Effectiveness of Powerful Tools for Caregivers on Caregiver Burden and on Care Recipient Behavioral and Psychological Symptoms of Dementia: A Randomized Controlled Trial

Abstract

Objective:

The aim of this study was to evaluate the effectiveness of a psychoeducational intervention, Powerful Tools for Caregivers (PTC), for family caregivers of individuals with dementia.

Design:

A pragmatic, two-arm randomized controlled trial compared the PTC intervention, as delivered in practice, to usual care. Participants randomized to usual care functioned as a control group and then received the PTC intervention.

Interventions:

PTC is a six-week manualized program that includes weekly two-hour classes in a group setting facilitated by two trained and certified leaders. The educational program helps caregivers to enhance self-care practices and manage emotional distress.

Setting and Participants:

Two stakeholder organizations delivered the intervention in community settings. Participants were family caregivers of individuals with dementia recruited from the community in Florida.

Methods:

Primary outcomes were caregiver burden and behavioral and psychological symptoms of dementia of the care recipient. Secondary outcomes included caregiver depressive symptoms, self-efficacy, self-rated health, and life satisfaction. Measures were collected at baseline (n= 60 participants), post-intervention (n= 55), and at 6-week follow-up (n= 44).

Results:

Intent-to-treat analyses found PTC reduced caregiver burden (d= −0.48), depressive symptoms (d= −0.53), and increase self-confidence (d= 0.68), but found no significant benefit for behavioral and psychological symptoms of dementia in care recipients. PTC was rated highly by participants and program attrition was low, with 94% of caregivers completing at least four of the six classes.

Conclusions and Implications:

Although no significant effects were found for behavioral and psychological symptoms of dementia, this trial supports the effectiveness of PTC to improve caregiver outcomes as delivered in the community.

Brief summary

Powerful Tools for Caregivers is an effective psychoeducational program as delivered in practice settings by community stakeholders.

Alzheimer’s disease and related dementias are highly prevalent and disabling conditions that pose significant burden on individuals with dementia and their families¹. Informal family caregivers provide most of the care for the individuals with dementia, and while caregivers can benefit from caregiving², some are at high risk for burden, depressive symptoms, and poor health outcomes³. Interventions developed to address the challenges of caregiving have shown small to moderate favorable effects for families⁴. Multicomponent, psychoeducational interventions that require active participation of caregivers have shown broad effects⁵, with improved outcomes for the caregiver and care recipient⁶. For example, a 6-session support and counseling intervention for family caregivers significantly delayed nursing home placement of the care recipient with dementia⁷. Most research to date, however, has focused on developing and testing the efficacy of new interventions. Less research has considered the effectiveness of these interventions as implemented in practice by community stakeholders. A review published in 2015 suggested that only 3% of over 200 interventions for caregivers have been examined in a translational context⁴. Treatments that are sustainable as implemented by front-line community stakeholders can potentially bridge the large research-practice gap⁴.

To advance knowledge on a caregiver intervention currently implemented in social and clinical settings, we examined the effectiveness of Powerful Tools for Caregivers (PTC)⁸ as delivered by community organizations. PTC is a psychoeducational intervention for caregivers of patients with chronic illnesses that focuses on caregiver self-care, management of emotion, improved communication, self-confidence in coping with caregiving demands, and use of community resources⁸. The program is delivered in a group format, once a week over six consecutive weeks, and is free of cost for caregivers. PTC was developed from the widely disseminated Chronic Disease Self-Management Program (CDSMP)⁹, and both are theoretically grounded in Bandura’s social cognitive theory¹⁰. The Administration on Aging/Administration for Community Living has recognized PTC as an evidence-based disease prevention and health promotion program. Indeed, about ten studies indicate that PTC is efficacious in improving caregiver confidence and indicators of psychological well-being^11-19. However, few of these previous studies were randomized controlled trials, and most did not focus on dementia caregivers or test the effect of PTC on care recipients’ behavioral and psychological symptoms of dementia (BPSD). BPSD are common behavioral expressions that often pose a significant burden on caregivers.

The aim of this study was to assess the effectiveness of PTC as implemented by community stakeholders through a pragmatic²⁰ randomized controlled trial. Our primary outcomes were caregiver burden and the BPSD in the care recipient. Based on previous evidence^11-19, we hypothesized that compared to usual care, caregivers participating in PTC would experience decreased burden and similar improvements on related secondary outcomes, such as depressive symptoms and self-rated health. Our hypothesis about BPSD is more tentative, but caregiver factors play an important role in BPSD and are potentially modifiable, which make them a target for interventions to improve the health and well-being of the care recipient with dementia^21,22. Indeed, reductions in problematic behaviors have been achieved with multimethod interventions centered on caregivers^6,23. While PTC does not specifically address BPSD, two studies have found PTC to reduce caregiver reactions to BPSD^11,14. We hypothesized PTC would reduce BPSD in the care recipient as a secondary benefit from the caregivers’ reduced burden, improved communication and recognition of emotions, and the development of coping skills that may reduce manifestation and avoid worsening of BPSD.

Methods

Study Design

A pragmatic, two-arm randomized controlled trial examined the effectiveness of PTC as compared to usual care. The control group received a delayed PTC intervention (supplementary Figure). We evaluated PTC as implemented by community organizations to obtain evidence of the program’s effectiveness in real-life community settings. The study was registered with ClinicalTrials.gov (Identifier: NCT02697721) and approved by the Florida State University Institutional Review Board (Human Subjects Committee Number: 2017.22631).

Procedures

The study and the PTC program were advertised in the community using local newspapers and magazines, announcements at local events, and flyers at memory disorder clinics, libraries, places of worship, retirement communities, and senior centers. Potential participants were screened over the phone using the inclusion/exclusion criteria described below. Caregivers who met the inclusion criteria and were interested in participating in the study met with a research assistant who explained the study and obtained written consent.

All study participants were placed on a list of individuals interested in PTC. Participants in the study were in the same list as non-study participants, who could be individuals not interested in the research or did not meet the inclusion criteria. The community organizations maintained information on preferred time, day, and location for each caregiver and worked to find an average of ten caregivers who could participate in the program at a specific time, day, and location. The program was conducted at various locations in the community. Respite was available to facilitate attendance and PTC classes were free for all caregivers.

Inclusion and Exclusion Criteria

The target participants were informal caregivers of persons with dementia who exhibited BPSD. Inclusion criteria: caregivers 18 years or older who live with or provide care for the care recipient for an average of at least four-hours per day; caring for persons with a physician diagnosis of dementia (mild to moderate stage of dementia); caring for a care recipient who experiences BPSD that bothers or upsets (at least “moderate”) the caregivers as assessed with six-items adapted from the Revised Memory and Behavior Problems Checklist (RMBPC)²⁴. Exclusion criteria: care recipients who were bed-bound or had advanced dementia; Functional Assessment Staging²⁵ stage seven; caregivers with cognitive, hearing, visual, or other physical impairments that lead to difficulty with informed consent process, assessment, or participation in the intervention; caregivers who were involved in another study that addressed caregiver burden or behavioral expressions or who had already participated in PTC.

Randomization

Participants were randomized to either the experimental or control group with a 1:1 allocation ratio (block size of four and two strata to keep balance within each of the community organizations). The group assignment was concealed from the research team (except the project statistician) and participants until an envelope was opened after providing informed consent; there was no blinding thereafter.

Data Collection

The experimental group was assessed (1) within one-week before the start of the intervention, (2) within one week after the end of the intervention, and (3) at least six-weeks after the completion of the intervention. The control group’s participation in PTC was delayed for at least six-weeks to allow for the first two assessments of the study. The control group was assessed (1) at the time they signed the consent, (2) six weeks after the first assessment, and (3) within one-week after the end of the intervention (supplementary Figure).

The assessments were completed through a combination of a face-to-face interview and a self-reported questionnaire. To help with the retention and to compensate for the time spent completing the assessments, participants received $30 each for the first and second assessments and $40 for the third assessment.

Intervention

PTC is a program implemented across the United States and other countries by collaborations with community-based organizations. Two trained and certified PTC leaders facilitate the six classes by following the manualized program. The main themes of each weekly two-hour class are: (1) Taking Care of You; (2) Identifying and Reducing Personal Stress; (3) Communicating Feelings, Needs, and Concerns; (4) Communicating in Challenging Situations; (5) Learning from Our Emotions; and (6) Mastering Caregiving Decisions. Qualitative research indicates that the program is well received, meets participants’ needs, and is culturally appropriate for African-Americans and Latinx^12,26. More information on the program and on the training of PTC leaders is available at https://www.powerfultoolsforcaregivers.org/.

The PTC intervention was delivered by PTC certified leaders who were staff at the community organizations. About 80% of participants received PTC from the Alzheimer Project, Inc., a non-profit organization that serves caregivers by providing services at no cost, including respite, support groups, and counseling. About 20% of our study’s participants took part in PTC organized at a local Continuing Care Retirement Community.

Measures

The primary outcome measures were caregiver burden and care recipient’s BPSD. Caregiver burden was measured with the 22-item Zarit Caregiver Burden Scale²⁷. The BPSD were measured with the 29-item Cohen-Mansfield Agitation Inventory (CMAI)²⁸ and the 24-item frequency and reaction scales of the RMBPC²⁴. Secondary outcomes assessed several aspects of caregiver functioning, including: Caregiver depressive symptoms measured with the 10-item version of the Center for Epidemiologic Studies Depression Scale (CESD)²⁹; Caregiver confidence with the 11-item Caregiving Self-Efficacy Scale¹²; Caregiver self-rated health and life satisfaction assessed with single-item measures; and the 12-item Neuropsychiatric Inventory (NPI)³⁰. The baseline assessment included questions on basic sociodemographic information. The post-PTC assessment included a treatment satisfaction questionnaire.

Data Analysis

Sample size calculations were based on the assumption of a 15% attrition rate by the second assessment. We estimated that a total sample size of 60 (30 experimental and 30 control) would provide 80% power (α=0.05) to detect a medium effect size (d=0.5) for a repeated measure statistical test and >80% power to detect a large effect (d=0.8) for a between-groups statistical test. We stopped recruitment when we reached 60 individuals who completed the first assessment, as planned. No interim analyses for efficacy or futility were done.

T-test and χ² were used to compare the baseline characteristics between groups. Repeated measures analysis of variance was used to analyze the effect of time, group (experimental, control), and the interaction between the two factors. Effect sizes were computed as Cohen’s d³¹. The main study endpoints were the changes between time 1 and 2, but we also report the changes between time 2 and 3. Similar to previous studies on caregiver interventions^32-34, we used 0.5 SD of time 1 as a criterion of improvement from time 1 to 2 to evaluate clinical significance. For each outcome, we calculated the net improvement (percentage of participants who improved ≥0.5 SD - percentage of participants who worsened ≥0.5 SD). We included all participants with follow-up data in the analyses, following the intent-to-treat principles.

Results

Of the 169 caregivers screened, 73 (43%) met eligibility criteria, were interested in participating in the study, and were randomly assigned (Figure 1, CONSORT study flow diagram). Of the 37 caregivers randomized in the experimental group, 24 enrolled in PTC and completed the first assessment, which was done within one week from the first PTC class. All 36 caregivers randomized to the control group completed the first assessment, which was done immediately after the randomization and explains the lack of attrition at first assessment. Of the 60 participants who completed the first assessment, 55 (92%) completed the second assessment, and 44 (73%) completed the third assessment. Of the 52 caregivers who started the intervention, 49 (94%) attended four or more classes (M=5, in both groups), and three participants quit after the first or second class. The participants were enrolled in the study from February 2016 to January 2018. Most participants were lost from data collection before they started the intervention (13 in the experimental and 8 in the control group). Scheduling conflicts were the most common reason for not receiving the intervention (one participant in the control group provided a health-related reason for not participating in the intervention).

Figure 1.

Consort diagram illustrating the flow of participants through each stage of the study. “Discontinued PTC” refers to participants who completed only one or two of the six PTC classes. “Analyzed” refers to the number of participants with primary outcome data between Assessment 1 and Assessment 2.

Table 1 presents the demographic and clinical characteristics of the 60 participants who completed the first assessment. There were no statistically significant differences on the baseline characteristics between the two groups.

Table 1.

Baseline descriptive data of caregivers in total sample and by groups.

	Total (n=60)	Experimental (n=24)	Control (n=36)	P-value
Caregiver
Age	66.66 (11.94)	68.17 (11.37)	65.66 (12.36)	.43
Education	15.77 (2.83)	15.92 (2.86)	15.67 (2.85)	.74
Sex (female)	46 (77%)	19 (75%)	27 (79%)	.71
European-American	49 (82%)	20 (83%)	29 (81%)	.79
Rural residence	10 (17%)	3 (13%)	7 (19%)	.63
Married	47 (78%)	21 (87%)	26 (72%)	.51
Employed (full or part time)	17 (30%)	6 (25%)	11 (32%)	.16
Same household	42 (70%)	20 (83%)	22 (61%)	.07
Relationship (spouse)	33 (55%)	16 (67%)	17 (47%)	.14
Relationship (child)	22 (37%)	8 (33%)	14 (39%)	.66
Caregiving length: >1 year	51 (86%)	18 (78%)	33 (92%)	.14
Providers of personal care	34 (57%)	14 (58%)	20 (56%)	.38
Help received	2.52 (1.54)	2.63 (1.50)	2.44 (1.59)	.66
Satisfaction with help	3.04 (1.54)	3.26 (1.71)	2.88 (1.41)	.37
Outcome variables [range]
Burden (Zarit) [0-88]	47.33 (15.36)	48.89 (15.58)	46.29 (15.34)	.53
Depressive symptoms (CESD) [0-30]	10.35 (5.13)	10.99 (5.47)	9.95 (4.93)	.45
Self-efficacy/Confidence [11-110]	73.22 (16.52)	70.15 (18.44)	75.27 (15.04)	.24
Life-Satisfaction [1-5]	3.56 (1.15)	3.61 (1.23)	3.53 (1.11)	.79
Self-Rated Health [1-5]	3.10 (0.92)	3.25 (0.99)	3.00 (0.86)	.30
Total (CMAI) [29-203]	53.92 (17.87)	50.06 (18.07)	56.49 (17.52)	.17
Physically Aggressive (CMAI) [11-77]	13.47 (5.04)	13.13 (3.69)	13.70 (5.81)	.67
Physically Non-Aggressive (CMAI) [10-70]	21.23 (8.09)	18.88 (8.33)	22.84 (7.62)	.06
Verbally Aggressive (CMAI) [3-21]	5.00 (3.13)	4.96 (3.43)	5.03 (2.96)	.93
Verbally Non-Aggressive (CMAI) [5-35]	14.58 (6.02)	13.10 (5.75)	15.56 (6.08)	.12
RMBPC Frequency [0-96]	39.38 (15.18)	35.03 (17.57)	42.28 (12.80)	.07
RMBPC Reaction [0-96]	24.15 (16.34)	20.33 (18.78)	26.59 (14.32)	.15
Neuropsychiatric Symptoms (NPI) [0-12]	5.92 (2.75)	5.08 (2.89)	6.50 (2.53)	.051
Care recipient with dementia
Age	80.67 (8.63)	78.71 (9.64)	81.97 (7.75)	.15
Sex (female)	31 (52%)	9 (38%)	22 (61%)	.07

Note: Values in the table are numbers and (percentages) or mean and (SD). Same household and relationship variables are in relation to care recipient. Daily personal care includes dressing/undressing, or bathing/ showering. Help received (including from family members, friends, and paid providers) and satisfaction with help received were on a scale 0 to 5, wit 5 indicating getting “a lot of help” and “very satisfied”. CESD = Center for Epidemiologic Studies Depression Scale. RMBPC = Revised Memory and Behavior Problems Checklist; CMAI = Cohen-Mansfield Agitation Inventory; NPI = Neuropsychiatric Inventory.

Main Outcomes

Table 2 presents the mean and SD at each assessment point. Table 3 presents effect sizes for comparisons across times and groups. Significant improvements were found between time 1 and 2 in the experimental group with medium to large decreases in caregiver burden (d=−0.85), depressive symptoms (d=−0.70), and increases in caregiver self-efficacy (d=0.94). These changes in the experimental group were significantly different as compared to the changes in the usual care control group. The group labelled as “control” participated in PTC classes between time 2 and 3 (Figure 1 and supplementary Figure) and we observed medium increases in caregiver self-efficacy (d=0.48) and life-satisfaction (d=0.66), but smaller and not significant changes on burden (d=−0.18) and depressive symptoms (d=−0.39). Contrary to our hypothesis, we found no significant changes on measures of care recipients’ BPSD.

Table 2.

Descriptive statistics for the outcomes across times and groups.

Group Time Intervention	Experimental Group			Control Group
	Time 1	Time 2	Time 3	Time 1	Time 2	Time 3
	PTC	Usual Care		Usual Care		PTC
Caregiver
Burden (Zarit)	51.78 (14.03)	41.36 (15.46)	41.51 (15.15)	47.18 (15.33)	45.17 (14.86)	42.28 (12.58)
Depressive symptoms (CESD)	11.72 (5.15)	8.03 (4.79)	9.11 (4.70)	10.15 (4.96)	9.96 (5.15)	8.19 (4.66)
Self-efficacy/Confidence	69.88 (13.64)	85.33 (10.28)	82.71 (14.76)	74.72 (15.28)	73.09 (15.06)	80.00 (16.55)
Life-Satisfaction	3.50 (1.24)	3.50 (1.10)	3.61 (1.20)	3.47 (1.11)	3.24 (1.26)	3.92 (0.85)
Self-Rated Health	3.33 (1.02)	3.43(0.87)	3.44 (0.78)	2.94 (0.85)	2.91 (0.97)	3.04 (0.92)
Care recipient symptoms
Total (CMAI)	50.71 (19.00)	48.43 (15.93)	50.22 (15.57)	56.64 (17.65)	55.38 (15.68)	53.38 (12.74)
Physically Aggressive (CMAI)	13.05 (3.67)	13.28 (3.22)	12.52 (2.73)	13.68 (5.94)	13.86 (4.57)	13.23 (3.99)
Physically Non-Aggressive (CMAI)	19.00 (8.64)	17.97 (7.62)	19.88 (8.72)	22.74 (7.67)	22.05 (7.98)	21.85 (8.36)
Verbally Aggressive (CMAI)	5.24 (3.59)	5.10 (2.84)	4.94 (2.56)	5.12 (3.02)	4.85 (2.35)	4.23 (2.01)
Verbally Non-Aggressive (CMAI)	13.43 (6.02)	12.08 (5.40)	12.88 (4.81)	15.76 (6.02)	14.85 (5.45)	14.08 (4.82)
RMBPC Frequency	35.80 (18.35)	35.98 (17.45)	37.90 (17.25)	41.18 (12.26)	38.95 (12.17)	36.66 (12.68)
RMBPC Reaction	22.02 (20.92)	22.84 (20.35)	21.87 (22.12)	26.18 (14.09)	21.52 (16.17)	17.56 (12.24)
Neuropsychiatric Symptoms (NPI)	5.19 (2.99)	5.76 (3.06)	6.44 (3.73)	6.53 (2.61)	6.39 (2.54)	6.08 (2.76)

Note. For time 1 and 2, the n=34 for control and n=21 for experimental. For time 3, the n=26 for control and n=18 for experimental. CESD = Center for Epidemiologic Studies Depression Scale; RMBPC = Revised Memory and Behavior Problems Checklist; CMAI = Cohen-Mansfield Agitation Inventory; NPI = Neuropsychiatric Inventory.

Table 3.

Effect sizes (Cohen’s d) of changes across times and groups.

	Time 1 and Time 2			Time 2 and Time 3
	Experimental	Control	Difference	Experimental	Control	Difference	Combined a
Caregiver
Burden (Zarit)	−0.85*	−0.22	0.63*	−0.18	−0.18	0.00	−0.48
Depressive symptoms (CESD)	−0.70*	−0.04	0.66*	0.07	−0.39	0.46	−0.53
Self-efficacy/Confidence	0.94*	−0.11	1.05*	0.00	0.48*	0.48	0.68
Life-Satisfaction	0.00	−0.21	0.21	0.05	0.66*	0.61	0.37
Self-Rated Health	0.10	−0.05	0.15	−0.09	0.22	0.31	0.17
Care recipient symptoms
Total (CMAI)	−0.34	−0.14	0.20	0.29	−0.17	0.46	−0.24
Physically Aggressive (CMAI)	0.11	0.05	0.06	−0.20	−0.25	0.06	−0.09
Physically Non-Aggressive (CMAI)	−0.23	−0.13	0.10	0.47	0.05	0.42	−0.08
Verbally Aggressive (CMAI)	−0.09	−0.18	0.08	0.10	−0.39	0.49	−0.26
Verbally Non-Aggressive (CMAI)	−0.37	−0.24	0.13	0.28	−0.19	0.47	−0.27
RMBPC Frequency	0.02	−0.18	0.21	0.22	−0.33	0.55	−0.17
RMBPC Reaction	0.06	−0.32	0.37	−0.11	−0.21	0.10	−0.09
Neuropsychiatric Symptoms (NPI)	0.30	−0.06	0.36	0.16	−0.18	0.34	0.04

Note: See note of Table 2 for abbreviations and sample sizes.

^aThe combined effect is the weighted average of effects in the intervention group between times 1 and 2 and effects in the control group between times 2 and 3 (i.e., pre and post the Powerful Tools for Caregivers in both the intervention and control groups).

^*p< 0.05 from repeated measure ANOVA.

Table 4 presents the proportions of caregivers who improved or worsened and the net improvement between time 1 and 2. In the experimental group, over 50% of participants showed improvements ≥0.5 SD on caregiver burden, depressive symptoms, and self-efficacy and 14% or less worsened on these three outcomes. There were no similar improvements in the control group and the net improvement difference was 41% for caregiver burden, 32% for depressive symptoms, and 70% for self-efficacy. There were only small differences in net improvement for the other outcomes.

Table 4.

Clinical Significance of changes across Time 1 and Time 2.

	Experimental (n = 21)			Control (n = 34)			Net Improvement difference
	Improved	Worsened	Net Improvement	Improved	Worsened	Net Improvement
Burden (Zarit)	57%	5%	52%	32%	21%	11%	41%
Depressive symptoms (CESD)	52%	14%	38%	32%	26%	6%	32%
Self-efficacy/Confidence	76%	10%	67%	23%	26%	−3%	70%
Life-Satisfaction	19%	19%	0%	12%	24%	−12%	12%
Self-Rated Health	33%	24%	10%	15%	15%	0%	10%
Total (CMAI)	33%	24%	10%	38%	35%	3%	7%
Physically Aggressive (CMAI)	10%	14%	−5%	6%	9%	−3%	−2%
Physically Non-Aggressive (CMAI)	33%	24%	10%	47%	32%	15%	−5%
Verbally Aggressive (CMAI)	10%	0%	10%	6%	0%	6%	4%
Verbally Non-Aggressive (CMAI)	5%	0%	5%	6%	0%	6%	−1%
RMBPC Frequency	19%	29%	−10%	29%	15%	14%	−24%
RMBPC Reaction	19%	24%	−5%	47%	12%	35%	−40%
Neuropsychiatric Symptoms (NPI)	10%	29%	−19%	23%	26%	−3%	−16%

Note: Clinically significant improvement and worsening was defined as a change of 0.5 SD or more from time 1 to time 2 assessment. Net improvement is the proportion of participants who improved minus the proportion of participants who worsened. Net Improvement difference is the difference between the experimental and control group. See note of Table 2 for abbreviations.

In post-PTC assessments, participants rated highly the overall program (M=9.11; range: 1=poor to 10=excellent) and each weekly class, ranging from 8.26 for week 4 to 8.95 for week 3. Both the classes and program were rated to be about the right duration for 80% of participants; 20% found the classes or the program too short, and none found it too long. When asked to select which of the “tools” they felt will really help, the top choices were “Action plan” (selected by 76%), “Positive self-talk” (70%), and “Relaxation tools” (64%). Most caregivers strongly agreed that support from other class participants was helpful (M = 4.76; range 1-5). Finally, caregivers reported high post-intervention confidence in dealing with various aspects of caregiving (M>4; range 1-5), particularly “Identify ways to locate community resources”, “Manage stress”, and “Ask for help”.

Discussion

This pragmatic, randomized controlled trial examined the effectiveness of a psychoeducational program on caregiver and care recipient outcomes. We found that PTC was effective in reducing caregiver burden (d=−0.48) and depressive symptoms (d=−0.53), increasing self-efficacy (d=0.68), but had no significant effect on care recipients’ BPSD. These findings support the value of PTC for the well-being of caregivers as implemented by community organizations.

Our results for the caregiver outcomes were largely consistent with previous PTC findings based on different study designs^11-19. For example, a study without a control group found significant pre-post reductions in depressive symptoms (d=−0.57) and increases in self-efficacy (d=1.07)¹². Our results for PTC were also generally consistent with other caregiver interventions, such as for depressive symptoms^35,36. As in previous studies, we examined the proportion of individuals with clinically significant changes. We found that PTC was associated with 52% and 38% net improvements on caregiver burden and depressive symptoms. By comparison, a multicomponent intervention in white, African-American, and Latinx caregivers found net improvements between 9% and 31% for burden and between 22% and 39% for depressive symptoms³². Similarly, a telephone-based intervention found a net improvement of 19% for burden and 25% for depressive symptoms³⁴. While we found small and non-significant changes on life satisfaction (d=0.37) and self-rated health (d=0.17), overall PTC is at least as effective as other interventions, and the benefits are evident as implemented in practice.

A primary aim of our study was to test whether PTC would have indirect benefits for the care recipient, particularly for agitation-related BPSD. Personality and behavioral changes are common with the progression of dementia^37-39, and BPSD are particularly challenging for families and paid care providers^1,40. Agitation-related symptoms often lead to the use of pharmacological interventions (e.g., antipsychotics) that have an unfavorable risk-benefit profile^1,41. Similar to other non-pharmacological interventions^23,42, we expected PTC to reduce or minimize BPSD overtime, but we found no significant effects of PTC on BPSD. Our null results contrast with two previous PTC studies that found some improvements on RMBPC reaction scores^11,14. Besides our modest sample size, this null finding could be due to increased awareness and reduced stigma in reporting BPSD after completion of PTC. It is also possible that potential benefits may emerge over time and we were unable to detect them with our short follow-up. Conversely, the null findings may suggest that the broad curriculum of PTC attend less to this aspect of caregiving (note that PTC was not specifically designed to address BPSD) and more to caregiver self-care. Thus, medium-to-large benefits are unlikely. Tailored interventions are likely to produce better results for BPSD, but these might require more resources and increase the challenges of implementing them into practice^42,43.

Our pragmatic trial allowed us to evaluate other aspects of PTC as delivered in real-life settings. Of note, about 94% of those who initiated the program attended most classes and were highly satisfied with the program. While this low attrition is a major attractive feature of PTC, almost 30% of caregivers who initially enrolled did not initiate the program. Many caregivers pointed to the difficulty of attending the classes. Even for those who were able to attend the program, it generally took weeks before they found a class that fit their schedule. This underscores the burden faced by caregivers and the difficulties of seeking help, a problem that is likely to be particularly severe in rural communities. Online or telephone interventions might be better able to reach caregivers who have difficulty attending traditional in-person group classes^34,44,45.

Limitations and Future Directions

Among the limitations of our study was the relatively modest sample size and the relative low proportion of caregivers from minority groups or with low education, which potentially reduces the generalizability of the findings. Larger multi-site studies are needed to provide more robust evidence, but our pragmatic trial reached findings broadly consistent with previous PTC research. Data on the effectiveness and sustainability of the program need to be gathered from comparative effectiveness research in which PTC is compared not only to usual care, but to other active interventions⁴⁴. While we found that the benefits were mostly retained at least six weeks post intervention, studies with longer follow-up are needed. Furthermore, we did not include outcomes such as placement in a nursing home or healthcare utilization that are of major relevance to payers. Finally, cost-effectiveness of the program needs to be evaluated. PTC was developed from CDSMP, which was estimated to cost $350 per participant and produced significant health care savings⁴⁶. Future research should evaluate whether PTC is similarly cost-effective.

Conclusions and Implications

The results of this pragmatic trial confirmed the effectiveness of PTC in decreasing caregiver burden, decreasing depressive symptoms, and increasing caregiver confidence. Although the behaviors and health of caregivers and care recipients are intimately intertwined, we did not find PTC to have significant benefits on care recipients’ BPSD. Still, because of the program’s broad appeal, PTC remains relatively accessible, it has been sustainable across a wide range of communities over the past 20 years, and provides an effective and viable approach toward addressing the needs of the growing caregiver population.