Impact of the FindMyApps program on people with mild cognitive impairment or dementia and their caregivers; an exploratory pilot randomised controlled trial

Abstract

Purpose

FindMyApps is a web-based selection-tool and errorless learning training program to help people with mild dementia/Mild Cognitive Impairment (MCI) and caregivers find user-friendly apps. In preparation of a definitive trial, the impact and feasibility of the FindMyApps intervention on self-management and engagement in meaningful activities, including social participation, was explored.

Materials and methods

An exploratory pilot randomised controlled trial (trial registration approval number: NL7210) with pre/post measurements was conducted with community-dwelling people with mild dementia/MCI and their caregivers (n = 59) in the Netherlands. Dyads in the experimental group (n = 28) received training to use the tablet and FindMyApps, and the errorless learning method was taught to their caregivers. Control group dyads (n = 31) received a tablet (without FindMyApps), instructions on tablet-use and links to dementia-friendly apps. Validated instruments were used to assess person with dementia’s self-management, meaningful activities and social participation, caregiver’s sense of competence and both their quality of life.

Results and conclusions

No statistical significant group differences on the outcomes were found. Small to moderate effect-sizes in favour of the FindMyApps group were found for self-management and social participation. Caregivers tended to have more positive care experiences. Subgroup analyses showed that people older than 70 benefitted more from FindMyApps regarding self-management and higher educated people benefitted more regarding social participation. FindMyApps is feasible for the target group and may have potential to improve self-management and social participation. For a future definitive effectiveness trial a larger sample size is recommended, as well as taking into account the possible impact of education and age.

IMPLICATIONS FOR REHABILITATION

• The FindMyApps program is feasible for people with dementia/MCI and may have potential to improve their self-management and social participation.

• Informal caregivers may develop a more positive care experience when utilizing FindMyApps to find suitable apps for self-management and social participation for the person with dementia/MCI.

Keywords:

• Dementia
• touch screen technology
• psychosocial interventions
• self-management
• social participation
• mild cognitive impairment

Introduction

Dementia is one of the major causes of dependency and disability in older people worldwide [1]. Because of the increasing number of people living with dementia, it is becoming more difficult to meet their need for care and support, particularly in people with early-stage dementia, as dementia care is currently focussed on the later stages of the disease [2]. Additionally, there is a focus on medical treatment, whereas the equally important counselling and educational aspects of dementia care are often overlooked [3,4]. This is unfortunate, because especially during the early stages of dementia, coping with the diagnosis of dementia and its immediate consequences can be very difficult [5]. This challenge calls for effective solutions. The use of technology is seen as one of the tools to improve care for people with early-stage dementia [6–8].

Self-management and social participation are two domains of social health in which people with dementia most frequently experience unmet needs [9–11]. Many applications (apps) have been developed to support people in managing their daily life, engaging in activities, and staying in touch with their network. Apps could support people with dementia for the same purposes [12–16]. This includes apps to support memory and structure in daily life [17], apps to play games for cognitive stimulation [18] or as a pleasant and meaningful activity [19], and apps to view art to improve people’s wellbeing [20]. Whilst many potentially useful apps already exist, it may be difficult for people with dementia to find apps that suit their interests, needs and abilities [13,14,21]. Additionally, not all people with dementia are familiar with the use of tablet computers (tablets) and need support to learn how to use touch screen devices [15,16,19]. Therefore, a tool to select suitable and usable apps and to support people with early-stage dementia in using the tablet and apps would be of important added value.

The FindMyApps intervention, consisting of a training session and personalised selection-tool, has been designed in collaboration with people with dementia and their caregivers [22,23]. FindMyApps aids users in operating a tablet and selecting apps that match their interests, needs and abilities. Specifically, users are supported to find apps for self-management and social participation. It is expected that the quality of life of people with dementia will improve if they are enabled to engage in social and meaningful activities and perform activities of daily living (ADLs) and as a consequence may experience greater self-efficacy and autonomy. This may benefit caregivers as well: effective support for self-management and social participation for the person with dementia may lead to a more positive experience of caregiving for the caregiver [24]. This could lead to an increased sense of competence and better quality of life for caregivers.

The FindMyApps program was developed in accordance with the Medical Research Council (MRC) framework for developing and evaluating complex interventions [25,26]. First, two studies were conducted to develop the FindMyApps program by means of user-participatory development [22,23]. Next, a first pilot study evaluated the feasibility of the program for people with dementia [27]. The implementation of the FindMyApps program and mechanisms of impact were investigated in a process evaluation in a second pilot trial [28], in which we also explored the effectiveness of the updated version of FindMyApps. In this paper we will report on this explorative effect study. In addition we evaluated the feasibility of the trial procedures in preparation for a definitive effectiveness trial.

The following research questions were addressed:

1. Does FindMyApps after 3 months of use have potential impact compared to usual tablet use with regard to:

   1. The person with dementia/MCI’s ability to self-manage, engagement in activities and social participation, self-efficacy, experienced autonomy and quality of life?

   2. Their caregivers’ care experience, sense of competence and quality of life?

2. Are there any issues regarding recruitment, study attrition, sample size, and the administration time of outcome measures and procedures?

3. How many participants will be needed in a definitive trial to ensure sufficient statistical power to demonstrate moderate to large clinical differences in the outcome measures?

In addition, we explored to what extent age, level of education, and tablet experience prior to the intervention are associated with our outcome measures. These findings would provide us with valuable insight into the target group most suited for the FindMyApps intervention for a definitive trial.

Materials and methods

Design

For this explorative study a two-arm, non-blinded, randomised controlled trial design was applied. Participant dyads consisted of a person diagnosed with mild dementia [29] or Mild Cognitive Impairment (MCI) [30] and their informal caregiver. Dyads were randomly assigned to the experimental or the control group, stratified by diagnosis (dementia/MCI) and whether or not they cohabited. Data for this study regarding the person with dementia/MCI’s self-management, social participation, self-efficacy, experienced autonomy and quality of life were collected through standardised questionnaires with the participants at baseline (T0) and after three months intervention (T1). From caregivers, data were collected on their sense of competence, positive care experience and quality of life.

The study was approved by the medical ethical committee of the VU University medical centre (VUmc) in Amsterdam, The Netherlands, who declared this study exempt from the Medical Research Involving Human Subjects Act (registration number: 2017.401). The trial was registered and approved in the Netherlands Trial Register (NL7210; https://www.trialregister.nl/trial/7210). In our reporting we followed the CONSORT guidance.

Participants and setting

Participants included people with mild dementia with a Mini Mental State Examination score (MMSE) [31] of 18–24 or a Global Deterioration Scale score (GDS) [32] of 3–4, or MCI (diagnosed by a neurologist at a memory clinic, MMSE ≥25, GDS 2–3) and their caregivers. MCI is a syndrome defined as “cognitive decline greater than expected for an individual’s age and education level but that does not interfere notably with activities of daily life” [33]. More than half of the people with MCI progress to dementia within 5 years, which is why both people with dementia and MCI were included in this study.

Participant dyads were recruited in the Netherlands from September 2018 until January 2020. We intended to recruit a total of 80 dyads (40 experimental/40 control) to be able to show moderate to large effect sizes (d = 0.75; power 0.80, alpha = 0.025, dropout 15%). Participants were recruited through various healthcare institutions involved in the care for people with dementia/MCI, such as memory clinics, and through Meeting Centres for people with dementia and caregivers, Alzheimer cafes, and public announcements. Dyads were included if the person with dementia/MCI cohabited with their caregiver, or when the caregiver visited the person with dementia/MCI at least twice a week. Additionally, access to Wi-Fi was required. Participants were excluded if they had moderate to severe dementia (MMSE < 18; GDS5-7), severe eyesight problems or if they participated in other interventional research studies that were expected to influence the outcomes measured in the FindMyApps study.

Intervention

Experimental group

Dyads in the experimental group received the FindMyApps intervention, consisting of the FindMyApps selection tool and an introductory training session in how to use the tablet and the selection tool to find apps that match their personal needs and interests. The introductory training session lasted approximately 1–1.5 h. This included training of the caregiver in how to guide the person with dementia/MCI during the three months intervention period. A telephone and email helpdesk was available for consultations during the intervention period.

FindMyApps and caregiver training

In the FindMyApps training session both the person with dementia/MCI and the caregiver received instructions on the use of a tablet and the FindMyApps selection tool. Participants also received a written manual containing the information covered during the training session. Training included basic functions such as charging the tablet and more complex functions such as making folders for downloaded apps.

During the training, caregivers were additionally taught how to support the person with dementia/MCI in using the tablet and FindMyApps, so that they could continue to provide training for the person with dementia/MCI throughout the intervention period. This is a vital part of the intervention, as proper coaching and support are essential for older adults to adopt technology [34]. The training of the person with dementia was based on the errorless learning method, which has been used effectively with people with different types of dementia [35,36]. Caregivers received a brief explanation of how to use the errorless learning approach to support the person with dementia/MCI learning to use the tablet. This included instructions for how to implement an errorless learning approach in practice, such as using a stepwise approach, encouraging participants not to make guesses about how to perform an activity, and repeating steps frequently while preventing mistakes [37].

FindMyApps selection tool

The FindMyApps selection tool is a web application that can be used on tablets with either the Android or iOS operating system. Using the web application, users have access to a database containing pre-selected apps (an “app library”). The apps in the FindMyApps library focus on self-management and meaningful activities, including social activities, and are all assessed on dementia-friendliness. The criteria used to include apps in this library were based on research conducted with people with dementia and their caregivers [14,22,23]. The FindMyApps selection tool is intended to be used by persons with early-stage dementia with support of their caregiver [23].

Participants created a user profile in which they specified their preferences for features of apps (Figure 1(A)). The preferences offered were: large font size; less text, many pictures; only Dutch language apps; real photos (instead of drawings or animations); simple to operate (with respect to gesture-control); and the comprehensiveness of instructions offered, via a help function for example. The FindMyApps database is divided into three main categories in which users can look for apps: “in and around the house,” “contacts” and “leisure time” (Figure 1(B), which refer to self-management, social participation and meaningful activities, respectively. These categories are divided into subcategories which show several apps (Figure 1(C)). Every app has a description and shows by means of a score how that app matches the pre-set user preferences (Figure 1(D)). The app descriptions include a link to the Apple App Store (iOS) or the Google Play Store (Android) depending on the operating system of the tablet. Every page within the FindMyApps app shows an explanation button in case help is needed. A personalised “MyApps” page showed an overview of apps that had been opened by the user in the Apple App Store or Google Play Store (Figure 1(E)). A demonstration video on FindMyApps covering the functions of the tablet and FindMyApps was available on the tablet and remained available after the introductory training session, for the full duration of the intervention period (Figure 1(F)).

Figure 1. Different pages of the FindMyApps selection tool. (A) Personal Settings (B) “Discover” page with app categories “In and round the house,” “Contacts” and “Leisure time.” (C) List of apps within a category (D) App description page (E) “My Apps” page (F) “Training” page with instruction videos.

Control group

In the control group participants received general tablet training (of approximately one hour) and did not receive FindMyApps. Dyads were shown a video only demonstrating the functions of the tablet, which was viewable on the FindMyApps project website, for the full duration of the intervention period. In addition, participants were given a written manual with the information shared during the tablet training, which also included a list of references/links to websites that suggested useful apps for people with dementia/MCI. A telephone and email helpdesk was also available to the control group.

Instruments

Individuals with dementia/MCI

If participants received a dementia/MCI diagnosis more than three months before inclusion of the study, the Brief Cognitive Rating Scale (BCRS) [32] was administered with the caregiver to determine the severity of cognitive decline and to confirm that participants fulfilled the inclusion criteria. The Guidelines for the Rating of Awareness Deficits (GRAD) [38] was used to measure the self-awareness of cognitive deficits, at baseline. Participant characteristics based on The Older Persons and Informal Caregivers Survey Minimum Dataset (TOPICS-MDS) [39] were assessed at baseline by the caregiver.

Data on the outcome measures were collected at baseline (T0) and after three months (T1). A face-to-face interview with the person with dementia/MCI took place in their home using the following standardised questionnaires:

Self-management was measured by the Self-Management Ability Scale – Short Version (SMAS-S), which is a reliable instrument (18 items, α = 0.78) [40]. The SMAS-S is divided into six subscales: (1) taking initiative (self-motivated in improving wellbeing; (2) investment behaviour (for long-term benefits); (3) variety (maintain and gain various resources for wellbeing); (4) multi-functionality (maintain and gain resources to serve multiple dimensions of wellbeing; (5) self-efficacy (maintain and gain trust in personal abilities to achieve wellbeing; and (6) to gain a positive frame of mind [40]. The subscales are based on the idea that these self-management abilities are needed to manage external resources to maintain and restore physical and social wellbeing [41].

Social participation was measured by the Adult Social Care Outcomes Toolkit (ASCOT) [42], social activity question, administered to the person with dementia/MCI. In addition, the Maastricht Social Participation Profile (MSPP) [43] was administered with the caregiver. This scale is subdivided into three subscales which all have acceptable construct validity (26 items, α > 0.60): (1) consumptive participation (benefiting from society by e.g. visiting a restaurant; (2) formal social participation (contributing to society by e.g. volunteer work); and (3) informal social participation (benefiting from or contributing to society, or both by contact with friends, family or acquaintances).

The self-reported Dutch General Self-efficacy Scale [44] was used to assess self-efficacy in a reliable manner (10 items, α = 0.85) [45]. Experienced autonomy was measured by the 12-item Experienced Autonomy questionnaire [46]. Reliability and validity of this instrument have not been investigated previously.

Quality of life was measured by the 30-item Dementia Quality of Life Instrument (DQoL), with an internal-consistency reliability ranging from 0.67 to 0.89 in people with mild to moderate dementia [47].

Primary caregiver

The caregivers received an email with questionnaires to fill out online.

For sense of competence the 7-item Short Sense of Competence Scale (SSCQ, α = 0.76) [48] was used. The Positive Experience Scale (PES, α = 0.74) [49] was administered to measure positive care experiences, and for quality of life of the informal caregiver the one-item on quality of life from the TOPICS-MDS-informal carer version [39] was administered.

Procedure

Interested dyads were informed in person or by telephone by one of the researchers. They all received written study information by email or by post. All participant dyads in the trial provided written informed consent. To check the eligibility of a participant a telephone interview was conducted with the caregiver, in which data were collected concerning the diagnosis, severity of dementia/MCI (if necessary the BCRS), background characteristics and the GRAD. If the participant was eligible, further background characteristics of the person with dementia/MCI and caregiver and baseline measurements for the caregiver were collected by questionnaires, sent by email to the caregiver (via Castor Electronic Data Capture (EDC)). Additionally, an appointment was then set for the training in the use of the tablet and FindMyApps and the baseline measurement at the home of the person with dementia/MCI. Dyads were offered the possibility to borrow a tablet during the intervention period if they did not own one yet, or if they had a tablet that was too old to use FindMyApps. This first home visit was done by trained interviewers (junior researchers, bachelor and master students neuro- and clinical psychology and brain and cognitive science, and nursing students). During the intervention period participants were free to use the tablet and/or FindMyApps in the frequency they preferred. Participant dyads received follow-up phone calls after two and after six weeks to support them in using the tablet and provide help if there were any technical problems or other difficulties preventing them from using the tablet. Participants could also contact the helpdesk themselves by phone or email. After the three month intervention period the same questionnaires were sent by email to the caregiver and a second appointment was made to administer the questionnaires face-to-face with the person with dementia/MCI. However, due to the coronavirus disease 2019 (COVID-19) pandemic, the post-test questionnaires with the final six persons with dementia/MCI that were administered after the end of March 2020 were performed by telephone with the presence of the informal caregiver.

Data analysis

The background characteristics of participants were summarised using descriptive statistics (means, standard deviations). Differences in background characteristics between experimental and control groups were tested by using t-tests for interval variables and Pearson Chi-Square tests for nominal or categorical variables. Descriptive statistics were also computed for the outcomes of the standardised questionnaires. Outcome variables on self-management, engagement in activities and social participation, self-efficacy, experienced autonomy and quality of life of the person with dementia/MCI and their caregivers’ care experience, sense of competence and quality of life at post-test were compared between the experimental and control group using analyses of covariance (ANCOVAs), adjusting for GDS. Both completers-only and intention-to-treat analyses were conducted. For all outcome measures, effect sizes (partial eta squared; η_p²) were calculated.

The following post-hoc exploratory analyses were performed: Using classification factors, we investigated whether age (≤70 versus >70 years old, based on the median), education (primary, secondary and further vocational education versus higher vocational and academic education, also based on the median) or tablet use before the intervention started (they did use a tablet versus they did never) were associated with the outcome variable. For variables that showed at least small effect sizes with improved outcomes in the experimental group, the possible influence of these classification factors was investigated by running two-way ANCOVAs. Statistical analyses were performed in SPSS24.0.

With regard to the feasibility of the intervention and RCT issues regarding recruitment, study attrition, sample size, and the administration time of outcome measures and procedures were inventoried and based on the calculated effect sizes on the outcomes a power analysis was conducted to determine how many participants would be needed in a future RCT.

Results

Study population

Fifty-nine dyads were initially included as eligible for the study and randomised into the experimental group (n = 28) and the control group (n = 31) (Figure 2). However, before the baseline measurement took place five participants (one person with dementia/MCI in the experimental group; three people with dementia/MCI and one caregiver in the control group) dropped out (Figure 2) resulting in 27 persons with dementia/MCI and 28 caregivers participating in the FindMyApps intervention and 28 persons with dementia/MCI and 30 caregivers in the control group. Attrition during the intervention period was 16.8%, with seven dyads dropping out, one additional person with dementia/MCI and four caregivers. The experimental and control groups did not differ in background characteristics at baseline, such as age and education level (Table 1). However, in the control group 4 participants had Lewy body dementia, whereas in the experimental group none of the participants had this diagnosis.

Figure 2. Flow chart of the participants of this pilot RCT.

Table 1. Background and clinical characteristics of participants of the experimental (exp.) and control (cont.) group in the pilot RCT at baseline.

Persons with dementia/MCI	Exp. Group (n = 28)	Cont. Group (n = 31)	Difference test	p
Sex, n			Chi^2 = 0.336	.602
Female	12	11
Male	16	20
Age, M (SD), [min–max]	72.71 (7.783), [62–92]	71.74 (9.644), [51–86]	t = − .432	.674
Type of dementia, n			Chi^2 = 5.177	.395
Alzheimer’s disease	17	16
Frontotemporal	3	1
Vascular dementia	3	4
Lewy body dementia	0	4
Other	2	2
MCI, n	3	4
Severity of cognitive impairment (BCRS/GDS) [1–7]			Chi^2 = 1.070	.586
2. Very mild cognitive decline	4	2
3. Mild cognitive impairment (MCI)	19	22
4. Mild dementia	5	7
Awareness (GRAD score), n			Chi^2 =2.596	.458
Intact	7	11
Mildly impaired	9	6
Moderately impaired	0	1
Severely impaired	1	2
Not administered	11	11
Education level, n			Chi^2 = 4.003	.676
Primary education	–	1
Secondary education (vocational)	8	6
Secondary education (academic)	1	1
Further education (vocational)	7	9
Higher education (vocational)	8	12
Higher education (academic)	4	2
Use of tablet at start of the study, n			Chi^2 = 2.876	.411
Every day	6	9
Once a week	4	3
Once or twice every month	4	1
Once in my life	1	2
Never used	13	16
Caregivers	Exp. Group (n = 28)	Cont. Group (n = 31)	Difference test	p
Sex, n			Chi^2 = 1.146	.284
Female	20	18
Male	8	13
Age, M(SD), [min–max]	65.61 (10.196) [38–81]	68.03 (11.675) [41–85]	t = .846	.401
Relationship with person with dementia, n			Chi^2 = 1.170	.557
Partner	23	27
Child	4	4
Professional caregiver	1	0
Education level, n			Chi^2 = 3.191	.671
Primary education	–	–
Secondary education (vocational)	6	6
Secondary education (academic)	5	3
Further education (vocational)	5	6
Higher education (vocational)	7	13
Higher education (academic)	5	3
Living situation, n			Chi^2 = 2.102	.551
Cohabiting with the person with dementia/MCI	18	19
Living alone	2	1
Living together with partner and possibly children	7	11
Partly with parent, partly somewhere else	1	–
Use of tablet at start of study, n			Chi^2 = 5.301	.151
Every day	12	17
Once a week	7	2
Once in my life	1	–
Never used	8	12

Note. Between-group differences were tested using Pearson Chi-square test for categorical/nominal variables and t-tests for interval variables. MCI: Mild Cognitive Impairment; BCRS: Brief Cognitive Rating Scale; GDS: Global Deterioration Scale; GRAD: Guidelines Rating of Awareness in Dementia.

Results outcome measures

A data entry check of 10% of the data showed an error rate below 1%, which were all corrected. From the maximum of 10,904 responses (47 dyads, 232 items), only 4 responses (0.03%) were missing. The missing items corresponded to two participants, with one missing 1 item and one missing 3 items. It was concluded that these responses were missing completely at random. The values were imputed using the Expectation-Maximization algorithm [50]. Additionally, for one participant all data relating to the DQOL instrument was missing, because this participant was unable to complete the questionnaire. The data was checked for outliers, none were more than three standard deviations from the mean. Because completers-only and intention-to-treat analyses showed similar results, only the completers-only outcomes are reported here.

After 3 months (T1), no significant differences on measures of self-management, social participation and engagement in activities, self-efficacy, experienced autonomy or quality of life of the person with dementia/MCI were observed between the experimental and control group (Table 2). Additionally, no significant differences were observed for the caregivers’ care experience, sense of competence or quality of life. However, non-significant effect sizes on the different outcome measures were also examined to trace tendencies, which could be useful information for a definitive trial to be conducted in the future. According to Cohen (1988; 49) η_p² = .01 indicates a small effect size; η_p² = .06 a medium effect size and η_p² = .14 a large effect size.

Table 2. Numbers of outcome measures at baseline (T0) and post-test (T1) and results of ANCOVAs for persons with dementia/MCI and their caregivers in the experimental (E) and control group (C) from completers-only analysis with baseline score and GDS as covariate.

	Baseline (T0)		Post-test (T1)		Adjusted means
	mE (SD) (n = 25)	mC (SD) (n = 22)	mE (SD) (n = 25)	mC (SD) (n = 22)	mE	mC	F	p	ηp^2
Outcome measures – person with dementia/MCI [range]
Self-management
SMAS-S, [0–15]
Taking initiative	7.8 (2.3)	9.7 (2.1)	8.7 (2.7)	9.0 (2.4)	9.1	8.6	.61	.44	.01
Investment behaviour	9.6 (2.7)	10.2 (2.4)	9.5 (2.2)	9.9 (2.5)	9.6	9.7	.04	.84	.00
Variety	7.8 (2.5)	9.5 (2.4)	7.6 (2.9)	8.4 (2.3)	7.9	8.1	.11	.75	.00
Multi-functionality	8.4 (1.8)	9.0 (1.3)	8.6 (2.6)	8.2 (1.9)	8.7	8.1	1.05	.31	.02
Self-efficacy	9.0 (2.0)	9.4 (2.4)	9.1 (2.4)	9.0 (1.7)	9.2	8.9	.22	.64	.00
Positive frame of mind	8.5 (2.6)	8.2 (1.9)	8.4 (3.4)	8.4 (2.7)	8.3	8.4	.00	.96	.00
Total, [0–90]	43.3 (7.6)	46.5 (7.0)	44.2 (10.2)	44.5 (7.4)	45.0	43.5	.37	.55	.01
Social participation
ASCOT^a, [1–4]	2.1 (.8)	1.8 (.7)	1.8 (.8)	1.7 (.6)	1.8	1.8	.01	.92	.00
MSPP, [0–3]
Consumptive participation	.65 (.36)	.67 (.47)	.66 (.41)	.71 (.43)	.7	.7	.16	.70	.00
Formal social participation	.45 (.49)	.35 (.40)	.57 (.52)	.42 (.43)	.5	.5	.45	.51	.01
Informal social participation (acquaintances)	.95 (.62)	1.18 (.44)	.92 (.47)	1.13 (.39)	1.0	1.1	1.34	.25	.03
Informal social participation (family)	.89 (.67)	1.04 (.49)	.84 (.47)	1.03 (.35)	.9	1.0	1.49	.23	.03
Self-efficacy
Dutch GSE, [10–40]	31.0 (6.5)	30.7 (4.6)	28.7 (6.9)	30.6 (4.6)	28.7	30.7	1.75	.19	.04
Experienced autonomy
Experienced Autonomy questionnaire^a, [12–60]	31.9 (5.6)	32.3 (5.0)	33.1 (4.5)	33.0 (4.0)	33.2	32.8	.14	.71	.00
Quality of life
DQoL, [1–5]
Self-esteem (4 items)	3.7 (.6)	3.6 (.6)	3.6 (.7)	3.5 (.5)	3.5	3.5	.06	.80	.00
Positive affect/humor (6)	3.7 (.6)	3.6 (.5)	3.6 (.5)	3.6 (.5)	3.6	3.6	.13	.72	.00
Negative affect^a (11)	2.4 (.6)	2.7 (.6)	2.4 (.7)	2.4 (.7)	2.5	2.3	1.02	.32	.02
Feelings of belonging (3)	3.9 (.5)	3.7 (.5)	3.8 (.4)	3.7 (.5)	3.8	3.7	.10	.75	.00
Sense of Aesthetics (5)	3.7 (.6)	4.0 (.6)	3.5 (.7)	3.9 (.6)	3.6	3.8	2.11	.15	.05
General quality of life (1)	3.6 (.8)	3.0 (.9)	3.3 (.9)	3.1 (.7)	3.2	3.2	.02	.89	.00
Outcome measures – caregiver [range]
Feeling of competence
SSCQ, [0–7]	4.3 (1.9)	4.4 (1.4)	4.0 (2.1)	4.2 (1.9)	4.1	4.2	.06	.80	.00
Positive care experience
PES, [0–16]	11.6 (3.1)	11.8 (2.8)	11.8 (2.7)	11.0 (3.6)	11.8	10.9	1.70	.20	.04
Quality of life
TOPICS-MDS
General^a, [1–5]	3.1 (.9)	2.9 (.9)	3.2 (.9)	3.1 (.9)	3.2	3.2	.00	.97	.01
Grade, [0–10]	7.0 (1.1)	7.0 (1.3)	7.2 (.9)	7.1 (1.3)	7.1	7.1	.08	.78	.00
General – compared to a year ago^a, [1–5]	3.2 (.8)	3.5 (1.0)	3.2 (.8)	3.1 (.9)	3.3	3.0	1.42	.24	.03

Note. SMAS-S: Self-Management Abilities Scale – Short version; ASCOT: Adult Social Care Outcomes Toolkit; MSPP: Maastricht Social Participation Profile; Dutch GSE: Dutch General Self-Efficacy Scale; DQoL: Dementia Quality of Life; SSCQ: Short Sense of Competence Scale; PES: Positive Experience Scale; TOPICS-MDS: The Older Persons and Informal Caregivers Survey – Minimum DataSet. ^a = lower scores are better, for other measures higher scores are better. Significant numbers are p ≤ .0.05. In bold are the effect sizes that show improved results in the experimental group, η_p² = .01 indicates a small effect size; η_p² = .06 a medium effect size and η_p² = .14 a large effect size [51].

Further exploring the observed differences in the adjusted means between the groups, for the persons with dementia/MCI small effect sizes were observed on the SMAS-S subscales “taking initiative” (η_p² = .01) and “multi-functionality” (η_p² = .02) and the total score (η_p² = .01), with the FindMyApps group improving and the control group declining (Table 2). Another small effect size (η_p² = .01) was observed for the MSPP subscale “formal social participation” with larger improvements in the FindMyApps group than in the control group. On the other hand, the Dutch General Self-Efficacy Scale, MSPP subscale “informal social participation” (for both acquaintances and family), and the subscale “sense of aesthetics” of the DQoL showed small negative effect sizes ranging between −.01 and −.05, suggesting some tendency of decline at post-test compared to baseline in both groups, and sometimes to the detriment of the experimental group. Finally, a small effect size for the subscale “negative affect” of the DQoL was observed (η_p² = .02), with the experimental group remaining the same and the control group tending to improve somewhat (less negative affect).

For the caregiver, a small effect size was found for the PES (η_p² = .04), with some improvement in the experimental group and some decline in the control group. The general quality of life (QoL) of caregivers (TOPICS-MDS) seemed to have improved somewhat (η_p² = .01) in both groups after the three-month intervention period. However, the general quality of life compared to a year ago as measured by the TOPICS-MDS remained the same in the experimental group and declined somewhat in the control group (η_p² = .03).

Secondary subgroup analyses (two-way ANCOVAs) showed a statistically significant interaction (p = .01) between the groups (FindMyApps/usual tablet use) and the level of education of the people with dementia/MCI on the formal social participation of the person with dementia/MCI, whilst controlling for baseline scores of formal social participation, F(1, 41) = 8.72, η_p² = 0.18. Persons with dementia/MCI with higher education improved in the FindMyApps group and declined in the control group, whereas persons with dementia/MCI with lower education improved more in the control group compared to the FindMyApps group. No other statistically significant interaction effects were found (Table 3). Larger improvements were found on the subscale “taking initiative” (SMAS-S) in the FindMyApps group compared to the control group among people with dementia/MCI aged over 70 years old and those higher educated, although this did not reach significance. For the subscale “multi-functionality,” small to medium positive effect sizes (η_p² = .01–.08) were also found in persons with dementia/MCI over 70 years old, higher educated and with prior tablet experience, with some improvements in the FindMyApps group (almost statistically significant for those older than 70) and some declines in the control group. Also for the total score of the SMAS-S, it was found that persons with dementia/MCI in the FindMyApps group who had used a tablet before the intervention showed larger improvements than persons with dementia/MCI who had never used a tablet before.

Table 3. Results of subgroup analyses: two-way ANCOVAs conducted on the outcome measures which showed positive effect sizes ≥.01, with improved outcomes in the experimental group, with age, education and tablet use before the intervention as classification factors.

	n		Baseline (T0)		Post-test (T1)		Adjusted means
	E	C	mE (SD)	mC (SD)	mE (SD)	mC (SD)	mE	mC	F	p	ηp^2
Outcome measures – person with dementia/MCI [range]
Self-Management
SMAS-S, [0–15]
Taking initiative
Age ≤ 70	9	11	7.7 (3.2)	9.5 (1.4)	8.4 (3.2)	9.4 (2.5)	9.0	8.9	.63	.43	.02
Age > 70	16	11	7.9 (1.8)	9.8 (2.8)	8.9 (2.5)	8.7 (2.5)	9.3	8.1
Lower education	14	12	8.8 (2.7)	9.3 (2.2)	9.1 (2.9)	8.3 (1.9)	9.1	7.9	.61	.44	.01
Higher education	11	10	6.5 (0.9)	10.1 (2.0)	8.2 (2.4)	10.0 (2.7)	9.3	9.3
No tablet-use	11	12	7.5 (2.8)	10.4 (1.9)	8.6 (3.2)	9.5 (2.5)	9.2	8.6	.00	.98	.00
Tablet-use	14	10	8.0 (2.0)	8.8 (2.1)	8.8 (2.4)	8.5 (2.4)	9.1	8.4
Multi-functionality
Age ≤ 70	9	11	8.8 (1.0)	8.9 (1.2)	8.0 (2.7)	8.7 (1.8)	8.0	8.6	3.5	.07	.08
Age > 70	16	11	8.3 (2.1)	9.1 (1.4)	8.9 (2.6)	7.7 (2.0)	9.2	7.5
Lower education	14	12	8.9 (1.6)	8.8 (1.4)	8.4 (2.5)	8.0 (2.0)	8.2	8.0	.72	.40	.02
Higher education	11	10	7.8 (1.9)	9.3 (1.2)	8.8 (2.9)	8.5 (1.9)	9.4	8.1
No tablet-use	11	12	8.2 (1.6)	9.3 (1.2)	7.9 (2.8)	8.4 (1.9)	8.2	8.0	.46	.50	.01
Tablet-use	14	10	8.6 (2.0)	8.6 (1.3)	9.1 (2.4)	8.0 (2.0)	9.1	8.1
Total, [0–90]
Age ≤ 70	9	11	43.8 (6.2)	45.5 (5.9)	45.7 (10.4)	45.5 (8.1)	46.3	45.1	.09	.77	.00
Age > 70	16	11	43.0 (8.5)	47.5 (8.0)	43.4 (10.3)	43.4 (6.9)	44.4	41.8
Lower education	14	12	46.9 (7.1)	44.9 (6.2)	45.1 (10.0)	42.8 (6.4)	43.8	42.7	.07	.79	.00
Higher education	11	10	38.6 (5.6)	48.4 (7.6)	43.0 (10.8)	46.5 (8.3)	46.8	44.3
No tablet-use	11	12	45.1 (7.3)	47.5 (5.9)	43.5 (13.5)	45.2 (8.4)	43.3	43.6	.53	.47	.01
Tablet-use	14	10	41.9 (7.8)	45.3 (8.2)	44.8 (7.2)	43.6 (6.3)	46.5	43.3
Social participation
MSPP, [0,3]
Formal social participation
Age ≤ 70	8	11	.42 (.53)	.24 (.45)	.42 (.53)	.27 (.36)	.41	.36	.00	1.00	.00
Age > 70	16	11	.48 (.49)	.45 (.34)	.65 (.52)	.58 (.45)	.61	.55
Lower education	14	12	.45 (.50)	.31 (.39)	.48 (.50)	.61 (.47)	.45	.67	8.72	.01*	.18
Higher education	10	10	.47 (.50)	.40 (.44)	.70 (.55)	.20 (.23)	.67	.20
No tablet-use	11	12	.40 (.47)	.42 (.47)	.55 (.48)	.47 (.46)	.55	.47	.00	.98	.00
Tablet-use	13	10	.51 (.52)	.27 (.31)	.59 (.58)	.37 (.40)	.53	.44
Outcome measures – caregiver [range]
Positive care experience
PES, [0,16]
Age ≤ 70	15	14	12.6 (3.0)	12.1 (2.6)	12.1 (2.3)	11.1 (3.7)	11.5	10.9	.17	.68	.00
Age > 70	10	8	10.2 (3.1)	11.3 (3.3)	11.2 (3.1)	10.6 (3.6)	12.3	11.0
Lower education	14	11	10.8 (3.3)	11.7 (3.2)	11.2 (2.8)	10.3 (3.4)	11.8	10.3	.79	.38	.02
Higher education	11	11	12.7 (2.8)	11.9 (2.6)	12.5 (2.3)	11.6 (3.8)	11.8	11.5
No tablet-use	8	9	10.3 (3.6)	12.1 (3.3)	10.1 (3.6)	11.9 (2.8)	11.0	11.7	2.56	.12	.06
Tablet-use	17	13	12.3 (2.9)	11.6 (2.6)	12.5 (1.7)	10.3 (4.0)	12.2	10.4

Note. E: experimental group; C: control group; SMAS-S: Self-Management Abilities Scale – Short version; MSPP: Maastricht Social Participation Profile; PES: Positive Experience Scale. Significant numbers are p ≤ 0.05 and are marked with an asterisk. In bold are the effect sizes that show improved results in the experimental group.

With regard to the caregivers’ positive care experience, a small effect size was found for those in the FindMyApps group with lower education (η_p² = .02) and a moderate effect size for those with tablet experience (η_p² = .06), with the FindMyApps participants somewhat improving and participants in the control group somewhat declining.

Reported use of the intervention during the trial

From the follow-up calls 2 and 6 weeks after the start of the intervention period, and from the semi-structured interviews held with participant dyads for the process evaluation we found that in approximately half of the persons with dementia/MCI (13 out of 28) in the experimental group, used FindMyApps and/or the tablet at least once a week. In the control group 14 out of 31 persons with dementia/MCI used the tablet at least once a week.

For the completers-only analysis this means that in the experimental group 13 out of 25 persons with dementia/MCI (52%) used FindMyApps and/or the tablet at least once a week, and in the control group 14 out of 22 (64%).

Discussion

Overall results

The purpose of this pilot study was firstly to explore the added value and feasibility of the FindMyApps intervention compared to usual tablet use for persons with dementia/MCI and their caregivers. Secondly, we aimed to evaluate the feasibility of the trial procedures as preparation for a future definitive trial into the effectiveness of the FindMyApps intervention. The study results show that people with dementia/MCI that received FindMyApps after three months intervention did not differ significantly from controls in terms of their ability to self-manage, engagement in activities and social participation, nor in their self-efficacy, experienced autonomy or quality of life. Also, no statistically significant differences between the FindMyApps group and controls were found with respect to the caregivers’ care experience, sense of competence and quality of life. Possible explanations for this could be that FindMyApps had no added value for the users in this study, that the intervention period of three months was insufficient to show results, or that the sample size was too small to demonstrate statistical significant effects. Furthermore, the control group included 4 people with Lewy body dementia, whereas the experimental group had none. This raises the question whether differences in type of dementia may have influenced the outcomes and if training methods for learning to use the tablet and FindMyApps need to be adjusted for people with different types of dementia because of differences in capabilities and deficits.

Small positive effect sizes tended to show small positive tendencies in the FindMyApps group for self-management and social participation. These trends are in line with expectations, based on the findings on these outcomes in a previous small scale pilot study into FindMyApps, and which concluded that the program has the potential to positively influence self-management and social participation in people with dementia [27]. In addition, findings from the process evaluation carried out as part of this pilot trial, demonstrated that the FindMyApps selection tool was usable for people with dementia/MCI and positively valued by those who used it regularly [28]. These findings of FindMyApps are in line with the study of Joddrell et al. who concluded that touch screen technology is usable for people with dementia, and offers potential for meaningful occupation, fun, entertainment and independent activities [14].

For the caregiver, a small, non-significant, positive effect size was found for positive care experiences, showing a tendency towards a more positive care experience in those who used FindMyApps. This is an improvement compared to the previous small-scale feasibility study [27], which showed a significant effect on positive care experience in the control group. This could suggest that the improved version of FindMyApps (developed after the first small scale pilot), which includes the introduction of demonstration videos and the personalised “MyApps” page, tended to have better supported the person with dementia/MCI and as a consequence improved the care experience of caregivers in the experimental group [27]. An earlier study also suggests that technology can reduce the demands on caregivers, and as such can even contribute to diminish the economic costs of dementia [18].

The statistically significant interaction between type of intervention and education on “formal social participation” suggests that persons with mild dementia/MCI who were higher educated benefitted more from FindMyApps on this aspect. On the other hand, people with lower education seemed to have benefitted more from usual tablet use. A possible explanation for this could be that the FindMyApps training was better received by the higher educated people with dementia/MCI.

People over age 70 and those higher educated who used FindMyApps for three months tended to show larger improvements on self-management. Furthermore, for self-management, participants who had used a tablet before the intervention seemed to have benefitted more from FindMyApps than from usual tablet use. Our results suggest that on the domain of self-management persons with mild dementia/MCI aged over 70, with higher education and those with some tablet experience who used FindMyApps seem to have benefitted more than usual tablet users. On the other hand, we found that caregivers with lower education and some tablet experience before the intervention benefitted more from FindMyApps than from usual tablet use: they showed more improvement in positive care experience than those with higher education and no tablet experience.

Strengths and limitations

The involvement of persons with mild dementia/MCI both during the user-participatory development of the FindMyApps intervention and during this explorative evaluation is one of the strengths of this study, as is the randomised controlled study design. Active involvement of persons with dementia in the development of technological interventions has been found to improve the acceptability and usefulness [52]. Although several earlier literature reviews emphasised the importance of actively involving persons with dementia in the development of technological interventions, only a limited amount of studies have realised this so far [8,52,53]. Furthermore, subgroup analyses (age, education and previous tablet use) provided insight in factors that may have influenced the outcomes, which is of major importance for designing a definitive trial.

This study also had some limitations. Fewer participant dyads (n = 59) were recruited than intended (n = 80) to be able to show moderate to large effect sizes. One of the reasons is that the target group included people with mild dementia/MCI, whom often do not yet receive care or have a diagnosis, which makes it more difficult to find them. Additionally, because this pilot study was underpowered to demonstrate statistically significant differences of outcomes with small to medium effect sizes, final conclusions on the effectiveness of FindMyApps on these outcomes are not yet possible. Indeed, this was not the intention of this explorative pilot study. We therefore want to emphasise that generalising the current outcomes should be done with great caution. In a future definitive trial, a larger sample will be needed to evaluate the effects of the FindMyApps program on these outcomes. Based on t-tests, at least 64 dyads per group will be needed for outcomes with moderate effect sizes (d = .05) to show statistically significant differences with a power of 0.80 and an alpha of 0.05 [51]. Additionally, an a priori power calculation was performed with G*Power version 3.1 [54] to allow for a recommendation regarding a sample size based on ANCOVAs. The calculation indicated a sample size of 79 dyads per group to detect moderate effect sizes, with a power of 0.80, and an alpha of 0.05, based on two groups, 2 degrees of freedom, and 2 covariates (baseline scores and GDS). These calculations are based on moderate effect sizes. For smaller effect sizes, larger numbers would be needed to have sufficient power. Of course the clinical relevance of outcomes with (very) small effect sizes needs to be considered first. Also, when the intervention is used more intensively, these effects could increase and smaller groups may be able to show these effects as well.

A limitation that needs improvement for a definitive trial is the lack of information about the amount of time that participants spent on using the tablet/FindMyApps during the intervention period. Originally we intended to use analytics data for this from every participant. However, due to the way the anonymised data was recorded, it was not possible to extract these data for analysis. Reasons for this were, for instance, that several participants had registered with the same username (their first name), and some participants (inadvertently) made several new accounts during the intervention period. Furthermore, the system did not register a distinction between the test accounts used by the trial participants themselves and the accounts used by trainers/interviewers during the introductory training and at other moments. For a future definitive trial, it would therefore be advisable to use unique usernames for each participant, with only one account possible per user.

From the semi-structured interviews executed in the process evaluation and follow-up calls, it appeared that approximately half of the persons in the experimental group and 38% of the persons in the control group did not or barely used the tablet/FindMyApps. In the first months of the trial, the follow-up calls were not always executed according to protocol, which caused some participants to forget to use the tablet. Little or no usage by some of the participants, especially in the experimental group, may have negatively influenced the study outcomes, as the process evaluation showed that participants who did use the FindMyApps selection tool, generally experienced it as user-friendly and useful [28]. Therefore, the implementation should be improved to stimulate usage more and consequently improve the potential effects the intervention might have on self-management and social participation.

Our data showed that if the participants with dementia/MCI in the FindMyApps group had some amount of experience with a tablet prior to enrolment in the study, their self-management tended to increase more in the intervention period, compared to those persons with dementia/MCI who had not previously used a tablet. One hypothesis that could be tested in future is that the three months intervention period may have been too short to learn to use the intervention properly, for those who had no experience at all and consequently they may have benefitted less from it.

Due to the COVID-19 pandemic, we were limited in collecting data from home visits from mid March 2020: The final post-test interviews with 6 (12.8%) persons with dementia/MCI had to be administered by phone instead of face-to-face, as the researchers were not allowed to visit them. These participants received the questionnaires before the interview, which provided them with the option to prepare the interview together with their caregiver. These procedural changes, in combination with the social self-isolation of people during the pandemic (and sometimes less visits of their informal caregiver), may have influenced the outcomes at post-test for some people and thus the study outcomes.

Scientific and practical relevance

This explorative pilot study confirms that persons with mild dementia/MCI are able to provide valuable research input. This is a step forward in dementia research, as experiences and opinions of people with dementia are still often neglected [55]. In addition, the process evaluation of this pilot study, submitted elsewhere [28], and a previous pilot study [27], showed that, based on the experiences by participants, FindMyApps has the potential to improve social health (self-management and social participation) in people with dementia/MCI.

The findings of the explorative controlled effect evaluation provide insight into the potential effects of a personalised tablet intervention on the person with dementia/MCI’s ability to self-manage, engagement in activities and social participation, as well as their self-efficacy, experienced autonomy and quality of life. It also provides insight into the potential effects of such an intervention on their caregivers’ care experience, sense of competence and quality of life. And finally, this study provided first insights into the characteristics of possible subgroups that may benefit most of FindMyApps. This extends the scope of current research on technological interventions, which often focuses on compensation for cognitive decline, rather than the equally important social and emotional domains, which are often neglected [56]. These insights are also relevant for the design of a future definitive trial as well as for further implementation of FindMyApps in practice.

Recommendations for a future definitive trial

In this study, the Guidelines for the Rating of Awareness Deficits (GRAD) question was not administered to all participants. For the definitive trial, however, we recommend to administer the GRAD with all caregivers. This will allow to personalise the approach of study participants with regard to self-awareness of their cognitive deficits. For instance, to people who are more aware of these deficits the intervention may be best presented as a helpful tool to compensate for their disabilities, whereas to those who are less aware of their cognitive disabilities the intervention could be best presented as a means for easily finding meaningful activities without referring to the person’s disabilities. This is supported by earlier research that found a correlation between perceived utility and adoption of assistive technology in people with dementia [57]. Consequently, such a personalised approach may improve the acceptance of FindMyApps and the motivation of people with dementia/MCI to use FindMyApps. In addition, it might be useful to investigate whether the FindMyApps training needs to be more personalised based on the level of education and type of dementia/MCI of the user.

The process evaluation of FindMyApps suggested that with more effective and consistent implementation of the training components by the researchers, and with a clearer explanation of the helpdesk function, people might be able to use the FindMyApps selection tool and the tablet more frequently and effectively [28]. It is recommended to advice participants to use the intervention at least several times per week, rather than leaving it to the participants to decide how often to use it. If the training would be improved this may also improve the outcomes for participants who do not have tablet experience before starting to use FindMyApps.

It is recommended that some changes will be conducted for the FindMyApps app. The current FindMyApps program is a web application, which is limited and less stable than a native app. This led to some technical problems during the trial, such as buttons that did not respond because of poor internet connection. It is recommended to solve such problems before further evaluating the effectiveness [58]. A solution to this would be to develop a native app, which is more stable. Furthermore, it would be useful if every participant would get a unique username that is only known to the researchers in order to be able to collect useful analytics data for research purposes while maintaining anonymity, such as time spent on the tablet/FindMyApps, videos played and page views. This way it would also prevent that users register multiple times if they accidentally log out.

Even though the data input check demonstrated very little missing data and very few transcription errors in the outcome measures, it is recommended to input the data directly into an electronic data capture system, such as Castor EDC, rather than transcribing data from paper forms. This may help to prevent unnecessary errors. It is also recommended that the questionnaires are reconsidered and shortened because many participants gave feedback that the questionnaires were sometimes difficult to interpret, were too much repetitive and tiring [28]. Additionally, because some tendencies were found for the impact of age, education and tablet use prior to the intervention, it is recommended this is further explored in a definitive trial.

Conclusion

Based on this explorative effect evaluation, the process evaluation reported previously [28] and the conducted feasibility study, we can conclude that the FindMyApps intervention seems to be a feasible intervention for people with mild dementia/MCI and their caregivers. The outcomes of this pilot study need to be considered with caution as because of the relatively small sample size the study was not sufficiently powered to show statistically significant small to moderate effects. The study design and methods were feasible, although some improvements are recommended for a definitive effectiveness trial. A larger sample size will be needed to investigate the effectiveness of FindMyApps on the self-management and social participation of people with mild dementia/MCI and the sense of competence and care experience of their caregivers. Additionally, it is recommended to take into account the possible impact of education level in the evaluation of the effect of FindMyApps on social participation of people with mild dementia/MCI.

Acknowledgements

On behalf of all authors, we would like to express our gratitude to all the persons with dementia/MCI and their caregivers who participated in the FindMyApps study. We would also like to thank the student interviewers who were involved in the study from VU University, Saxion University of Applied Sciences, Amsterdam University of Applied Sciences and University of Amsterdam.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Marie Curie Innovative Training Network (ITN) action H2020-MSCA-ITN-2015 under Grant 676265; Foundation of Support VCVGZ under Grant 250; and Foundation Hofje Codde & Van Beresteyn.