Family Cancer Caregiver Use of and Benefit from an Internet-delivered Insomnia Intervention: Results from a Single-Group Feasibility Trial

Abstract

Background:

Cancer caregivers are more likely to report clinically significant symptoms of insomnia than cancer patients and the general population, yet research has been limited regarding cognitive-behavioral therapy for insomnia (CBT-I) among this population.

Methods:

To better understand cancer caregivers’ engagement with and benefit from CBT-I, cancer caregivers were enrolled in a nonrandomized pilot feasibility trial of an evidence-based Internet-delivered insomnia program. Thirteen caregivers completed mixed-methods assessments prior to receiving the insomnia program and after the nine-week intervention period.

Results:

Compared to the five caregivers who did not complete any intervention Cores, the eight caregivers who completed at least one of the intervention Cores tended to report more sleep impairment (insomnia symptom severity; minutes of sleep onset latency and wake after sleep onset), less physical and emotional strain from caregiving, and less maladaptive sleep beliefs at the baseline assessment. These caregivers who used the program also showed large improvements in their insomnia symptoms. Caregivers’ qualitative feedback about their experience with the program identified potential areas that might be modified to improve caregivers’ engagement with and benefit from Internet-delivered insomnia programs.

Conclusions:

Findings suggest that family cancer caregivers can use and benefit from a fully-automated Internet-delivered CBT-I program, even without caregiving-specific tailoring. Further rigorous research is needed to better understand whether and how program modifications may allow more caregivers to initiate and engage with this program.

BACKGROUND

Between 50–75% of family members who provide unpaid care to a person with cancer report clinically significant insomnia symptoms.^1,2 Cognitive behavioral therapy for insomnia (CBT-I) is the first-line treatment for insomnia and has been extensively studied in cancer patients.³ The limited trials of CBT-I among cancer caregivers demonstrate mixed evidence on its uptake and efficacy among this unique population. Carter⁴ tested a brief CBT-I protocol adapted specifically for cancer caregivers. The in-person and telephone-based program was well-received by caregivers, who appreciated the flexible scheduling of appointment time and location; however, overall adherence was suboptimal and there were not overall reliable improvements in sleep. Applebaum and colleagues⁵ tested a traditional CBT-I protocol of eight 1-hour in-person visits. Recruitment for this trial was challenging, with time demands of the sessions and travel limiting enrollment. Once enrolled, however, caregivers receiving CBT-I described the treatment as “incredibly effective.”⁵

Delivering CBT-I by the Internet to caregivers holds potential to overcome common barriers to caregivers’ access to this highly effective treatment; however, there have been no trials of Internet-delivered CBT-I program among caregivers. Sleep Healthy Using the Internet (SHUTi) is a fully-automated Internet intervention that has been found effective in adults with insomnia in general populations^6,8 and cancer survivors.⁹ To begin to better understand the potential of Internet-delivered CBT-I to meet the significant need for insomnia treatment among cancer caregivers, we completed a small, single-group feasibility trial of SHUTi among cancer caregivers to examine potential influences of the caregiving context on program engagement and benefit, feasibility of a fully-automated eHealth approach, and caregivers’ preliminary benefit from and feedback about the SHUTi program.

METHODS

This study was a nonrandomized pilot feasibility trial of SHUTi among cancer caregivers (University of Virginia IRB #3809; NCT04661306). Given that feasibility of the fully-automated, Internet-delivered CBT-I approach with caregivers had not been established, this design permitted us to determine its practicality.

Participants and Procedure

Adults with regular Internet access (i.e., ≥2 times/week) were eligible to participate if they: (1) self-reported providing unpaid care (e.g., practical, medical, and/or emotional support) to an adult family member or “family-like” close individual who was either currently undergoing any cancer treatment (e.g., surgery, chemotherapy, radiation, for either curative or palliative intent) or completed treatment within the past 2 years, and (2) scored ≥10 on the Insomnia Severity Index (ISI) indicating clinically significant insomnia symptoms. Exclusion criteria included having another untreated sleep, medical, or psychiatric disorder; current insomnia therapy; or recent change in sleep medication. Participants were recruited online (e.g., via email newsletter and social media postings) from across the U.S. and from September 2020-December 2021. Interested individuals completed eligibility screening online and follow-up eligibility confirmation by phone. Participants provided informed consent prior to data collection at enrollment, then completed baseline assessment (questionnaires and 10 days of prospectively-collected sleep diaries). Upon completion, all participants received SHUTi for 9 weeks, then completed post-assessment.

Intervention

SHUTi is a fully-automated and interactive online CBT-I program.^6,7 Content is delivered through six intervention “Cores,” modeling the weekly sessions of traditional face-to-face therapy. Technical, but not clinical, support is available from study staff. There is no caregiving-specific content in SHUTi. See Supplement for further details.

Measures

The overall practicality of the fully-automated Internet-delivery approach to CBT-I among caregivers was examined through caregivers’ use of the SHUTi program (number of Cores completed), as well as their open-ended feedback regarding acceptability and their overall experience with the program. Caregivers reported caregiving-related variables (e.g., physical and emotional strain of caregiving, see Table 1) at baseline, as well as sleep outcome variables at both baseline and post-assessment. Sleep outcomes were participants’ self-reported insomnia symptom severity (7-item ISI)^10,11 and sleep diary metrics of minutes of sleep onset latency (SOL) and wake after sleep onset (WASO) averaged from 10 days of daily diaries.¹² Last, participants rated their sleep beliefs (16-item Dysfunctional Beliefs and Attitudes about Sleep scale [DBAS]).¹³ See Supplement for further details.

Table 1.

Sleep Outcomes and Associations with Caregiving Context Variables in SHUTi Users

		Sleep Outcomes		Sleep Beliefs
	ISI	SOL	WASO	DBAS
Pre (M±SD)	14.75±6.25	30.93±20.05	93.57±41.43	5.35±1.31
Post (M±SD)	9.34±4.60	16.71±8.98	76.86±34.50	3.85±1.57
Pre-Post difference (Cohen’s d)	1.26	1.12	1.26	1.46
Caregiving Context Variables (Cohen’s d & M)
Diagnosed in past year	0.99	0.27	0.14	1.50
No (n=4)	−3.50	−15.63	−17.50	−1.89
Yes (n=4)	−7.25	−12.33	−15.67	−0.80
Spouse of CR	0.27	0.80	0.14	0.10
No (n=3)	−6.00	−20.25	−15.75	−1.29
Yes (n=5)	−5.00	−11.80	−17.10	−1.38
Time spent caregiving	0.69	0.06	1.64	0.11
≤20 hours/week (n=4)	−4.00	−14.50	−9.88	−1.39
≥21 hours/week (n=4)	−6.75	−13.83	−25.83	−1.30
r
Caregiving physical strain	.41	−.27	.12	.47
Caregiving emotional strain	−.38	.12	.26	.61

Note. N=8 (except SOL/WASO N=7); ISI = Insomnia Severity Index; SOL = Sleep Onset Latency (from sleep diary data); WASO = Wake After Sleep Onset (from sleep diary data); DBAS = Dysfunctional Beliefs and Attitudes about Sleep; CR = Care recipient; Diagnosed in past year: 0 = No, 1 = Yes; Spouse of CR: 0 = No, 1 = Yes; Time spent caregiving: 0 = 0 to 20 hours per week, 1 = 21 hours or more per week

Statistical Analysis

Given the small sample size, effect sizes are presented, and medium or larger effects (i.e., d≥0.5, |r|≥.3) are interpreted, with caution. Baseline differences between SHUTi users and non-users were examined using Cohen’s d of independent samples t-tests. Change in outcomes among caregivers who used SHUTi was computed as the Cohen’s d for the paired samples t-test. Associations between caregiving context variables and raw change scores in outcome variables were examined using Cohen’s d of the independent samples t-tests and Pearson’s correlations. Participants’ open-ended responses were reviewed for common themes.

RESULTS

Fourteen caregivers were enrolled. One participant withdrew due to the patient’s death. Participants were almost all non-Hispanic/Latina white women (n=12, 92%), were middle-aged (M=63, range: 44–70), and most had completed a college degree (n=10, 76.9%).

Program Use

Of the 13 caregivers included in analyses, eight completed ≥1 SHUTi Core (i.e., SHUTi users, 57%, M=4.2 Cores). Six caregivers (46%) completed ≥4 Cores, which is considered treatment completion as all primary CBT-I components have been received. At baseline, caregivers who later used SHUTi, relative to those who did not (N=5), tended to report more sleep impairment (ISI M=15.0 vs. 13.75, d=0.31; SOL M=32.1 vs. 26.3, d=0.49; WASO M=88.28 vs. 66.88, d=0.71), but less physical (M=1.38 vs. 2.75, d=1.78) and emotional caregiving strain (M=3.38 vs. 4.50, d=1.34) and maladaptive sleep beliefs (DBAS M=5.32 vs. 5.77, d=0.45).

Program Experiences

Most caregivers who expressed satisfaction with the program specifically addressed how both the cognitive and behavioral therapeutic components were helpful (see Table 2 for representative quotations). Regarding cognitive components, caregivers appreciated how the program helped them “think differently” about their sleep, by “dispelling many myths” they held about sleep. These cognitive shifts helped “lessen their obsession” and anxiety about sleep. Caregivers described behavioral components as helpful but challenging. Multiple caregivers reported appreciating the “structured approach to resetting sleep habits,” with the daily sleep diaries providing accountability and awareness of “routines and thought patterns” that perpetuated insomnia. On the other hand, caregivers also found these daily diaries “a bit of a pain.” Sleep restriction was also noted as challenging by multiple caregivers.

Table 2.

Themes and Representative Quotations from SHUTi User Open-Ended Responses

Topic / Theme	Example quotations
Components
Cognitive components helpful	One participant appreciated learning that their “sleep/wake cycle [was] not uncommon” “It educated me about my insomnia, which helped lessen my obsession about lack of sleep. I’ve long felt that I needed 8 hours per night and had developed sleep anxiety. It became like a vicious cycle of no sleep, anxiety about no sleep, causing more lack of sleep. The core suggestions gave information that calmed [me] down and helped in that I gradually worried less about my sleep.”
Behavioral components helpful	“Structured approach to resetting sleep habits” “It clearly outlined courses of actions” “I also became more aware of my personal habits that interfered with my ability to sleep well. The daily updates helped me to track unhealthy patterns in my bedtime routine. It helped me to create a plan to modify my behavior related to sleep preparation and habits.”
Behavioral components challenging	“The restrictive sleep cycle was not helpful. I have nothing I can do to stay awake ‘til the middle of the night.... Trying to stay awake ‘til a certain time usually results in my not being able to sleep at the time I am allowed to do so” “I just found I didn’t like filling out the questions everyday” “The daily journal was a bit of a pain, but definitely worth the time to do, because it made me look closely at my behaviors.”
Caregiving tailoring
Insomnia is different for caregivers	“My wife completed cancer treatment... The only part that is still affecting her on a regular basis are the hormone suppressants. She has hot flashes several times a night and is uncomfortable but stays in bed lying awake. Blankets are pulled on or off, etc. That has a big impact on my quality of sleep.” “[SHUTi] did not specifically address [caregivers’ sleep] needs, particularly if awakened to give care to someone du ring the night. Staying awake until late at night was not doable for me. The modules may be helpful for someone having ordinary sleep difficulties.” “If I were a caregiver for someone in the throes of cancer treatment rather than someone in recovery such a s my husband, forcing myself to stay awake to achieve a defined sleep window would be counterproductive. I would be better served sleeping when I could to better care for him when he needed me, which would very likely be outside the sleep window.”
Concerns about impact for household	“If I am setting an alarm to rise at an hour designated by the sleep study, I’m going to be waking my partner also. I should not have to disturb their sleep to fulfill my obligation.” “My partner did not like me getting out of bed when I couldn’t sleep or getting up immediately on waking”
Address anxiety	Suggested to add content related to “how to address anxiety issues when caring for someone with cancer... I used meditation recordings with ear buds to help me relax through worries and fall asleep.” “I think it would be ideal to have a weekly counseling session to deal with anxiety, the root cause of my current sleep struggles.”
Format of delivery
Benefit of the Internet	“The online format was perfect for me. I could access it when I had time.” “Being able to complete it via computer made it very simple and attractive... [there was] no pressure since it’s Internet-based rather than in-person.”
Recommendations to ease burden	“Stretch some lessons out so wouldn’t take so much time reading when you have so much else to do.” “The format worked for me, but maybe an audio component would be helpful to others, so they could multi-task (which they’re already doing)”

Several caregivers also addressed the unique aspects of caregiving that may require tailored content. Recommended content included addressing when the patient is the caregiver’s bedpartner and has long-term disruptive nighttime side effects, as well as how to adjust CBT-I treatment recommendations when “awakened to give care to someone during the night.” Caregivers also raised concerns about how the behavioral treatment components may “disturb” the care recipients’ sleep. Two caregivers also suggested more directly addressing “anxiety issues while caring for someone.”

Program Sleep Outcomes

Caregivers who used SHUTi showed large improvements on their self-reported insomnia severity, sleep diary metrics, and sleep beliefs. These different constructs showed a mixed pattern of associations with caregiving context factors (see Table 1).

DISCUSSION

This single-group feasibility study affords preliminary, but novel, information about the use, acceptability, and potential efficacy of a fully-automated Internet-delivered CBT-I program among cancer caregivers. Caregivers’ program engagement was lower than that in prior SHUTi trials among the general population⁶ and cancer survivors.⁹ Caregivers who used SHUTi reported more severe sleep disturbance at baseline relative to caregivers who did not use the program, while non-users reported greater caregiving strain. This suggests caregiving strain may undermine caregivers’ ability to initiate Internet-delivered CBT-I, while insomnia symptom severity may promote uptake. Findings corroborate past trials of therapist-provided CBT-I for cancer caregivers.^4,5

Most caregivers were satisfied with SHUTi even without caregiver-specific tailoring. Caregivers did share where tailoring may be beneficial to caregivers–such as how to adjust behavioral treatment components to accommodate nighttime caregiving. Feedback was, however, largely applicable to how the program might be improved for all individuals. These findings are important, given there are multiple existing digital health interventions that could address pertinent psychosocial concerns for caregivers, so developing novel interventions specifically for this population may not always be necessary.

While it is important not to over-interpret findings given the limited sample, results support the potential of Internet-delivered CBT-I for reducing insomnia symptoms among cancer caregivers. Caregivers who used SHUTi showed substantially improved insomnia, with effect sizes comparable to those reported in past SHUTi trials.^6,9 Importantly, caregivers tended to report more benefit when the care recipient had been diagnosed within the past year (ISI), provided more than 20 hours of care per week (ISI, WASO), and greater emotional strain from caregiving (ISI). This suggests that these factors may not prevent even high-intensity caregivers from benefitting from Internet-delivered CBT-I once engaged in the program.

Limitations and Future Directions

Given the small sample with limited sociodemographic diversity, conclusions should be drawn cautiously. The sample largely comprised highly-educated women; future studies must secure larger and more diverse samples that better represents the complete population of cancer caregivers. Recruitment was slower than expected, at least in part due to the COVID-19 pandemic. Future trials should also consider including caregivers for any health condition, rather than cancer alone, given that cancer caregivers’ feedback was largely not cancer-specific.

Conclusions

Findings from this first feasibility trial of Internet-delivered CBT-I among family cancer caregivers suggest that caregivers may be able to use and benefit from a fully-automated online CBT-I program, even when there is no caregiving-specific tailoring. Caregivers who engaged with the Internet-delivered CBT-I program reported a large reduction in their insomnia symptoms, although they also suggested program modifications that may help more caregivers to initiate, complete, and benefit from the program.

Abbreviations:

CBT-I

Cognitive-Behavioral Therapy for Insomnia

DBAS

Dysfunctional Beliefs and Attitudes about Sleep

ISI

Insomnia Severity Index

SHUTi

Sleep Healthy Using the Internet

SOL

Sleep Onset Latency

WASO

Wake After Sleep Onset

Availability of data and materials:

Data collection materials are available by contacting the corresponding author. Data are not available for distribution as this permission was not explicitly granted as part of the informed consent agreement process.