Single-Group Trial of an Internet-Delivered Insomnia Intervention Among Higher-Intensity Family Caregivers: Rationale and Protocol for a Mixed Methods Study

Background Family caregivers are more likely to experience insomnia relative to noncaregivers but have significant barriers to accessing gold standard cognitive behavioral therapy for insomnia treatment. Delivering interventions to caregivers through the internet may help increase access to care, particularly among higher-intensity caregivers who provide assistance with multiple care tasks over many hours per week. Although there are existing internet interventions that have been thoroughly studied and demonstrated as effective in the general population, the extent to which these interventions may be effective for caregivers without tailoring to address this population’s unique psychosocial needs has not been studied. Objective The goal of this trial is to determine what tailoring may be necessary for which caregivers to ensure they receive optimal benefit from an existing evidence-based, internet-delivered cognitive behavioral therapy for insomnia program named Sleep Healthy Using the Internet (SHUTi). Specifically, we will test the association between caregivers’ engagement with SHUTi and their caregiving context characteristics (ie, caregiving strain, self-efficacy, and guilt) and environment (ie, proximity to care recipient; functional status, cognitive status, and problem behavior of care recipient; and type of care provided). Among caregivers using the program, we will also test the associations between change in known treatment mechanisms (sleep beliefs and sleep locus of control) and caregiving context factors. Methods A total of 100 higher-intensity caregivers with significant insomnia symptoms will be recruited from across the United States to receive access to SHUTi in an open-label trial with mixed methods preassessments and postassessments. At postassessment (9 weeks following preassessment completion), participants will be categorized according to their engagement with the program (nonusers, incomplete users, or complete users). Study analyses will address 3 specific aims: to examine the association between caregivers’ engagement with SHUTi and their caregiving context (aim 1a); to describe caregivers’ barriers to and motivations for SHUTi engagement from open-ended survey responses (aim 1b); and among caregivers using SHUTi, to determine whether cognitive mechanisms of change targeted by SHUTi are associated with differences in caregiving context (aim 2). Results Institutional review board approvals have been received. Data collection is anticipated to begin in December 2021 and is expected to be completed in 2023. Conclusions Findings will inform the next research steps for tailoring and testing SHUTi for optimal impact and reach among caregivers. Beyond implication to the SHUTi program, the findings will be translatable across intervention programs and will hold significant promise to reduce inefficiencies in developing digital health interventions for caregivers while also increasing their impact and reach for this underserved population. Trial Registration ClinicalTrials.gov; NCT04986904; https://clinicaltrials.gov/ct2/show/NCT04986904?term=NCT04986904 International Registered Report Identifier (IRRID) PRR1-10.2196/34792

Phone screening: This screening includes a brief medical history review and structured clinical interview to ensure that caregivers with potential counterindications to cognitive-behavioral therapy for insomnia are not enrolled. By removing clinical assessments not strictly assessing contraindications, we reduced the estimated time for completion of this screening by 10 minutes (25%) to approximately 30 minutes.
Baseline questionnaire assessment: This assessment battery includes constructs fundamental to our proposed aims, and uses brief, validated scales to reliably assess these constructs. By carefully reducing included constructs and identifying shorter validated scales for included constructs where possible, we reduced the estimated completion time for the baseline assessment by 15 minutes (33%) to approximately 30 minutes.
Post-assessment questionnaires for SHUTi users: SHUTi users will complete an online questionnaire battery of both validated self-report scales and free-response items assessing constructs fundamental to our proposed aims and their evaluation of SHUTi. We have added 3 items to assess participants' use and interest in wearable activity monitors for reducing assessment burden in future trials. By carefully reducing included constructs and identifying shorter validated scales for included constructs where possible, we reduced the estimated completion time for the baseline assessment by 15 minutes (33%) to approximately 30 minutes.
We have also sought to balance assessment burden with capturing actionable feedback from caregivers regarding potential tailoring needs. Thus, we have retained the battery of open-ended items that assess this information. Based on feedback from reviewers, we are also willing to include optional phone interviews during the synthesized member checking phase of the study with caregivers who are interested and willing to provide more in-depth information about their experiences using SHUTi as a high-intensity caregiver.
Post-assessment questionnaires for SHUTi non-users: SHUTi non-users will complete a brief online questionnaire battery of both multiple choice and free-response items assessing constructs fundamental to our proposed aims and their barriers to initiating SHUTi. We have added 3 items to assess participants' use and interest in wearable activity monitors for reducing assessment burden in future trials. By carefully reducing included constructs and identifying shorter validated scales for included constructs where possible, we reduced the estimated completion time for the baseline assessment by 10 minutes (50%) to approximately 10 minutes.
Passive sensing to reduce assessment burden: We continue to consider the use of passive sensing by wearable devices (e.g., fitbit) or ambient sensors (e.g., RFID) to reduce or eliminate sleep diary completion as part of SHUTi. We are working to build out these technical capabilities and test the impact on intervention engagement and efficacy, given it is unknown whether the completion of sleep diaries serves as an engagement and/or cognitive intervention component. While these studies are ongoing, we added items to our assessment battery capturing caregivers' ownership and use of wearable devices, as well as their level of interest in using passive sensors to reduce daily diary commitment. Our follow-up studies will be informed by the joint findings of our ongoing work to develop technical capabilities to use passive sensing with SHUTi with our direct assessment of caregiver interest in these capabilities in the present proposal.

Cognitive-Behavioral Focus of Intervention
We also appreciate the consideration of factors outside of maladaptive cognitions and behaviors related to caregivers' insomnia. Here, we provide a more complete description of the SHUTi intervention to establish that this intervention addresses the broader, complex nature of insomnia beyond thoughts and behavior. We also emphasize how our primary aims of the proposal will help us pinpoint where a more targeted focus on environmental and/or caregiving factors may be necessary for high-intensity caregivers to maximally benefit from this intervention.

SHUTi program:
SHUTi is a self-guided, fully automated, interactive, and tailored Internet-based program modeled on the primary tenants of face-to-face cognitive-behavioral therapy for insomnia (CBT-I; sleep restriction, stimulus control, cognitive restructuring, sleep hygiene, relapse prevention). CBT-I is the firstline recommended treatment for insomnia. 1 The primary model of behavior change for SHUTi is based on Social Cognitive Theory (SCT). 2 SCT, the most widely used model of behavior change in the Internet intervention literature, 3 explains behavior change through the dynamic interaction of behavior, personal factors (e.g., cognitions), and the environment. SHUTi content is metered out over time through six "Cores." As demonstrated in the table of SHUTi content, the intervention addresses insomnia by targeting modifiable insomnia-perpetuating cognitive and behavioral factors, it also clearly addresses how non-modifiable life events influence insomnia.

Current proposal:
The central question of our proposal directly addresses the tension highlighted by NCATS leadership: What tailoring is necessary and sufficient to achieve optimal engagement with and efficacy of SHUTi for caregivers? On one hand, SHUTi has demonstrated robustly positive effects among medicallyand psychiatrically-diverse samples in multiple clinical trials, and our preliminary evidence suggests caregivers generally reported comparable satisfaction, fit, and usability of SHUTi compared to non-caregivers. On the other hand, there is reason to believe that caregiving has unique psychological and environmental factors that would affect use and impact of SHUTi, and our preliminary evidence suggests that caregivers did not report as significant treatment benefits as non-caregivers. By directly assessing to what extent caregiving user-and environmental characteristics are related to SHUTi uptake, engagement, and efficacy, we will establish to what extentand howcaregiving-specific tailoring is needed to maximize caregivers' benefit from SHUTi. Moreover, our analyses will determine the exent to which these findings extend to other evidence-based digital health interventions for family caregivers. We are grateful for reviewers' supportive comments, which commended our "important topic," and "strong and experienced team of investigators." Our modifications in this A1 application were described as "highly responsive," with a "re-designed approach to provide a solid plan for this [study] and future steps." We address the constructive critiques of our reviewers below:

Reviewers 1 & 3:
We are grateful for the strong support of R1 and R3. R1: We agree that insomnia is one of a host of psychological concerns of caregivers, and clarify that information gained by understanding what tailoring is necessary and sufficient to optimize the usability and efficacy of our online insomnia program for caregivers will also be pertinent to the tailoring of other evidence-based digital health programs for caregivers. Although we have retained open-ended questions to assess the entire sample for feasibility, we will include phone interviews with a subset of caregivers, who indicate that they are willing and interested, during the synthesized member checking process, per reviewer suggestion, to elicit further detail about potential tailoring needs. As suggested, we will also include a robust protocol for reaching non-completers to limit asymmetric attrition across levels of engagement. R3: We clarify how findings will practically inform tailoring: not by altering the caregiving context, but instead, by altering the content or delivery of the intervention to be more accessible or acceptable given the context. Examples include breaking Cores into smaller segments, or using more audio to allow multi-tasking. Our research team has published multiple papers validating elements of the Model for Internet Interventions. [4][5][6][7] Reviewer 2: We appreciate the critiques of R2. As in response to R1, we emphasize the importance of information gained from this proposal regarding tailoring to optimize usability and efficacy of SHUTi also being pertinent to the tailoring of other digital health and insomnia programs for caregivers, enhancing the translational significance of this work. We also clarify that, although the latest R01-level clinical trial of SHUTi exclusively recruited adults 55 and older, multiple clinical trials of SHUTi have recruited adults of all ages and demonstrated positive effect of the program. Alzheimer's disease/dementia and cancer are among the top 5 most common main care recipient problems (accounting for 11% and 6% of caregivers, respectively). 8 Thus, we expect caregivers for people with dementia or cancer will be highly represented in our sample. However, our recruitment methods (including research registry outreach, in-clinic recruitment, and online national recruitment methods) will target high-intensity caregivers across caregiving contexts. In particular, our primary recruitment sourcesthe National Rehabilitation Research and Training Center on Family Support and the University Center for Social and Urban Research Survey Research Program research registryare both agnostic to caregiving condition. We will also work with our CTSA recruitment support teams to ensure a distribution of caregiving contexts in our study. Regarding study measures, all study have been selected for their robust validation and use in our prior trials with caregivers. The reviewer raises an important consideration related to assessment burden for this population. We have carefully reviewed our batteries to use shorter validated scales where possible and remove any assessments that are not strictly necessary for completing and interpreting proposed study aims. We will emphasize with participants that they may take breaks in completing the questionnaire battery, returning to complete the surveys where they left off via features embedded in our online survey tools. In addition, both R1 and R2 asked for more detail about the six SHUTi Cores: we clarify that Cores act as an online analog for the weekly sessions typically used when delivering CBT for insomnia in a face-to-face format, covering content on: (1) program overview, (2 & 3) behavioral insomnia treatment strategies, (4) cognitive insomnia treatment strategies, (5) sleep hygiene education, and (6) relapse prevention. We direct reviewers to Thorndike et al., 2008 13 for complete program details.
In summary, we are grateful for the strong support of the reviewers. By addressing reviewers' critiques, the rigor of our proposed methods and significance of study findings are clarified and enhanced. As such, this proposal has even greater potential to address the central research question on tailoring interventions for caregivers that reviewers described as "important," "significant," and "innovative."