A Mindfulness‐Based Intervention as a Supportive Care Strategy for Patients with Metastatic Non‐Small Cell Lung Cancer and Their Spouses: Results of a Three‐Arm Pilot Randomized Controlled Trial

Abstract

Background

Although mindfulness‐based interventions have been widely examined in patients with nonmetastatic cancer, the feasibility and efficacy of these types of programs are largely unknown for those with advanced disease. We pilot‐tested a couple‐based meditation (CBM) relative to a supportive‐expressive (SE) and a usual care (UC) arm targeting psychospiritual distress in patients with metastatic lung cancer and their spousal caregivers.

Patients and Methods

Seventy‐five patient‐caregiver dyads completed baseline self‐report measures and were then randomized to one of the three arms. Couples in the CBM and SE groups attended four 60‐minute sessions that were delivered via videoconference. All dyads were reassessed 1 and 3 months later.

Results

A priori feasibility benchmarks were met. Although attendance was high in both groups, dyads in the CBM group indicated greater benefit of the sessions than those in the SE group (patients, CBM mean = 2.63, SE mean = 2.20, p = .003; spouses, CBM mean = 2.71, SE mean = 2.00, p = .005). Compared with the UC group, patients in the CBM group reported significantly lower depressive symptoms (p = .05; d = 0.53) and marginally reduced cancer‐related stress (p = .07; d = 0.68). Medium effect sizes in favor of the CBM compared with the SE group for depressive symptoms (d = 0.59) and cancer‐related stress (d = 0.54) were found. Spouses in the CBM group reported significantly lower depressive symptoms (p < .01; d = 0.74) compared with those in the UC group.

Conclusion

It seems feasible and possibly efficacious to deliver dyadic interventions via videoconference to couples coping with metastatic lung cancer. Mindfulness‐based interventions may be of value to managing psychological symptoms in the palliative care setting. Clinical trial identification number. NCT02596490

Implications for Practice

The current randomized controlled trial has established that a mindfulness approach to the management of patients’ and spouses’ psychospiritual concerns is acceptable and subjectively deemed more beneficial than a supportive‐expressive treatment for patients with metastatic non‐small cell lung cancer (NSCLC). We also revealed that videoconference delivery, here FaceTime, is an acceptable approach even for geriatric patients with metastatic NSCLC and that patients and their spousal caregivers prefer a dyadic delivery of this type of supportive care strategy. Lastly, this trial has laid the foundation for the role of mindfulness‐based interventions in the palliative care setting supporting patients with advanced NSCLC and their spousal caregivers.

Short abstract

The feasibility and efficacy of mindfulness‐based interventions for patients with advanced cancer have not been widely examined. This article describes a couple‐based Meditation intervention compared with a supportive expressive intervention and a usual care control group, targeting psychospiritual distress in patients with metastatic lung cancer and their spousal caregivers.

Introduction

Given the generally incurable nature of metastatic non‐small cell lung cancer (NSCLC), patients and their families, particularly their spouses, tend to be overwhelmed by the diagnosis and are at risk of experiencing psychological and spiritual distress (e.g., lack of meaning, hopelessness, alienation) [1, 2, 3, 4, 5, 6, 7]. Although families coping with an advanced cancer diagnosis express spiritual needs and desire spiritual care, their spiritual and existential concerns remain largely unmet [5, 8]. As spouses cope with their own anxiety and spiritual distress, providing emotional support to the patient may be difficult [9]. Moreover, spiritual well‐being is interdependent in couples so that spouses’ spiritual well‐being is associated not only with their own quality of life (QOL) but also with the patient's QOL and vice versa [10].

Besides the diminishing QOL, unmanaged psychospiritual distress may have important cancer care implications as depression predicts worse survival in patients including those with metastatic NSCLC [11, 12]. Moreover, a lack of spiritual well‐being is associated with patients’ desire for a hastened death and suicidal ideation independent of depression, social support, physical function, and cancer symptom burden [1, 3]. In fact, the American Society of Clinical Oncology has argued that spiritual distress is a cause of suffering and that quality and compassionate palliative care initiatives must address patients’ spiritual concerns regarding existential meaning [13]. Yet, evidence‐based interventions that target spiritual well‐being, particularly for patients with advanced cancer, are limited [14, 15]. Moreover, although the psycho‐oncology literature has increasingly recognized the need for providing supportive care services to family caregivers, only a few dyadic randomized controlled trials (RCTs) in NSCLC exist, and none of them has reported spiritual well‐being as an intervention target or outcome [16, 17, 18]. In fact, only one dyadic pilot trial reported significant improvements in depression and anxiety for patients and caregivers who received a six‐session telephone counseling intervention compared with those in a usual care control group [18]. Relative to an education or support comparison group, dyadic coping skill training may not appear to improve mood in patients with lung cancer [16, 17].

In addition to conventional psychosocial care approaches, Eastern‐based mind‐body medicine such as yoga and meditation has been shown to improve mood and spiritual well‐being in patients with cancer [19, 20, 21, 22, 23]. Because the majority of previous work has focused on women with breast cancer, we have tailored these types of interventions for NSCLC patients and caregivers dyads [24, 25, 26]. Seeking to address important knowledge gaps and meet the specific needs of couples coping with metastatic NSCLC, we developed a brief intervention that integrates meditation training with emotional sharing and communication exercises [27].

In the current study, our primary goal was to examine the feasibility of implementing an RCT of this couple‐based meditation (CBM) intervention via videoconference delivery. Because facilitating supportive communications regarding cancer‐related concerns may reduce distress in couples, we sought to compare the CBM intervention with a supportive‐expressive (SE) control group in addition to a usual care (UC) control group [28, 29]. Based on previous dyadic trials, we hypothesized that at least 60% of eligible couples would consent to participate in the three‐arm RCT, 75% of those randomized to the CBM and SE groups would attend at least three of the four sessions, and at least 60% of couples would be retained at the 3‐month follow‐up assessment. Additionally, we hypothesized that, compared with a usual care (UC) control group, the CBM intervention would yield at least medium effect sizes regarding improvements in patients’ and spouses’ depressive symptoms, cancer‐related distress, and spiritual well‐being outcomes. Additional between‐group comparisons (CBM vs. SE and SE vs. UC) were considered exploratory.

Materials and Methods

Participants

Patients were eligible if they (a) were diagnosed with stage IV NSCLC, (b) were currently receiving treatment (e.g., radiotherapy, chemotherapy), (c) had an Eastern Cooperative Oncology Group (ECOG) performance status [30] of less than or equal to 2, and (d) had a romantic partner with whom they had resided for a minimum of 6 months. Both patients and spouses had to (a) be at least 18 years old, (b) be able to read and speak English, and (c) be able to provide informed consent. We excluded patients and spouses who (a) regularly (self‐defined) participated in psychotherapy or a formal cancer support group and (b) were not oriented to time, place, or person as determined by the clinical team.

Procedures

Research staff identified potential participants via the institution's electronic clinic appointment system. During clinic visits, the staff members approached the patients and spouses, confirmed their eligibility, and obtained written informed consent prior to data collection. If a spouse was not present during a clinic visit, we asked for the patient's permission to contact him or her to obtain consent. Patients completed the National Comprehensive Cancer Network (NCCN) Distress Thermometer at the time of consent [31]. Then, both patients and spouses completed baseline survey measures prior to randomization. Follow‐up assessments were administered 4 weeks and 12 weeks later. For each completed assessment, patients and spouses received a $20 gift card (totaling $120 per couple). The MD Anderson Institutional Review Board approved all procedures prior to participant enrollment.

Randomization

Couples were randomized to either the CBM, SE, or UC group through a form of adaptive randomization called minimization, ensuring that the groups were balanced based on patients’ prognostic factors of age, sex, ECOG performance status, and the NCCN distress score using a computerized system (Filemaker).

Intervention Groups

Patients and spouses together attended one session per week for 4 weeks (60 minutes each; total of 240 minutes). All sessions were administered via FaceTime by a master's‐licensed psychological counselor intern (LPC‐I). Based on their preference, participants used their own or a loaner device (i.e., iPad Mini 4). (Although FaceTime was compliant with the Health Insurance Portability and Accountability Act at the time of trial implementation, it was no longer compliant at the time of the submission of this article.)

CBM Group

The CBM program is rooted in the constructs of interconnection, mindfulness, and compassion and includes both intrapersonal (i.e., meditations) and interpersonal (i.e., emotional sharing) exercises. Briefly, during session 1, couples received an overview of the program and were experientially introduced to mindfulness meditation techniques by guiding their awareness to their current experience (i.e., self‐regulation of attention) with a motivation of acceptance as described by Bishop et al. [32]. Then, participants were asked to bring this state of nonjudgmental and fully present awareness to sharing their experiences with each other. During session 2, couples participated in two meditations that focused their attention on their interconnectedness and their feelings of compassion for the self and their spouse [33]. After each visualization exercise, couples were asked to take turns to mindfully share their reflections and experiences with each other, thereby continuing to cultivate a state of compassion for each other. In session 3, couples were asked to reflect on things, events, and people for which they were grateful via a guided gratitude meditation followed by mindful and compassionate sharing [34]. Lastly, session 4 was influenced by the value‐based living process of Acceptance Commitment Therapy, assisting couples in answering the question “What do you want your life to stand for?” Couples completed first a values worksheet and then a reflection exercise to help identify their core values. The counselor then assisted couples to identify strategies that ensure that their lives reflect their self‐identified values. The interventionist reviewed all the practices and tools the couple had learned over the course of the program and supported them in proactively developing a plan for continued implementation. Couples received audio files of the meditations and printed materials of the session content and instructions on how to continue with this program when no longer meeting with the interventionist.

SE Group

The SE sessions were modeled after the social support intervention used by Breitbart and colleagues [35]. However, the original protocol was modified to develop an appropriate condition for comparison with the CBM program, mainly pertaining to the dyadic rather than group‐based delivery. Sessions 1 and 2 focused on discussing cancer‐related concerns that couples tend to experience when coping with cancer, such as communicating with health care providers; coping with family and friends; caregiving issues; sexual intimacy concerns; relationship changes; coping with physical functioning concerns; fears about future physical or psychological changes, and mortality; and future plans. Sessions 3 and 4 focused on concerns that were not necessarily cancer‐specific, including maintaining and improving family relationships, retirement and career changes, spirituality, health behaviors, life goals, and hobbies. Using a reflective listening approach, the interventionist encouraged patients and spouses to share their concerns with each other. Using a worksheet, the interventionist addressed with couples how to use supportive communication with each other and redirected if a patient or spouse became unsupportive. Because no additional therapeutic tools were offered, the interventionist did not probe for deep emotional disclosure.

Quality Control

To ensure treatment fidelity, all CBM and SE sessions were audio‐recorded (with the participants’ permission obtained during the informed consent process) and reviewed on an ongoing basis using a fidelity checklist.

Manipulation Check

To compare the extent to which couples in the CBM and SE groups disclosed emotions, we performed linguistic word count analyses of the professionally transcribed audio recordings of all CBM and SE sessions. We used the Linguistic Inquiry and Word Count (LIWC) software to examine group differences on participants’ language regarding the LIWC categories of anxiety, sadness, anger, and positive emotions [36].

Control Group

The UC group received the usual care as provided by their health care team.

Demographic and Medical Factors

Demographic items (e.g., age, race and ethnicity) were included in the baseline questionnaires. Patients’ medical data were extracted from their electronic medical records.

Feasibility Data

We documented eligible participants’ consent rates and study attrition as well as class attendance and homework completion for dyads in the intervention groups. Participants in the CBM and SE groups completed an evaluation of the intervention to assess their satisfaction with the session content, dyadic delivery, and their perceptions of the overall usefulness and benefit of the program (0–3 scale for questions). Dyads in the CBM arm were asked to rate their preference regarding an online delivery preference.

Outcome Measures

Both patient and spouse completed paper‐and‐pencil questionnaires at baseline and again 1 and 3 months later assessing the following: (a) depressive symptoms with the Center for Epidemiologic Studies Depression Scale (CES‐D) [37], a 20‐item self‐report measure focusing on the affective component of depression; (b) cancer‐related stress symptoms with the Impact of Events Scale (IES), a 15‐item scale assessing intrusive thoughts and avoidance behaviors related to the cancer diagnosis [38]; and (c) spiritual well‐being with the Functional Assessment of Cancer Therapy–Spiritual Well‐Being Scale, a 12‐item instrument assessing the spiritual dimensions of peace, faith, and meaning [39].

Data Analysis Plan

To examine feasibility, we calculated descriptive statistics of consent, session attendance, assessment completion, and program satisfaction. To establish preliminary efficacy, we used an intent‐to‐treat analysis when performing multilevel modeling (MLM) using PROC MIXED (version 9.4; SAS Institute, Cary, NC). PROC MIXED uses a likelihood‐based estimation method for missing data so that attrition is less of a concern [40]. We used separate analyses for patients and spouses and controlled for baseline level of the given outcome. Assessment time was included as categorical variable. For patients, we included the randomization factors in the MLM models. We used contrast statements within the mixed procedure to test for the hypothesized group differences (i.e., CBM vs. UC). The additional comparison (i.e., SE vs. UC and CBM vs. SE) were considered exploratory. Because the current study is a pilot trial and not adequately powered, we supplemented the inferential statistics with effect sizes (Cohen's d) associated with each between‐group comparison interpreting effects as small (d = 0.2), medium (d = 0.5), and large (d = 0.8) [41]. A medium effect size between the CBM and UC group was considered hypothesis confirming.

Results

Participant Characteristics

Baseline characteristics by group and role and are shown in Table 1. Overall, patients’ mean ± SD age was 65.0 ± 10.4 years (range, 30–93), and spouses’ mean ± SD age was 63.9 ± 10.3 years (range, 30–93). Approximately half of patients (51%) and spouses (51%) were female, and there were three same‐sex couples. A majority of patients (85%) and spouses (79%) were non‐Hispanic and white. The sample was relatively well educated with a little more than half of patients (55%) and spouses (58%) having at least a college degree. About half (52%) of patients but only about one third (35%) of spouses reported that they never smoked. Patients were about 1 year after their initial diagnosis (mean ± SD, 1.11 ± 1.40 years), and about half of the patients had an ECOG performance status of 1 (55%) and at least one brain metastasis (51%). Couples were married or cohabitating a mean ± SD of 30 ± 17.0 years.

Table 1.

Baseline characteristics of patients and spouses by group assignment

Characteristics	Patients (n = 75), n (%)			Spouses (n = 75), n (%)
	Couple‐based meditation (n = 26)	Supportive‐expressive (n = 24)	Usual care (n = 25)	Couple‐based meditation n = 26)	Supportive‐expressive (n = 24)	Usual care (n = 25)
Gender
Male	13 (50)	12 (50)	12 (48)	11 (42)	12 (50)	13 (52)
Female	13 (50)	12 (50)	13 (52)	15 (58)	12 (50)	12 (48)
Age, mean ± SD (range), years	65.46 ± 8.8 72–77)	62.88 ± 11.3 (40–93)	65.40 ± 11.3 (30–85)	65.58 ± 10.6 (48–89)	61.58 ± 11.1 (47–93)	63.56 ± 9.5 (30–80)
Ethnicity
Non‐Hispanic white	23 (88)	18 (75)	23 (92)	21 (81)	16 (68)	22 (88)
Latino or Hispanic	2 (8)	2 (8)	0 (0)	4 (15)	4 (16)	1 (4)
Asian	0 (0)	0 (0)	2 (8)	0 (0)	0 (0)	2 (8)
Black	1 (4)	3 (13)	0 (0)	1 (4)	4 (16)	0 (0)
More than one	0 (0)	1 (4)	0 (0)	0 (0)	0 (0)	0 (0)
Education: some college or more	23 (88)	20 (83)	16 (64)	22 (85)	16 (67)	18 (72)
Household income: $75,000 or more	15 (79)	10 (59)	16 (84)	15 (78)	10 (59)	16 (84)
Employment status
Full‐time	3 (12)	3 (13)	5 (20)	5 (20)	9 (41)	6 (24)
Retired	14 (54)	11 (48)	15 (60)	13 (52)	7 (32)	11 (44)
Part‐time	1 (3)	3 (13)	1 (4)	0 (0)	2 (9)	1 (4)
Disability	4 (16)	3 (13)	1 (4)	1 (4)	1 (5)	0 (0)
Home maker	3 (12)	3 (13)	2 (8)	5 (20)	3 (13)	6 (24)
Other	1 (3)	0 (0)	2 (8)	1 (4)	0 (0)	1 (4)
ECOG performance status
0	10 (38)	8 (33)	7 (28)
1	14 (54)	13 (54)	15 (60)
2	2 (8)	3 (13)	3 (12)
Radiation	21 (84)	19 (80)	23 (92)
Chemotherapy	17 (68)	19 (79)	21 (84)
Surgery	4 (16)	5 (21)	4 (16)
Time since diagnoses, mean ± SD (range), weeks	52.96 ± 54.55 (3–182)	44.26 ± 54.68 (4–220)	68.36 ± 96.48 (1–357)

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

Feasibility

Recruitment and Retention

We approached 204 patients, and 130 (64%) were deemed eligible, of whom 82 consented (63% consent rate). Ineligibility was mainly due to not having an eligible partner or spouse (n = 54). Refusals were due to coping well (n = 24), being too busy (n = 17), and not being interested (n = 7). Of those who consented, two became ineligible (transferred to hospice), three died, and two withdrew consent prior to randomization. Of the 75 dyads who were randomized, 65% were retained through the 3‐month follow‐up assessment. Attrition was due to patients’ death (n = 10), disease progression (n = 6), being impacted by Hurricane Harvey (n = 6), and passive or active withdrawals (n = 5). Although there was no differential dropout based on group assignment (p = .91), in dyads who were not retained, the patients were marginally younger (p = .09) and more likely to have a brain metastasis (p = .005), and the spouse reported significantly higher spiritual well‐being (p = .01) compared with dyads who completed the study. See Figure 1 for the CONSORT chart.

Figure 1.

CONSORT chart.Abbreviations: T2, 4‐week follow‐up; T3, 12‐week follow‐up.

Session Attendance and Acceptability

Dyads attended a mean ± SD of 3.12 ± 1.4 sessions (range, 0–4) and 3.08 ± 1.6 (range, 0–4) sessions in the CBM and SE groups, respectively, with 71% of couples attending four sessions and 76% attending at least three sessions. Session attendance did not differ as a function of group (p = .92). Both patients and spouses in the CBM group rated the intervention significantly more beneficial than those in the SE group (0–3 scale: patients, CBM mean = 2.63, SE mean = 2.20, p = .003; spouse, CBM mean = 2.71, SE mean = 2.00, p = .005). All patients and spouses in the CBM group and 93% of patients and 87% of spouses in the SE group indicated that dyadic delivery was their preference (compared with “individual is my preference” or “no preference”). Regarding online preference, 38% of patients and 35% of spouses preferred online; 38% of patients and 41% of spouses had no preference, and 25% patients and 24% of spouses stated that they preferred all sessions in person.

Manipulation Check

The language of couples in the CBM contained significantly more positive emotions (F = 13.77, p < .001) and less anxiety (F = 3.96, p = .05) compared with those in the SE group. The two active groups did not differ in language revealing sadness (F = 0.91, p = .35) and anger (F = 1.79, p = .19).

Preliminary Efficacy Results

Because the presence of brain metastases was significantly associated with differential dropout as well as study outcomes, we included brain metastasis (yes vs. no) in addition to the aforementioned covariates in the patient models. In the spouse models, we controlled for age and sex as they were associated with the outcomes at p < .05.

Depressive Symptoms

Based on the MLM analysis including covariates, the group main effect for patients’ CES‐D scores was not significant (F = 0.05, p = .95; least square means [LSMs]: CBM = 10.92; UC = 11.31; SE: 10.75). Yet, we found a significant group × time interaction (F = 3.26, p < .05) with planned contrast comparisons revealing a significant group difference at the 3‐month follow‐up between the CBM and the UC groups (F = 4.07, p = .05; d = 0.53) and a marginally significant group difference between the CBM and the SE group (F = 3.36, p = .07; d = 0.59; LSMs: CBM = 8.64; UC = 11.54; SE = 11.80). The difference between the SE and UC group was not significant (F = 0.05, p = .82; d = 0.04). For spouses, there was a significant group main effect (F = 3.70, p < .05; LSMs: CBM = 6.73; SE = 9.32; UC = 11.57) with planned comparisons revealing a significant difference between the CBM and UC group (F = 7.39, p < .01; d = 0.74), a medium effect but not significantly different between the CBM and SE group (F = 1.99, p = .16; d = 0.65), and no significant differences between the SE and UC group (F = 1.53, p = .22; d = 0.08). These results are depicted in Figure 2.

Figure 2.

Least square means for patients and spouses for depressive symptoms. (A): Patients. (B): Spouses. The patient model controlled for baseline level of the CES‐D, age, sex, Eastern Cooperative Oncology Group performance status and brain metastases; the spouse model controlled for baseline level of the CES‐D, age, and sex. Higher scores denote worse symptoms.Abbreviations: 1MFU, 1‐month follow‐up; 3MFU, 3‐month follow‐up; CBM, couple‐based meditation; SE, supportive‐expressive; UC, usual care; CES‐D, Center for Epidemiologic Studies Depression Scale.

Cancer‐Related Stress Symptoms

The overall MLM model did not reveal significant group differences for patient's IES scores (F = 0.16, p = .85, LSMs, CBM: 15.48 vs. UC: 17.17 vs. SE: 16.66). Post hoc analyses revealed at the 3‐month follow‐up a medium effect size when comparing the CBM with the UC group (F = 3.55, p = .07, d = 0.68) and the CBM with the SE group (F = 2.25, p = .14, d = 0.54; LSMs: CBM = 12.00; UC = 18.38; SE = 17.11). No effect was found between the SE and UC group (F = 1.14, p = .71, d = 0.14). For spouses, we also did not find a significant group effect in the overall model (F = 0.32, p = .73; LSMs: CBM = 17.76; SE = 17.32; UC = 19.60). Post hoc analyses revealed at the 3‐month follow‐up a small effect when comparing the CBM with the UC group (d = 0.27) and the CBM with the SE group (d = 0.36) and no effect between the SE and UC group (d = 0.08). None of these comparisons was significant (LSMs: CBM = 17.04; UC = 21.10; SE = 20.38). These results are depicted in Figure 3.

Figure 3.

Least square means for patients and spouses for cancer‐related stress symptoms. (A): Patients. (B): Spouses.The patient model controlled for baseline level of the IES, age, sex, Eastern Cooperative Oncology Group performance status and brain metastases; the spouse model controlled for baseline level of the IES, age and sex. Higher scores denote worse symptoms.Abbreviations: 1MFU, 1‐month follow‐up; 3MFU, 3‐month follow‐up; CBM, couple‐based meditation; SE, supportive‐expressive; UC, usual care; IES, Impact of Events Scale.

Spiritual Well‐Being

We did not find a significant group main effect difference based on the MLM analyses (F = 0.58, p = .57). Planned comparisons revealed a small effect in favor of both the CBM (d = 0.26) and SE (d = 0.33) groups relative to the UC group. There was no difference between the two intervention groups (LSMs: CBM = 37.16; UC: 35.78 vs. SE: 37.56). Similarly for spouses, no significant group differences were found in the overall model (F = 1.79, p = .17); however, planned comparisons revealed a marginally significant difference with a medium effect size between the CBM and UC group (F = 3.44, p = .07, d = 0.67) and a small but not significant effects between the CBM and SE groups (d = 0.22) and the SE and UC groups (d = 0.43, LSMs, CBM: 39.08 vs. UC: 36.33 vs. SE: 37.56). These results are depicted in Figure 4.

Figure 4.

Least square means for patients and spouses for spiritual well‐being. (A): Patients. (B): Spouses.The patient model controlled for baseline level of the FACT‐SP, age, sex, Eastern Cooperative Oncology Group performance status and brain metastases; the spouse model controlled for baseline level of the FACT‐SP, age, and sex. Higher scores denote better well‐being.Abbreviations: 1MFU, 1‐month follow‐up; 3MFU, 3‐month follow‐up; CBM, couple‐based meditation; SE, supportive‐expressive; UC, usual care; FACT‐SP, Functional Assessment of Cancer Therapy–Spiritual Well‐Being Scale.

Discussion

The goal of this pilot RCT was to examine the feasibility and preliminary efficacy of the CBM intervention for patients with metastatic lung cancer and their spouses targeting psychospiritual outcomes. The results revealed that the trial was feasible as it met our a priori feasibility criteria regarding consent, retention, and adherence rates. Of note, almost three fourths of dyads in the CBM and SE groups attended all four sessions. Although there were no significant group differences between the two active groups in session attendance and intervention engagement (e.g., outside of session discussions), participants in the CBM group rated the sessions significantly more beneficial than those in the SE group. Regarding acceptability, we were particularly interested in the dyadic and videoconference delivery approaches. Almost all patient and spouses in both active groups indicated that they preferred dyadic delivery, and the majority of participants indicated that they either had no preference or preferred videoconference over in‐person delivery. Thus, we are confident that both dyadic and videoconference delivery are acceptable for couples coping with metastatic lung cancer, which is encouraging as several participants were in their 80s and even 90s.

Regarding evidence for preliminary efficacy, the findings of the current trial suggest a clear signal of the CBM intervention relative to a UC group and even a strong active comparison group controlling for the effects of attention, support, and emotional processing of the cancer experience. Although the trial was not adequately powered to reveal between‐group differences, we generally supported our hypothesis of revealing medium effect sizes. More specifically, at the 3‐month follow‐up, patients in the CBM group reported lower depressive symptoms relative not only to the UC but also to the SE group. In contrast, no effect was found for participants in the SE group compared with the UC group regarding depressive symptoms. For spouses, the effect was stronger such that at both follow‐up time points, the CBM group reported lower depressive symptoms relative to the SE and UC group. Of note, there was also no effect between the SE and UC group for spouses. Similarly to depressive symptoms, at the 3‐month follow‐up, patients in the CBM group reported lower cancer‐related stress compared with the UC and SE groups, with no effects between the SE and UC groups. Although we revealed some evidence of an intervention signal for spouses’ cancer‐related stress, our a priori medium effect size benchmark was not met. Lastly, regarding spiritual well‐being, we only detected a small effect for patients in the CBM relative to the UC group. In contrast, we found a medium effect size across the follow‐up period for spouses in the CBM compared with the UC group.

When interpreting these findings, a few key issues must be considered. The primary aim of this trial was to examine feasibility of implementing the current RCT methods. As such, the psychospiritual outcomes were assessed to identify an initial signal for intervention efficacy relative to the UC and SE comparison groups. The trial was not designed to confirm efficacy as it was underpowered, and no adjustment to the alpha level for multiple comparisons was made. Thus, caution is warranted. At the same time, these findings convincingly support the further investigation of the CBM intervention as a supportive care strategy. Moreover, the CBM program is a brief intervention, led by master's‐level professionals, and delivered via videoconference for the patients’ and caregivers’ convenience. Moreover, most psychosocial interventions in the psycho‐oncology literature only show small effect sizes, particularly when compared with an active control group. Here, we demonstrated a signal relative to a highly relevant comparison group. Lastly, although the improvements for spiritual well‐being seem to have been modest for patients, within‐group differences of the well‐known eight‐session Meaning Centered Group (or Individual) Therapy developed by Breitbart and colleagues produced small to medium effect sizes for spiritual well‐being [42]. Thus, the CBM intervention may be on par (particularly for spouses) with Breitbart's eight‐session, psychiatrist‐ or psychologist‐led intervention. Yet, an efficacy trial with a larger sample is needed to establish more stable effect sizes.

Although this trial has laid an important foundation of the CBM intervention regarding the feasibility and acceptability of intervention content and delivery approach as well as preliminary efficacy, our study is limited by the sample's fairly homogenous characteristics, particularly regarding race and ethnicity and socioeconomic status. An efficacy trial with a larger sample is needed to facilitate subgroup analyses and identify participant characteristics associated with differential treatment responses. Subgroup analyses may also examine if participants’ computer skills and/or session delivery preference moderate the treatment response. It is also important to note that patients were not selected for significant symptom burden, and only about half of patients and spouses met the screening criteria for depression and/or posttraumatic stress disorder. Based on meta‐analyses, there is reason to believe that those with elevated symptoms are more responsive to psychosocial interventions; however, it is unknown at this point how participants meeting diagnostic criteria for a mood and/or anxiety disorder would fare on this treatment [43, 44].

Conclusion

On the basis of the findings of this three‐arm pilot RCT, it is feasible to deliver dyadic interventions via video conference to couples coping with metastatic lung cancer. Moreover, preliminary evidence shows that a mindfulness‐based intervention may be of value to managing psychological symptoms in the palliative care setting.

Author Contributions

Conception/design: Kathrin Milbury, Yisheng Li, Zhongxing Liao, Anne S. Tsao, Cindy Carmack, Lorenzo Cohen, Eduardo Bruera

Provision of study material or patients: Zhongxing Liao, Anne S. Tsao

Collection and/or assembly of data: Yisheng Li, Sania Durrani

Data analysis and interpretation: Kathrin Milbury, Yisheng Li, Lorenzo Cohen, Eduardo Bruera

Manuscript writing: Kathrin Milbury, Yisheng Li, Sania Durrani, Lorenzo Cohen

Final approval of manuscript: Kathrin Milbury, Yisheng Li, Sania Durrani, Zhongxing Liao, Anne S. Tsao, Cindy Carmack, Lorenzo Cohen, Eduardo Bruera

Kathrin Milbury, Yisheng Li, Sania Durrani, Zhongxing Liao, Anne S. Tsao, Cindy Carmack, Lorenzo Cohen, Eduardo Bruera

Disclosures

Anne S. Tsao: Bristol‐Myers Squibb, Eli Lilly & Co., Genentech, Roche, Novartis, Ariad, EMD Serono, Merck, Seattle Genetics, AstraZeneca, Boehringer‐Ingelheim, Sellas Life Science, Takeda, Epizyme, Huron (SAB), Eli Lilly & Co., Millennium, Polaris, Genentech, Merck, Boehringer‐Ingelheim, Bristol‐Myers Squibb, Ariad, Epizyme, Seattle Genetics, Takeda, EMD Serono (RF). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board