Enhanced Coping and Self-Efficacy in Caregivers of Stem Cell Transplant Recipients: Identifying Mechanisms of a Multimodal Psychosocial Intervention

Abstract

Background:

In a recent trial, a six-session intervention (BMT-CARE) that integrated medical information with cognitive-behavioral strategies improved quality of life (QOL), mood, coping skills, and self-efficacy in family/friend caregivers of hematopoietic stem cell transplantation (HCT) recipients. We examined whether improvements in coping and self-efficacy mediated the intervention effects on QOL and mood.

Methods:

From 12/2017 to 4/2019, we enrolled 100 caregivers of HCT recipients in a randomized clinical trial of BMT-CARE versus usual care. Caregivers completed self-report measures of QOL (CareGiver Oncology QOL questionnaire), depression and anxiety symptoms (Hospital Anxiety and Depression Scale), coping skills (Measure of Current Status), and self-efficacy (Cancer Self-Efficacy Scale-Transplant) at enrollment (prior to HCT) and 60 days post-HCT. We used causal mediation regression models to examine whether changes in coping and self-efficacy mediated intervention effects on QOL, as well as depressive and anxiety symptoms.

Results:

Improvements in 60-day QOL in patients assigned to BMT-CARE were partially mediated by improved coping and self-efficacy (indirect effect=6.93, SE=1.85, 95% CI [3.71, 11.05]). Similarly, reductions in 60-day depression and anxiety symptoms were partially mediated by improved coping and self-efficacy (indirect effect depression=−1.19, SE=0.42, 95% CI [−2.23, −0.53]; indirect effect anxiety=−1.46, SE=0.55, 95% CI [−2.52, −0.43]). Combined improvements in coping and self-efficacy accounted for 67%, 80%, and 39% of the total intervention effect on QOL and depression and anxiety symptoms, respectively.

Conclusions:

Coping and self-efficacy are essential components of a brief psychosocial intervention that improves QOL and mood for caregivers of HCT recipients during the acute recovery period.

Precis:

A brief multimodal psychosocial intervention improved quality of life and reduced depression and anxiety symptoms for caregivers of patients undergoing stem cell transplant, compared to a usual care control group. These effects were partially mediated by improvements in coping skills and self-efficacy for caregivers assigned to the intervention.

Lay Summary:

We previously reported that a six-session program (BMT-CARE), focused on providing medical information, caregiving skills, and self-care and coping strategies, improved quality of life (QOL) and mood of caregivers of hematopoietic stem cell transplantation (HCT) recipients, compared to a group of caregivers who received care as usual. Using statistical models, our findings suggest that learning coping skills and improving self-efficacy were the most essential components of this program that likely led to better QOL and mood for those caregivers.

Introduction

Hematopoietic stem cell transplant (HCT), a common treatment for patients with hematologic malignancies, places substantial burden on family and close friend caregivers. With the patient’s prolonged hospitalization, intense recovery period, and adverse effects, caregivers are a critical source of practical and emotional support before, during, and after HCT. Caregivers monitor ongoing physical symptoms, administer complex medication regimens, provide emotional support, and attend numerous outpatient appointments, all while managing the uncertainty of the transplant outcome. Thus, caregivers of HCT recipients often experience tremendous caregiving burden, as well as decrements in quality of life (QOL) and mood.^1,2 Psychosocial interventions aiming to mitigate the negative effects of cancer caregiving have demonstrated some improvements in caregivers’ distress, mood, QOL, coping, and self-efficacy;³ however, results have been mixed.^4,5 These inconsistencies may be due to varying intervention content and timing employed across these trials. Moreover, we lack an understanding of the key mechanisms responsible for conferring benefit, which is necessary in order to maximize the scalability and potential for dissemination of such an intervention while minimizing clinician resources and participant burden.

Our group recently reported findings from a single-center randomized controlled trial evaluating the feasibility and preliminary efficacy of a brief multimodal psychosocial intervention (BMT-CARE) for caregivers of patients receiving HCT.⁶ BMT-CARE was informed by Van Houtven’s organizational framework for caregiving interventions,⁷ which outlines that a caregiver intervention may influence caregiver psychological and QOL outcomes via pathways of acquired caregiving skills in the clinical, management-related, or psychological realms. We found that BMT-CARE was feasible, and caregivers assigned to BMT-CARE reported better mood, QOL, self-efficacy, and coping skills, as well as reduced caregiving burden compared to the usual care control group. These findings suggest that it is possible to mitigate caregivers’ psychological distress at a time typically associated with substantial elevations.²

Given the positive effects of BMT-CARE on caregiver-reported outcomes, we sought to explore whether the psychosocial mechanisms targeted by our specific intervention components underlie these changes in mood and QOL. For example, to target caregiving self-efficacy, BMT-CARE provides psychoeducation of HCT throughout each phase to enhance a) preparedness for the transplant process, b) knowledge and expectations of symptoms, ongoing patient care needs, and overall treatment trajectory, and c) caregiving self-efficacy for performing tasks such as arranging medication and treatment schedules, managing difficult symptoms, and prioritizing self-care during stressful times. To target coping skills, BMT-CARE incorporates CBT-based skills such as behavioral activation, cognitive restructuring, relaxation training, and coping effectiveness training to promote psychological coping, reduce or tolerate distress, manage uncertainty; and optimize self-care via health behaviors and stress management. Per Van Houtven’s model,⁷ we hypothesized that the caregiver intervention (i.e., BMT-CARE) may influence caregiver outcomes (i.e., anxiety symptoms, depression symptoms, and QOL) through observed improvements in caregiving clinical and management skills (i.e., caregiving self-efficacy) and caregiving psychological skills (i.e., coping skills) from baseline to post-intervention.

Methods

Study Design

From 12/2017 to 4/2019, we recruited caregivers of patients undergoing autologous and allogeneic HCT at the Massachusetts General Hospital to participate in a non-blinded, randomized clinical trial of a multimodal psychosocial intervention (BMT-CARE) compared with usual HCT care alone (Clinicaltrials.gov identifier: NCT03328663). The Dana-Farber / Harvard Cancer Center (DF/HCC) Institutional Review Board approved the study, and all caregivers provided informed consent.

Participants

Eligible caregivers were a spouse, relative, or friend of a patient with a hematologic malignancy undergoing HCT who either lived with the patient or had in-person contact with them at least twice per week and was identified by the patient as the primary caregiver for HCT. Caregivers were also required to be at least 18 years old and able to read and respond to questions in English or with minimal interpreter assistance.

Procedures

Study staff reviewed the transplant clinic schedule to identify patients scheduled for an upcoming HCT consent appointment. With permission from the oncologist, the study staff approached consecutively eligible caregivers at the patients’ consent appointment to introduce the study and obtain written informed consent. Patients who did not have a caregiver present at this visit could identify a caregiver and give study staff permission to contact this person by telephone to discuss the study and obtain verbal consent. Within 72-hours following informed consent, enrolled caregivers completed baseline self-report measures via a HIPAA-compliant web-based survey tool or paper-based questionnaires. The DF/HCC Office of Data Quality then employed a computer-generated number sequence to randomize caregivers in a 1:1 manner, stratified by type of HCT (autologous or allogeneic). Caregivers repeated surveys at a mid-point follow-up at day +30 (+ 7-day window), and day +60 (+20-day window) post-HCT. For these analyses, we focused on the day +60 assessment following intervention completion.

The full trial and intervention procedures have been reported previously.⁶ Briefly, caregivers assigned to BMT-CARE participated in a six-session coping skills intervention that integrated HCT-related medical information with cognitive-behavioral strategies to enhance caregiver knowledge and skills across the transplant trajectory. Sessions took place approximately weekly for four weeks, followed by every other week for the remaining two sessions, with the first session occurring prior to HCT and the last occurring up to day +60 post-HCT. Each caregiver received a BMT-CARE workbook that included in-session and at-home practice exercises to reinforce caregiving skills (e.g., symptom and medical management, mobilization of social supports), cognitive-behavioral skills-based strategies (e.g., cognitive reframing, mindfulness, communication, and acceptance while living with uncertainty), and behavioral self-care strategies to promote caregiver physical and mental health. BMT-CARE was delivered by trained study therapists (i.e., licensed clinical psychologists, licensed social workers, clinical psychology fellows) and was administered in-person, via telephone, or via videoconference. Per institutional standards, caregivers assigned to usual care met with the transplant social worker (who was not one of the trained clinicians delivering the intervention) once prior to HCT to address their concerns. Following each intervention session, study therapists recorded caregiver attendance and the session length in minutes. Study therapists also rated the caregivers’ completion of home practice on a Likert-type scale of 1 (not complete) to 7 (complete).

Measures

Sociodemographic and clinical characteristics.

Caregivers reported their race, ethnicity, religion, marital status, income, and education level on a demographic questionnaire. Study staff collected clinical information such as the patients’ transplant type from the electronic health record.

QOL.

Caregivers reported their QOL using the Caregiver Oncology QOL questionnaire (CarGOQOL), a 29-item instrument that measures ten domains of QOL and generates a composite score (range 0–100), with higher scores indicating better QOL.⁸

Mood.

We used the Hospital Anxiety and Depression Scale (HADS)⁹ to assess anxiety and depression symptoms in the last week. The HADS is a 14-item measure consisting of two subscales for depression and anxiety symptoms, respectively. Each subscale ranges from 0 (no distress) to 21 (maximum distress), with higher scores indicating greater distress.

Coping skills.

We administered the Measure of Current Status Part A (MOCS-A) to assess current self-perceived status on several coping skills that are targeted by the intervention, such as relaxation, cognitive restructuring, and assertive communication. The 13 items in Part A are rated from 0 (strongly disagree) to 4 (strongly agree), with higher scores indicating better perceived coping ability and use of skills.¹⁰

Caregiving self-efficacy.

We used the Cancer Self-Efficacy Scale-Transplant (CASE-T) to examine caregivers’ expressed confidence in managing the impact of HCT. This 17-item measure ranges from 0–170, with higher scores indicating greater self-efficacy in transplant specific tasks and demands.¹¹

Statistical Analysis

Using STATA (version 9.3), we described caregiver sociodemographic and clinical characteristics with measures of central tendency. We calculated linear regressions to examine the effect of group assignment (BMT-CARE vs. usual care) on changes in the two potential intervention mechanisms (coping skills and caregiving self-efficacy) from baseline to day +60. We also examined the Pearson Product-Moment Correlation between change in coping and change in self-efficacy in order to assess collinearity. We then implemented causal mediation regression models with 5,000-iteration bias-corrected bootstrapping¹² to test whether the effects of BMT-CARE on QOL, depression symptoms, and anxiety symptoms at day +60 were mediated through changes in coping skills and self-efficacy from baseline to day +60.

Using causal mediation, we tested a multiple mediation model for each of the three outcomes with group assignment as the independent variable, change in both coping skills and self-efficacy as mediator variables, and either QOL, depression symptoms, or anxiety symptoms as the dependent variable. We adjusted each model for the baseline criterion score of the dependent variable. We interpreted the unstandardized regression coefficients based on a two-sided alpha of 0.05 and evaluated them as a measure of effect size.¹³ Bootstrapped estimates of the specific indirect effects and the total indirect effect (coping skills and self-efficacy) were interpreted with bias-corrected 95% confidence intervals (CI), in which any CI that did not contain zero was considered a significant indirect effect. The total indirect effects were used to test the mediation hypotheses; the magnitude of the total indirect (mediated) effect was interpreted as the amount by which QOL, depression symptoms, or anxiety symptoms were expected to increase or decrease indirectly through both coping and self-efficacy for caregivers assigned to BMT-CARE versus usual care. While the current study is an exploratory analysis, the overall study sample size (n=100) was based on the feasibility of obtaining preliminary data to determine effect sizes for caregiver outcomes and to inform sample sizes for a future multi-site study.⁶ When exploring preliminary intervention effects, prior studies suggest that a minimum of 30 participants are needed in each group in order to estimate a parameter with a power analysis.^14,15

Results

Caregiver Characteristics

We approached 138 caregivers for the study and enrolled a total of 100 (72.5%). Of these 100, eight caregivers became ineligible due to the patient no longer undergoing HCT (e.g., aborted transplant plan or patient death prior to planned transplant). Accounting for missing data on the MOCS-A and CASE-t, our total effective sample for analyses was 82 caregivers. Most caregivers were white (92.3%, 85/92), female (69.6%, 68/92), married to the patient (81.5%, 75/92), and had a median age of 61 years old (range 22–93). There were no meaningful differences in caregiver demographics between study groups at the baseline assessment (see Table 1). Sessions were a median of 55 minutes long (range=25–80).⁶ Of those assigned to the BMT-CARE intervention, 30 caregivers (67%) attended all 6 sessions, six (13%) attended 3–5 sessions, and two (4.4%) attended 1–2 sessions. Therapists rated caregivers’ home practice as mostly complete, on average (Median=5, Range=1–7).

Table 1.

Baseline Caregiver and Patient Characteristics

Caregiver Characteristics	BMT-CARE (N = 45)	Usual Care (N = 47)
Age, Median (range)	58 (23–75)	62 (22–93)
Female Sex, n (%)	31 (68.9%)	33 (70.2%)
Relationship to Patient, n (%)
Married	36 (80.0%)	39 (83.0%)
Child	5 (11.1%)	1 (2.1%)
Parent	3 (6.7%)	3 (6.4%)
Sibling	1 (2.2)	3 (6.4%)
Divorced	0 (0.0)	1 (2.1%)
Race, n (%)
White	42 (93.3%)	43 (91.5%)
American Indian	0 (0.0)	2 (4.3%)
Black	2 (4.5%)	1 (2.1%)
Other	1 (2.2%)	1 (2.1%)
Hispanic Ethnicity, n (%)	1 (2.2%)	1 (2.1%)
Education, n (%)
High school	10 (22.2%)	11 (23.4%)
College	24 (53.3%)	27 (57.4%)
Post-Graduate Degree	11 (24.5%)	9 (19.2%)
Employment, n (%)
Employed	22 (48.9%)	27 (57.4%)
Retired	14 (31.1%)	14 (29.8%)
Caring for Home or Family	4 (8.9%)	4 (8.5%)
Unemployed	2 (4.4%)	2 (4.3%)
Disabled	3 (6.7%)	0 (0.0%)
Patient Diagnosis, n (%)
Acute Leukemia	17 (37.8%)	15 (31.9%)
Lymphoma	10 (22.2%)	11 (23.4%)
Multiple Myeloma	10 (22.2%)	7 (14.9%)
Myelodysplastic/Myeloproliferative Neoplasms	8 (17.8%)	14 (29.8%)
Transplant Type, n (%)
Allogeneic	28 (62.2%)	31 (65.9%)
Autologous	17 (37.8%)	16 (34.1%)
QOL (CarGOQOL), mean (SD)	73.24 (10.34)	77.66 (12.61)
Caregiving burden (CRA), mean (SD)	55.27 (11.93)	47.28 (11.43)
Anxiety Symptoms (HADS-A), mean (SD)	7.80 (3.92)	6.40 (4.09)
Depression Symptoms (HADS-D), mean (SD)	4.92 (2.80)	3.51 (3.43)
Self-Efficacy (CASE-T)	138.52 (24.60)	146.78 (23.33)
Coping Skills (MOCS-A)	30.48 (8.73)	33.80 (8.63)

Note: All values were collected at the baseline assessment prior to study randomization.

BMT-CARE Intervention Effects on Changes in Caregiver Coping Skills and Self-Efficacy

Linear regression models revealed that, compared to caregivers in the usual care group, caregivers who received BMT-CARE reported a significant increase in coping skills (B=8.51, SE=1.89, 95% CI [4.75, 12.28], Cohen’s D = 1.02, p<.001) and self-efficacy (B=9.13, SE=3.94, 95% CI [1.29, 16.98], Cohen’s D = 0.75, p=.023) at day +60, adjusting for baseline values. The correlation between change in coping and change in self-efficacy was small (r=0.34). Caregivers in the BMT-CARE group had a mean increase in use of coping skills (M=5.27, SE=1.59, 95% CI [2.04, 8.49] compared to a mean reduction in coping skills among caregivers receiving usual care (M=−4.67, SE=1.20, 95% CI [−7.10, −2.23]; M_diff =9.94, SE_diff= 1.99). Caregivers in the BMT-CARE group also had a mean increase in caregiving self-efficacy (M=16.22, SE=4.08, 95% CI [7.96, 24.47]) compared to a slight mean increase in self-efficacy among caregivers receiving usual care (M=1.70, SE=3.11, 95% CI [−4.59, 8.01]; M_diff =14.52, SE_diff= 5.14). Given the significant effects of BMT-CARE on changes in coping skills and self-efficacy from baseline to day +60, we conducted causal mediation regression models to determine whether the previously reported effects⁶ of BMT-CARE on QOL, depression symptoms, and anxiety symptoms, were mediated by these changes.

Effects on QOL Partially Mediated by Changes in Coping Skills and Self-Efficacy

First, compared to usual care, BMT-CARE was associated with significant improvements in day +60 QOL (B=0.41, SE=0.10, 95% CI [0.21, 0.61], Cohen’s D = 0.90 p<.001), controlling for baseline QOL. Second, increases in both coping skills (B=0.52, SE=0.12, p<.000, 95% CI [0.29, 0.74]) and self-efficacy (B=0.15, SE=0.05, p=.001, 95% CI [0.06, 0.24]) from baseline to day +60 were significantly associated with higher QOL at day +60, controlling for QOL at the baseline assessment. Finally, with both group assignment and changes in coping skills and self-efficacy simultaneously entered in the model, the effect of BMT-CARE on day +60 QOL became non-significant (B=3.45, SE=2.35, p=.142, 95% CI [−1.16, 8.07]), indicating a potential mediation effect.

The causal mediation model assessing the effect of BMT-CARE on day +60 QOL revealed a significant partial mediation effect simultaneously through both changes in coping skills and self-efficacy, such that QOL increased by 6.93 points on the CarGOQOL through both improved coping skills and self-efficacy for caregivers assigned to BMT-CARE. The specific indirect effect for each mediator (changes in coping skills and self-efficacy) was significant. The total indirect effect accounted for 67% of the total effect of BMT-CARE on QOL. See Table 2 for estimates of the indirect effects, bootstrapped standard errors, and bias-corrected 95% CIs.

Table 2.

Mediation Effects

Indirect Effects of BMT-CARE Intervention on Quality of Life
	Effect	Bootstrapped SE	Bias-Corrected 95% CI
Indirect Effect: Coping Skills Δ	4.95*	1.57	[2.30, 8.47]
Indirect Effect: Self-Efficacy Δ	1.98*	1.07	[0.43, 4.87]
Total (Multiple) Indirect Effect	6.93*	1.85	[3.71, 11.05]
67% of the total effect on QOL can be attributed to the indirect effect
Indirect Effects of BMT-CARE Intervention on Depression Symptoms
	Effect	Bootstrapped SE	Bias-Corrected 95% CI
Indirect Effect: Coping Skills Δ	−0.82*	0.35	[−1.60, −0.22]
Indirect Effect: Self-Efficacy Δ	−0.37*	0.23	[−1.11, −0.05]
Total (Multiple) Indirect Effect	−1.19*	0.42	[−2.23, −0.53]
80% of the total effect on depression symptoms can be attributed to the indirect effect
Indirect Effects of BMT-CARE Intervention on Anxiety Symptoms
	Effect	Bootstrapped SE	Bias-Corrected 95% CI
Indirect Effect: Coping Skills Δ	−1.06*	0.49	[−2.07, −0.16]
Indirect Effect: Self-Efficacy Δ	−0.40	0.29	[−1.10, 0.002]
Total (Multiple) Indirect Effect	−1.46*	0.55	[−2.52, −0.43]
39% of the total effect on anxiety symptoms can be attributed to the indirect effect

Note:

Δ = change;

^*p <.05

Effects on Depression Symptoms Partially Mediated through Changes in Coping Skills and Self-Efficacy

First, compared to usual care, BMT-CARE was associated with a significant reduction in day +60 depression symptoms (B=−1.55, SE=0.59, 95% CI [−2.71, −0.38], Cohen’s D = 0.64, p=.010), controlling for baseline depression symptoms. Second, increases in both coping skills (B= −0.09, SE=0.03, p=.004, 95% CI [−0.14, −0.03]) and self-efficacy (B= −0.03, SE=0.01, p=.012, 95% CI [−0.05, −0.01]) from baseline to day +60 were significantly associated with fewer depression symptoms at day +60, controlling for depression symptoms at the baseline assessment. Finally, with both group assignment and changes in coping skills and self-efficacy simultaneously entered in the model, the effect of BMT-CARE on day +60 depression symptoms became non-significant (B= 0.30, SE=0.61, p=.625, 95% CI [−1.49, 0.89]), indicating a potential mediation effect.

The causal mediation model assessing the effect of BMT-CARE on day +60 depression symptoms revealed a significant partial mediation effect simultaneously through both changes in coping skills and self-efficacy, such that depression symptoms decreased by 1.19 points on the HADS-depression subscale through both improved coping skills and self-efficacy for caregivers assigned to BMT-CARE. The specific indirect effect for each mediator (changes in coping skills and self-efficacy) was significant (Table 2). The total indirect effect accounted for 80% of the total effect of BMT-CARE on depression symptoms.

Effects on Anxiety Symptoms Partially Mediated through Changes in Coping Skills and Self-Efficacy

First, compared to usual care, BMT-CARE was associated with a significant reduction in day +60 anxiety symptoms (B=−3.61, SE=0.81, 95% CI [−5.21, −2.00], Cohen’s D = 0.89, p<.001), controlling for baseline anxiety symptoms. Second, increases in coping skills (B= −0.11, SE=0.04, p=.010, 95% CI [−0.20, −0.03]) from baseline to day +60 were significantly associated with fewer anxiety symptoms at day +60, controlling for anxiety symptoms at the baseline assessment, while the association between increases in self-efficacy and anxiety symptoms at day +60 did not meet the threshold for statistical significance (B= −0.03, SE=0.02, p=.066, 95% CI [−0.66, 0.002]). Given that the effect of changes in self-efficacy on anxiety symptoms approached significance, we chose to retain the model as we had previously tested it. With both group assignment and changes in coping skills and self-efficacy simultaneously entered in the model, the effect of BMT-CARE on day +60 anxiety symptoms is attenuated, as seen by the 1.28 reduction in the coefficient (B= −2.33, SE=0.87, p=.008, 95% CI [−4.04, −0.62]); this indicates a potential mediation effect, albeit small.

The causal mediation model assessing the effect of BMT-CARE on day +60 anxiety symptoms revealed a significant partial mediation effect simultaneously through both changes in coping skills and self-efficacy, such that anxiety symptoms decreased by 1.46 points on the HADS-Anxiety subscale through both improved coping skills and self-efficacy for caregivers assigned to BMT-CARE. However, while the specific indirect effect through changes in coping was significant, the indirect effect through changes in self-efficacy was not significant (Table 2). The total indirect effect accounted for 39% of the total effect of BMT-CARE on anxiety symptoms.

Discussion

This exploratory study aimed to elucidate the mechanisms by which a psychosocial intervention (BMT-CARE), compared to usual care, improves QOL and mood for caregivers of HCT recipients across the transplant trajectory. As described previously, we observed that caregivers assigned to BMT-CARE experienced significant improvements in coping and self-efficacy. The current data suggest that positive changes in coping and self-efficacy were associated with significant improvements in QOL and reductions in depression and anxiety symptoms. While other psychosocial interventions for caregivers of HCT recipients have shown promising effects,^16–18 interventions have varied, and the mechanisms driving these benefits are not well understood. A thorough understanding of mechanisms conferring benefit is important for maximizing intervention effects, scalability, and cost effectiveness, as well as minimizing patient burden and clinician resources. Findings from the present study extend available evidence by suggesting that coping skills and self-efficacy are modifiable constructs and active components of BMT-CARE that likely influence overall QOL and mood.

Caregivers in BMT-CARE showed improvements in coping skills targeted by the BMT-CARE intervention. Participants learned and practiced relaxation, cognitive restructuring, and assertive communication, as well as the ability to choose an effective coping technique that accounts for the controllability of the stressor (problem-focused coping vs. emotion-focused coping).¹⁹ While studies show that proficiency with skills, such as engaging mindful awareness and self-compassion, may be linked to better management of distress among caregivers in other cancer populations,²⁰ evidence of the impact of psychosocial interventions on coping efforts in the HCT caregiving setting is lacking and mixed.^21,22 To our knowledge, ours is the first to demonstrate improved coping in caregivers in the HCT setting. A meta-analysis of family caregiver interventions in oncology did show improvement in coping skills with small to medium effect sizes,²³ and a more recent systematic review showed improvements in coping following psychological interventions for partners of patients with cancer.²⁴ In the broader caregiving literature, few studies have examined the effects of interventions on coping for informal caregivers of people with dementia; however, one trial identified improvements in emotion-focused coping as a mediator for intervention effects on depression.²⁵ The current finding that reductions in anxiety and depression symptoms and improvements in QOL were partially mediated by enhanced coping skills for caregivers who received BMT-CARE suggests that incorporation of skills to boost adaptive coping for caregivers in the acute transplant period is warranted.

Our finding that improvements in QOL and reductions in depression symptoms may have occurred indirectly through improvements in self-efficacy for caregivers randomly assigned to BMT-CARE extends our understanding of the potential relationships between caregiver self-efficacy and psychosocial outcomes.²⁶ A sense of mastery and caregiving competence have been associated with reduced caregiving burden, anxiety, and depression symptoms in other caregiving populations; however, few studies have explored intervention mechanisms in the HCT caregiver setting.²⁷ In one study, a problem-solving education intervention among caregivers of patients who received allogeneic HCT led to improvements in self-efficacy, distress, and fatigue.¹⁸ A meta-analysis and a systematic review of family caregiver interventions in oncology showed improvement in self-efficacy with small to medium effect sizes.^23,24 In caregivers of patients with Alzheimer’s Disease and other major neurocognitive disorders, interventions were effective in improving self-efficacy, albeit with small effect sizes.²⁸ In informal caregivers of patients with different types of dementia, interventions have been mixed in improving self-efficacy;^29,30 however, it has also been proposed as a possible therapeutic mechanism for improving other psychological outcomes.³¹ Our findings extend this evidence by exploring specific mechanisms and suggesting that building mastery and confidence for managing the demands of transplantation may influence QOL and depression symptoms for HCT caregivers.

Importantly, while the indirect effects of coping and self-efficacy on QOL and depression symptoms were robust, accounting for 67% and 80% of the total effect of BMT-CARE on these outcomes, respectively, the indirect effects on anxiety symptoms only accounted for 39% of the total effect of BMT-CARE. In addition, our results showed that the association of changes in self-efficacy with improvements in anxiety symptoms approached, but did not reach, statistical significance. It is possible that building confidence and mastery related to caregiving tasks does not mitigate the overwhelming uncertainty of the transplant outcome for caregivers during the acute recovery period. Rather, change in coping skills, such as the ability to relax, communicate one’s needs, and reframe cognitions, accounted for the observed indirect effect on anxiety. It is well established that caregiver self-efficacy and anxiety during HCT are strongly related.³² Notably, cancer caregivers reporting less self-care practices, including stress management, have been shown to have lower preparedness and decision-making self-efficacy. In turn, lower engagement in self-care practices were associated with worse caregiver anxiety.³³ Based on this, it is possible that despite improvements in self-efficacy, the caregivers in our sample continued to struggle with engagement in self-care practices, which impeded the effect of improved self-efficacy on anxiety. Future work could explore additional mechanisms that might contribute specifically to reductions in anxiety.

The results of the current study should be interpreted within the context of several limitations. First, the generalizability of the findings is limited given that this was a single-site trial at a tertiary academic medical setting with a homogenous sample of caregivers. In addition, due to the sample size, we were not able to control for additional factors that may influence caregiver-reported outcomes in these exploratory analyses. Therefore, future studies should replicate these findings with a larger sample size and adjust for additional empirically and theoretically supported factors. We also plan to incorporate longer-term follow-ups in a future multi-site study to assess sustained intervention and mediation effects. Importantly, we are unable to infer definitive causality of whether improved coping and self-efficacy led to better QOL and less depression and anxiety symptoms, or vice versa, given that the mediators and outcomes were assessed at the same time. Future work could incorporate multiple time points to establish directionality of these mechanisms of change. While coping and self-efficacy may be considered to have overlapping constructs, our results showed the change in these constructs was not collinear, with a small Pearson Product-Moment Correlation coefficient, suggesting that the change in each construct represents a discrete skillset. Finally, while BMT-CARE is a brief intervention, future work might explore even briefer, self-administered interventions to further maximize scalability and accessibility.

While the unmet needs of caregivers of HCT recipients are increasingly recognized, relatively few interventions target their specific needs and no interventions span the transplant course. BMT-CARE was designed to teach coping skills with the goal of helping caregivers manage stress and masterfully navigate caregiving for someone close to them during HCT. The current findings provide important insights into what skills caregivers may benefit from most to buffer negative effects on QOL and depression and anxiety symptoms throughout the transplant process. Moreover, these findings may be informative for caregiver-directed interventions in other cancer populations. Therapists’ reports that the assigned homework was ‘mostly complete’ suggests that caregivers benefited from practicing learned intervention skills between sessions. Identifying mechanisms by which an intervention may confer benefit is a crucial step towards maximizing scalability and implementation in clinical care settings, where time and resources may be limited. Future interventions could capitalize on these skills primarily to promote caregiver emotional well-being. In addition, given the interdependence of depression and anxiety in patients and caregivers with advanced cancer,³⁴ future research may examine whether interventions targeting coping and self-efficacy may improve patient-reported outcomes.

These findings provide insight into the salient intervention mechanisms that may drive improvements in psychosocial outcomes for caregivers of HCT recipients. Caregivers of patients undergoing HCT endure substantial declines in QOL and are at ongoing risk for depression and anxiety during and after the patient’s transplant. As previously reported, BMT-CARE is a feasible and acceptable approach to manage stress and burden throughout the transplant process, thereby improving QOL and reducing depression and anxiety symptoms. This study provides evidence suggesting that the improvements in QOL and mood experienced by caregivers assigned to BMT-CARE are, in part, explained by an acquisition of effective coping skills and confidence in their ability to provide care for their loved one.

Figure 1.

BMT-CARE Intervention Effects on Quality of Life Indirectly Through Adaptive Coping Skills and Caregiving Self-Efficacy.

Note: Adjusted for baseline Quality of Life

Figure 2.

BMT-CARE Intervention Effects on Depression Symptoms Indirectly Through Adaptive Coping Skills and Caregiving Self-Efficacy.

Note: Adjusted for baseline depression symptoms

Figure 3.

BMT-CARE Intervention Effects on Anxiety Symptoms Indirectly Through Adaptive Coping Skills and Caregiving Self-Efficacy.

Note: Adjusted for baseline anxiety symptoms