Introduction

Alzheimer's Disease International reported that more than 9.9 million new cases of dementia have been diagnosed worldwide, with Asia having the largest number (49%) of new cases. The report further projects that there will be 38.58 million cases by 2030 and 67.18 million by 2050 (Gauthier et al., 2021). In Taiwan, 7.64% of the total aged population is believed to experience dementia (Taiwan Alzheimer Disease Association, 2022), with 65.7% of nursing home residents experiencing the syndrome (Taiwan Dementia Society, 2016). Dementia is characterized by a cluster of symptoms that include both cognitive and noncognitive impairments. The latter are collectively referred to as behavioral and psychological symptoms of dementia (BPSD), with symptoms including aggression, agitation, wandering, verbal outbursts, delusions, hallucinations, apathy, and anxiety (Chakraborty et al., 2019). A survey has shown that the prevalence of BPSD is 67.9% among those diagnosed with dementia, whereas 30% of patients with dementia (PwD) have exhibited three or more indicators of BPSD (Vaingankar et al., 2017).

BPSD are usually the main cause of family caregivers' distress in coping with their aged loved ones and are the primary reason for early nursing home placement (Deardorff & Grossberg, 2019). BPSD are associated with various unmet needs (Alzheimer's Society Canada, 2017). If care providers cannot understand the meanings behind these problematic behaviors, it will inevitably lead to worsening communication and increasing frustration with the dementia sufferer's problematic behaviors, leading to increased stress for the care providers (Cloak & Khalili, 2022). In general, care providers in long-term care (LTC) have not received sufficient dementia-related educational training, especially in terms of strategies for dealing with BPSD (Hunter et al., 2016). As a result, most cannot understand the primary causes of BPSD or the BPSD-related needs of residents with dementia (RwD), leading them to inadequately empathize with and fail to appropriately address the needs of RwD. At a loss for how to deal with these patients, caregivers tend to follow medical orders in terms of administering sedatives and implementing protective restraints or to simply ignore these patients. Consequently, the patients' problems are neither solved nor alleviated, making it even more difficult to deliver effective care, which often leads to RwD repeatedly seeking medical attention and consuming medical resources (Cloak & Khalili, 2022).

Gerlach and Kales (2018) have argued that a well-designed, nonpharmacological therapeutic strategy that manages BPSD effectively can reduce medical costs and improve care quality. A systematic literature review summarizing the implementation of relevant measures to manage BPSD in RwD in nursing homes revealed that educational training for care providers can effectively improve BPSD (Bessey & Walaszek, 2019). However, some related education and training programs do not have a followable manual, whereas others are complicated, are difficult to remember, or lack detailed descriptions of their educational content. Moreover, the diversity in the responses and reactions of residents experiencing dementia complicated by BPSD is extremely heterogeneous. In the current climate of healthcare workers already being burdened by heavy caseloads, it is necessary to develop a simple and memorable management model that assists care providers in clinical application. Therefore, a new care and management model for BPSD named “the need-centered Watch–Assess–Need intervention–Think (WANT) care and management model for behavioral and psychological symptoms of dementia” (WANT model) was developed by the research team (Yang, 2020). However, before this new model may be applied in Taiwan and, possibly, globally in the future, it is necessary to verify its feasibility and develop empirical-based scientific care measures to monitor its effectiveness. Therefore, an education and training program based on the WANT model was proposed for testing. The purpose of this research is to evaluate the feasibility of the WANT education and training program for managing the BPSD in residents living in an LTC facility.

Methods Research Design

This study used a mixed-method design that combines qualitative and quantitative research to increase the breadth and depth of understanding and corroboration (Schoonenboom & Johnson, 2017). Data were collected from July 1, 2019, to September 30, 2019.

Settings and Participants

The WANT education and training program was implemented at a large, 111-bed, privately owned, for-profit nursing home in southern Taiwan. Participants included care providers (hired nonprofessional care workers) and their corresponding RwD. Formal power calculations are not typically required or conducted in feasibility testing studies (Arain et al., 2010). The researcher invited 40 participants (20 care providers and 20 of their RwD) to participate in this study. The inclusion criteria for care providers were as follows: (a) formally hired and (b) have at least 6 months of experience in dementia care. The inclusion criteria for the RwD were as follows: (a) diagnosed by a physician and a Mini-Mental Status Examination score of < 26 (high school education and above) or 17 (below high school education) or Clinical Dementia Rating Scale ≥ 1 as various types of dementia and (b) exhibited at least one BPSD symptom based on the Chinese version of Cohen-Mansfield Agitation Inventory (CMAI). The only exclusion criterion was for RwD with reversible dementia (e.g., as a side effect of infection, malnutrition).

Outcome Measures

Whereas qualitative data are important to understand the participants' perspectives and experiences in using this model, quantitative data can provide preliminary evidence of model effectiveness. The following tools discussed below were used to collect quantitative data.

Chinese version of Cohen-Mansfield Agitation Inventory (for residents with dementia)

RwD-related agitation was determined using a modified form of the CMAI. The original scale, developed by Cohen-Mansfield et al. (1989), is the most commonly used instrument to measure agitation. The long form includes 29 types of agitated behaviors. The assessment score ranges from 1 (no emergence of symptoms) to 7 (many examples per hour), with higher scores correlated with more severe agitated behavior. The CMAI institutional version was translated by L. C. Lin et al. (2007). The content validity index (CVI) of the Chinese version of CMAI is .99, the test–retest reliability with 2 weeks apart showed an intraclass correlation coefficient of .69–.91, and interrater reliability with four observers showed an intraclass correlation coefficient of .63–.88, with a Cronbach's alpha of .70 (L. C. Lin et al., 2007). The Cronbach's alpha was .70 in this study.

The Cornell Scale for Depression in Dementia (for residents with dementia)

The Cornell Scale for Depression in Dementia (Alexopoulos et al., 1988) is a 19-item scale used to measure depressive symptoms in PwD. The Chinese version was translated by J. N. Lin and Wang (2008), and good reliability and validity have been established. The scale includes five subscales: mood-related signs, behavioral disturbances, physical signs, cyclic functions, and ideational disturbances. Each item is measured on a 3-point continuum of “absent” (0), “mild or intermittent” (1), and “severe” (2), with an overall score range of 0–38. The CVI is .92, and concurrent validity with the Geriatric Depression Scale-Short Form is .322 (p < .001); the interrater agreement for each item is significant (kappa = .43–.89); and the Cronbach's alpha for internal consistency reliability is .84 (J. N. Lin &Wang, 2008). The Cronbach's alpha was .81 in this study.

Attitude towards Dementia Care Scale (for care providers)

The Attitude towards Dementia Care Scale developed by Cheng (2007) was used in this study to measure care providers' attitudes toward dementia and care for RwD. This scale is a semantic differential scale, and each item is paired with opposite semantic adjectives. Positive and negative attitudes include easy–difficult, fulfilled–frustrated, meaningful–worthless, patient–bored, challenging–uninteresting, safe–scary, individual differences–immutable, and manageable–powerless. For each item, the most positive attitude is scored as 10 and the most negative attitude is scored as 0. The total scale score ranges from 0 to 80, with higher scores indicating more positive attitudes toward RwD. The CVI is .70, and the Cronbach's alpha .75. The Cronbach's alpha was .57 in this study.

Dementia Behavior Disturbance Self-efficacy Scale (for care providers)

The Chinese version of Dementia Behavior Disturbance Self-efficacy Scale developed by Uei and Yang (2013) was modified from the Dementia Behavior Disturbance Scale (Baumgarten et al., 1990), which includes 28 items. Each item is scored on a 5-point Likert scale of 0 (strongly no confidence), 1 (no confidence), 2 (Acceptable confidence), 3 (confidence), and 4 (strong confidence). The total score range is 0–112 points, with higher scores indicating more confidence in managing BPSD. The CVI is .92, and the Cronbach's alpha is .89 (Uei & Yang, 2013). The Cronbach's alpha was .30 in this study.

Qualitative interview guide

The interview guide was developed by the research team and includes two main questions: (a) After receiving the WANT model education and training program, what are your feelings about it? (b) Is the program effective in helping you deal with behavioral problems stemming from RwD?

The Watch–Assess–Need Intervention–Think Education and Training Program

An education and training program based on the WANT model (Yang, 2020) was proposed in this study. “WANT” is the abbreviation of Watch–Assess–Need intervention–Think, with “Watch” referring to the need to pay attention to the frequency and status of behaviors of RwD, “Assess” referring to the need to assess relevant background and proximal factors for such behaviors and deduce the needs behind, “Need intervention” referring to the need to implement care measures according to residents' needs, and “Think” referring to the importance of reflecting on and sharing mutual experiences with peers. WANT is a “need-centered” and “need-oriented” care approach that applies logical reasoning to guide care providers in dealing with BPSD in PwD. This 12-week program includes two phases. In the first phase (Weeks 1–4), a 1-month workshop that includes lectures, discussions, and case sharing is held. After the initial 3 hours of lecture in Week 1, the researcher invites each participant (care provider) to select an individual case, complete the designated reflective diary assignment, and share and discuss their thoughts during Weeks 2–4. In the second phase (Weeks 5–12), the participants are instructed to use the program manual to practice and to record their particular situations. During this phase, the researcher visits the study site to hold weekly focus groups during Weeks 5–11 and to lead the participants in sharing their reflective diaries and discussing problems encountered during practice. The evaluation is held on Week 12 (Table 1). The total timeframe was 16 hours, with 6 hours in Phase 1 and 10 hours in Phase 2.

Table 1 - The WANT Model Education and Training Program Time Curriculum Teaching Methods Week 1 Introduction of dementia, BPSD and related factors and needs, WANT model, reflexive diary assignment Lecture + discuss + the first-time assignment practice (3 hours) Weeks 2–4 Case discussion based on WANT model reflexive diary Sharing + open for discussion + assignment practice (1 hour) Weeks 5–8 Case report based on WANT model reflexive diary Sharing + open for discussion (1 hour) a Week 12 Evaluation of the feasibility of WANT model Sharing + open for discussion (3 hours)

Note. WANT = Watch–Assess–Need intervention–Think; BPSD = behavioral and psychological symptoms of dementia.

a No assignment.

Procedure

After obtaining Research Ethics Committee approval, the researcher contacted the manager of the study site and explained the study. The manager then helped select and invite eligible care providers to participate. Data collection and intervention began after obtaining consent from each participant. Data were collected by the researcher at three time points: T1 (preintervention), T2 (1 month after intervention), and T3 (3 months after intervention). The training program is workshop based, and the qualitative data were collected throughout the entire intervention period.

Ethical Considerations

This study was approved by the Research Ethics Committee (No. 106-182), and participation was voluntary. After the researcher explained the research, informed consent forms were obtained from each participant (including care providers and their corresponding RwD). A family surrogate signed the consent if an RwD was unable to do so. All of the participants had the right to withdraw from the study at any time. All interviewee data were presented anonymously and digitally encoded.

Data Analysis

Quantitative data were analyzed using SPSS Statistics 17.0 (SPSS, Inc., Chicago, IL, USA). General linear model repeated measures were used to compare the means of three measurements. Qualitative data were analyzed using content analysis. First, the researcher reviewed the entire verbatim data content repeatedly, confirmed the meaningful messages in the data, and selected partial excerpts for coding. After this, the excerpts were categorized according to the themes and specific research questions to be discussed (Abrams et al., 2020; Haven & Van Grootel, 2019). In terms of rigor, the researchers made constant observations at the site, which were recorded in reflexive diaries. Moreover, discussions and peer debriefings were held among research team members to increase the credibility of the interpretations. After completing data collection, the interviewees were asked to review the data content (member check) to avoid interpretation gaps (Haven & Van Grootel, 2019).

Results

The demographic characteristics of the participants (including both care providers and RwD) are shown in Tables 2 and 3.

Table 2 - Demographics of Care Providers (N = 20) Variable n % Age (years; M and SD) 33.55 9.94  20–35 12 600  36–45 6 30  46–65 2 10 Gender  Male 3 15  Female 17 85 Religion  None 2 10  Buddhism 6 30  Taoism 7 35  Christian 3 15  Others 2 10 Educational level  Senior high school 9 45  College 11 55 Years in current unit  ≤ 1 6 30  ≤ 2 8 40  ≥ 3 6 30 Had attended course in dementia care  No 8 40  Yes 12 60 Years in dementia care  ≤ 1 4 20  ≤ 2 6 30  ≥ 3 10 50 Experience dealing with BPSD  No 4 20  Yes 16 80

Note. Age range: 20–55 years. BPSD = behavioral and psychological symptoms of dementia.

Table 3 - Demographics of Residents With Dementia (N = 20) Variable n % Age (years; M and SD) 79.75 9.26 Gender  Male 6 30  Female 14 70 Religion  None 3 15  Buddhism 5 25  Taoism 8 40  Christian 2 10  Others 2 10 Educational level  Illiteracy 11 55  Elementary school 4 20  Junior high school 1 5  Senior high school 2 10  College 2 10 Marital status  Widowed 14 70  Have a wife/husband 6 30 Have taken dementia medication  No 7 35  Yes 13 65 Have taken antipsychological medication  No 8 40  Yes 12 60 Restraint (previous 3 months)  No 13 65  Yes 7 35 Restraint (subsequent 3 months)  No 13 65  Yes 7 35

Note. Age range: 57–93 years.

Effects of the Watch–Assess–Need Intervention–Think Model Education and Training Program

Significant variance in the mean agitated behavior of RwD was found across the three measurements (F = 6.01, p = .01), with post hoc comparisons showing a significant difference in the mean of agitated behaviors of RwD between T1 and T2 (p = .003), T2 and T3 (p = .022), and T1 and T3 (p = .022), and that T3 was lower than both T1 and T2 and that T2 was lower than T1. These results indicate that, at 1 and 3 months postintervention, agitated behaviors in the RwD had decreased from preintervention levels (Table 4).

Table 4 - Comparison of Outcome Variables Over Three Time Points (N = 20) Outcome Time 1 (T1) Time 2 (T2) Time 3 (T3) GLM-RM Post Hoc Mean SE 95% CI Mean SE 95% CI Mean SE 95% CI F p Patient  Agitated  behaviors 80.10 8.71 [61.85, 98.34] 64.00 4.75 [54.04, 73.95] 54.25 4.36 [45.11, 63.38] 6.01 .01* T2 < T1, T3 < T2, T3 < T1  Depressive  symptoms 14.40 1.57 [11.1, 17.70] 12.50 1.41 [9.55, 15.45] 9.65 1.15 [7.24, 12.05] 6.52 .007** T2 < T1, T3 < T2, T3 < T1 Care providers  Self-efficacy 66.10 3.13 [59.53, 72.66] 73.35 2.33 [68.46, 78.23] 77.50 3.40 [70.37, 84.62] 2.49 .11  Attitude 47.30 1.75 [43.63, 50.96] 54.15 2.74 [48.40, 59.89] 58.75 2.88 [52.72, 64.78] 5.74 .01* T2 > T1, T3 > T2, T3 > T1

Note. GLM-RM = general linear model–repeated measures.

*p < .05. **p < .01.

The results indicate that the mean score for depressive symptoms in the RwD differed significantly across the three measurements (F = 6.01, p < .01), with post hoc comparisons showing a significant difference in mean depressive symptoms in the RwD between T1 and T2 (p = .004), T2 and T3 (p = .004), and T1 and T3 (p = .007), and that T3 was lower than T1 and T2 and that T2 was lower than T1. These results indicate that, 1 and 3 months after the intervention, depressive symptoms in the RwD had decreased from preintervention levels.

The results show that the care providers' mean attitude toward dementia varied significantly across the three measurements (F = 5.74, p = .01), with post hoc comparisons showing a significant difference in the mean attitude of care providers toward dementia between T1 and T2 (p = .002), T2 and T3 (p = .034), and T1 and T3 (p = .034), and that mean attitudes were higher at T3 than at T1 and T2 and higher at T2 than at T1. These results indicate that, at 1 and 3 months postintervention, care providers' attitudes toward dementia had improved compared with preintervention levels.

The mean self-efficacy of the care providers did not significantly vary across the three measurements (F = 2.49, p = .11), with post hoc comparisons showing a significant difference in the mean self-efficacy of care providers between T1 and T2 (p = .038) and T2 and T3 (p = .038), but not between T1 and T3 (p = .076). These results indicate the self-efficacy of care providers increased slightly from T1 to T3.

Care Providers' Perspective of the Program

The following four themes were extracted from the qualitative data collected from the care providers.

Self-perceived improvement in self-efficacy in managing the behavioral and psychological symptoms of dementia

After the intervention, care providers felt they were paying greater attention to what the RwD were trying to do when their BPSD appeared. After understanding the meanings underlying the problematic behaviors of RwD, care providers were able to manage BPSD more effectively to prevent negative outcomes.

I've started to pay more attention to their needs. In the past, I didn't intend to interact with them and paid less attention to their needs…. At present, I am willing to spend more time paying attention to them. (Caregiver 12)

I've started to realize that in order to provide residents with appropriate care, it is necessary to associate with their past lifestyle. (Caregiver 5)

Able to view problems from a more need-centered perspective

Care providers perceived that they were able to better associate the problematic behaviors of RwD with their unmet needs. They noted that they associated residents' past experience and current situation with their problematic behaviors and thus better understood their inner needs. When residents displayed problematic behaviors, care providers had a better intuitive grasp of their needs.

Participating in this program helped me think about how to respond to the residents quickly…. I've started to think more about residents' needs and thinking about what's better for them. (Caregiver 13)

Compared to the past, I've learned that residents' behavior is associated with their needs and that they need care. The elderly with dementia are unable to explain their needs, but they yell out, which means they have a need…. (Caregiver 6)

Changed attitudes toward dementia and altered behavior

Care providers found that their attitude toward RwD became more positive and their patience with residents increased. Care providers learned that the problematic behaviors of RwD are linked to their inability to express their needs. Thus, the care providers became less frustrated and less likely to blame residents for their outbursts and actions. Their attitude changed toward residents' behavior. They no longer merely provided routine care, and their attitudes toward dementia and care for RwD became more accommodating and accepting.

I feel that I've become more patient and more likely to listen to residents and cooperate with them. In the past, I used to ask residents to obey my instructions at specific timepoints. Now, I seldom do so. I've slowed down and asked residents more patiently. I am not as impatient as I used to be.... (Caregiver 4)

Reduce care burden and relieve stress

After the training, care providers perceived their care burden as significantly lessened. Care providers previously found it stressful when RwD led to problematic behaviors. Because care providers learned to be more sensitive to the problematic behaviors of RwD, they developed better interactions with residents and became more positive. It became easier for them to take care of residents.

Now I can better understand these residents, their needs, what they want to do, and the fact that their problematic behaviors are caused by a need for assistance. Therefore, I will figure out how to deal with their problems. It has become more time saving and less stressful to take care of them…. (Caregiver 8)

Discussion

The results showed that, after participating in the WANT model education and training program, care providers perceived a gradual decrease in agitated behaviors in the RwD under their care. This is attributed to the care providers developing a needs-centered and needs-oriented approach that improved their logical thinking ability and helped them discover the causes of patient problems and find solutions to patients' needs, ultimately reducing agitated behavior in RwD associated with unmet needs. This result is consistent with Ballard et al. (2018), who discovered that need-centered educational training effectively diminishes agitated behaviors in RwD, and with da Silva Serelli et al. (2017), who found that applying theoretical teaching skills to clinical care effectively improved the ability of care providers to manage aggressive and nonaggressive behaviors in PwD. Moreover, this result is also consistent with the findings of a randomized controlled trial conducted by Fossey et al. (2014), who found using integrated analysis that need-centered training intervention measures reduced agitation and the frequency of antipsychotics use.

In addition, depressive symptoms in the RwD gradually decreased after the WANT-based program. This may be because the need-centered and need-oriented focus of the training encouraged care providers to pay more attention to and address residents' needs. This method enabled residents to feel they were respected and valued, which effectively reduced the depressive symptoms of RwD. Similar results were found by Ballard et al. (2018), who discovered that need-centered educational training can increase social interactivity in RwD and effectively reduce their emotional issues (Blake et al., 2020). In addition, these results are consistent with da Silva Serelli et al. (2017), who asserted that care providers who believe they are able to deal with dementia-induced problematic behaviors are more willing to interact with RwD, which can effectively alleviate depression in RwD.

From the care providers' perspective, the research results showed that the WANT-based program helped care providers gradually improve their attitudes. This may be attributed to their new appreciation that problematic behaviors of RwD are associated with unsatisfied needs and that residents are not intentionally disobeying care providers' instructions. This may have resulted in care providers being more likely to treat the causes of problematic behaviors of RwD with empathy and adopting a more positive attitude toward dementia. This research finding is consistent with Villar et al. (2018), who found educational training to have a positive effect on care providers in terms of improving their attitudes and behaviors toward treating dementia and improving their quality of care. This finding is also consistent with that of Ho et al. (2021), who suggested that educational training encourages care providers to develop a more positive attitude toward dementia. A positive attitude and need-centered care strategies can lead to increased ability to provide dementia care. The findings of past studies support that participation in educational training has a positive effect on care providers, in terms of improving their behavior toward RwD and becoming more proactive and helpful (Elpers et al., 2017; Mullan & Sullivan, 2016; Villar et al., 2018). As a result, care providers may no longer ignore RwD and actively work to improve their interactions with RwD to further improve the quality of life of sufferers (Muralidharan et al., 2019).

Although the care providers' self-efficacy after program participation did not show a significant change, the score at T3 was somewhat improved from T1. Moreover, in the qualitative data, care providers self-reported that they had experienced improved self-efficacy in managing BPSD. These findings may be compared with two related studies. Conway and Chenery (2016) reported that the use of communication strategies to implement educational training can immediately enhance self-efficacy and the ability to deal with dementia-induced problematic behaviors (DeSouza et al., 2020). However, McCabe et al. (2015) showed a different result in their finding that a 6-month educational training improved care providers' self-efficacy and reduced their stress, with the effectiveness lasting for up to 3 months. The difference in results may be attributable to differences in the content and duration of programs in this and the previous two studies.

After implementation of the WANT-based program, care providers' care burden and stress were reduced, their RwD-care-related anxiety was stabilized, and their interactions with RwD were more pleasant. Furthermore, their competence in dealing with problematic behaviors was improved, and they could more smoothly complete daily care tasks, which decreased their work burden. Similar to this study, McCabe et al. (2015) found that educational training helps reduce care stress. However, da Silva Serelli et al. (2017) found that dementia-related educational training does not decrease care providers' care stress and burden. Similarly, Conway and Chenery (2016) also found that dementia-related educational training does not reduce care providers' work stress because receiving additional educational training added to their already busy workload, increasing their stress. The result in this study differs from the above conclusions.

Limitations of This Study

Several potential problems with this study must be acknowledged. First, several of the participants experienced difficulty in logical reflexive thinking, and several lacked learning motivation. A well-trained seed trainer (i.e., a trainer of trainers) would probably be helpful and is recommended in future implementa