Family Caregiver Training Program (FCTP): A Randomized Controlled Trial

Research Design

A delayed-intervention, randomized design with a control group was used to test the effects of the FCTP on caregiver knowledge and confidence and the effects of changes in knowledge and confidence on burden, depression, QOL, occupational performance, and satisfaction. After screening and informed consent, baseline clinical and demographic data were collected by questionnaires distributed by either the U.S. Postal Service, fax, email, or in person just before the intervention or control. Thirty-six numbers were entered into Research Randomizer (Version 4.0; Urbaniak & Plous, 2013), which provided two random groups of numbers to represent the experimental (FCTP) and control (standard care and delayed intervention) groups. Participants in the experimental group completed a 2-hr presentation (one session) about the FCTP. Participants in the control group received one 90-min standard care presentation (general information and resources). Outcomes data were collected pre- and postintervention and at 3-mo follow-up.

Sample Size and Power Analysis

On the basis of a power analysis performed with results from the ADL Knowledge Test collected as part of the FCTP feasibility study (DiZazzo-Miller, Samuel, Barnas, & Welker, 2014), a sample size of 18 per group (36 total) had 0.80 power to detect a Cohen’s d of 0.97, allowing for a Type I error rate of 0.05.

Participants

Participants included family caregivers of people with dementia who lived in the metropolitan Detroit area. Inclusion criteria required participants to be (1) primary caregivers identified by providing at least 4–8 hr of care (5 days per week), (2) at least 18 yr old, and (3) fluent in English. Care recipients needed to have an Alzheimer’s or related dementia diagnosis. Professional and paid caregivers were excluded. Table 1 provides participant demographics.

Instruments

Caregiver Training and Data Collection

Data Analysis

For Aim 1, we performed two mixed analyses of variance (ANOVAs) to test for significant between-group differences on caregiver knowledge and confidence outcomes at two time points: immediately posttest and 3 mo posttest. The dependent or outcome variable for the first ANOVA was knowledge (i.e., measured by the ADL Knowledge Test); the independent variable was participation in the FCTP. Potential covariates (i.e., pretest knowledge, depression, burden, QOL scores, and the level of function of the care recipient [BADLS score]) were not found to be significant. Fisher’s least significant difference (LSD) post hoc analysis was performed to determine the specific variable relationships with significant differences. The second mixed ANOVA was conducted with confidence as the dependent variable.

For Aims 2 and 3, paired t tests were used to test for within-group differences between pretest and 3-mo follow-up for the burden, depression, QOL, occupational satisfaction, and occupational performance outcomes. Data were tested for normal distribution and significant outliers (Laerd Statistics, 2013; Munro, 2005). Within-group analyses were performed for only the intervention group because this analysis was exploratory (i.e., we were exploring potential relationships for future studies). IBM SPSS Statistics (Version 22; IBM Corp., Armonk, NY) was used for all analyses.

Caregiver Demographics

A separate one-way ANOVA was performed to determine whether there were preexisting differences between the experimental and control groups, and it revealed no significant differences between the groups. In general, participants were female, were White, were between age 55 and 64 yr, had a bachelor’s degree, and had power of attorney for their care recipient. Care recipients’ functional status scores, measured with the BADLS, indicated that care recipients required minimal to moderate assistance.

Aim 1: Caregiver Knowledge

The mixed ANOVA indicated a significant Group × Time interaction for knowledge, F(2, 68) = 12.68, p < .001. The simple effects follow-up for group revealed no significant differences between groups at pretest, F(1, 34) = 0.08, p = .78; however, there were significant differences between groups at posttest, F(1, 34) = 12.08, p = .001, and at 3 mo posttest, F(1, 34) = 27.49, p < .001. Fisher’s LSD pairwise comparisons for the control group revealed nonsignificant changes in knowledge across all three time points: pretest to posttest (p = .80), posttest to 3 mo posttest (p = .78), and pretest to 3 mo posttest (p = .92; Table 2).

Aim 1: Caregiver Confidence

The Group × Time interaction failed to reach significance for caregiver confidence, F(2, 68) = 0.14, p = .87, and there was not a group effect, F(1, 34) = 0.46, p = .50. However, there was a significant main effect for time, F(2, 68) = 10.19, p < .001, indicating an increase in caregiver confidence across time regardless of group assignment. The simple effects follow-up for group revealed no significant differences between groups at pretest, F(1, 34) = 0.17, p = .68; at posttest, F(1, 34) = 0.77, p = .39; or at 3 mo posttest, F(1, 34) = 0.29, p = .59 (see Table 2).

Aim 2: Burden, Depression, and Quality of Life

The results of the within-group paired t tests revealed no significant difference for burden or depression between the pretest and 3-mo posttest scores. Of the four QOL domains measured, only Domain 1 (physical health), t(17) = −8.74, p < .001, was significant; therefore, caregivers who attended the intervention group reported a significant increase in the physical health domain of their quality of life. There were no significant findings for the remaining three domains: Domain 2 (psychological), t(17) = 11.05, p = .30; Domain 3 (social relationships), t(17) = 0.37, p = .71; and Domain 4 (environment), t(17) = −0.27, p = .79 (Table 3).

Aim 3: Performance and Satisfaction

The results of the within-group paired t tests revealed no significant difference for occupational performance or satisfaction between the pretest and 3-mo posttest scores.

Caregiver Knowledge

Findings from this study reveal a significant improvement in knowledge and maintenance of knowledge 3 mo posttest with the intervention group and corroborate the findings of Done and Thomas (2001), the only known study in which caregiver ADL knowledge was examined. Although in this study we used the ADL Knowledge Test, Done and Thomas developed the Assessment of Awareness about Communication Strategies (AACS) test, which was used to assess caregiver knowledge of communication strategies. Both DiZazzo-Miller et al. (2014) and Done and Thomas reported a significant change (p < .001) in knowledge of communication strategies. The ADL Knowledge Test, however, expands on the AACS test by testing additional areas of ADL knowledge as well as communication. These initial posttest findings are similar and are likely because content delivered in an organized training is apt to result in newly gained knowledge. Moreover, these findings show that content related to communication, eating and feeding, and nutrition is not present in what has been deemed standard care. Therefore, even when caregivers receive standard care, information on ADLs is limited.

Caregiver Confidence

Confidence findings reveal that caregivers in both the experimental and control groups demonstrated a significant increase in confidence immediately posttest; however, neither group maintained confidence 3 mo posttest. These findings are similar with those of Lewis et al. (2010), in which 90% of caregivers reported an increase in confidence. Lewis et al., however, used a single-group pretest–posttest design with an Internet version of the Savvy Caregiver Training Program and did not test for any follow-up effects. In addition, using a single-group design, Graff, Vernooij-Dassen, Hoefnagels, Dekker, and de Witte (2003) revealed nonsignificant findings in caregivers’ sense of competence after 5 wk of in-home occupational therapy services. Hepburn, Lewis, Sherman, and Tornatore (2003) did not find a significant change in caregiver mastery after a 12-hr workshop over the course of 7 wk that focused on caregiver roles, knowledge, skills, attitudes, and self-care issues. However, using a randomized controlled trial design, Graff et al. (2007) found a significant increase in caregiver mastery at 6 and 12 wk after 10 in-home sessions of occupational therapy over 5 wk. These home sessions focused on cognitive–behavioral interventions. Given, Sherwood, and Given (2008) stated that best practices for enhancing caregiver knowledge should be addressed with interventions focused directly on caregivers. Reducing uncertainty on how to provide care can relieve stress, leaving caregivers feeling better informed, thereby improving confidence (Given et al., 2008).

It should be noted that although the control group received standard care, they still received an intervention of sorts that provided information from a person who was physically present, listened, and showed interest in their questions and concerns. Rather than providing caregivers in the control group with a packet of information and then sending them home, the researcher provided a presentation on standard care information and demonstrated how some of the websites and resources provided in the standard care packet were accessed. In this situation, both methods of providing information for the intervention and control groups were presented in the same manner, which allowed for a comparison of information rather than a comparison on which form of media was used. Moreover, the subject matter of the control group was relevant to the participants’ current situation, which may also provide an explanation for the increase in confidence for the control group.

In fact, although the information was defined as “standard care,” when asked informally how many people received the packet used for the standard care group from a physician, not one caregiver raised his or her hand. This information was new to all caregivers. The lack of knowledge and resources reported by participants and the desire to obtain more information is noted throughout the literature. DiZazzo-Miller, Samuel, & Pociask (2013) discussed how one caregiver classified her approach to seeking information about dementia and caregiving as being in an “endless loop” where there were resources, but she did not know where or how to find good quality information. Although information is indeed available online and through professional associations, families and caregivers of people with dementia are often sent home with the diagnosis to learn as they go, which may explain why the control group also felt an immediate increase in confidence along with the intervention group. Nevertheless, a lack of interaction within the 3-mo follow-up period may have allowed for a regression to the pretest mean scores for both groups.

Quality of Life

QOL is more widely studied in caregiver literature than burden or depression, although these three areas combined compose the most studied outcomes in dementia caregiver research (Schulz & Martire, 2004). In the occupational therapy literature, Dooley and Hinojosa (2004) and Graff et al. (2007) found significant improvement in QOL after in-home occupational therapy intervention and at follow-up. In contrast, Ávila et al. (2004) did not find significant changes in QOL after a 14-wk support group in a clinical setting.

In these studies, the length of the intervention did not appear to have an impact. However, the two significant findings throughout the literature related to caregiver QOL occurred as a result of in-home services from occupational therapists as opposed to a support group setting at a facility. The significant finding from the in-home services alone is interesting to note because this current study took place in a group setting, although the subject matter was related specifically to ADLs, which typically take place within the home setting. Therefore, this study provides novel information within the caregiving literature that supports improved QOL after providing caregiver training in a workshop setting. The significant change in the physical health domain is supported by findings from de Vugt and Verhey (2013), suggesting that teaching role management techniques, which include a large component in ADLs, will in fact foster self-care abilities and promote physical health and well-being. It is also important to note that ADLs, the focus of the study, is one of the main facets of the WHOQOL–BREF physical health domain, which was the only domain that showed a significant improvement 3 mo posttest in the intervention group.

Burden and Depression

Burden and depression are factors that are prevalent within the dementia caregiving literature (Alzheimer’s Association Women and Alzheimer’s Poll, 2010, as cited in Alzheimer’s Association, 2013); yet, evidence in effecting change on burden and depression for caregivers of people with dementia in the occupational therapy literature is limited. The current study is the first known study to test whether knowledge provided to caregivers mediates burden and depression. Findings suggest that although knowledge improved significantly for the intervention group, depression was maintained and, therefore, did not increase. Depression scores for the intervention group remained as a mean score of 1, which indicates minimal depression.

Caregiver burden in the current study was maintained and, like depression, did not increase over a 3-mo follow-up period. The mean burden score of the intervention group remained unchanged and averaged a score of 2, which indicates moderate burden. Because most caregivers in the current study reported the minimal scores available on both the depression and the burden measures, most of the sample had little to no burden and normal to mild depression. Schulz et al. (2002) suggested that when caregivers participate in research with pretest scores near normal, it is not surprising if improvement is not found. Therefore, although it was unlikely for burden or depression scores to improve, it was positive that neither of these factors increased, which was the stated hypothesis.

Occupational Performance and Satisfaction

The COPM has been used to access occupational performance and satisfaction in the caregiver population by both McGrath, Mueller, Brown, Teitelman, and Watts (2000) and Graff et al. (2003). Although McGrath et al.’s qualitative exploratory study demonstrated the clinical utility of the COPM tool with this population, Graff et al. used the occupational performance and satisfaction scales as outcomes of their caregiver training study and found significant improvements in both areas. Participants in Graff et al.’s study had a median performance and satisfaction score of 5 at pretest and 7 at posttest (out of a possible 10 points) compared with a median COPM score of 8 at both pretest and posttest in the current study. Because our participants had higher scores at baseline, there was less potential to improve (Table 3).

Limitations

Because the sample was from the metropolitan Detroit area, consisted of only 36 participants, had a limited representation of diverse ethnic populations, and had an overwhelmingly larger number of women than men, results are not generalizable to the general caregiver population of people with dementia. The sample is too small to determine any potential differences with the limited diversity represented. Moreover, the ADL Knowledge Test was not a standardized measure. Although there was no significant difference between the intervention and control groups, the low baseline scores in burden and depression and high scores in performance and satisfaction allowed little room for improvement.

Future Research

Although previous studies have suggested that improving self-care abilities may promote physical health and well-being, the physical health component of QOL should be further explored because significance was found 3 mo posttest with the intervention group. Incorporation of a required competency check (i.e., as opposed to simply a demonstration and optional practice session) would provide a structured hands-on practice approach that may affect occupational satisfaction and performance. After reexamining the three modules, communication is a different instrumental activity of daily living task from eating, feeding, and nutrition and, therefore, should be evaluated as a separate area on the ADL Knowledge Test rather than together in the same module. Moreover, early, middle, and late stage caregiver stratification may counter the ceiling effect present with outcome scores of burden, depression, QOL, and occupational performance and satisfaction. Finally, incorporating a third group that receives standard care in booklet form only and a fourth group that receives the intervention along with follow-up phone calls or check-ins may better examine confidence and whether the presentation alone contributed to the improved confidence of the control group.