A Feasibility Study of Multi-Component Fall Prevention for Homebound Older Adults Facilitated by Lay Coaches and Using a Tablet-Based, Gamified Exercise Application

Abstract

Although homebound older adults face high risk for falls, they are unable to utilize community-based fall prevention programs due to their mobility limitations. In this article, we report a feasibility study of a four-session, multicomponent fall prevention program for low-income homebound older adults using pre, post, mixed-method design. The manualized program was delivered by lay coaches who were trained and supervised by a physical therapist. The program also used an iPad-based gamified strength and balance exercise app (called KOKU) that was operable without the need to connect to the internet. Participants (N = 28) in this study were highly receptive to the program and approved all components: psychoeducation, the KOKU app, home-safety checks, safe ambulation training, and medication review. The study showed that a brief, multicomponent fall prevention program for homebound older adults is feasible and acceptable. Further research is needed to evaluate its effectiveness.

Introduction

According to the Centers for Disease Control and Prevention (CDC, 2017), more than one of four community-dwelling older adults (aged 65+) fall each year, and one of five falls causes serious injuries such as broken bones or a head injury. Homebound older adults who tend to have high medical and psychiatric illness burden face even higher rates of falls, because difficulties with walking and balance, lower body weakness, and polypharmacy, including psychotropic drugs, are significant fall risk factors (CDC, 2017; Choi, Sullivan, & Marti, 2019; Watanabe, 2016). In Choi, Sullivan, & Marti’s study (2019), 41% of 2,224 individuals aged 50+ who received home-delivered meals reported past-year falls, and frequently used emergency medical services for lift assists. Actual fall rates are likely to be higher as older adults tend to underreport falls and fall-related injuries even to their health care providers (CDC, 2017; Hoffman et al., 2018).

A number of evidence-based fall prevention (FP) programs are available in the community (National Council on Aging, n.d.), and a systematic review showed that exercise programs targeting balance, resistance, and muscle strength reduce fall rates by 23% (Sherrington et al., 2019). However, homebound older adults, especially low-income homebound older adults who lack transportation, have difficulty accessing these programs. Although in-home FP programs have been implemented for older adults to increase level of physical activity/exercise and home safety, some of these programs include multiple home visits by physical therapists, occupational therapists, and/or registered nurses (Miller et al., 2010; Müller et al., 2019; Shubert et al., 2017; Szanton et al., 2019). While visits by these professionals may be ideal, it is too costly to be scalable and sustainable for a growing number of homebound older adults. The professional workforce shortage (Alliance for Physical Therapy Quality and Innovation, 2019) is also a barrier to making such programs accessible to many older adults.

With the rapid population aging, the numbers of homebound older adults are projected to grow rapidly (Xiang et al., 2020). Scalable and sustainable FP programs that can reach a large number of homebound older adults are needed. Given multiple fall risk factors among homebound older adults, effective FP programs for these older adults should also use a multicomponent approach including balance and strength exercise programs; home-safety checks; training in safe ambulation, mobility aid use, safe showering/bathing; enhanced nutrient intake; and a medication review. Most falls and fall injuries occur at home due to loss of balance/dizziness, slipping, and tripping (Choi, Choi, et al., 2019) and relatively simple home-safety assessment/improvement and home-modifications have been found to be cost-effective fall injury prevention interventions (Keall et al., 2017; Kunigkeit et al., 2018). Mobility device use among the U.S. older-adult population has increased over the past decade (Gell et al., 2015), and training in safe use of mobility aids, safe ambulation, and bathroom use is likely to be especially helpful for mobility-impaired older adults.

In this article, we report a feasibility study of a four-session, multicomponent FP intervention that was designed to be scalable and sustainable in social service agencies serving low-income, homebound older adults. Our intervention components are largely based on the CDC’s STEADI (Stopping Elderly Accidents, Deaths & Injuries; Stevens & Phelan, 2013) tool kit that is theory-driven and applies concepts from the Chronic Care Model (Wagner, 1998) with emphasis on self-management support and the Transtheoretical Stages of Change Model (Prochaska & Velicer, 1997) for behavior change. We made some content and procedural adaptations based on our clinical experience with low-income homebound older adults. The intervention’s two scalability and sustainability features are (a) facilitation by lay coaches trained by a physical therapist and (b) the “Keep on Keep up (KOKU),” a free, tablet-based, gamified strength and balance exercise application. The KOKU app was specifically designed for older adults at risk of falls with all safety features and progressive intensity levels based on the Otago Exercise Program (Campbell & Robertson, 2003).

The primary aim in this article was to describe participants’ acceptance of the overall intervention and the KOKU app and their perception of helpfulness of each intervention component as well as overall intervention experience. The secondary aim was to examine preliminary efficacy, that is, changes between pre- and post-test scores for exercise frequency and efficacy, disability, depressive symptoms, sleep quality, and self-ratings of overall physical and mental health, fear of falls, and falls. The findings may provide insights into a brief, potentially cost-efficient, FP program that may be able to reach a growing number of homebound older adults.

Method

Study Design and Participants

This feasibility study used a one-group, pre, post, mixed-method design. Case managers of a Meals on Wheels program in Central Texas referred 30 potentially eligible participants (age 50+ years; English-speaking; self-reported past-year fall or strength/balance problems; and cognitively intact) between September 1, 2019, and March 15, 2020. In addition, four study participants referred five friends/neighbors to the study team. Of these 35 referred individuals, 29 individuals were found eligible at the time of telephone screening using the five-item Falls Risk Assessment Tool (FRAT; Nandy et al., 2005) and agreed to participate in the study. However, one person did not continue in the program due to post-enrollment cancer diagnosis and chemotherapy. All 28 participants completed four intervention sessions and 6- and 12-week follow-up assessments. The first author’s university institutional review board (IRB) approved the study.

After each participant provided IRB-approved written informed consent, the FP coach who was assigned to work with the participant completed a baseline assessment as part of the intervention. Each participant received a folder containing copies of the CDC’s STEADI brochures and an iPad preloaded with the KOKU app. (Once uploaded onto an iPad, KOKU is operable without the need to connect to the internet.) An independent assessor (not the coach who worked with the participant) conducted follow-up assessments at 6 weeks and 12 weeks after baseline. Due to Covid-19 social distancing practice, a quarter of 6-week and half of 12-week follow-up assessments were done by telephone rather than in-person.

Intervention

The intervention manual was co-developed by the first, second, and the last authors, with the following six components that are designed to be delivered in four 90-min, weekly in-home sessions. Two 15-min telephone check-in calls, each in the fifth and sixth week, followed the in-home sessions to encourage participants to apply their FP knowledge/skills in daily life.

1. Fall risk assessment using the CDC’s STEADI Stay Independent brochure which is based on the Fall Risk Questionnaire (FRQ; Rubenstein et al., 2011) and identification of the need for physical evaluation by health care providers (Session 1);

2. Psychoeducation using the following CDC STEADI brochures: Fact Sheet: Talking About Fall Prevention With Your Patients; Checklist, Fall Risk Factors; What You Can Do to Prevent Falls; and Postural Hypotension: What It Is and How to Manage It (Session 1);

3. Demonstration of the KOKU app and weekly KOKU (3 times a week, 30 min each with 30 repetitions for each type of exercise) and/or other exercise plan development (Session 2);

4. Home-safety checks and safety improvement plan development (Session 3);

5. Coaching and practice for safe ambulation and mobility aid use, showering/bathing, and transfer from/to bed and chair (Session 3); and

6. Filling out the medication review form, referrals to a pharmacist-led medication review available through a local Area Agency on Aging (AAA), and healthy eating habits (Session 4). We included healthy eating habits as low-income homebound older adults have high rates of cardiovascular disease and diabetes, which are significant fall risk factors (Manemann et al., 2018; Yang et al., 2016).

Due to Covid-19 social distancing practice that began in mid-March 2020, three participants received one to two sessions via telephone, and one participant received all four sessions via telephone.

In telephone sessions, coaches provided verbal descriptions for KOKU, safe ambulation, and other safety skills, and asked participants to describe the status of each room in their homes for safety plan developments.

Lay FP Coach Training

Two bachelor’s-level individuals with 3+ years of aging-service practice experience were trained on the intervention manual. The first author trained them on overall FP programs, CDC STEADI brochures, and medication review. The last author, a physical therapist and FP expert, trained FP coaches on home-safety checks, safe ambulation and mobility aid use, showering/bathing, and transfer from/to bed using didactic training sessions and videos. The second author and her team, developers of the KOKU app, trained FP coaches on KOKU use. Then, the FP coaches practiced their FP skills with two to three practice cases. The didactic and practice-case-based training took approximately 50 hr for each FP coach. Coaches received as-needed supervision throughout the project using the FP Coach Adherence Scale that was developed to monitor and ensure coaches’ fidelity (on a 6-point scale: 0 = very poor, 5 = very good) to intervention components in each session. All sessions received four or five ratings.

Measures.

Overall intervention acceptability was assessed at 6 weeks with the 11-item, 7-point, modified Treatment Evaluation Inventory (TEI; Landreville & Guerette, 1998). The TEI was originally developed to evaluate psychosocial treatment for older adults; however, the wording was generic enough that we were able to use it to evaluate the intervention in this study by replacing the word “treatment” with “program.” Higher scores indicate higher acceptability, with a score of 44 indicating moderate acceptability.

Acceptability of the KOKU App was assessed at 6 weeks with two questionnaires that were developed and validated by the second author and her team to measure technology acceptance and system usability specifically for the KOKU app. The 15-item technology acceptance questionnaire is based on technology acceptance models among older adults (see Portz et al., 2019) and includes four subscales: four-item Perceived Ease of Use; four-item Perceived Usefulness; four-item Attitude Toward Using; three-item Intent-to-Use on 7-point scale (1 = strongly disagree to 7 = strongly agree). The 10-item, 5-point (1 = strongly disagree to 5 = strongly agree) system usability questionnaire includes specific questions about KOKU’s functionalities.

Perceived helpfulness of the intervention was assessed with participants’ qualitative feedback that was obtained from a semi-structured interview (30–45 min) with each participant at 6 weeks. We solicited participants’ opinions on helpfulness of each intervention component (e.g., “Did you find discussion of fall risks [home-safety checks and plan development] useful?” “How about the tablet-based KOKU App? Did you like it; did you enjoy it? If not, why?”) and overall experience (e.g., “What have you noticed in terms of your strength building and other aspects of FP since you participated in the study?”). We also asked them about facilitators and barrier to keeping their motivation for exercise and how the program may be improved.

Preliminary efficacy was assessed with the following measures at baseline and 12 weeks: (a) exercise frequency (number of days in a typical week engaged in moderate/vigorous exercise) and exercise efficacy (nine-item Outcome Expectations for Exercise Scale [OEES]; Resnick et al., 2000); (b) disability (12-item World Health Organization Disability Assessment Schedule [WHODAS 2.0; (World Health Organization, 2010)]) and functional impairment (number of impairment in activities and instrumental activities of daily living (ADL/IADL) during the past month; (c) depressive symptoms (the Patient Health Questionnaire–9 [PHQ-9]; Kroenke et al., 2001); (d) sleep patterns (hours of sleep, sleep quality [1 = very bad to 4 = very good], and frequency of sleep medication use [0 = not during the past month, 1 = less than once, 2 = 1–2 times a week, 3 = 3+ times a week]); (e) self-ratings of physical and mental health (1 = poor to 5 = excellent); and (f) fear of falling (16-item Falls Efficacy Scale–International [FES-I]; Helbostad et al., 2010).

We assessed the number of falls (“an unexpected event in which the participants come to rest on the ground, floor, or lower level”; Lamb et al., 2005) at baseline and monthly thereafter for 3 months to ensure that the intervention did not increase falls. At baseline, participants self-reported the number of falls during the preceding 3, 6, and 12 months. Following their enrollment, participants were instructed to complete and return three monthly (counting from baseline) fall/injury calendars. For any faller, a follow-up call was made to determine the circumstances and clinical outcomes of the fall(s).

Analysis.

We used univariate analysis to describe the participants’ characteristics and their acceptance of the overall intervention and the KOKU app. We used thematic analysis (Nowell et al., 2017) for qualitative data to explore common themes related to perceived helpfulness. The first, third, and fourth authors reviewed all interview transcriptions independently and then discussed the themes (intercoder reliability = .98) to reach consensus. The last author reviewed the methods and themes to meet the trustworthiness criteria (Shenton, 2004). To examine preliminary efficacy, we used paired-sample t tests for exercise efficacy, disability, ADL/IADL impairments, depressive symptoms, sleep patterns, self-rated health, and fear of falls, and the Wilcoxon signed rank tests for exercise frequency and fall counts. Statistical significance was set at p < .05.

Results

Participant Characteristics

Table 1 shows that participants were, on average, 74 (SD = 9.7) years and had four chronic illnesses, 71% women, and 82% Black or Hispanic; 50% lived alone; and 86% reported income ≤US$35,000. The baseline FRQ score (M = 9.0, SD = 3.5) showed high fall risk (i.e., ≥4), and 71% reported any fall and 50% reported 2+ falls in the past year.

Table 1.

Participant Characteristics at Baseline (N = 28).

Variable	Baseline
Age, M (SD)	73.5 (9.7)
Gender, n (%)
Female	20 (71.4)
Male	8 (28.6)
Race/ethnicity, n (%)
Non-Hispanic White	5 (17.9)
Black/African American	14 (50.0)
Hispanic	9 (32.1)
Income (in US$), n (%)
Up to $35,000	24 (85.7)
$35,001–$50,000	2 (7.1)
Refuse	2 (7.1)
Living alone, n (%)	14 (50)
No. of medical conditionsa, M (SD)	4.0 (1.9)
Fall risk (FRQb), M (SD)	9.0 (3.5)
No. of times fallen last year, n (%)
0	8 (28.6)
1	6 (21.4)
2	7 (25.0)
3	5 (17.9)
4	1 (3.6)
12	1 (3.6)

^aArthritis, hypertension, diabetes, lung, heart disease, stroke, cancer, kidney, liver, incontinence, and hypotension.

^bFRQ ≥ 4 indicates high fall risk.

Acceptability of the Overall Program and KOKU

Table 2 shows TEI score of 73 (SD = 4.8) of the maximum possible 77. Participants also rated the KOKU app high on ease of use, usefulness, attitudes toward using, intention to use, and system usability; and 61% reported that they used it at least once in the past week. A majority of participants who did not use KOKU reported walking to take advantage of nice seasonal weather (and planned to use KOKU more when the Texas heat wave prevented them from going outside).

Table 2.

Acceptability of the Overall Program and the KOKU App (N = 28).

Variable	6 week
Overall program acceptability (TEIa), M (SD)	72.7 (4.8)
KOKU acceptabilityb
Perceived Ease of Use, M (SD)	24.5 (4.7)
Perceived Usefulness, M (SD)	25.9 (2.5)
Attitude Toward Using, M (SD)	20.4 (1.9)
Intention to Use, M (SD)	19.6 (2.2)
System Usability Scale, M (SD)	43.9 (6.1)
Days used KOKU in the past week, n (%)
0	11 (39.3)
1	2 (7.1)
2	7 (25.0)
3	6 (21.4)
7	2 (7.1)

Note. KOKU = Keep on Keep up; TEI = Treatment Evolution Inventory.

^aMaximum possible score is 77 for TEI. Higher scores indicate greater acceptance.

^bMaximum possible scores are 28 for Perceived Ease of Use, Perceived Usefulness, and Attitude Toward Using; 21 for Intention to Use; and 50 for System Usability Scale. Higher scores indicate more positive ratings.

Perceived Helpfulness: Qualitative Feedback

Psychoeducation.

All participants stated that psychoeducation was useful in raising awareness of the importance of prevention.

KOKU.

All participants reported a positive response to KOKU. The most common themes were its structure (e.g., progressive intensity level), convenience, and the enjoyment and usefulness of the avatar coach (who looked like an older man and spoke with British accent). A woman (76 years) said,

I didn’t have to leave my home, I can do it right here comfortably in my home and not feel as awkward as I do at the gym. Also, if I get tired during a workout I can rest comfortably … Exercises gradually increased in difficulty and I was excited to see what new exercises were going to be given to me.

Another woman (75 years) said,

The tablet was better than trying to do exercises that are on a paper. My PT gave me exercises to do, but I’d forget how to do them because I only had a paper to look at. The little guy gave me a constant reminder to what I needed to do and how to do it. I felt that it was good to have a visual how to do the exercise and it was easy to use.

Participants showed affection toward the avatar, calling him with an endearing name. A man (52 years) said,

I like the little cartoon man. That man, he motivates me every time he comes out and blows that whistle. I’d call him Mr. Bob, the workout man. I look forward to him every time I turn on the program, to start my session and wait for him to blow that whistle.

A woman (74 years) said,

The avatar needs a shave [with a chuckle], but I liked it. He talks the whole time and repeats and repeats how to do the exercise. It’s good support because sometimes you might forget how to do the exercise and he keeps repeating to help you remember.

Another woman (68 years) said, “Who would have known that a little man with a British accent would get me to exercise?” In fact, some liked it so much that they complained that KOKU’s 30 maximum allowed repetitions were too low and recommended them to be increased to 50 or 100 and to also include more high-intensity exercises.

Home-safety checks.

The common themes were the helpfulness of the safety-related knowledge and the pride in their ability to develop and implement safety plans. A few participants did not allow their FP coach to enter a particular room in their home citing privacy concerns or embarrassment about clutter; however, all of them really liked walking through all or part of their home and checking safety. A man (89 years) said,

That was a very good and useful element. I still have a lot of things to work on and have a to-do list of the things I need to correct for my safety. It made me to become more aware of the importance of environmental safety.

Another man (84 years) said, “I have changed my behavior and environment. I use bathmat with rubber backing now, freed up some space by moving furniture, and taped the rug down.”

The participant whose sessions were all via telephone said,

Because of Covid-19 we did the best we could, we talked on the phone and described. She [her FP coach] took me on the phone to every room … kitchen, bedroom, bathroom … She asked me to describe the rug in front of the bath. I’m very aware now that I was tripping with the edges of the rug. There were other tripping hazards in my house that I didn’t think of as risks. I used carpet tape to glue bathroom rug to the tiles and it is working great. Four weeks of talking to that angel once a week completely changed the way I look at things. We made the best of the best.

Practice of safe ambulation, mobility aid use, showering/bathing, and transfer.

The most common themes were related to more caution and taking preventive measures. Many stated that they now have their canes/walker nearby and use them more, wear safer footwear, and use safer ways to get in/out of shower as they practiced with their FP coaches.

Medication review and healthy eating habits.

All participants appreciated that their FP coaches helped them fill out the medication review form and made referrals to the local AAA’s medication review program. However, they did not express enthusiastic approval for medication review compared with other intervention components, due possibly to not having received the review results at the time of 6-week follow-up. They were pleased to have a colorful chart about healthy eating habits and the diet-related discussion.

Overall evaluation of the program and suggestions for improvement.

All participants stated that the program was very beneficial, and they enjoyed working with their FP coach. A 70-year-old woman summed it up:

As the old saying goes, “two heads are better than one.” With somebody telling you and pointing out at things helped me a lot. I have an alarm button, but if you can do anything to keep from falling, that is better.

A woman (75 years) said,

The program was real good. I enjoyed it and learned a lot. It was good for me to do different exercises than what I already knew. My back and knee are not hurting as much, and I am more active now! I appreciated home safety observations and made all those suggested changes.

Many also mentioned the helpfulness of the follow-up telephone calls for keeping their motivation up and allowing more opportunities to talk with their coaches.

All participants who continued with their exercise plans reported that they themselves and their family members noticed and were delighted by improvement in their physical functioning—getting up from a chair with ease, climbing the stairs better, going on longer distances on walks, improved balance, decreased muscle stiffness, having more energy, having less bodily pain (“my pain went away”), breathing better, and resting and sleeping better. They invariably reported exercises made them “feel good,” “feel better mentally,” and “being less foggy and more positive.”

Almost all improvement suggestions were about KOKU. One 84-year-old man with vision problems reported that the iPad screen for KOKU was too small for him. One participant also suggested putting a message of congratulations at the end of daily KOKU exercise routine to motivate people who may not hear such encouragement often. A couple of participants also stated that the introductions and safety warnings before actually letting people to start each exercise were a bit too much, which may not be helpful for people who are not very motivated about exercising. A few also experienced frustration as KOKU froze, although they understood the technological problems may not be completely avoided. Some also suggested a longer-term intervention because they enjoyed interacting with their FP coaches.

Preliminary Efficacy: Changes Between Baseline and 12 Weeks

Table 3 shows statistically significant improvement between baseline and 12 weeks in exercise frequency and efficacy, sleep quality, and self-rated physical health and significant reduction in disability (WHODAS 2.0) and sleep medication use frequency. At baseline, 39% of the participants reported no exercise; at 12 weeks, the percentage fell to 11% (n = 3; they reported eye infection, flared up Crohn’s disease, and other illnesses as barriers). The proportions of those who exercised 3+ days increased from 43% to 57%. However, there was no statistically significant changes in ADL/IADL impairments, depressive symptoms, sleep hours, self-rated mental health, and fear of falling, and fall counts, although some scores showed improvement. For example, two of three participants who reported three falls in 3 months prior to baseline reported no fall and one reported one fall at 12 weeks.

Table 3.

Preliminary Efficacy (N = 28).

Variable	Baseline	12 week	t/Z value	p valuea
Exercise efficacy (OEESb), M (SD)	37.6 (6.3)	40.6 (3.5)	−2.624	.014
Exercise frequency (days in a week), n (%)			−2.363	.018
0	11 (39.3)	3 (10.7)
1–2	5 (17.9)	9 (32.1)
3–7	12 (42.9)	16 (57.1)
Disability (WHODAS 2.0c), M (SD)	11.2 (6.7)	9.5 (6.4)	2.346	.027
No. of ADL/IADL impairmentsc (0–12), M (SD)	2.5 (2.0)	2.3 (2.3)	0.529	.601
Depressive symptoms (PHQ-9c), M (SD)	5.4 (4.2)	4.5 (3.9)	1.015	.319
No. of hours of sleep, M (SD)	6.5 (1.7)	6.7 (1.5)	−0.463	.647
Self-ratings of sleep qualityb (1–4), M (SD)	2.9 (0.7)	3.4 (0.6)	−3.300	.003
Frequency of sleep medication use (0–3)d, M (SD)	1.3 (1.3)	0.7 (1.1)	2.566	.016
Self-ratings of physical healthb (1–5), M (SD)	3.0 (0.8)	3.4 (1.0)	−2.073	.048
Self-ratings of mental healthb (1–5), M (SD)	3.4 (0.9)	3.5 (1.0)	−0.626	.537
Fear of falling (FES-Ic), M (SD)	34.6 (9.4)	35.6 (12.0)	−0.584	.564
Fall counts in the past 3 months, n (%)			−0.504	.614
0	20 (71.4)	19 (67.9)
1	5 (17.9)	7 (25.0)
2	0 (0)	2 (7.1)
3	3 (10.7)	0 (0)

Note. OEES = Outcome Expectations for Exercise Scale; WHODAS = World Health Organization Disability Assessment Schedule; ADL = activities of daily living; IADL = instrumental activities of daily living; PHQ = Patient Health Questionnaire.

^aThe p values are from paired-sample t tests, except for exercise frequency and fall counts in the past 3 months, which are from the Wilcoxon signed rank tests.

^bHigher OEES scores and self-ratings of sleep quality, physical health, and mental health indicate more positive ratings.

^cHigher WHODAS 2.0, ADL/IADL impairments, PHQ-9, and FES-I scores indicate greater problem severity.

^d0 = not during the past month; 1 = less than once; 2 = 1–2 times a week; and 3 = 3+ times a week.

Discussion

In this article, we report acceptability of a brief FP program for low-income homebound older adults that was delivered by trained lay FP coaches and included a tablet-based, gamified exercise app. Participants were highly receptive to the overall program and the KOKU app. They valued all safety-related knowledge, home-safety checks, and training in safe ambulation and transfer. Their qualitative feedback also showed that participants took pride in their improved FP knowledge and capacity to take steps to reduce fall risk.

As in previous pilot studies of tablet- or mobile phone–based exercise programs for older adults (Arkkukangas et al., 2020; Taylor et al., 2020), our study participants liked the KOKU app for its convenience and structure. It provided easy accessibility to and availability of exercise routines. KOKU was also a great source of motivation as they enjoyed the gamified program with a relatable (in terms of age and body shape) avatar who encouraged and supported them to exercise. All of them anthropomorphized the avatar as if it were their personal trainer. They felt accomplished/empowered when performing the exercises on their own at home.

Interestingly, however, their KOKU approval and enjoyment were not translated into consistent use, as 39% reported not using KOKU at 6-week follow-up. For some participants, it appeared that KOKU-based exercising provided confidence to venture out to have walks in the neighborhood during the relatively warm Texas winter and nice spring. Other participants may have needed more than 4 weeks of intervention for keeping up their motivation, habit formation, and adherence (Lally & Gardner, 2013). Common barriers to adhering to exercise programs among older adults are a lack of motivation and poor health (Arkkukangas et al., 2018). A reminder to do exercise, some kind of prompts in the app, may facilitate their motivation and adherence (Arkkukangas et al., 2020).

With or without KOKU, the significant increase in exercise frequency since baseline is notable. For many participants, the immediate physical and mental health benefits and the sense of accomplishment/empowerment that they experienced from exercising motivated them to keep exercising. Their family and/or friends also noticed the changes and provided support; however, poor health (acute illness and flaring up of chronic illnesses) was a barrier to exercising for a few people. In the relatively short feasibility study, we could not measure long-term adherence to exercising routines. Measures of and strategies to improve adherence are needed in future studies.

This feasibility study’s primary aim was not to evaluate the intervention efficacy. However, it is worth noting that despite the small sample size, significant changes between baseline and 12 weeks in exercise frequency and efficacy, disability, sleep quality, and self-rated physical health hold promise for effectiveness of the intervention with a sufficiently powered sample. Lack of a significant reduction in falls was likely due to the short observation period. Lack of reduction in fear of falling was likely due to increased fall risk awareness, as many talked about how the program reminded them to be more cautious. A meta-analysis found that exercise interventions to reduce fear of falls in community-dwelling older people immediately following the interventions had limited effectiveness, and whether they subsequently reduce fear of falls or increase physical activity is not known (Kendrick et al., 2014). On the contrary, Stanmore et al. (2019) found residents (aged 55+) of U.K. assisted living facilities who participated in a 12-week exergame intervention had reduced fear of falling compared with those who received standard care (physiotherapy advice and leaflet). Boongird et al. (2017) also reported that a 12-month, simple home-based exercise program reduced fear of falling.

In sum, a brief, lay-coach-delivered intervention including the KOKU app is acceptable to low-income homebound older adults. Because the KOKU app was developed to encourage exercising without in-person coaching, it was an optimal tool that older adults could use during the Covid-19 lockdown. Although in-home sessions may be desirable in ensuring safety of some older adults, telephone sessions went well, too. Based on our telephone delivery experience under Covid-19, we have developed a videoconferencing delivery (including use of video training clips) of all intervention components. A recent pilot study of a telehealth-delivered Otago Exercise Program for low-income older adults reported a high level of satisfaction among participants (VanRavenstein et al., 2020). In future programming, we also plan to do in-house medication review (by a PharmD student trained in medication review for older adults with multi-morbidity) to have the results available for discussion during telephone check-in calls.

This study had some limitations. The sample size was small, the follow-up assessments were short-term, and lacked a validated measure for exercise type and frequency. It is necessary to evaluate the intervention’s effects on falls and fall worries in a future randomized controlled trial with a large sample and longer-term follow-ups using well-established outcome measures. The study reported here provides a strong justification for such a trial to evaluate real-world effectiveness of lay-coach delivery and use of a tablet-based exercise program. The study participants’ high receptivity to the intervention suggests that it has a high probability of adoption in aging-service agencies once effectiveness is established.