Can peer education improve beliefs, knowledge, motivation and intention to engage in falls prevention amongst community-dwelling older adults?

Get Your Custom Essay Written From Scratch
We have worked on a similar problem. If you need help click order now button and submit your assignment instructions.
Just from $13/Page
Order Now

Linda A. M. Khong1 • Richard G. Berlach2 • Keith D. Hill3 • Anne-Marie Hill3

Published online: 3 January 2017

� Springer-Verlag Berlin Heidelberg 2016

Abstract The aim of the study was to evaluate the effec-

tiveness of delivering a contemporary peer-led falls pre-

vention education presentation on community-dwelling

older adults’ beliefs, knowledge, motivation and intention

to engage in falls prevention strategies. A two-group quasi-

experimental pre-test–post-test study using a convenience

sample was conducted. A new falls prevention training

package for peer educators was developed, drawing on

contemporary adult learning and behaviour change princi-

ples. A 1-h presentation was delivered to community-

dwelling older adults by peer educators trained with the

new package (intervention group). Control group partici-

pants received an existing, 1-h falls prevention presentation

by trained peer educators who had not received the adult

learning and behaviour change training. Participants in

both groups completed a purpose-developed questionnaire

at pre-presentation, immediately post-presentation and at

one-month follow-up. Participants’ levels of beliefs,

knowledge, motivation and intention were compared across

these three points of time. Generalised estimating equations

models examined associations in the quantitative data,

while deductive content analysis was used for qualitative

data. Participants (control n = 99; intervention n = 133)

in both groups showed significantly increased levels of

beliefs and knowledge about falls prevention, and intention

to engage in falls prevention strategies over time compared

to baseline. The intervention group was significantly more

likely to report a clear action plan to undertake falls pre-

vention strategies compared to the control group. Peer-led

falls prevention education is an effective approach for

raising older adults’ beliefs, knowledge and intention to

engage in falls prevention strategies.

Keywords Accidental falls � Peer group � Health education � Health promotion


Falls amongst older adults are a serious health and socio-

economic problem (Peel 2011). The direct cost of falls-re-

lated hospitalisations in Australia was estimated to be over

$648 million in 2007–2008 (AIHW: Bradley 2012). There is

strong evidence that interventions including strength and

balance exercise, cataract surgery, medication review and

multifactorial strategies can reduce falls (Dean-

drea et al. 2010; Gillespie et al. 2012). However, older adults

have been found to have low levels of uptake and engagement

in falls prevention strategies, suggesting that there is a gap in

translating these research findings into practice (Nyman and

Victor 2012; Yardley et al. 2006, 2007).

Qualitative research findings demonstrate that many

older adults possess low levels of knowledge, and believe

that falls prevention is not personally relevant to them or

have low motivation to engage in falls prevention strategies

(Dickinson et al. 2011; Haines et al. 2014; Hill et al. 2011).

Concepts of health behaviour change suggest that provid-

ing people with knowledge and motivation is critical for

Responsible editor: H.-W. Wahl.

& Linda A. M. Khong Linda.Khong1@my.nd.edu.au

1 School of Physiotherapy, Institute for Health Research, The

University of Notre Dame Australia, 19 Mouat Street,

PO Box 1225, Fremantle, WA 6959, Australia

2 School of Education, The University of Notre Dame

Australia, PO Box 1225, Fremantle, WA 6959, Australia

3 School of Physiotherapy and Exercise Science, Curtin

University, GPO Box U1987, Perth, WA 6845, Australia


Eur J Ageing (2017) 14:243–255

DOI 10.1007/s10433-016-0408-x



achieving health behaviour change. Studies that have pro-

vided older adults in hospitals with individualised level

falls prevention education interventions have demonstrated

positive changes in behaviour (Haines et al. 2011; Hill

et al. 2013, 2015; Michie et al. 2011). However, there has

been limited translation of this educational approach into

the community setting.

One review has proposed peer education as a potentially

valuable approach that could influence health-related

behaviour amongst peer participants (Peel and Warburton

2009). Peer education encompasses a range of learning

approaches where information, skills and values are con-

veyed amongst people who share common characteristics

such as age or shared experience (Simoni et al. 2011). A

peer educator deemed as a credible source and positive role

model can play a pivotal role in promoting self-confidence

and influencing health-related behaviour amongst their

peer participants (Peel and Warburton 2009). Evidence

from a systematic review of 17 studies (7442 people) using

peer education found that providing peer education resulted

in positive health behaviour outcomes for the recipients

(Foster et al. 2007).

There is limited empirical research investigating the

impact of peer education in the area of falls prevention,

especially where an older individual peer delivers a pre-

sentation to a group of other older adults. A large sys-

tematic review of falls prevention studies in the community

setting (159 trials) included only four studies that evaluated

education interventions and only one of these was a peer

intervention study. The review found that evidence for

education interventions was inconclusive (Gillespie et al.

2012). Previous findings suggest that there is uncertainty

about the efficacy of peer-led falls prevention education as

facilitated by presentation, lecture or discussion (Allen

2004; Deery et al. 2000; Kempton et al. 2000). The limi-

tations of these studies included not describing the content

of the intervention clearly and not describing pedagogical

and underlying theory that had guided the design and

implementation of the interventions. Researchers recom-

mended that studies that evaluate behavioural interventions

should define the framework chosen to design the inter-

vention and include description of the content and mode of

delivery, but importantly also describe the active compo-

nents of the intervention that are intended to facilitate

behaviour change and the behaviour change techniques

used (Abraham and Michie 2008; Michie and Johnston

2012). Use of theories to inform health behavioural change

interventions has been advocated because it provides a

matrix of enablers, barriers and mechanisms to explain and

predict health behaviour (Improved Clinical Effectiveness

through Behavioural Research Group (ICEBeRG) 2006).

Therefore, provision of a peer-led presentation should

ideally be underpinned by adult learning principles

(Merriam and Bierema 2014) and behavioural change

framework such as the behaviour change wheel theory

(Michie et al. 2011). This may improve beliefs, knowledge,

motivation and intention which could facilitate behaviour

change, namely the uptake of falls prevention strategies by

older adults.

The aim of this study was to evaluate the effect of

delivering a peer-led falls prevention presentation on

community-dwelling older adults’ beliefs and knowledge

about falls prevention, and their motivation and intention to

engage in falls prevention strategies. The study compared

the effect of delivering a contemporary presentation by an

individual older adult to a group incorporating adult

learning principles and behaviour change strategies against

delivering an existing peer-led falls prevention presentation

that did not incorporate these principles or behaviour

change strategies.


Study design

A two-group quasi-experimental pre-test–post-test study

design using a convenience sample was conducted. At the

initial control group stage (Phase 1), participants received

the existing peer-led presentation. In the subsequent

intervention group stage (Phase 2), participants received

the new contemporary peer-led presentation (Fig. 1).


The study was approved by the University of Notre Dame

Australia’s Human Research Ethics Committee (Reference

014134F and 015013F). All participants provided written

informed consent.

Participants and setting

Participants were community-dwelling older adults who

were attending a peer-led falls prevention presentation.

Inclusion criteria for both control and intervention groups

consisted of being aged 60 years or older, attending a peer-

led falls prevention presentation during the study phases

and being able to complete a questionnaire. Older adults

who resided in residential care facilities or were hospi-

talised were excluded.

The presentations were organised by a large not-for-

profit community organisation that promoted injury pre-

vention and community safety in Western Australia. The

community engagement officer from the community

organisation was a qualified health promotion professional,

who managed their peer-led education programmes. The

244 Eur J Ageing (2017) 14:243–255




community engagement officer was the key person who

recruited and trained new volunteer older adult peer edu-

cators to present and deliver the peer-led falls prevention

education programme, which aimed to raise awareness of

falls prevention amongst community-dwelling older adults.

These peer educators’ ages ranged from 65 to 85 years and

most were retired and possessed diverse working experi-

ence before retirement.

Fig. 1 Flow diagram of the recruitment of participants and data collection process for the study

Eur J Ageing (2017) 14:243–255 245





A convenience sample was recruited for both the control

and intervention phases of the trial.

Peer-led presentations were organised by the community

engagement officer who advertised the falls prevention

presentation to existing older adult social groups in Wes-

tern Australia, retirement village associations and other

seniors’ networks through mailed flyers or newsletters five

months prior to conducting each phase of the study. The

community engagement officer was the organisation’s

contact person for these groups and played an active role in

the scheduling of the falls prevention presentations to each

group, as well as providing support for the programme.

Control conditions

The control conditions consisted of participants receiving

the existing peer-led presentation during Phase 1 (2014).

This was a 1-h presentation delivered by five volunteer

peer educators that has been delivered regularly for

approximately 10 years. The existing peer-led falls pre-

vention presentation consisted of the peer educators shar-

ing falls-related content knowledge such as risk factors for

falls and strategies for reducing risk of falls, including

managing one’s medications, improving balance by

undertaking exercises, checking feet and footwear and

completing environmental modifications (Deandrea et al.

2010; Gillespie et al. 2012). The training for these volun-

teer peer educators, conducted by the community engage-

ment officer, consisted of a 5-h session which provided

them with this information (Table 1). The content was

regularly reviewed by the organisation and focused on

providing the best available strategies that could be used by

older adults to reduce their falls risk. However, the training

did not include information about the principles of adult

learning and health behaviour change. Peer educators were

also provided falls prevention support materials such as a

videotape, booklet and flyers to use during presentations, to

aid in conveying the falls prevention message to the

community groups of older adults. These existing peer

educators were experienced presenters all aged over

60 years who had delivered the presentations for between

two and ten years. The training for both existing and new

peer educators delivering the presentations to the control

and intervention groups is presented in Table 1.


A contemporary falls prevention peer-led education pro-

gramme was designed by the research team to be used in

Phase 2 (2015). The programme consisted of providing

training and resources for new volunteer peer educators to

also deliver a 1-h peer-led falls prevention presentation to

groups of community-dwelling older adults. The aim of the

presentation was to improve the older community-dwelling

adults (1) self-belief that taking measures to reduce their

risk of falls would be useful, (2) knowledge about falls and

falls prevention strategies and (3) motivation and intention

to engage in falls prevention strategies.

The development and implementation of the presenta-

tion was informed by previous studies conducted by the

present authors, whereby key stakeholders were consulted,

including community-dwelling older adults (Khong et al.

2016) and experts in the area of education and falls pre-

vention. Feedback was also sought from the peer educators

who were delivering the existing presentations (Khong

et al. 2015). The design and implementation of the con-

temporary presentation was based on the framework of the

behaviour change wheel theory (Michie et al. 2011) and

was also informed by educational and adult learning prin-

ciples (Anderson et al. 2001; Merriam and Bierema 2014).

Six new volunteer peer educators were recruited via

daily advertisements run on a community radio whose tar-

get audience was older adults. These six volunteers com-

pleted their training but only two were available to deliver

the presentation during the intervention phase of the trial.

The first day (5 h) of the peer educator training was con-

ducted by the community engagement officer who imparted

falls-related content knowledge such as the definition of a

fall, statistics about the nature and incidence of falls in the

community, and the risk factors contributing to falls

(Table 1). This training session was identical to the one

provided to those peer educators who were delivering the

existing presentations to the control groups. The new peer

educators were also provided with the same falls prevention

support materials (a videotape, booklet and flyers) to deliver

their presentations to the intervention groups. Subsequently,

an additional 4-h training session was conducted by the

researchers for the new peer educators using purpose-de-

veloped education resources. These resources consisted of a

facilitator–trainer guide and instructional aids, a training

video and a peer educator guidebook, including a pro-

gramme fidelity checklist (Bellg et al. 2004). Principles of

adult learning, behaviour change techniques (such as goal

setting) and pedagogical skills, including suggestions on

how to conduct an interactive presentation, were shared

with the new peer educators (Table 1) (Abraham and

Michie 2008; Anderson et al. 2001; Fleming 2008; Merriam

and Bierema 2014). Peer educators were trained to establish

themselves as a credible source of information when they

delivered a presentation and were encouraged to share

personal insights regarding falls prevention to engage and

foster their peers’ learning and self-confidence. Each new

peer educator was provided with a guidebook consisting of

information imparted during the training session. A training

246 Eur J Ageing (2017) 14:243–255




video with prompts involving an experienced university

educator was created. This video modelled the contempo-

rary falls prevention presentation to a live audience. Sub-

sequently, the video was developed as an online resource

for training new peer educators.

Following the training, each new peer educator con-

ducted an initial falls prevention presentation with support

from the organisation and a fellow peer educator. After

delivering a presentation, the peer educator completed the

programme fidelity checklist (Bellg et al. 2004), which was

used as a guide for self-reflection and feedback and to

promote adherence to the intervention delivery.

Data collection and procedure

Data collection followed the same procedure during both

phases of the trial. The peer educator arrived at the local

community group when a presentation was organised. Prior

to the delivery of the presentation, the older adults who

attended were invited to participate in the trial and those

who provided written consent were recruited. Each par-

ticipant completed a purpose-developed questionnaire prior

to the peer educator delivering the falls prevention pre-

sentation and following the presentation. The follow-up

questionnaire was mailed out to each participant 1 month

after the presentation.

The design of the questionnaire items was based on other

studies that designed questionnaires specifically to evaluate

behaviour change or evaluated behaviour change regarding

falls prevention (Cane et al. 2012; Hill et al. 2009; Huijg

et al. 2014). The overall design of the questionnaire was

based on the framework of behaviour change wheel theory

(Michie et al. 2011), namely capability (awareness and

knowledge), opportunity and motivation (Michie et al.

2011). There were seven closed items (see Table 3) which

were rated on a five-point Likert scale (Strongly Agree,

Agree, Neutral, Disagree, Strongly Disagree). The final

open-ended item (item 8) asked each participant to list up to

three measures that they could take in the next month which

would help them avoid falling or reduce their risk of falling

(Table 5). The post-presentation and one-month follow-up

questionnaires were modified slightly in terms of wording

of the questionnaire items, so that the wording was in the

context of having attended the peer educators’ presentation.

At the one-month follow-up, telephone calls were made to

each participant to advise them to expect a questionnaire,

which was subsequently mailed out with a prepaid enve-

lope. A single mail or telephone call was made to remind

those who did not respond within two weeks of the deadline

to return the questionnaire.

The first seven questionnaire items are shown in

Table 3. The four outcomes measured using the

Table 1 Training sessions undertaken to prepare peer educators of existing and contemporary programmes to deliver peer-led falls prevention education presentations

Training sessions for peer educators Existing

programme a


programme b

Training session (5 h): conducted by community engagement officer 4 4

Learning objectives: introduction to epidemiology of falls-related content knowledge, e.g. falls information

including incidence of falls in the community, risk factors for falling, evidence-based falls prevention

strategies c

4 4

Training activity provided: demonstration and lecture 4 4

Activity supporting material: lecture notes 4 4

Peer-led falls prevention presentation support material: video, booklet and flyers 4 4

Additional training session (4 h): conducted by research team 7 4

Learning objectives: develop an awareness of learning styles; describe basic principles of adult learning and

apply them in delivering falls prevention presentations; identify and integrate relevant behaviour change

techniques into falls prevention presentations d

7 4

Training activity provided: learning style questionnaire, online video links, discussion, group work and

interaction, and mock presentation practice

7 4

Activity supporting material: peer educator guidebook and online training video; programme fidelity e


self-reflection guide

7 4

a Peer educators were trained and already had two to ten years of experience delivering the existing peer-led falls prevention education preceding

the research period b

Newly recruited volunteer peer educators who were trained to deliver the contemporary peer-led falls prevention education c

Deandrea et al. (2010) and Gillespie et al. (2012) d

Abraham and Michie (2008), Anderson et al. (2001), Fleming (2008) and Merriam and Bierema (2014) e

Bellg et al. (2004)

Eur J Ageing (2017) 14:243–255 247




questionnaire were: (i) beliefs about falling and falls pre-

vention (measured using items 1 and 2), (ii) levels of

knowledge about falls prevention (measured using items 3

and 5), (iii) motivation to reduce risk of falling by engaging

in falls prevention strategies (measured using item 4) and

(iv) intention and a plan to undertake falls prevention

strategies (measured using items 6 and 7). The final item

(8) is a question that aimed to understand the participants’

knowledge, intention and plan to undertake falls prevention

strategies, as shown in Table 5.

Other information collected at baseline was participants’

sociodemographic characteristics, including age, gender,

socio-economic index (Australian Bureau of Statistics

2013), self-rated health, number of prescribed medications

taken per day, history of falls in the past 12 months and

level of mobility.

Prior to the commencement of the main trial, a conve-

nience sample of community-dwelling older adults who

attended social walking groups was enrolled to evaluate the

test–retest reliability of the questionnaire. Subsequently,

the questionnaire was pilot-tested with older adults from

two other social groups completing the questionnaires

across three points of time, after which slight changes were

made to the format of the questionnaire and to the

instructions given for completing it in order to clarify the

procedure for participants.

Data analysis

Baseline characteristics of the two groups’ participants

were compared using t test for continuous data, and Pear-

son’s Chi-square and Fisher’s exact tests were used for

comparison of categorical data. The test–retest reliability

of the questionnaire was established using intraclass cor-

relation (ICC) and Cohen’s kappa coefficient (kappa). A p-

value \.05 was considered significant for all analyses. Participants’ responses to the seven closed items (de-

pendent variables) measuring beliefs, knowledge, motiva-

tion and intention outcomes were compared within and

between the intervention and control groups using gener-

alised estimating equation (GEE) modelling (Liang and

Zeger 1986). The GEE approach was considered appro-

priate because it was able to account for correlations

amongst the participants’ outcomes and was able to include

more than one covariate (either continuous or categorical)

(Liang and Zeger 1986; Williamson et al. 1996). The

independent variables were participants’ sociodemographic

information. Final GEE models included only significant

independent variables (p \ .05). Results were reported using odds ratios (OR) with accompanying 95% confidence

intervals and p-values. Quantitative data were analysed

using statistical package SPSS �

(Statistical Package for

Social Sciences, version 22 for Windows).

Qualitative data obtained from both groups’ open-ended

responses (item 8 in the questionnaire) were transcribed

verbatim and exported to NVivo 10 for Windows (QSR

International Pty Ltd 2012). These data were analysed using

deductive content analysis, which is based on using previ-

ous knowledge around the research topic (Elo and Kyngas

2008). A categorisation matrix was constructed using

Australian recommendations for falls prevention for com-

munity-dwelling older people (Australian Commission on

Safety and Quality in Healthcare 2009) and systematic

reviews which summarised the evidence for falls prevention

strategies for community-dwelling older people (Deandrea

et al. 2010; Gillespie et al. 2012). The main category was

participants’ knowledge about falls prevention as evidenced

by the measures identified in their plan (see Table 5). The

primary researcher read through transcripts to gain a sense

of the content. Participants’ responses about their falls

prevention measures were coded by theme and assigned

according to the predetermined categories within the

matrix. New categories were generated for responses that

could not be categorised within the matrix. Two researchers

discussed the data but identified their corresponding generic

and sub-categories independently. Frequency counts were

also undertaken of each category or sub-category. Final

findings of the two independent researchers were compared

and triangulated to enhance trustworthiness of the findings.

Sample size

For conducting the test–retest reliability, for an estimated

reliability index of 80%, with an alpha level of 5% and

power of 80%, a minimum sample of 46 participants were

required (Walter et al. 1998). Regarding sample size for the

main trial, as previous trials in this area had not been

conducted, a minimum number of 100 participants were

chosen for the control group to gain sufficient data to

calculate the sample size for the subsequent Phase 2. The

control phase of the study used data from participants (pre-

and post-presentation measurements for each participant)

and measured differences over time. These data from the

control group indicated that when examining the mean

differences of each of the seven items, the minimum dif-

ference in the responses from the participants from pre- to

post-presentation was normally distributed with a standard

deviation of 0.44. If the true difference in the mean

response was 0.155, then 65 participants (with paired pre-

and post-presentation data) needed to be enrolled in the

intervention group to be able to reject the null hypothesis

that this response difference was zero with probability

(power) 0.8. The type I error probability associated with

this test of this null hypothesis was 0.05. Since in the

control group trial there was a dropout rate of 17% between

248 Eur J Ageing (2017) 14:243–255




baseline and one-month follow-up, the aim was to enrol at

least 80 participants for Phase 2 of the study.


The content and face validity of the questionnaire was

evaluated by health professionals and community-dwelling

older adults and the questionnaire was revised based on

their feedback. The final questionnaire was pre-tested with

16 older adults. Forty-nine older adults (aged 60 and over)

subsequently participated in the test–retest reliability trial

of the questionnaire. There was moderate to substantial

agreement across items (Kappa = .585–.765) (Landis and

Koch 1977). On further analysis, compared to the rest of

the items, the kappa for questionnaire item 5 assessing ‘‘I

am confident that if I wanted to, I could reduce my risk of

falling’’ was the lowest at 0.585 (moderate agreement).

Percentage agreement ranged from 73.5 to 87.8% and the

ICC for the participants’ mean score of outcome measures

between retest occasions was 0.88, which was considered a

good level of agreement (Portney and Watkins 2009).

There were n = 141 participants who enrolled and of

those n = 99 participants (70%) completed Phase 1 (con-

trol) of the trial. For the intervention trial, n = 196 enrolled

and n = 133 participants (67%) completed Phase 2 (in-

tervention). The flow of participants through the study is

shown in Fig. 1. The main reasons for not providing any

response to the post-presentation or follow-up question-

naire included participants needing to leave the presenta-

tion venue prior to the post-presentation questionnaire

being administered, being unwell, away on holiday or

unable to be contacted at the one-month follow-up. Par-

ticipants were excluded from the analysis if they did not

complete the questionnaire after the presentation or at the

one-month follow-up. There were no significant differ-

ences in the demographic characteristics between partici-

pants who dropped out compared to participants who

completed the follow-up questionnaire.

Participant characteristics from both groups are pre-

sented in Table 2. Intervention group participants were

significantly more likely to be male (p = 0.006) and come

from higher socio-economic areas (p = 0.002).

Participants’ levels of beliefs, knowledge about falls and

falls prevention, motivation and intention to reduce their risk

of falling at baseline and after the presentations are presented

in Table 3. Participants in both control and intervention

groups showed increased levels of self-perceived knowl-

edge, increased self-belief that falls prevention would be

useful and increased levels of motivation to prevent falls at

post-presentation and at one-month follow-up. Participants

in both groups also reported higher levels of intention

(control median 4.4, intervention median 4.5) and clear plans

(control median 4.3, intervention median 4.3) in falls pre-

vention strategies following the presentations.

For the GEE modelling (Table 4), the Likert scores of

the seven items were found to be bimodal and therefore

were recoded into a dichotomised variable. Rating of

‘‘Strongly Agree’’ and ‘‘Agree’’ were recoded to ‘‘Agree’’

or 1 and ‘‘Neutral’’, ‘‘Disagree’’ and ‘‘Strongly Disagree’’

were recoded to ‘‘Disagree’’ or 0. Participants within both

the control and intervention groups demonstrated signifi-

cantly increased levels of beliefs that falls prevention

measures would be useful and that knowledge about falls

prevention strategies increased intention to take measures

to prevent falls. Both groups also reported a clear action

plan to engage in falls prevention strategies at post-pre-

sentation or at one-month follow-up (Table 4) compared to

baseline. Despite participants’ improved levels of motiva-

tion to reduce their risk of falling across the three points of

time within both the control and intervention group, there

was no significant between-group difference when inves-

tigated in the GEE modelling. Multivariate analysis

demonstrated that the intervention group was significantly

more likely to report that they had developed a clear action

plan which they intended to implement to reduce their risk

of falling compared to the control group [OR = 1.69, 95%

CI (1.03–2.78)], but there were no significant differences

between groups regarding beliefs and knowledge about

falls prevention, and levels of intention to engage in falls

prevention strategies.

Female participants in both groups were significantly

more likely to believe that taking measures to prevent falls

was useful [OR = 3.99, 95% CI (1.08–14.68)]; to report

increased levels of knowledge about falls prevention after

the presentation [OR = 2.34, 95% CI (1.09–5.13)]; to

report increased intention to take measures to prevent falls

[OR = 1.82, 95%CI (1.02–3.270]; and to report a clear

action plan to reduce their risk of falling [OR = 2.47, 95%

CI (1.51–4.02)] (Table 4). Participants who reported that

they had previously discussed falls prevention with their

doctor or health professional or received falls prevention

information were significantly more likely to report an

increased knowledge of falls risk [OR = 3.07, 95% CI

(1.09–8.66)] and to develop a falls prevention action plan

[OR = 2.12, 95% CI (1.19–3.78)].

Deductive content analysis of the written responses of

both control and intervention groups’ participants to the

open-ended item (item 8) is displayed in Table 5. Partici-

pants identified measures that they considered they could

take that would help reduce their risk of falling, which

were coded into three generic categories: (1) evidence-

based strategies of which there were seven sub-categories,

(2) non-evidenced strategies and (3) no strategies. The

latter two categories were new categories generated from

data that did not fit into the predetermined categories.

Eur J Ageing (2017) 14:243–255 249




Table 5 shows the measures that participants identified as

being helpful for reducing their risk of falling. Summative

responses from both control and intervention groups’ par-

ticipants within each generic and sub-category are sum-

marised in Table 5.

Knowledge about environmental modification measures

was the largest sub-category represented, which included

comments about adaptation of the internal and external home

environment. One participant described ‘‘shortened electric

blanket cords beside bed … so I would not fall over it’’. The environmental aids sub-category represented

responses that described using mobility aids such as a

walking stick. The balance and mobility sub-category

included measures relating to posture, balance and gait but

excluded exercises. Examples included ‘‘Walking rather

than shuffling; Make a conscious effort to lift my feet when

walking’’. The other sub-categories described and coded


Exercise: Continued with tai-chi; Balance exercises;

Did quad [quadriceps] strengthening exercises; See-

ing a physiotherapist to help me with my strength.

Feet and Footwear: Podiatrist; Got rid of loose fitting

shoes. Medication: Health check with doctor and

using correct medications

Table 2 Participants’ baseline characteristics

Characteristic Control

n = 99


n = 133


Age (years), M (SD) 77.9 (6.9) 79.2 (7.0) .142 b

Number of prescribed medication taken per day, Mdn (IQR) 4.0 (5.0) 4.0 (5.5) .606 b

Number of people who had fallen in the past 12 months, n (%) 40 (40.4) 45 (33.8) .304 a

Gender, n (%) .006 a *

Female 71 (71.7) 72 (54.1)

Socio-economic area, n (%) .002 a *

Higher 59 (59.6) 104 (78.2)

Self-rated health, n (%) .261 a

Poor/fair 25 (25.3) 22 (16.5)

Good 52 (52.5) 79 (59.4)

Very good 22 (22.2) 32 (24.1)

Self-rated difficulty with walking, n (%) .115 a

No 61 (61.6) 95 (71.4)

Use of walking aid inside of house, n (%) .182 c

Nil aids 83 (83.8) 122 (91.7)

Walking stick 11 (11.1) 8 (6.0)

Walking frame 5 (5.1) 3 (2.3)

Use of walking aid outside of house, n (%) .612 a

Nil aids 72 (72.7) 104 (78.2)

Walking stick 15 (15.2) 17 (12.8)

Walking frame 12 (12.1) 12 (9.0)

Ambulatory distance without rest on level ground, n (%) .182 a

\400 m 21 (21.2) 17 (12.8) 400–800 m 23 (23.2) 35 (26.3)

801 m–1.6 km 13 (13.1) 29 (21.8)

1.7–3.2 km 15 (15.2) 24 (18.0)

3.3 km or more 27 (27.3) 28 (21.1)

Previously discussed issue of falls with health professional/doctor or

received falls prevention information from them? n (%)

.232 a

Yes 34 (34.3) 36 (27.1)

M mean, SD standard deviation, Mdn median, IQR inter quartile range a

Determined by using Chi-square test b

Determined by using t test c

Determined by using Fisher’s exact test

* Significant at p \ .05

250 Eur J Ageing (2017) 14:243–255




Participants in both groups also provided responses, in

addition to the falls prevention measures they listed, that

appeared to reflect their increased beliefs about the need to

reduce their risk of falling. This was evidenced by

comments that demonstrated recognition of the need to

change or modify their behaviour, with one participant

stating ‘‘[I] truly believe I need to change’’. Other

responses indicated that participants accepted that the

topic was personally relevant to them, with statements such


Awareness of the likelihood of falling at my age;

Your presentation reinforced my current behaviour to

prevent falls; I made a deliberate attempt to analyse

my [falls] risks in my small unit.

Some responses were categorised as being not evidence

based and some participants stated ‘‘none’’ or ‘‘nil’’ when

asked to list measures they planned to take to reduce their

risk of falls. Measures that were categorised as not being

evidence based included ‘‘Slow down and take [your] time;

Being careful always; Slower walking; Watching more’’.