Sleep to Reduce Incident Depression Effectively (STRIDE): study protocol for a randomized controlled trial comparing stepped-care cognitive-behavioral therapy for insomnia versus sleep education control to prevent major depression

The present study is a large-scale stepped-care clinical trial in the primary care setting that utilizes a sequential, multiple assignment, randomized trial (SMART) design to determine the effectiveness of dCBT-I alone and in combination with clinician-led CBT-I for insomnia and the prevention of MDD incidence and relapse. The study sponsor is the National Institute of Mental Health (NIMH), which had no role in the study protocol or study design. Any proposed protocol amendments will first be approved by NIMH and then the local IRB before being implemented. Protocol deviations will be documented using a breach report form.

Study setting

This is a RCT testing the effectiveness of a stepped-care model of insomnia treatment as implemented in primary care for the treatment of insomnia and prevention of MDD. As part of the Perfect Depression Care initiative, Henry Ford Health (HFH) routinely collects Patient Health Questionnaire (PHQ-9) [49] data from primary care patients. As part of our implementation into HFH primary care, patients who endorse sleep disturbances on the PHQ-9 will be referred to our services for further assessment and study eligibility determination. HFH primary care includes over 30 locations in southeastern Michigan.

Patient recruitment sources

STRIDE patients are primarily recruited from multiple sources. Primarily, this study uses data from the Perfect Depression Care initiative in HFH primary care. Specifically, we identify individuals with PHQ-9 total scores below the clinical cutoff of 10 [49], but who report ≥ 1 on item #3 (“Trouble falling or staying asleep, or sleeping too much several days or more in the last 2 weeks”), which has high sensitivity (82.5%) and specificity (84.5%) for identifying insomnia symptoms in primary care. We will supplement recruitment efforts as necessary by recruiting HFH patients who have an insomnia diagnosis listed in the EMR and via study advertisements included in HFH clinics, online (e.g., HFH wellness program), in the community, and directed toward patients who previously participated in insomnia studies in our center. Interested individuals who consent online to eligibility screening will complete a battery of online surveys.

Eligibility information will be derived from electronic medical records (EMRs) and from patient reports on an online screening survey. The inclusion criteria are as follows: (1) Insomnia Severity Index (ISI) [50] score of 15 or higher to reflect clinically severe insomnia severity and (2) no clinically significant depressive symptoms as reflected by a score of 10 or lower on the Quick Inventory of Depressive Symptomatology (QIDS) [51]. The exclusion criteria are as follows: (1) age < 18 years, (2) current use of antidepressants for depression, (3) bipolar or seizure disorders, (4) untreated sleep disorders other than insomnia (e.g., obstructive sleep apnea, narcolepsy, restless legs syndrome), and (5) known diagnosis of major depression at baseline. Note that patients may start any new medication and/or therapy during participation without penalty. Eligible patients will then be randomized to step 1 treatment.

Allocation

Patients are randomized 1:1 to CBT-I or control at each step using block randomization (step 1 uses 50-person blocks; step 2 uses six-person blocks). Only the study coordinator accesses the allocation sequence and assigns patients to groups.

Blinding

Patients are blinded to the active therapy. Therapists are not blinded. Outcomes will be linked to a blinded group variable, which will be unblinded after primary analyses.

Patient flow

See Fig. 1 for the full patient flow diagram.

Fig. 1figure 1

Flowchart from recruitment through treatment randomization and 2-year follow-up

Step 1

After baseline assessment, participants are randomized into dCBT-I or online sleep education control, each with a group size of n = 500. The dCBT-I group receives treatment via the Sleepio platform (www.sleepio.com), which includes all six standard components of CBT-I delivered weekly through an automated online health program. The sleep education control group receives weekly electronically delivered information regarding sleep education and sleep hygiene. Patients then complete a study outcome assessment after step 1 treatment.

Step 2

Patients who do not remit with dCBT-I (ISI > 7; anticipated ~ 50%) will be randomized to clinician-led CBT-I or sleep education control in step 2 of the SMART design. Treatment will be provided by personnel with CBT-I training (psychologists and nurses). Upon completing step 2, we will assess insomnia and depressive symptoms.

Follow-up assessments

At 1 and 2 years after the initial randomization, we will conduct follow-up assessments, which include measures of insomnia and depressive symptoms, as well as assessment of DSM-5 MDD incidence and relapse (via the SCID-5) occurring at any point in the previous 12-month period before the follow-up assessment. Rumination (aim 3) will be assessed at baseline, post-step 1, post-step 2, and at 1- and 2-year follow-ups.

Study interventionsStep 1: Digital cognitive behavioral therapy for insomnia

Patients randomized to digital CBT-I in step 1 will complete the Sleepio program via the Internet (www.sleepio.com, Big Health Inc.). Sleepio is among several currently available digital CBT-I programs and was selected for this study because it is evidence-based, standardized, and fully automated. Patients are asked to complete six sessions; each session will be unlocked on a weekly basis, and patients are advised to complete one session per week. The intervention covers behavioral components (sleep restriction, stimulus control), cognitive components (e.g., cognitive restructuring, paradoxical intention), progressive muscle relaxation, and sleep hygiene. Sessions are directed by an animated “virtual therapist” who reviews and guides the progress of the patient based on the submitted sleep data. Patients complete daily sleep diaries to help monitor treatment progress and adherence to behavioral sleep strategies.

Step 2: Clinician-led CBT-I

As CBT-I is standardized, clinician-led CBT-I includes six standard sessions that cover the same material as digital CBT-I. However, working one-on-one with a therapist allows for greater personalization of behavioral sleep strategies and cognitive therapy, as well as personalized addressing of treatment barriers such as difficulty adhering to sleep schedules and difficulty completing homework. Patients complete daily sleep diaries to help monitor treatment progress and adherence to behavioral sleep strategies.

Step 1: Digital sleep education control

Patients randomized to the online sleep education condition received six weekly emails based on the National Institutes of Health guide to healthy sleep [52]. Information was provided on the basics of sleep regulation; relationships between sleep and health problems such as obesity, diabetes, and cardiovascular disease; effects of sleep-disruptive substances such as caffeine, nicotine, and alcohol; and tips on creating a sleep-conducive bedroom environment. Psychoeducation and sleep hygiene were selected because they are common in clinical practice, especially in primary care [53, 54], and also because they are commonly used as an attention control in clinical trials of insomnia. Importantly, neither sleep education nor sleep hygiene is considered an effective standalone treatment for insomnia [55].

Step 2: Sleep education control

Digital CBT-I non-remitters who are randomized to attention control in step 2 will receive sleep education. This includes six telemedicine sessions that cover the same material outlined above in the digital sleep education control section.

Data collection schedule

We will assess the study outcomes and other relevant sociodemographic and health-related information at baseline, post-step 1, post-step 2, and 1- and 2-year follow-ups. Please refer to the schedule in the study flowchart in Fig. 1. NIMH, the study sponsor, had no role in the data collection. This study does not involve collecting biological specimens.

Study measures

Primary end-points include (aim 1) the Insomnia Severity Index (ISI) [50], (aim 2) diagnosis of major depressive disorder via clinical interview using the structured clinical interview for DSM-5 disorders [56], and (aim 3) the Pre-Sleep Arousal Scale’s cognitive factor (PSASC) to measure nocturnal rumination [57].

Screening and baseline assessment occur at the same time point and are captured using online surveys. For screening purposes, patients first complete the ISI and the Quick Inventory of Depressive Symptomatology (QIDS) [51]. Patients who endorse insomnia and deny current clinical depression will continue with the rest of the baseline assessment, which includes the Pre-Sleep Arousal Scale’s cognitive factor (PSASC) to measure nocturnal rumination in addition to sociodemographic and other health-related information.

Post-step 1 assessment occurs 1 week after completing step 1 intervention (dCBT-I or control) and includes the ISI, QIDS, and PSASC among other clinical measures. The ISI serves as our primary end-point for insomnia symptoms as tested in aim 1. Patients who report insomnia remission (operationalized as ISI ≤ 7 at post-step 1) will no longer receive insomnia treatment and will enter the follow-up data collection phase of the study described below.

On the other hand, patients who do not remit (ISI ≥ 8 at post-step 1) will then be elevated to step 2 treatment, which involves clinician-led CBT-I or sleep education control conducted via in-person or telemedicine video (the procedure was face-to-face before the COVID-19 pandemic, then switched to telemedicine video during the pandemic due to changes in clinic operations).

Post-step 2 assessment occurs 1 week after completing step 2 intervention (CBT-I or control) and includes the ISI, QIDS, and PSASC among other clinical measures. All step 2 patients then enter the follow-up data collection phase of the study.

Follow-up data collection will last for 2 years and consists of both online surveys and telemedicine video clinical interviews. Both the surveys and interviews are administered at 1 and 2 years after completing the final treatment. Surveys include the ISI, QIDS, and PSASC among other clinical symptom measures. The interviews include the SCID-5 module to diagnose major depression, which is oriented to assess over the preceding year; the MDD diagnosis serves as our primary end-point for MDD incidence and relapse for aim 2. In addition, we will assess new involvement in psychotherapy and/or pharmacotherapy for depression, which may reflect new-onset depression.

Data management

All data will be collected using Qualtrics and will be downloaded from their servers for analysis. Downloaded data will be anonymized and stored on secured network folders that are encrypted and password-protected to preserve confidentiality.

Patient retention

To maximize retention, participants will receive reminders for follow-ups via email, phone, and text using the patient portal already in use throughout our primary care networks. In addition, we will pay participants a stipend per assessment and follow up with them throughout the study to ensure participant retention, particularly during the 1- and 2-year follow-up periods. Our group has been successful in this regard (> 85% retention) in prior NIH and industry trials of > 1–2 years duration. Similar to previous long-term studies in our center, between post-treatment and each follow-up session, we will hold 2–3 raffles per year where a participant who updated their contact info during any given 4-month period wins a prize. In addition, we will send out “thank you” notes for each assessment and a regular newsletter of study progress and eventually results to those who have completed the trial.

Assessment of safety

CBT-I is considered safe when delivered in clinician-led and digital formats; thus, the likelihood of serious adverse events occurring during this trial is low. Even so, we will record all occurrences of serious adverse events in patients, which are defined as deaths, suicide attempts, motor vehicle collisions, and complaints about the interventions. Although CBT-I is safe, we will discontinue treatment at patient request and/or if a therapist believes CBTI to be harmful to a patient.

As insomnia increases the risk for depression and anxiety, it is possible that patients may contact the research team seeking additional mental health services. When this occurs, the research team will help guide patients to connecting with the appropriate mental health services.

Insomnia is a risk factor for suicidal thoughts. Therefore, we will regularly assess suicidal thoughts and behaviors. For patients who express imminent suicidal intent, we will guide patients to seek an evaluation in their nearest hospital emergency services and/or contact their local emergency services to perform a wellness check.

Data monitoring committee

An independent data monitoring committee (DMC) consisting of independent scientists was appointed. The DMC met in the first study year to review study protocols and trial safety. The DMC will then meet on an as-needed basis to review safety and trial conduct.

The steering committee is led by Drs. Drake (PI), Kalmbach (co-I), and Cheng. The steering committee, under the primary supervision of Dr. Drake, is responsible for facilitating patient recruitment, ensuring systems are in place to guarantee institutional compliance with US laws and NIH policies, reviewing trial conduct, translating the research proposal into operational plans and procedures, and overseeing the dissemination of study findings. The steering committee met weekly during the first 6 months, then biweekly thereafter to monitor the progress of the study objective, review interim analyses, and disseminate study findings. The remaining study team members, including the study coordinator and research assistants, support the study by conducting recruitment, consenting, and data collection at each site as well as data tracking in Qualtrics and reporting on progress.

Analytic plan

Detailed descriptive analysis of all quantitative data will be performed. We will use information from this preliminary investigation to (1) describe univariate and bivariate sample distributions of the data, (2) identify the interrelationships among variables (i.e., need for covariate adjustment), and (3) check for violation of assumptions underlying identified statistical techniques (e.g., independence, linearity, homoscedasticity, and normality). The study design involves comparing the study arms to evaluate their relative changes from pre-treatment to post-treatment. Consistent with appropriate clinical trial methodology, we will perform intent-to-treat analyses utilizing all data points with mixed modeling. We will use full information maximum likelihood estimation to handle data missingness in our mixed models. NIMH, the study sponsor, had no role in the analysis plan and will have no role in the data analysis or interpretation.

Aim 1: The immediate focus of the analysis will be on the post-treatment impact on sleep outcomes (ISI) of the Internet-based dCBT-I, followed by the clinician-led CBT-I group, compared to the sleep hygiene control group. A linear mixed model will be used to examine the post-treatment sleep values using the baseline values as covariates. The model is flexible and will allow the inclusion of other covariates as appropriate (medication, alcohol use, caffeine, etc.).

Aim 2: MDD prevention will be assessed using a generalized mixed-effects logistic model that compares the incidence of MDD at year 1 and 2 follow-ups across treatment conditions. Covariates will also be tested, including baseline depression (QIDS), familial history of depression, and relevant demographics (e.g., sex, race). The effectiveness of the stepped-care model will be assessed via two planned comparisons. The first will compare the rate of depression between those in the full stepped-care condition (dCBT-I to clinician-led CBT-I) and those in the step 2 control (dCBT-I to control). The second will assess the overall effectiveness of the stepped-care model by comparing MDD between those who were eligible to receive stepped-care (i.e., those randomized to dCBT-I at step 1) and those who received the sleep education control at step 1. A secondary analysis will also be conducted with QIDS scores using a linear mixed-effects model to examine the differences in depression severity.

Aim 3 (test rumination as a mediator of CBT-I on depression): Mediation significance testing will involve three models: (1) the direct effect of treatment condition on the treatment outcome (e.g., depression), (2) the effect of the treatment condition on the proposed mediator (i.e., change in rumination from pretreatment to posttreatment; this is the α pathway), and (3) the effect of the mediator (change in rumination) on the treatment outcome (e.g., depression) while controlling for treatment condition (this is the β pathway). The indirect (i.e., mediated) effect of the predictor on the outcome variable will be tested using the product of the α and β parameter estimates. Significance testing of the indirect path will be conducted using confidence intervals estimated using the PRODCLIN method implemented in R using RMediation. 

Sample size justification and power: Using current recommendations for effectiveness trials, the current study is powered using clinically relevant effect sizes (moderate) for sleep and incident MDD [58]. We hypothesize a reduction of depression incidence of 44% consistent with our preliminary data. The sample size for the project will be determined from aim 2, the most conservative primary analysis for the proposal. For the aim 2 analysis comparing the control group to the aggregated treatment group(s) on MDD incidence or recurrence, we require 405 subjects per group to detect a decrease of 44% (9.0% vs. 16.2%) with 80% power (Table 1). The number of subjects increases to 453 per group with a similar decrease but a lower incident rate in each group (8% vs. 14.4%; Table 1). Thus, targeted enrollment for aim 2 is a minimum of 405 subjects per group (group 1: all patients who received any form of active treatment in step 1 and/or step 2 vs group 2: all patients who were randomized to control in step 1; refer to the rightmost side of Fig. 1 for reference). To achieve our target sample size, we will enroll 1000 patients in step 1. Even with 19% attrition (higher than anticipated), we will retain 405 patients in each group.

Table 1 Final sample sizes for a 35% decrease in depression incidence/relapse at 3 conservative levels of depression onset

留言 (0)

沒有登入
gif