv erse health ev ents and high pr ogram satisfaction w er e reported. Frailty pr e valence in the contr ol gr oup incr eased wher eas it decr eased in the SFL gr oup . Danilovich et al, 31 (USA) Randomized contr olled trial Determine the feasibility of train-ing HC As to lead the SFL pr o-gram with their Community Car e Pr ogram clients 42; 83%; 74.8 years (8.8) Inter vention =24; 83%; 74.1 years (9.9) Contr ol =18; 83%; 75.6 ye ars (8.1) Ye s Community-dwelling older adults receiving HC A ser vices thr ough the Community Car e Systems, Inc. in Illinois Client and HC A pr ogram satis-faction rates, HC A SFL training e valuation information and SFL fi delity rates 12 w eeks It is feasible for HC As to imple-ment SFL safely with clients. Participants vie w ed SFL as highly satisfactor y and HC As w er e able to adapt SFL for their clients. HC As ha ve high job satisfaction, and leading SFL enhances w ork achie ve ment and pride ( Continued )

(5)

T able 2 (Continued). Refer ence (coun-tr y) study design Study purpose Sample siz e; % female; a g e (y ears) (SD) comparator P opulation Outcomes and measur e-ment; follo w-up Inter v ention effect King et al, 32 (Ne w Zealand) Cluster ran-domized contr olled trial Evaluate the impact of a restorativ e home car e ser vice, compared with usual care for community-dwelling older people 186; 74%; 79 years Restorative =93; 69.9%; 78.4 years (6.5) Contr ol =93; 77.4%; 80.5 ye ars (6.3) Ye s Older people receiving assistance fr om a home care agency in Ne w Zealand SF-36, Nottingham Extended ADL, TUG, Master y scale , DSSI 4 and 7 months Signi fi cant bene fi t in health-related quality of life (SF-36) at 7 months compared to usual car e, particularly in the mental health component. A signi fi cantly higher number of older people in the inter vention gr oup w er e identi-fi ed for reduced paid care giver support or discharge Parsons et al, 33 (Ne w Zealand) Cluster ran-domized contr olled trial T o determine the impact of a re storative model of home care on social support and ph ysical function among community-dw elling older people Inter vention =108; 71.3%; 79.1 years (6.9) Usual car e =97; 60.8%; 76.9 years (7.6) Ye s Community-dwelling people older than 65 ye ars (55 years if Maori or Paci fi c Islander) w er e eligible if the y w er e ne w referrals to the car e coor dination agency for home care in Ne w Zealand SPPB, DSSI 6 months Inter vention gr oup had a gr eater mean incr ease in ph ysical function ov er time than the usual car e gr oup as determined by ov erall SPPB Renehan et al, 36 (Australia) Randomized con-tr olled trial Determine feasibility and accept-ability of a community-deliv er ed post-hospital falls pr e ve ntion pr o-gram, incorporating ex er cise, medication re vie w , and education 13; 53.8%; 82.2 years (9.7) Inter vention =7; 57.1%; 84.0 years (6.4) Contr ol =6; 50% 75.8 ye ars (11.6) Ye s Recently discharged fr om a hospital within metr opolitan Melbourne, Australia after being admitted for a fall and a ne w or existing client of the community home care organization Quality of life (A QoL-8D), falls, fall risk (FR OP-Com), FES-I, Fried Frailty Criteria, Katz ADL, quali-tative falls inter vie w . Inter vention only: fi ve times sit-to-stand, 4 test balance scale (feet together , semi-tandem, tandem and one leg stand), TUG 6 months The ex er cise component of the multifactorial inter ve ntion was highly valued by inter vention par -ticipants. Five falls during inter -vention period: inter vention gr oup: 3; contr ol gr oup: 2. No signi fi cant changes in ph ysical outcome measur es. Saeterbakk en et al, 37 (Norwa y) Randomized contr olled trial Determine the effects of a 10-w eek home-based str ength train-ing pr ogram on ph ysical function tests, le vel of ph ysical activity , muscle str ength, and RFD among frail old adults receiving home car e ser vices 23; 73.9% Inter vention =11; 86.5 ye ars (6.4) Contr ol =12; 83.5 years (5.7) Ye s Older adults living at home and in need of home care due to func-tional disabilities and/or medication Leg extension and elbow fl exion maximal str ength and RFD , pr e-ferr ed/max walking speed, stair climb, 5x chair rise and ph ysical activity (acceler ometer) 10 w eeks Inter vention gr oup signi fi cantly impr ov ed peak RFD in leg extension ( Continued )

(6)

T

able

2

(Continued).

Refer

ence

(coun-tr

y)

study

design

Study

purpose

Sample

siz

e;

%

female;

a

g

e

(y

ears)

(SD)

comparator

P

opulation

Outcomes

and

measur

e-ment; follo

w-up

Inter

v

ention

effect

Ste

vens-Lapsley

et

al,

34

(USA)

Randomized

contr

olled

trial

T

o

determine

whether

a

pr

ogre

ssiv

e

multicomponent

ph

ysical

therap

y

inter

vention

in

the

home

setting

can

impr

ove

functional

mobility

for

decondi-tioned

older

adults

following

acute

hospitalization

compared

to

usual

car

e

rehabilitation

22;

63.6%;

85.4

years

(7.8)

Inter

vention

=12;

70%;

83

ye

ars

(6.8)

Usual

car

e

=9;

58.3%;

87

ye

ars

(8.8)

Ye

s

Older

adults

with

multiple

chr

onic

conditions,

who

w

er

e

discharged

fr

om

the

Acute

Car

e

for

the

Elderly

unit

of

the

Univ

ersity

of

Colorado

Hospital,

with

ph

ysician

re

ferrals

for

home

health

ph

ysical

therap

y

4

m

walking

speed,

modi

ﬁ

ed

Ph

ysical

P

erformance

Test,

SPPB,

six

minute

walk

test

30

and

60

da

ys

Inter

vention

gr

oup

had

signi

ﬁ

-cantly

gr

eater

impr

ov

ements

in

walking

speed,

modi

ﬁ

ed

ph

ysical

performance

test,

and

SPPB

scor

es

compared

with

the

usual

car

e

gr

oup

at

60

da

ys

T

untland

et

al,

35

(Norwa

y)

Randomized

contr

olled

trial

In

ve

stigate

the

effectiv

eness

of

reablement

(multicomponent

home-based

re

habilitation)

in

older

adults

compared

with

usual

car

e

in

relation

to

daily

activities,

ph

ysical

functioning,

and

health-related

quality

of

life

Inter

vention

=31;

71%;

79.9

years

(10.4)

Contr

ol

=30;

63.3%;

78.1

ye

ars

(9.8)

Ye

s

P

eople

applying

for

,

or

referr

ed

to

,

home-based

ser

vices

due

to

functional

decline

in

one

or

mor

e

ADL

in

a

rural

municipality

in

Norwa

y

COPM

was

used

to

measur

e

self-per

ceiv

ed

activity

performance

and

satisfaction

with

perfor

-mance.

Health-r

elated

quality

of

life

(COOP/W

onka),

TUG,

grip

str

ength

3

and

9

months

Signi

ﬁ

cant

impr

ov

ements

in

mean

scor

es

fa

voring

reablement

in

COPM

performance

at

3

and

9

months

and

ov

erall

tr

eatment,

and

for

COPM

satisfaction

at

9

months

and

ov

erall

tr

eatment

Abbre

viations:

LiFE,

Lifestyle-integrated

Functional

Ex

er

cise

pr

ogram;

TUG,

Timed

Up

and

Go;

FES,

Falls

self-Ef

ﬁ

cacy

Scale;

ABC

,

Activities

speci

ﬁ

c-Balance

Con

ﬁ

dence

Scale;

SFL,

Str

ong

For

Life;

HC

A,

home

car

e

aide;

ADL,

activities

of

daily

living;

DSSI,

Duk

e

Social

Support

Index;

SPPB,

Short

P

erformance

Ph

ysical

Batter

y;

FR

OP-Com,

Falls

Risk

for

Older

P

eople

–

Community

Setting;

FES-I,

Falls

self-Ef

ﬁ

cacy

Scale

–

International;

RFD

,

Rate

of

Fo

rce

Dev

elopment;

COPM,

Canadian

Occupational

P

erformance

Measur

e.

(7)

T able 3 Summar y of inter vention implementation for randomized contr olled trial (RCT) studies Refer ence Implementation strateg y Inter ve ntion Bur ton et al, 28 , 29 L iF E an d str uc tu re d ex e rc is e p rog ra m w e re de liv e re d b y ca re ma na ge rs as p art of a res tor at iv e h om e ca re se rvi ce . C lie n ts w e re vis ite d ev ery 10 – 1 4 da ys b y th e ir ca re manager (a verage 3 vi si ts) to offer suppo rt , e nco u ragem e nt an d to pr o gr ess e x e rc is es. E x e rc is e w as ju st on e as p e ct of th ei r se rvi ce tha t w as d is cu ss e d d ur in g th es e vi sit s LiFE inter vention =7 balance and 6 low er limb str ength ex er cises incorporated into ev er yda y activities. Ex er cises w er e explained by car e manager and a descriptiv e manual was giv en to clients. Contr ol gr oup: structure d ex er cises =3x per da y 5 re petitions of 6 str ength and 2 balance ex er cises (15 – 20 minutes). Ex er cises w er e pr ovided on a sheet illustrating ex er cises. Danilovich et al, 30 , 31 HC As attended a half-da y SFL training session. HC As completed SFL with their clients 2x per w eek for the fi rst 2 w eeks of the pr ogram with clients completing a thir d session independently . After the initial 2-w eek period, HC As pr ovided their clients motivation at each usual care visit (2 – 3x per w eek) to perform SFL 3 times a w eek for an additional 10 w eeks. 12 w eeks SFL; 35-minute D VD that includes a warm up , 11 upper and low er extr emity ex er cises using therabands for re sistance , and a cool down. King et al, 32 P ai d ca re giv e rs u nd e rto ok 2 tr ain in g pr o gra m s b as e d o n a re sto ra tiv e ca re p hi lo so ph y d e sig n e d to o p ti mi ze ind e p en d e nc e in o ld e r p eo p le , ta ki n g ap p ro x ima te ly 5 and 1 8 w e ek s. Co or d in at or (ex p e ri e n ce d re gi ste re d n u rs e ) u n de rt oo k an in it ia l in -de p th as se ss m en t wi th th e p art ic ip an t u si n g T A R G ET , le ad in g to a su p p ort pl an fo r p ai d ca re gi ve rs to im p le m e n t at vis it s w ith p art ic ip an ts , ra n gin g fr om da ily to fo rtn ig ht ly . C oo rdi n ato rs pr ov id e d e n h an ce d su pe rvi sio n for p aid ca reg iv e rs by u n de rta ki n g co m -p u ls ory 2 -h ou r fort n ig h tly me e ti n gs an d al so pr ov id e d fo llow -u p co n su lt ati on s w it h p art ic ip ant (ph o n e ca ll o r vis it) e ve ry 3 m on th s, an d a fu ll re -a ss e ss m e nt at 1 2 mo n th s Repetitive ADL ex er cises designed to optimize independence and to further support par ticipant in re aching their goals Parsons et al, 33 Inter vention gr oup completed a goal facilitation tool (T ARGET) with a needs asses-sor to determine their needs and to establish the aims for the episode of care. Ser vices w er e structur ed accor ding to the principles of re storative home care (independence focused with individually tailore d activity pr ograms). HC As then completed tasks outlined in the support plan for each individual. Ho me ca re se rv ice s in clu de d d om e sti c ta sks ,pe rs on al ca re, sh op p in g, an d in d iv id ua lize d act iv it ie s (w alk in g or e x e rc is e p ro gr ams /o th e r act iv it ie s ai me d at im p ro vi n g fu n ct io n al ab ilit y) . In the in te rv e n tio n gr ou p 6 1 .7 % (n = 6 6) o f th e su p po rt p la n s de sc ri b ed su ch in di vi du al iz e d ac ti vit ie s as op p o se d to 1 5 % (n = 1 4 ) in th e u su al ca re gr o u p . Renehan et al, 36 Participants w er e appr oached 1 – 2 w eeks after discharge, inter vention included a medication re vie w , falls pr e vention education, and home-based ex er cise individually pr escribed and modi fi ed accor ding to pr e-existing frailty by ex er cise ph ysiologist (visits at w eeks 1, 2, 4, and 8 for ex er cise pr ogr ession), participants w er e pr ovided diagrams of ex er cises and instruction on ex er cise safety . Participant ’ s primar y nurse who regularly visited for other matters pr ovided encouragement. Thre e to fi ve times per w eek 20 – 30 minutes graduated low er limb str engthening and balance ex er cises (Otago Ex er cise Pr ogramme) Saeterbakk en et al, 37 A pr ofessional training instructor was pr esent at ev er y training session to mak e sure of the corr ect technique, intensity , and numbers of sets. S u per vi sed pr o gr essi ve st re ngth tr aini n g pr ogram m e 60 m inut e s 2x w e ek fo r 1 0 w eeks, 10 – 12 re p e ti ti o n s 2 – 3 set s squats, b o x lift s, seat e d ro w , chest p re ss, an d b ic ep s cur l. ( Continued )

(8)

T able 3 (Continued). Refer ence Implementation strateg y Inter ve ntion Ste vens-Lapsley et al, 34 Patients placed in the p rogr essiv e mu lticomp onent inter vention gr oup w er e referr e d to h o me health agency an d w er e see n by a sin gle ph ysical therapist who w as trained to p rovide th e pr ogr e ssiv e inter ve ntion. Tw o to thr ee ph ysical therap y visits p er w e ek for 3 0 da ys. Pr og re ss iv e mu lti co mp o ne n t p hys ica l th e ra p y in te rv e nt io n co n sis te d o f A DL tr ain ing (b at h ro om tr an sf e rs , b e d m ob ilit y, an d ca r tra n sf er s) , e vi de n ce -b as e d m ob ili ty tr ain ing (i n do or w alk in g, ga it tra in in g on fl at gr ou n d an d st air s, an d o ut do or w alk in g tr an si ti o n in g to an in de p e n de n t d ail y w al k in g p ro gr am) an d pr og res siv e str e n gt h tr ain in g (3 se ts 8 rep e tit ion s m ax su p in e le g p re ss , sta n di n g h ip e x te n sio n , b o dy -w e ig h t re si ste d pl an ta r fl exio n , se at e d pr ess and seat e d ro w u si n g the Shut tl e M in i P ress). P ar ti ci p ants w er e giv e n a h om e e x e rci se pr og ra m to co n ti n ue b as ic e x e rc is e s af te r di sch ar ge. T untland et al, 35 O cc u p ati on al an d p h ys ic al th e ra p is ts u se d C O P M to id en tif y ac ti vi ty lim ita tio ns p e r-ce iv e d as im p ort an t b y th e pa rti ci pa nt an d d e ve lo p e d a re h ab ilit ati on p la n w ith a m ax im u m re h ab ilit ati on pe rio d of 3 m o nt h s. T h e ra p is ts su p e rv is e d H C A w h os e fo cu s w as o n sti mu la ti n g th e p arti ci pa nt s to pe rf or m th e d ail y ac ti vi ti es th e ms el ve s, ra th er th an le tt in g o th e rs d o it fo r th e m, ph ys io th e ra p is t an d /o r o cc u p ati on al th e ra p is t vi sit ed p art ic ip ant s for a mi n im u m o f 1 h ou r e ac h w e ek . A ll H CAs att ain e d tra in in g b e fo re the in te rv en tio n wa s im p le m en ted . T h e ra p is ts h ad w ee k ly in fo rma l lu n ch m e e ti n gs w ith H CAs in o rde r to e ns ur e go od co m mu n ic at io n and fol lo w -u p o f in d iv id u al p art ic ip an ts Individualized re ablement pr ogram including training in daily activities (dr essing, food pr eparation, vacuuming, bus transport), adaptations to the en vir onment or the activity to simplify activity performance, and ex er cise pr ograms such as indoor or outdoor walking with or without walking aids, climbing stairs, transferring, and performing ex er cises to impr ove str ength, balance or fi ne motor skills. The ex er cises w er e incorporated into daily routines and participants w er e give n a manual explaining each of the ex er cises and encouraged to train on their own Abbre viations: LiFE, Lifestyle-integrated Functional Ex er cise pr ogram; SFL, Str ong Fo r Life; HC A, home car e aide; T ARGET , Towar ds Achie ving Realistic Goals in El ders Tool; ADL, activities of daily living; COPM, Canadian Occupational P erformance Measur e.

(9)

effects model applied and using the inverse variance

DerSimonian and Laird method.27 Heterogeneity was

assessed using the I2statistic and by visual inspection of

the forest plots. Statistical signiﬁcance was set atp<0.05

for all analyses.

Results

Study selection

Eight thousand seven hundred and ﬁfty eight papers were

found in theﬁve databases. After removing duplicates from

each database and then across databases, 1,679 articles remained. Articles were screened by title and 1,266 were

excluded. Reasons for exclusion are presented inFigure 1–

the study ﬂow chart. Article abstracts were then read and

a further 288 excluded, leaving 125 articles to be screened by full text. One hundred and seven studies were excluded and 18 studies were accepted to be included in the review.

Study characteristics

The 18 studies in the review included ten RCTs;28–37 six

single group pre- and post-test studies;38–43one feasibility

study;44and one retrospective study.45

There were 1,118 participants across the 18 studies,

ran-ging from eight in a pilot study39to 228 in the retrospective

study.45The average age of all participants was 80.4 (±3.3)

years and ranged from 74.5–85.4 years; 71.8% of the

parti-cipants were female. Seven studies were conducted in the

United States;30,31,34,38,41,42,45six in Australia;28,29,36,39,43,44

two in New Zealand32,33 and Norway;35,37 and one in

Hong Kong.40

Interventions

RCTs

The average age of the 656 participants in the ten RCTs

was 80.4 (±3.7) years, ranging from 74.5–85.4 years and

included an average of 69.6% females. A number of dif-ferent interventions were trialed. However, the majority

focused on strength and balance training.Table 2presents

a summary of the included RCT studies and Table 3

out-lines the interventions implemented.

The two Burton et al28,29 RCTs utilized the

Lifestyle-integrated Functional Exercise (LiFE) program which incor-porates seven balance and seven strength exercises into an

older person’s daily living activities.46,47 The intervention

was compared with a structured exercise program that focused on strength and balance also in an RCT over two

time periods; 8-week intervention28 and a 4-month

maintenance period.29 It must be noted the control group

was active and received the structured exercise program that was delivered within the usual restorative care service. Whereas, Danilovich et al created the Strong For Life (SFL) resistance training program, which included a 35-minute

DVD to assist home care aides to deliver the program.30,31

The exercise intervention in King et al’s study32 used

exercises that optimized independence, incorporating

repe-titive functional ADLs. Parsons et al33 used a similar

exercise intervention, again with the aim of optimizing independence and improving functional ability by incor-porating individualized activities, 62% of the support plans described included these activities with their participants compared to 15% for the control group (ie, usual care).

Stevens-Lapsley et al34utilized a progressive

multicom-ponent physical therapy intervention for 60 days to improve functional mobility for older adults who had just left

hospi-tal. Tuntland et al35trialed a reablement intervention, which

was individualized for each participant. The exercise pro-grams typically included indoor or outdoor walking, climb-ing stairs, transferrclimb-ing, engagclimb-ing in strength and balance

training, and improving ﬁne motor skills. Similar to the

LiFE program utilized by Burton et al,28 many of the

exercises were incorporated into ADLs. In a further study

by Saeterbakken et al,37 they conducted a 10-week

fully-supervised strength training intervention which was deliv-ered twice a week to participants.

Other studies

The other eight studies that were not RCTs included 462 participants, with an average age of 80.5 (±3.0) ranging

from 76.8–83.8 years and included an average of 74.8%

females. Table 4 presents a summary of the included

studies that are not RCTs and Table 5 the interventions

implemented. Bamgbase and Dearmon38reported

deliver-ing a 6-month individualized home exercise program with no other detail, and the main aim was preventing falls by using a multi-factorial intervention.

Burton et al39conducted a feasibility study utilizing the

LiFE program (previously described) to determine whether this intervention type could be delivered more broadly within a restorative care service. In a more recent study,

Burton et al44 conducted a feasibility study that explored

whether community care support workers who are non-allied health trained could deliver the LiFE exercise program to community care clients. It was determined that there were no adverse events and it was possible for this population to

deliver this falls prevention exercise program.44

(10)

The home care clients in Gallagher et al’s study45 received an individualized home exercise program which

was designed to address clients’ identiﬁed impairments

and falls hazards in the home. This study found that the Missouri Alliance for Home Care tool (MAHC-10) and interdisciplinary program was effective in identifying and

managing those who had fallen in their home.45

The 6-month intervention delivered by Kwok and

Tong40 was a multicomponent exercise program,

consisting of ﬂexibility, strength, balance, and aerobic

exercises, either delivered by a physiotherapist in

a center or by a care worker in the home. The physiother-apy-led center-based intervention was found to improve physical function, quality of life, and fall incidence, whereas the home-based care worker led program had no

effect on physical function and self-rated health.40

Muramatsu et al41 conducted a motivational

enhance-ment and three chair-bound moveenhance-ments program delivered

Literature search databases:

Duplicates removed within database databases:

All articles combined into one spread sheet and duplicates removed (N=1679)

Accepted

articles screened on the basis of title included (N=413)

Excluded (N=1266) reasons for exclusion: not community (N=88) cognitive impairment (N=146) not exercise intervention (N=496)

not home care (N=208) not older people (N=212) Protocol paper/systematic review (N=79)

therapist examined (N=37)

Excluded (N=288) reasons for exclusion:

not community (N=8) cognitive impairment (N=4) not exercise intervention (N=73)

not home care (N=129) not older people (N=60) Protocol paper/systematic review (N=5)

therapist examined (N=9)

Excluded (N=107) reasons for exclusion:

not community (N=3) cognitive impairment (N=5) no exercise intervention details (N=16)

no exercise outcomes (N=6) not home care (N=55) not older people (N=18) Protocol paper/systematic review (N=2)

therapist examined (N=1) no results (N=1) Accepted

articles screened on the basis of ful text (N=18) Accepted

articles screened on the basis of abstract included (N=125) MEDLINE (N=1306)

CINAHL (N=250) PubMed (N=2026) PsycINFO (N=4964) SPORTDiscus (N=212)

MEDLINE (N=488) CINAHL (N=190) PubMed (N=430) PsycINFO (N=989) SPORTDiscus (N=58) Total articles (N=8758) Total articles (N=2155)

Figure 1Flow chart of study selection.

(11)

T

able

4

Summar

y

of

included

non-randomized

contr

olled

trial

(RCT)

studies

Refer

ence

(coun-tr

y)

Study

design

Study

purpose

Sample

siz

e;

%

female;

a

g

e

(y

ears)

(SD)

comparator

P

opulation

Outcomes

and

measur

e-ment

follo

w-up

Inter

v

ention

effect

Bamgbade

and

Dearmon,

38

(USA)

Quasi-experimental

sin-gle

gr

oup

study

pr

e

and

post-test

Evaluate

the

impact

of

a

fall

assessment

tools

and

patient-speci

ﬁ

c

fall

pr

e

vention

pr

ograms

in

re

ducing

the

inci-dence

of

falls

and

fall-r

elated

inju-ries

among

older

patients

receiving

home

HHS

30;

sex

unknown;

aged

65

and

ov

er

No

Receiving

nursing

care

ser

vices

on

a

long-term

basis.

Suffer

var

ying

degr

ees

of

ph

ysical,

mobility

,

or

cognitiv

e

impair

-ments,

such

as

diabetes,

str

ok

e,

and

urinar

y

incontinence

Number

of

falls

and

falls

with

injuries

6

months

Reduction

of

falls

and

major

injuries.

Bur

ton

et

al,

39

(Australia)

Quasi-experimental

single

gr

oup

study

pr

e-and

post-test

(pilot)

Determine

whether

a

LiFE

was

suitable

for

deliv

er

y

in

a

restorativ

e

home

car

e

ser

vice

8;

75%;

80.8

years

(5.9)

No

Curr

ent

restorativ

e

home

car

e

clients,

living

in

P

erth

(W

estern

Australia)

Functional

re

ach,

chair

sit-to-stand

(1

and

5),

TUG,

T

andem

walk,

FES,

ABC

Scale,

Vitality

Plus

Scale,

The

Late

Life

Function

and

Disability

instruments,

PASE,

Acceler

ometer

,

LiFE

Assessment

tool, 8w

eeks

Dif

ﬁ

culties

using

the

acceler

-ometer

and

LiFE

Assessment

tool.

Signi

ﬁ

cant

impr

ove

ment

in

FES,

tandem

walk,

Late

Life

Function

total

scor

e,

and

PASE

Bur

ton

et

al,

44

(Australia)

F

easibility

study

To

in

vestigate

the

feasibility

of

community

care

w

ork

ers

deliv

er

-ing

a

falls

pr

e

vention

ex

er

cise

pr

ogram

to

older

clients,

at

low

or

medium

risk

of

falling,

as

part

of

an

existing

ser

vice

pr

ovision

29;

65.5%;

82.7

(8.7)

years No

Older

adults

receiving

at

least

one

community

car

e

ser

vice

fr

om

a

participating

community

care

organization

or

an

RAS

assessment,

living

in

P

erth,

W

estern

Australia

FR

OP-Com,

Ph

ysical

Activity

Readiness

Questionnair

e

(P

ARQ)

and

LiFE

assessment

tool

mea-sur

ed

at

baseline.

Client

and

staff

pr

ogram

satisfaction

measur

ed

post-test. 8w

eeks

LiFE

ex

er

cise

pr

ogram

was

safely

deliv

er

ed

by

RAS

assessors

and

support

w

ork

ers,

with

no

adve

rse

e

ve

nts

reported.

The

majority

of

clients

(82%)

repor

ted

enjo

ying

the

ex

er

cises,

59%

felt

it

made

a

positiv

e

change

to

their

health.

Ex

er

cise

adher

ence:

4.8

(SD:

2.2)

da

ys

per

w

eek.

Gallagher

et

al,

45

(USA)

Retr

ospectiv

e

Determine

concurrent

based

validity

of

the

MAHC-10

with

the

P

erformance

Oriented

Mobility

Scale

(Tinetti),

and

determine

the

bene

ﬁ

t

of

the

Balanced

Appr

oach

rehabilitation

pr

ogram.

228;

sex

unknown;

aged

65

and

ov

er

No

Patients

re

ferre

d

for

home

car

e

ser

vices

to

Catholic

Home

Car

e

MAHC-10

and

Tinetti

scor

es

6

months

Ther

e

was

w

eak

but

signi

ﬁ

cant

negativ

e

re

lationship

betw

een

the

MAHC-10

and

pr

e-Tinetti

scor

es.

Signi

ﬁ

cant

differe

nces

in

Tinetti

score

s

betw

een

pr

e-and

post-inter

vention

(

Continued

)

(12)

T able 4 (Continued). Refer ence (coun-tr y) Study design Study purpose Sample siz e; % female; a g e (y ears) (SD) comparator P opulation Outcomes and measur e-ment follo w-up Inter v ention effect Henw ood et al, 43 (Australia) Single gr oup study pr e-and post-test De velop an HC A-led ex er cise pr ogram for older adults with aged care needs and e valuate the bene fi ts fr om both the perspec-tiv e of the HC A and the person receiving the pr ogram 50; 74%; 82.1 years (6.1) No Older adults receiving gove rn-ment-funded domestic assistance or personal car e in Brisbane, Australia SPPB, falls, hours of sleep , visits to a general or clinical practice health pr ofessional, emergency depart-ment/hospital, ex er cise compli-ance, SF-36 18 w eeks Signi fi cantly impr ove d SPPB scor e, 19% re duction in par ticipants classi fi ed as frail and a re duction in health car e ser vice access of 47% acr oss the inter vention Kw ok et al, 40 (Hong K ong) Quasi-experimental pr e-and post-test Compar e the effects of an ex er -cise pr ogram super vised by a ph ysiotherapist at a da y training center with an ex er cise pr ogram deliv er ed by a care w ork er at home on ph ysical function, quality of life , and fall incidence in com-munity-dw elling older adults Center -based =24; 70.8%; 84.4 years (6.9) Home-based =26; 73.1%; 83.2 years (7.0) 5 center -based and 4 home-based participants had MMSE <12 Ye s Clients admitted to the Enhanced Home and Community Car e Ser vice team, re siding in the W ong T ai Sin District (Hong K ong). Suffering fr om chr onic diseases with moderate or se ver e le vel of impairment Elderly Mobility Scale, BBS, Numeric Pain Rating Scale, SF-12, and fall incidence 6 months Center -based participants had sig-ni fi cant impr ov ement in all out-comes and less falls compared to home-based participants Muramatsu et al, 41 (USA) Single gr oup study pr e-and post-test Test and enhance the feasibility of HM pr ogram deliv er ed by HC As for community-dwelling older adults in a Medicaid-funded home car e setting 54; 78%; 77 (63 – 101) years; almost 1/3 had mild cognitive impair -ment No Receiving in-home support ser -vices thr ough the Illinois Departm ent on Aging Community Car e Pr ogram, Chicago Daily activity dif fi culties and dependency (HM6, BADL, IADL), performance of 3 HM mov ements, SSPB, self-rated general health and pain, FES-I, falls, and ex er cise re lated social support 4 months Clients ‘ daily activity function and health outcomes impr ov ed signi fi -cantly . The pr ogram was w ell-re ceiv ed by clients, high re tention rates among clients (93%) Park and Chodzk o-Zajk o , 42 (USA) Single gr oup study pr e-and post-test Assess the feasibility of imple-menting simple, safe, non-equipment ev idence-based mov e-ments (HM pr ogram) using an affor dable and sustainable HC A based deliv er y model that re aches the maximum possible number of frail older adults living at home but who ar e at risk of nursing home admission 13; 84.6%; 76.8 years (11.4) No Older adults re ceiving home car e ser vices pr ovided by Help at Home and Family Ser vice; home visit pr ograms funded by Older American ’ s Act in the State of Illinois Evaluation sur ve ys and inter vie ws with clients, HC As, and site dir ectors. Functional ability of cli-ents: 30-seconds chair stand, 30-seconds arm curl (5 pounds for females, 8 pounds for males) and chair sit and re ach 4 months Clients signi fi cantly impr ov ed number of arm curls in 30 sec-onds. No signi fi cant impr ov ement in chair stand or chair sit and re ach. Most participants had a positiv e per ception and high satisfaction with the pr ogram. HM could be safely and successfully disseminated to frail older adults Abbre viations: MAHC-10, Missouri Alliance for Home Car e tool; LiFE, Lifestyle-integrated Functional Ex er cise pr ogram; FES, Falls self-Ef fi cacy Scale; ABC , Activities speci fi c-Balance Con fi dence Scale; TUG, Timed Up and GoP ASE, Ph ysical Activity Scale for the Elderly; RAS, Regional Assessment Ser vice; HC A, home care aide; SPPB, Short P erformance Ph ysical Batter y; FR OP-Com , Falls Risk for Older P eople – Community Setting; MMSE, Mini Mental State Examination; BBS, Berg Balance Scale; BADL, Basic Activities of Daily Living; IADL, Instrumental Activities of Daily Living; HM, Health y Moves ; FES -I, Falls self-Ef fi cacy Scale – International.

(13)

T able 5 Inter vention implementation of included non-randomized contr olled trial (RCT) studies Refer ence Implementation strategy Inter v ention Bamgbade and Dearmon, 38 Staff pr oviding nursing car e to patients of the privately owned home health car e agency w er e educated on the fall pr e vention pr ogram MAHC-10 fall risk assessment tool, medication re vie w pr ocess and management plan, home safety assessment, patient-and family-speci fi c education, individualized home-based ex er cise regimen Bur ton et al, 39 LiFE pr ogram was deliv er ed by car e managers as part of a restorativ e home car e ser vice. Car e managers explained to the client the differ ent ex er cises and how these ex er cises could be incorporated into their daily routines, manual also pr ovided. Clients w er e visited e ver y 10 – 14 da ys by their care manager (a ve rage 3 visits) to monitor how w ell the client was performing the fi rst ex er cises agr eed to and to encourage the client to start doing others. Ex er cise was just one aspect of their ser vice that was discussed during these visits 7 balance and 6 low er limb str ength ex er cises incorporated into e ver yda y activities Bur ton et al, 44 Staff completed a 4-hour training session on deliv ering the LiFE ex er cise pr ogram. Each staff member re ceiv ed 3 client folders, a staff folder and a LiFE trainer ’ s manual. Community car e w ork ers follow ed-up with their clients either during their usual ser vices or on a fortnightly basis 7 balance and 6 str ength ex er cises to complete whilst undertaking usual daily activities Gallagher et al, 45 Staff was trained in assessing risk factors and in fall pr e vention measur es. The training sessions took place ov er a 6-hour period and w er e follow ed-up by performance-based competencies and case studies. Patients w er e e valuated with MAHC-10, score s ≥ 4 of 10 (indication of high risk for falls) starte d on the Balanced Appr oach rehabilitation pr ogram. The CHC falls scre ening and Balanced Appr oach re habilitation pr ogram was designed to re duce the risk of falls in a home care population using an inter disciplinar y appr oach T reatment fr om ph ysical and occupational therapists aims to impr ov e gait, balance, balance con fi dence and reduce fall risk thr ough individualized home ex er cise pr ogram designed to addre ss impairments and home safety/mobility adaptations. The home pr ogram is in place for as long as the patient demonstrates a bene fi to r until safety/independence is achie ved (a verage 8 visits, 45 – 60 minutes each) Henw ood et al, 43 HC As attended a 1-da y w orkshop and receiv ed a printed manual containing detailed instructions for each ex er cise, information on health y lifestyle beha viors, and an ov er vie w of common barriers and enablers to ex er cise particip ation. HC As visited clients we ekly or fortnig htly for 60 – 90 minutes with at least 10 minutes assisting and monitoring client par ticipation in the ex er cise pr ogram 5 str ength ex er cises (chair stand, seated leg extension, standing calf raise, wall push-ups, theraband seated rows, pr ogressing to 3 sets of 12 repetitions), 3 balance ex er cises (single leg stand for up to 20 seconds on each side , tight rope walk of 10 steps forwar d and back, 10 re petitions side to side step). Clients w er e encouraged to ex er cise on most, pr eferably all da ys Kw ok et al, 40 Parti cipants chose to receiv e ex er cise training at either a da y training center under super vision of a ph ysiotherapist (gr oups of 6 – 10 participants) or individually at home assisted by a car e w ork er . Ph ysiotherapists pr ovided training to all care w ork ers and monitore d their performance e ver y 2 – 3 months thr ough on-site super vision to ensure quality of the training pr ogram at home 6 month fl exibility , str ength, balance, and aer obic ex er cise pr ogram with pain management, 1 – 2 sessions per w eek lasting appr o ximately 60 minutes ( Continued )

(14)

by home care aides for their clients over a 4-month

interven-tion. Results showed a signiﬁcant improvement in physical

ﬁtness, self-rated health, pain interference, Short Physical

Performance Battery (SPPB) total score, SPPB balance

tests, and fear of falling.41Park and Chodzko-Zajko42also

conducted a 4-month intervention called“Healthy moves for

aging well”which incorporated a physical activity

interven-tion with a lifestyle behavior change counseling method called Brief Negotiation. The activity portion of the

interven-tion was the same as that used by Muramatsu et al.41Arm

curl was the only functional ability outcome that reported

improvement between pre- and post-testing (p=0.021).42

Henwood et al’s43 Active@Home 18-week program

included weight bearing and balance exercises and was tailored to the level of the participant. The exercise pro-gram was delivered by home care support workers, with at least 10 minutes of their service time spent monitoring the

exercises.43Signiﬁcant improvement was found for

func-tional capacity as measured by the SPPB and a 19%

reduction in participants classiﬁed as frail.43

Outcome measures

RCTs

There was a variety of outcome measures utilized by the

included studies. The main physical test utilized byﬁve of

the RCTs was the TUG test.28–30,32,35 Unfortunately

Danilovich et al30 could only provide medians and

inter-quartile ranges rather than means and SDs for the TUG,

and only the intervention phase data of Burton et al’s28

study were included to provide similarity with the other studies included which assessed the post-intervention

per-iod. Four of the RCTs28,33,37,48 used the sit-to-stand ﬁve

times test and these data have been included in a meta-analysis.

Two RCTs33,34 utilized SPPB but unfortunately, after

many emails, the authors were unable to obtain the data

required for the Stevens-Lapsley et al34study. Therefore,

no meta-analysis could be conducted. The grip strength

test30,35 and gait speed/walking tests were used in two

RCTs each respectively.30,34 The data from the

Stevens-Lapsley et al34study were not available, however, Tinetti

et al’s study which was included in the previous systematic

review, measured walking speeds and was included in the meta-analysis.

Other studies

The SPPB test was used in two of the non-RCT studies38,43

and ﬁve of these studies also collected data on

T

able

5

(Continued).

Refer

ence

Implementation

strategy

Inter

v

ention

Muramatsu

et

al,

41

HC

As

receiv

ed

4

hours

of

training

on

how

to

deliv

er

the

HM

pr

ogram.

At

the

initial

visit

(a

verage

27

minutes,

range

10

–

45

minutes)

HC

As

assessed

re

adiness

for

PA,

had

their

clients

set

personally

meaningful

goals,

and

taught

the

3

chair

-bound

mov

ements.

HC

As

re

minded

their

clients

of

their

personal

PA

goal/r

ou-tines

as

part

of

their

regular

home

care

and

logged

their

HM

activities.

Fr

equency

of

HC

A

home

visits

depended

on

the

client

’

s

care

needs

assessed

by

the

Illinois

Determination

of

Needs

instrument

(typically

3x

per

w

eek

or

12

hours

per

w

eek)

Each

da

y

clients

completed

15

arm

curls

2

times

(holding

a

1-pound

w

eight

supplied

by

the

pr

oject,

or

a

soup

can

or

water

bottle),

ankle

point

and

ﬂ

ex

up

to

30

seconds

on

each

foot

3

times

and

seated

step-in-place

up

to

1

minute

Park

and

Chodzk

o-Zajk

o

,

42

HC

As

attended

a

training

session

to

deliv

er

HM,

this

included

how

to

teach

the

clients

the

3

chair

bound

mov

ements

whilst

utilizing

Brief

Negotiation

techniques

as

a

motivational

tool.

HC

As

visited

the

client

in

the

home

and

w

er

e

ask

ed

to

deliv

er

the

HM

pr

ogram

on

a

re

gular

basis

for

a

4-month

time

period

Each

da

y

clients

completed

15

–

20

arm

curls

2

times

(holding

a

1-pound

w

eight,

either

soup

can

or

water

bottle),

ankle

point

and

ﬂ

ex

w

orking

up

to

30

seconds

on

each

foot

3

times

and

seated

step-in-place

w

orking

up

to

1

minute

Abbre

viations:

MAHC-10,

Missouri

Alliance

for

Home

Car

e

tool;

LiFE,

Lifestyle-integrated

Functional

Ex

er

cise

pr

ogram;

CHC

,

Catholic

Home

Car

e;

HC

A,

home

care

aid

e;

PA,

ph

ysical

activity;

HM,

Health

y

Mov

es.

(15)

falls,38,40,41,43,45as did Renehan et al36in their RCT. Other physical performance measures used by single studies included functional reach, sit-to-stand one time, tandem walk (including tandem walk errors), and the six minute walk test. Health-related quality of life was measured using the SF-36, SF-12, EQ5D, AQol-8D, and the COOP/Wonka.

Falls related assessments included the falls efﬁcacy scale,

activity-speciﬁc balance scale, the Falls Risk for the Older

Person living in the Community and the MAHC-10 falls assessment tool.

Dropout and adherence to exercise

interventions

Participant withdrawal rates ranged from 3.9%–3323.3%,37

with an average dropout rate of 13.5% (±7.0). Adherence

rates were reported in seven of the studies.28,29,37,38,41,43,44

Bamgbade and Dearmon38stated 17% (n=5) of their

parti-cipants performed the exercises as instructed over the 2

months. In Burton et al’s28,29RCT studies, the intervention

group reported completing the activities 4.9 times/week during the intervention period (8 weeks) and 4.1 times/ week across the 6-month study, and the control group 4.4 times/week during the intervention (8 weeks) and 3.7 times/ week across the 6 months. In the 8-week study, adherence to exercise interventions delivered by community care

sup-port workers was resup-ported at 4.9 (±2.2) times per week.44

Henwood et al43reported adherence of 5.1 (±0.6) sessions

per week across the 18-week intervention and Saeterbakken

et al37 noted 84% adherence in the 10-week intervention.

Muramatsu et al41 reported half of their participants had

completed 5–7 days training each week at the end of the

intervention (4 months), 32% (n=16) 2–7 days/week at the

end, 16% “low“(“low“not described) and 2% did not do

any of the exercises.

Quality of the studies

In general, the quality of the ten RCT studies was rated

“medium” to “high”, suggesting low risk of bias.

Assessment details for potential bias in each study is

presented in Table 6. The two Burton et al28,29 studies

and Saeterbakkan et al's study37did not blind their

parti-cipants, assessors or those delivering the intervention (ie, care managers) and were therefore deemed to be high risk

for blinding. Allocation concealment was unclear forﬁve

of the studies and some were deemed not free of bias by the reviewers due to uneven or small sample size groups.

Sequence generation and incomplete outcome data

appeared to be well described and were at low risk of bias for the majority of studies.

Similar to the RCTs, the quality of the eight other studies was between ++ and + on the assessments, where

it was deemed that most had satisﬁed the criteria for

internal and external validity. It must be noted however, that all but one study did not have a comparison group, which meant a number of criteria were consequently not

reported.Table 7presents the assessment of quality for the

non-RCT studies included in the review.

Effectiveness of intervention programs

Results from seven of the RCTs were included in the

meta-analyses.28,30,32,33,35,37,48,49Two authors responded to data

requests30,35and their data have been included in the

meta-analysis, one other author responded but did not provide

their data.34

Table 6Randomized controlled trial (RCT)–assessment of risk of bias of RCTs

Study Sequence

generation

Allocation concealment

Blinding Incomplete

outcome data

Selective out-come reporting

Free of other bias

Burton et al,28 ₊ ₊ _- ₊ ₊ _?

Burton et al,29 + + - + + ?

Danilovich et al,30 ₊ _? ₊ ₊ _?

-Danilovich et al,31 + ? + + ?

-King et al,32 + ? + + + ?

Parsons et al,33 _? _? ₊ ₊ ₊

-Stevens-Lapsley et al,34 + ? + + +

-Tuntland et al,35 + + ? + +

-Saeterbakken et al,37 ₊ ₊ _- _- _?

-Renehan et al,36 + + + + -

-Clinical Interventions in Aging downloaded from https://www.dovepress.com/ by 118.70.13.36 on 20-Aug-2020

(16)

Figures 2–5present the data in forest plots for the TUG,

sit-to-stand ﬁve time, grip strength, and walking speed

respec-tively. Heterogeneity for TUG (I2=62%) and sit-to-standﬁve

times (I2=65%) was high. The grip strength and walking speed

(meters per second) analyses had no heterogeneity (I2=0).

Sit-to-stand ﬁve times was also analyzed in subgroups because

there appeared to be two samples that had better function28,48

and two poorer function.33,37 The difference between those

with better function and those with poorer function was

sig-niﬁcant (χ2=4.05, df=1 (p=0.04),I2=75.3%). However,

hetero-geneity was high and there was no signiﬁcant improvement

overall between the intervention and control groups for

sit-to-standﬁve times across the four studies. Walking speed was the

only test which reported an overall signiﬁcant (MD: 0.02,

p<0.001) difference. In this case, it favored the usual care (ie,

control) group in Tinetti et al's49study. TUG, sit-to-standﬁve

times, and grip strength showed no statistically signiﬁcant

difference between the groups for these tests.

Discussion

This systematic review identiﬁed a further 18 studies, since the

previous systematic review was published,22which examined

the effectiveness of physical activity/exercise interventions for older adults receiving home care services. Previously there were no studies included from Australia, however, in the last 6 years, six studies have been conducted, providing evidence for Australian home care agencies and national and state governments who fund these services. Disappointingly, we found no published studies from a number of large countries who provide home care services to older adults such as the United Kingdom, European countries (ie, the Netherlands, Ireland, Germany), and Scandinavia (other than Norway), and no studies from low to middle income countries.

Lack of consistency in methodology and measures has limited the potential to conduct meaningful meta-analyses in this area. Of the four outcomes that were analyzed by

meta-analysis, only walking speed changed signiﬁcantly between

the intervention and control groups. On this occasion, the control group demonstrated faster walking speeds than the intervention group. This was predominantly due to Tinetti

et al’s49study, which reported a faster walking speed (m/s)

for the usual care group at 6 months than the intervention group. However, walking speeds were found to be the same for each group at 12 months, which was 6 months

post-intervention.49 Tinetti et al49 did note that the intervention

group was signiﬁcantly more likely to be using a walking

stick than the control group at the end of the 6-month inter-vention. The intervention group also had more visits from

T able 7 Randomized contr olled trial (RCT) – assessment of quality of non-RCT papers Study 1.1 1.2 1.3 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 4.3 4.4 4.5 4.6 5.1 5.2 Bamgbade and Dearmon, 38 + NR NR NA + NA NA NA NA NA ++ ++ ++ + ++ + ++ NA + NA NA NR NR + NR -+ Burton et al, 44 ++ + + NA ++ NA NA NA NA NA ++ ++ ++ ++ + ++ ++ NA ++ NA NA NR NR ++ ++ + ++ Henw ood et al, 43 ++ + + NA ++ NA NA NA NA NA + ++ + + + ++ ++ NA + NA NA NR NR ++ + -+ Kwok and T ong, 40 ++ ++ ++ + + -+ + ++ + ++ + ++ + ++ + ++ ++ + ++ ++ + ++ ++ + ++ + Muramatsu et al, 41 + + ++ NA ++ NA NA NA NA NA ++ ++ ++ + ++ + ++ NA + NA NA NR NR ++ + + + Park and Chodzk o-Zajk o , 42 ++ + + NA + NA NA NA NA NA ++ ++ ++ + ++ + + NA ++ NA NA NR NR ++ + + + Burton et al, 44 ++ + + NA ++ NA NA NA NA NA ++ ++ ++ + ++ ++ ++ NA ++ NA NA NR NR ++ NR + ++ Gallagher et al, 45 + + + NA + NA NA NA NA NA NA ++ ++ ++ ++ + ++ NA + NA NA ++ ++ ++ ++ + + Notes: 1.1: Is the sour ce population or sour ce ar ea w ell described?; 1.2 :Is the eligible population or ar ea re pr esentative of the sour ce population or ar ea? ; 1.3: Do the selected participants or areas repres ent the eligible population or area ?; 2.1: Allocation to inter vention (or comparison). How was selection bias minimized?; 2.2: W er e inter ventions (and comparisons) wel l describ ed and appr opriate?; 2.3: W as the allocation concealed?; 2.4: W er e participan ts or inv estigators blind to exposure and comparison?; 2.5: W as the exposur e to the inter vention and comparison adequate?; 2.6: W as contamination accept ably low?; 2.7: W er e other inter ventions similar in both gr oups?; 2.8: W er e all participants accounted for at study conclusion?; 2.9: Did the setting re fl ect usual UK practice?; 2.10: Did the inter vention or contr ol comparison re fl ect usual UK practice?; 3.1: W er e outcome measur es reliable?; 3.2: W er e all outcome measur ements complete?; 3.3: W er e all important outcomes assessed?; 3.4: W er e outcomes re le vant?; 3.5: W er e there similar follow-up times in expos ur e and comparison gr oups?; 3.6: W as follow-up time meaningful?; 4.1: W er e exposure and comparison gr oups similar at baseline? If not, wer e these adjusted?; 4.2: W as intention to tr eat analysis conducted?; 4.3: W as the study suf fi ciently pow er ed to detect an inter vention effect (if one exists); 4.4: W er e the estimates of effect size given or calculable?; 4.5: W er e the analytical methods appr opriate?; 4.6: W as the precis ion of inter vention effects giv en o r calculable? W er e they meaningful?; 5.1: Ar e the study re sults internally valid (ie, unbiased)?; 5.2: Ar e the fi ndings generalizable to the sour ce population (ie, externally valid)?