CBT-I or Zolpidem/Trazodone for Insomnia

Insomnia is a common health problem that causes distress, impaired function, and increased risk for other health problems. Chronic insomnia is defined by problems with the quality or amount of sleep, including difficulty falling asleep, frequent awakenings, and/or awakening early and being unable to return to sleep. In developing this application, patients, providers, payors, and the investigators identified both concerns and opportunities with current treatments for chronic insomnia. Medications and Cognitive-Behavioral therapy for insomnia (CBT-I; a treatment program to improve sleep through changes in behavior and thinking) are both effective for treating insomnia. Zolpidem, the most frequently prescribed insomnia medication, is widely available, but may cause side effects and dependency. Trazodone is increasingly prescribed off-label for treatment of insomnia, but evidence of efficacy and safety is more limited. CBT-I is the recommended first line treatment by many professional organizations, but it is not widely available in physicians' practices. Patients, providers, and payors face important unanswered questions: Which treatment should be used for the treatment of chronic insomnia? Is combination therapy more effective, and does it result in lower zolpidem use? Who responds best to which treatment? These dilemmas are particularly relevant to patients and providers in rural areas, where access to behavioral health specialists is limited, and concerns regarding use of controlled substances is particularly acute.

To address these questions, we propose the study, Comparative Effectiveness of Zolpidem/Trazodone and Cognitive Behavioral Therapy for Insomnia in Rural Adults (COZI). The investigators will use a well-tested Internet version of CBT-I, which is just as effective as in-person CBT-I, but more widely available. The aims of COZI are to: 1: Compare the effectiveness of medication preference (zolpidem or trazodone), CBT-I, and combination treatment for insomnia symptoms over 6 months. 2: Compare the effectiveness of medication preference, CBT-I, and combination treatment for other symptoms and problems, including health-related quality of life, mood, and health outcomes. 3: Compare the side effects of medication preference, CBT-I, and combination treatment. COZI will study 155 patients recruited from 7 health care systems across the country, each of which has established practices in rural areas. Patients will be treated with medication prescribed by their own physician (zolpidem or trazodone), Internet CBT-I, or a combination of the two. The investigators will evaluate treatment effects at 9 weeks, and 6 and 12 months (with limited additional assessments for adverse events and medication use at 1 and 9 months). Our study questions and design reflect early and ongoing engagement of key stakeholders including patients, providers, and payors, who are members of the Study Advisory Committee (SAC). SAC members will work with study investigators to recommend the best strategies for recruiting patients, measuring treatment effects, and disseminating study findings. The SAC will meet quarterly. Patients and other stakeholders have already provided guidance on which treatments are most worth studying, and which symptoms would best reflect the effects of treatment. Patients and other stakeholders will also be part of the Steering Committee, which is the main decision-making group for our study, and which will meet monthly. COZI will lead to answers for patients and providers addressing insomnia; improved health and function for millions of rural Americans; and sustainable changes in how insomnia is treated in rural practices.

• Zolpidem (Drug), also known as Ambien: Zolpidem, as prescribed by physician
• Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) (Behavioral), also known as SHUTi: 6-week internet CBT-I multi-component intervention (SHUTi) including sleep restriction, stimulus control, cognitive restructuring, sleep behaviors, and relapse prevention
• Trazodone (Drug), also known as Desyrel: Trazodone, as prescribed by physician

Primary outcomes

• Change in Insomnia Symptom Severity [Baseline, 6 months]: Change in Insomnia Severity Index score from baseline to follow-up. The Insomnia Severity Index is a 7-item self-report questionnaire that measures severity of insomnia symptoms during the last two weeks. Each item is scored 0 (no problem) to 4 (very severe problem) and total between 0-28, with higher scores indicating more severe insomnia symptoms. Change is estimated mean change from mixed-effects linear regression model. Fixed effects: Baseline ISI, Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.

Secondary outcomes

• Number of Participants With Treatment Response of Insomnia Symptoms [Baseline, 6 months]: Treatment response defined as a ≥6 point reduction in Insomnia Severity Index score from baseline to follow-up. The Insomnia Severity Index is a 7-item self-report questionnaire that measures severity of insomnia symptoms during the last two weeks. Each item is scored 0 (no problem) to 4 (very severe problem) and total between 0-28, with higher scores indicating more severe insomnia symptoms. Data presented as count of participants meeting treatment response criteria at 6 months.
• Number of Participants With Remission of Insomnia Symptoms [6 months]: Remission of Insomnia symptoms defined as Insomnia Severity Index <8 at follow-up. The Insomnia Severity Index is a 7-item self-report questionnaire that measures severity of insomnia symptoms during the last two weeks. Each item is scored 0 (no problem) to 4 (very severe problem) and total between 0-28, with higher scores indicating more severe insomnia symptoms. Data presented as count of participants meeting insomnia remission criteria at 6 months.

Other outcomes

• Change in Anxiety and Depression [Baseline, 6 months]: The Patient Health Questionnaire (PHQ-8) is an 8-item self-report measure for assessing depression severity. The items are summed to yield a score between 0 and 24, with higher scores indicating more frequent depression symptoms. The Generalized Anxiety Disorder (GAD-7) Scale is a 7-item self-report measure used for screening and assessing severity of GAD. The items are summed to yield a score between 0 and 21, with higher scores indicating more severe GAD symptoms. Change is estimated mean change in mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Change in Cognitive Function [Baseline, 6 months]: The PROMIS Cognitive Function Short Form (8a) is an 8-item self-report measure related to executive functioning. Individual item responses range from 1 (very often) to 5 (never) and the total raw score is converted to standardized T-scores using established United States general adult population with a mean of 50 and a standard deviation of 10. Higher T-scores signify better cognitive function. Estimated mean change in T-score from baseline to follow-up is reported from mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site. Positive change indicates improved cognitive function.
• Change in Patient-Reported Health [Baseline, 6 months]: The PROMIS Global Health is a 10-item self-report measure that assess general health and functioning. Item responses range from 1 (excellent/none) to 5 (poor/always) and a single pain item ranges from 0 (no pain) to 10 (worst pain). Physical and mental health sub-scores are calculated from 4 items each, and the total raw sub-scores (range: 4-20) are converted to T-scores using distributions standardized to mean 50 (SD 10) for the US general population. Higher score indicates greater health-related quality of life. The Medical Outcome Survey (SF-12) is a 12-item self-report measure of physical and mental health. The Physical Component Score (PCS) and Mental Component Score (MCS) sub-scores range 0-100; higher score indicates better health. Change is estimated mean change from baseline to follow-up in mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Change in Fatigue [Baseline, 6 months]: The PROMIS Fatigue instrument is a 7-item self-report measure that assesses recent fatigue and how fatigue interfered with daily activities. Scores range from 7 to 35, with higher scores indicating greater fatigue. Change is estimated mean change from mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Change in Pain Intensity and Interference [Baseline, 6 months]: The PROMIS Pain Intensity scale is a 3-item instrument used to assess how much pain a person has. Scores range from 3-15, with higher scores indicating greater pain intensity. The PROMIS Pain Interference scale is a 6-item instrument used to assess the consequences of pain in different domains of daily life. Scores range from 6-30, with higher scores indicating greater interference. Change is estimated mean change from mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Change in Pittsburgh Sleep Quality Index [Baseline, 6 months]: The Pittsburgh Sleep Quality Index (PSQI) is an 18-item self-report measure that assesses sleep quality in the past month. These items are combined into seven component scores (subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbance, use of sleep medication, and daytime dysfunction). The seven component scores are added to produce a global sleep quality score ranging from 0 to 21; higher total score indicates worse sleep quality. Change is estimated mean change from mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Change in Daytime Sleepiness [Baseline, 6 months]: The Epworth Sleepiness Scale (ESS) is an 8-item self-report measure that assesses daytime sleepiness by asking about the likelihood of dozing in certain situations. Individual items are summed to yield the ESS score between 0 (never doze) to 24 (doze a lot) with higher scores indicating more sleepiness. Change is estimated mean change from mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Change in Munich Chronotype [Baseline, 6 months]: The Munich Chronotype Questionnaire (MCTQ) is a self-report scale to assess sleep structure, patterns, duration, and quality. The chronotype sub-score represents the mid-point time of sleep which is converted to a numeric value by adding the hour plus minutes divided by 60. Greater values indicate later chronotype. Positive change values indicate a shift to later chronotype. Change is estimated mean change from mixed-effects linear regression model. Fixed effects: Treatment Arm, Visit and Treatment Arm*Visit interaction. Random effects: intercept for Subject, nested within Site.
• Dysfunctional Beliefs and Attitudes About Sleep [Baseline, 6 months]: Dysfunctional Beliefs and Attitudes about Sleep (DBAS) Scale is a 16-item instrument used to evaluate dysfunctional thoughts related to insomnia. Each item is scored 0-10 and the total is the average of the 16 item scores (range 0-10), with higher scores indicating greater dysfunctional beliefs about sleep.
• Number of Participants With Insomnia Treatment Side Effects [1 month, 9 weeks, 6 months]: Side effects (frequency, severity, impairment) are recorded via self-report adapted from the FIBSER. Scores range from 0 (no side effects; better) to 21 (worse). Side effects were analyzed if participant reported severity of greater than moderate and/or frequency greater than 50% of the time. Count of participants meeting the threshold for analysis up through the 6 month follow-up.
• Medication Adherence [1 month, 9 weeks, 6 months, 9 months, 12 months]: Medication use is assessed using self-reported prescription, over-the-counter, and supplement name and frequency. Count of participants reporting zolpidem or trazodone use for sleep is recorded for adherence to the Medication and Combination treatment arms.
• Self-reported Falls [1 month, 9 weeks, 6 months]: Participants self-reported number of falls including context, location, and consequences of falls. Count of participants with one or more falls up through the 6 month follow-up.

Sex: All, ages: 18 Years to 80 Years

Inclusion Criteria:

* Age 18-80
* Insomnia Severity Index score > 10
* Regular internet and computer access
* Receives primary care in a Non-metropolitan/Rural residence

Exclusion Criteria:

* Use of hypnotic medication >2 times in the past week
* Current cognitive or cognitive behavioral treatment for insomnia
* Psychotic disorder
* Bipolar disorder
* Current substance use disorder
* Severe Chronic obstructive pulmonary disease (COPD) or other severe pulmonary disease (such as asthma or lung fibrosis)
* Cognitive impairment or dementia
* History of spontaneous or hypnotic-induced complex sleep behavior
* Delayed sleep phase disorder (DSPD)
* Shift work that includes working the night shift (between the hours of 12:00 a.m. - 6:00 a.m.)
* History of fracture or injurious fall in the past 12 months
* Currently pregnant, planning to become pregnant, or breastfeeding
* Other severe or uncontrolled mental or physical disorders that would make participation difficult or unsafe

Participant flow

Outcome results

• Change in Insomnia Symptom Severity (Primary), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) -8.05 (-9.44 to -6.65); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -7.82 (-9.32 to -6.32); Combination -10.19 (-11.66 to -8.73)
• Number of Participants With Treatment Response of Insomnia Symptoms (Secondary), Count of participants, Participants: Medication (Zolpidem or Trazodone) 22; Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 21; Combination 31
• Number of Participants With Remission of Insomnia Symptoms (Secondary), Count of participants, Participants: Medication (Zolpidem or Trazodone) 16; Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 13; Combination 16
• Change in Anxiety and Depression (Other pre specified), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) -2.37 (-3.39 to -1.36); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -2.74 (-3.87 to -1.61); Combination -4.63 (-5.7 to -3.56)
• Change in Cognitive Function (Other pre specified), Least squares mean, T-score: Medication (Zolpidem or Trazodone) 2.87 (0.81 to 4.93); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 6.64 (4.36 to 8.93); Combination 4.58 (2.42 to 6.74)
• Change in Patient-Reported Health (Other pre specified), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) 1.22 (-0.52 to 2.95); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 2.43 (0.49 to 4.37); Combination 4.01 (2.18 to 5.83)
• Change in Fatigue (Other pre specified), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) -5.44 (-7.99 to -2.89); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -8.82 (-11.66 to -5.99); Combination -6.8 (-9.48 to -4.12)
• Change in Pain Intensity and Interference (Other pre specified), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) -0.55 (-2.7 to 1.6); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -1.72 (-4.13 to 0.68); Combination -3.84 (-6.11 to -1.58)
• Change in Pittsburgh Sleep Quality Index (Other pre specified), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) -3.42 (-4.46 to -2.38); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -4.52 (-5.67 to -3.37); Combination -5.23 (-6.32 to -4.15)
• Change in Daytime Sleepiness (Other pre specified), Least squares mean, change score on a scale: Medication (Zolpidem or Trazodone) -2.81 (-3.77 to -1.84); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -2.61 (-3.68 to -1.54); Combination -2.18 (-3.2 to -1.17)
• Change in Munich Chronotype (Other pre specified), Least squares mean, hours: Medication (Zolpidem or Trazodone) 0.08 (-0.16 to 0.33); Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) -0.02 (-.029 to 0.24); Combination 0.23 (-0.03 to 0.5)
• Dysfunctional Beliefs and Attitudes About Sleep (Other pre specified), Mean, score on a scale: Medication (Zolpidem or Trazodone) 6.1 +/- 1.8; Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 6.0 +/- 1.3; Combination 6.2 +/- 1.7
• Number of Participants With Insomnia Treatment Side Effects (Other pre specified), Count of participants, Participants: Medication (Zolpidem or Trazodone) 16; Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 13; Combination 15
• Medication Adherence (Other pre specified), Count of participants, Participants: Medication (Zolpidem or Trazodone) 38; Combination 28
• Self-reported Falls (Other pre specified), Count of participants, Participants: Medication (Zolpidem or Trazodone) 8; Internet Cognitive Behavioral Therapy for Insomnia (CBT-I) 3; Combination 7

Adverse events

Time frame: 12 months

Serious adverse events by term

Other adverse events by term

• Study Protocol and Statistical Analysis Plan (PDF, 2023-04-20)

• Katie L. Stone, PhD, Principal investigator, California Pacific Medical Center/Sutter Bay Hospitals
• Daniel J. Buysse, MD, Principal investigator, University of Pittsburgh