Exploring the Meaning of Cognitive Behavioral Therapy for Insomnia (CBT-I) for Patients with Chronic Pain

Abstract

Objective:

Insomnia is one of the most common, persistent, and distressing symptoms associated with chronic pain. Cognitive behavioral therapy for insomnia (CBT-I) is the first-line treatment for insomnia, but patient preferences and perspectives about CBT-I within the context of chronic pain are unknown. The current qualitative study sought to understand the experience of CBT-I among patients with chronic pain, including aspects of CBT-I that were found to be difficult (e.g., pain as a specific barrier to adherence/drop-out), changes in sleep and pain functioning after CBT-I, and aspects of CBT-I that were appreciated.

Design:

Qualitative semi-structured interviews.

Methods:

We conducted individual semi-structured interviews with 17 Veterans with chronic pain and insomnia who had recently participated in CBT-I, as well as their CBT-I therapists, and used thematic analysis to identify conceptual themes.

Results:

Results revealed that patients and CBT-I therapists found changing sleep habits during CBT-I challenging due to anxiety and temporary increases in fatigue, but did not identify major pain-related barriers to adhering to CBT-I recommendations; patients experienced better sleep, mood, energy, and socialization after CBT-I despite minimal changes in pain intensity; and patients highly valued CBT-I as a personalized treatment for sleep and strongly recommended it for other patients with chronic pain.

Conclusions:

Findings of improved sleep and functional outcomes support efforts to incorporate CBT-I into chronic pain treatment, including educating patients and providers about the strong feasibility of improving sleep and quality of life despite ongoing pain.

Introduction

Insomnia is commonly associated with chronic pain. Sleep disturbances are reported among 80–90% of treatment-seeking patients with chronic pain, with at least half estimated to meet criteria for a diagnosis of insomnia.(1–4) These disturbances include trouble falling asleep, staying asleep, and early-morning awakenings with associated daytime sleepiness, fatigue, lack of energy, and difficulty concentrating.(5) Sleep disturbances may be precipitated by pain, but frequently become an independent, chronic condition due to compensatory strategies and behaviors (e.g., napping and inactivity during the day).(4) Unsurprisingly, patients with chronic pain describe sleep disturbance as more troublesome and detrimental to quality of life than other chronic pain-related impairments, rating sleep improvement as one of the most important goals of chronic pain treatment.(5–7)

Cognitive behavioral therapy for insomnia (CBT-I) is a highly-efficacious tailored package of proven behavioral and cognitive techniques that rapidly improve sleep quality by strengthening circadian rhythms, increasing sleep drive, reducing sleep effort, and alleviating insomnia-related anxiety.(8) The basic components of CBT-I include: 1) sleep restriction, which involves limiting time in bed to consolidate sleep; 2) stimulus control, which involves using the bed/bedroom only for sleep; 3) cognitive restructuring, which reduces cognitive arousal by altering maladaptive thoughts and beliefs about sleep and enhancing a mindful acceptance of insomnia symptoms, and 4) sleep hygiene education to address behavioral habits and environmental factors that negatively impact sleep. CBT-I has also shown widespread benefits for improved mood, reduced inflammation, and reduced reliance on sedative-hypnotics, all common vulnerabilities among patients with chronic pain.(9–11)

Although pharmacological treatment of insomnia is the most common approach in clinical practice, sedative-hypnotic medications—especially benzodiazepines—are associated with serious adverse effects, including dementia, falls and fractures, worsening depression, overdose, and motor vehicle accidents.(12, 13) Guidelines for management of chronic insomnia state that CBT-I provides better overall value than pharmacologic treatment due to fewer harms and more proven benefits.(8, 12, 13) CBT-I is an especially promising treatment for patients with comorbid insomnia and chronic pain for whom high-risk combined opioid and hypnotic medication is often prescribed.(14, 15)

Meta-analyses have shown that CBT-I improves sleep among patients with chronic pain(16, 17) and there is some indication that CBT-I improves pain outcomes.(18) The majority of this work has been based on experimental studies and clinical trials; patient perspectives on CBT-I in the context of chronic pain are unknown. Understanding consumer preferences and opinions about sleep treatment is crucial for widespread utilization of evidence-based treatments for insomnia.(19, 20) Qualitative insights into barriers and facilitators of insomnia care is particularly crucial among patients with chronic pain for whom participation in and adherence to CBT-I recommendations may be especially difficult due to the ongoing and widespread influence of pain.(21–24) Pain coping behaviors conflict with CBT-I strategies to regularize circadian rhythms(25, 26) and, when unaddressed, may result in premature discontinuation or poor adherence.(25, 27) Pain catastrophizing conflicts with CBT-I strategies to promote relaxation prior to sleep.(28–31) Pain-related beliefs (e.g., it is impossible to get a good night of sleep unless pain is eliminated), if unacknowledged, may undermine acceptance of key CBT-I messages.(22, 28, 32, 33)

The current qualitative study sought to understand the experience of CBT-I among patients with chronic pain, as well as the perspectives of their CBT-I providers, to identify potential barriers and facilitators for increasing use of and adherence to CBT-I among patients with chronic pain. We used semi-structured interviews to address the following research questions:

1. Which aspects of CBT-I were found to be difficult among patients with chronic pain, and specifically, what pain-related challenges to treatment attendance and adherence showed up during therapy?

2. What changes did patients experience in sleep, pain and overall functioning following CBT-I?

3. Which aspects of CBT-I were appreciated among patients with chronic pain?

Methods

We conducted individual semi-structured interviews with two groups of participants: 17 Veterans with chronic pain and insomnia who had recently participated in CBT-I and 7 CBT-I providers who had worked with these patients. The local institutional review board approved all procedures and we obtained informed consent from all participants prior to conducting interviews. The main goal of this study was to explore the experience of CBT-I among patients with chronic pain and insomnia, but we also explored barriers/facilitators for CBT-I referrals and strategies for “selling” CBT-I; these findings are reported separately.(34)

Participants

Patients.

Patient participants were Veterans who had completed at least one therapy session of CBT-I (medical record included a templated note from the first CBT-I session) and had clinically relevant chronic pain (score of 5 or greater on the 3-item PEG scale assessing pain intensity (P), interference with enjoyment of life (E), and interference with general activity (G)).(35) We used a purposeful sampling strategy to capture diverse perspectives regarding CBT-I, including consideration of demographic characteristics (e.g., gender, race/ethnicity/age, service era) and patient adherence (e.g., completion or premature discontinuation of CBT-I as indicated by the CBT-I provider in the medical record; completion was defined as a templated note indicating that patients had attended a final CBT-I session, whereas premature discontinuation was defined as a lack of a final CBT-I session note). Patients were excluded if more than three months had elapsed since their last CBT-I session to facilitate accurate recall of treatment. The study coordinator sent invitation letters to potentially eligible patients and if patients were interested, screened for eligibility over the phone. Recruitment and enrollment continued until thematic saturation was achieved and further interviews did not provide additional substantial information about themes. There were 17 patient participants.

CBT-I Providers.

Patients who completed the study were asked permission for study staff to contact their CBT-I providers to discuss the course of their therapy; all patients agreed. The study coordinator invited providers by email to be interviewed by non-clinician members of the study team. In some cases, providers were interviewed on separate occasions referencing different patients to capture new information (i.e., one patient who completed treatment vs. another who prematurely discontinued). Each provider interview was counted separately, resulting in 11 interviews among 7 providers. All CBT-I providers had completed the Veterans Affairs (VA) national CBT-I training initiative or equivalent local training and were actively involved in ongoing consultation and management of CBT-I referrals. Therapy sessions were based on a CBT-I manual developed by VA.(36) The VA CBT-I protocol consists of a treatment intake and five treatment sessions, with patients attending fewer or more sessions if clinically indicated. The basic components of this protocol include the CBT-I components discussed earlier, including sleep restriction, stimulus control, and cognitive restructuring. Due to the study goal of examining perspectives on real-world CBT-I delivery, we allowed flexible delivery of CBT-I and so content and sequencing of CBT-I sessions were free to vary among therapists and patients.

Interviews

Interviews were conducted between January and September 2018 using a semi-structured interview guide that the study team developed and refined. The guide began with broad questions regarding the Veterans’ history of chronic pain and insomnia, followed by open-ended questions designed to elicit participant reflections of their experiences with CBT-I. Participants discussed both pain specific and general barriers/facilitators to treatment adherence, what was perceived as most or least helpful about CBT-I, suggestions for improvement or modification to CBT-I for patients with chronic pain, and perceived changes in sleep, pain and overall functioning following CBT-I. A parallel guide was used with CBT-I providers to obtain their perspectives on treatment experiences with individual Veterans.

Trained bachelor’s level research staff conducted most interviews, with the primary author also conducting several interviews; to minimize potential bias in participant responses and the interviewer follow-up questions, all provider interviews and most patient interviews (82%) were conducted by non-clinicians. Interviews with patients lasted approximately 45–60 minutes and 30 minutes with providers. Patients were compensated for their time at the end of the interview. Interviews were audio-recorded using an encrypted recorder and professionally transcribed verbatim.

Data Analysis

We used a thematic analysis approach in which common ideas were identified across interviews and then grouped into larger conceptual themes, with underlying subthemes to capture unique dimensions of the larger constructs.(37) We started with a-priori categories based on the interview guide and expanded these categories to encompass new findings. We analyzed interviews using NVIVO software to code, organize, and retrieve qualitative data. Three trained coders analyzed the data concurrently with data collection. We developed a preliminary coding scheme and refined it using an iterative process by which members of the analysis team independently read, annotated, and applied preliminary codes to transcripts from patient and CBT-I provider interviews; following independent review of each transcript, coders met and discussed the coding list to reach consensus on additions or modifications to the coding scheme.

Once the coding scheme was stable (i.e., no new codes emerged, no additional changes made to code list) and coders achieved high inter-rater consistency, transcripts were divided among coders and the final code list was applied to all transcripts, including those used for code development. Multiple codes were applied to passages of text if appropriate. During the coding process, 25% of interviews were randomly selected for recoding by one of the other coders on the team to ensure consistency and comprehensiveness of coding; discrepancies were resolved through consensus. The analysis team met weekly during the coding process to ensure consistency in coding and to encourage reflexivity (i.e., awareness of how the coders’ background, preconceptions, and biases may influence interpretation of the data). After all coding was complete, the analysis team met to organize coded data into major themes.

Results

Participants were 3 women and 14 men, with the majority falling within the age range of 55–64 (18% aged 35–44, 18% aged 45–54, 29% aged 55–64, 35% aged ≥ 65). Participants were primarily Caucasian (88%) and male (82%), which matches the predominant demographics of the VA population. Nearly half of participants were married or cohabiting with a partner (47%) and reported current employment (42%). Baseline Insomnia Severity Index (38) scores averaged 19 and baseline PEG scores averaged 6, both within the moderate range of severity. The most commonly reported pain condition was musculoskeletal pain, including back and joint pain. Per the medical record, 29% of patients were taking prescription sleeping medications at the start of CBT-I and 29% had diagnosed obstructive sleep apnea. Participants completed an average of 5 sessions of CBT-I (range from 2 to 10). Of patients interviewed, 10 had a final CBT-I session note indicating completion of therapy, whereas 7 did not have a final CBT-I session note. All CBT-I providers were psychologists and mean number of years delivering CBT-I was 4 (range from 1 to 8).

In the interviews, we identified three major themes: 1) Patients and CBT-I providers acknowledged that changing sleep habits during CBT-I is challenging due to anxiety and temporary increases in fatigue, but they did not identify significant pain-related barriers to adhering to CBT-I recommendations. Patients discussed the ability to obtain better sleep using CBT-I techniques, even in the face of ongoing pain. 2) Patients and CBT-I providers described substantial improvement in a wide range of functional outcomes, including activity levels and socialization, but did not note substantial improvement in pain. 3) Patients valued their experience with this treatment and recommended it for other patients with comorbid insomnia and pain.

Challenges with CBT-I

Patients were asked about their experience during CBT-I, including any barriers or facilitators to adhering to each treatment component (e.g., sleep restriction, stimulus control, relaxation). Almost all patients identified sleep restriction as the most challenging part of CBT-I due to the difficulty staying awake until the scheduled bedtime and anxiety about changing their schedule. Sleep restriction often involves a later bedtime to consolidate sleep and reduce time awake in bed.

I mean it was tough at first. […] I would be so tired by the time 2 o’clock in the morning came, because I think that’s in the beginning what time he had me going to sleep, I couldn’t keep my eyes open and a few times I actually went to bed before 2 (patient 1031).

He went on to describe how he was glad that he persisted with sleep restriction.

Then it started to get better and better and better and better and then he moved my time back to midnight or 1 and so I was staying asleep for 5 or 6 hours, which was good, and I thought that’s great. And then I finally was like, ‘We’re done.’ I just told him, ‘You know, I’m getting enough sleep now. I’m feeling good. I’m feeling better.’ I wasn’t tired. I wasn’t lethargic (patient 1031).

Other patients expressed similar experiences with sleep restriction.

I was a little worried about going to bed so late. I was just dwelling on that and she just said, ‘You’d be surprised.’ And we were surprised because I found that I could do it and that I did sleep better. I mean I still had moments where I’d wake up, but I slept longer and more restful. So, yeah, I was anxious about that a lot (patient 1054).

Despite difficulties with sleep restriction, patients commonly identified this component of treatment as the most effective for improving their sleep, often resulting in a “breakthrough” of high quality sleep within a few weeks.

I was going from being in bed from 10 to 11 hours only getting 2–1/2 hours sleep to being in bed for 6 hours and getting 4–1/2 to 5 hours of sleep. […] In the end it made a lot of sense but it was rough getting there. After the first couple of nights it made it a whole lot easier when I saw how much sleep I was actually getting. Because I mean to get twice as much sleep as I was getting before, overall it was amazing, and that made it easier (patient 1104).

CBT-I providers reported that social support and encouragement from patients’ family members helped with adherence to sleep restriction. For example, this provider said:

The biggest help for sleep restriction with [patient 1054] was her family. Her husband and her daughter, she told them about the treatment and they got on board and they helped her stay up because they would stay up until 10:00. […] They would just hang out together (provider 3).

Notably, most patients did not identify pain-related barriers that prevented them from scheduling and attending appointments and following CBT-I treatment recommendations; this was consistent with provider reports. Of the 7 patients who were identified as prematurely discontinuing treatment, all reported some degree of improvement in sleep and expressed appreciation of CBT-I. Reasons for discontinuing CBT-I varied, but did not include pain: satisfaction with current sleep (n=3); assumed that they had successfully completed CBT-I so didn’t schedule final appointment (n=2); too busy to attend medical appointments (n=1); and surgery that interfered with treatment (n=1). Compared to patients who successfully completed treatment, those who discontinued reported more barriers related to scheduling and transportation. CBT-I providers did not generally modify CBT-I for their patients with chronic pain, although some used activity pacing (i.e., alternating activity and rest) and pain-specific relaxation exercises. In fact, this CBT-I provider tried not to make pain the focus during CBT-I so as not to dilute the treatment effect:

[Patients] can get so really focused on the pain and wanting the perfect treatment for that. Then all the sudden you come along, and you can help them with their sleep, and help a little bit and lo and behold, it actually has a secondary benefit of helping their pain. I focus a lot more on the sleep than the pain. […] [Pain is] a concern and I keep my eye on it, but I feel like the more that we talk about the pain, the less likely that the sleep intervention is going to be helpful (provider 4).

Providers acknowledged that pain-specific treatment would be helpful in addition to CBT-I, either integrated into CBT-I or before or after CBT-I.

Perceived Treatment Outcomes

Most patients reported a sustained improvement in sleep quality due to their engagement in CBT-I and many noticed that they were getting several hours more sleep. These perceived outcomes were not clearly linked to number of sessions attended.

I am very, very thankful for [the CBT-I therapist] now that you might say she’s changed my life. I’m getting rest now. I’m getting seven hours sleep now and that is really good when you compare to three and a half, four and a half hours before (patient 1073).

Although isolated pain incidents did disrupt sleep occasionally, patients reported a consistent improvement in sleep quality and quantity following CBT-I.

Patients were surprised when asked about the effect of CBT-I on their pain; many did not expect to have reduced pain since CBT-I focuses on their sleep. Patients reported some temporary pain relief following a good night sleep, but few reported sustained and clinically meaningful changes in pain frequency and intensity following CBT-I. Most patients believed they had very little agency over their pain, in direct contrast to sleep and the improvements in sleep following CBT-I.

I mean the more sleep I get, the better my day is. I might be in a lot of pain, but I can deal with pain. I can’t deal with no sleep. […] The pain is always going to be there. You know you can’t do anything about it. I mean if there’s a magic wand, yeah, take it, but I would rather have sleep over [no] pain (patient 1093).

Although patients did not report a strong connection between better sleep and reduced pain, they did perceive a connection between better sleep and improved quality of life, including increased energy and better social functioning.

It’s been quite a drastic change as far as my social interaction because I’m actually getting out and doing stuff. […] Now sleep is under better control and I feel more in control so I’m adventuring more. I’m doing more. […] I’m still in pain but my enjoyment of life is better. There were weeks where I wouldn’t go and visit a friend and now I see them just about every weekend. So it’s been quite a benefit to my life (patient 1104).

Several patients connected better sleep with improved mental health.

I would say [CBT-I] improved my quality of life tremendously because I’m sleeping which means I’m more aware, which means I’m able to do more things, which to me is huge. I’m not a person that sits around in my pink fuzzy slippers and eats bon-bons. Even when I can, I don’t. Even when I’m sitting watching TV, I’m like, ‘Can I fold clothes? Can I do this? Can I do that?’ So for me, not to be able to do something is miserable. So being too tired to do anything was not good for me. That part of it was maybe a mental issue (patient 1111).

Most CBT-I providers were unsure if pain intensity changed after CBT-I for their patients because they did not assess pain routinely during and after the treatment. All CBT-I providers described improvements in sleep for patients with chronic pain and noted that better sleep led to improved energy and overall quality of life. This provider noted the challenges inherent in working with patients with pain, whose chronic pain conditions may not resolve after CBT-I:

I would say that [CBT-I] is less effective for sleep [for people with chronic pain] than it is for people without chronic pain in my experience, but I think that it is much more effective than nothing. […] It helps them. It might not be as satisfying sometimes as a provider because you’re not able to touch the other thing which seems big, which is the chronic pain or the distress about the pain, but I think it’s still helpful. […] I can think of a few case examples where that’s been an important part for them is that they’ve been able to do more, and the more we do the more it helps chronic pain (provider 6).

The CBT-I provider summed it up by saying:

You know when I’m working with folks with chronic pain, it’s not that the pain is going to go away, it’s that they’ll be able to take it more places (provider 6).

In this provider’s opinion, the ability to get restorative sleep after CBT-I resulted in greater engagement in life, even in the face of pain.

Beliefs about the Value of CBT-I

CBT-I providers and patients believed that CBT-I was valuable for patients with chronic pain. CBT-I providers spoke with colleagues and encouraged referrals to this treatment for patients with chronic pain. Patients enthusiastically recommended it for other Veterans with comorbid insomnia and chronic pain. Several patients were so pleased with their treatment outcomes that they told friends and relatives with insomnia to start CBT-I. For example, this patient was enthusiastic about CBT-I:

I’m definitely promoting -- I mean I’ve talked with even a couple of people who have -- my friends who have similar issues as mine that, you know there is something at the VA that can maybe assist you with this that I felt I learned a lot from. You know, it was helpful. Because I know so many that do suffer from a lot of sleep issues (patient 1054).

Many patients commented on the value of obtaining individualized sleep treatment from a supportive therapist who provided positive feedback and held them accountable to treatment recommendations.

Interviewer: So, overall, what is it you like most about CBT-I, the sleep therapy?

Patient 1051: Well, [the CBT-I therapist], he was, you know, extremely helpful and, you know, very professional and always tried to help me and put me in the right perspective.

Interviewer: Yeah -- so you guys had a good relationship? A good rapport?

Patient 1051: Yes. […] He was giving me some tough [recommendations], but, you know, we’d talk about it every time and, you know, he’d say, well, you’re doing better now.

Another patient summed up the value of obtaining tools to get better sleep in the context of a supportive environment.

The outcome is that I now get sleep, that would be the A#1 thing. I walked into it, tried to be very adult about it and told myself this [therapist] is trying to help me, don’t fight her and she had a lot of patience and so I was the winner (patient 1073).

Discussion

Treatment guidelines recommend CBT-I as the first-line treatment for insomnia, particularly among populations at increased risk of harm from sedative-hypnotics, including patients with chronic pain;(12) however, we know little about how patients and CBT-I providers feel about CBT-I in the context of chronic pain. Given that pain-related coping behaviors, cognitions, and emotional distress may present challenges to CBT-I adherence,(21–24) our goal was to explore the experience of CBT-I among patients with chronic pain and their CBT-I providers to identify barriers and facilitators to treatment adherence, perceived treatment outcomes, and overall perceived value of CBT-I. These qualitative insights can be used to inform efforts to increase the uptake of evidence-based sleep treatments for patients with chronic pain, including how to communicate the value of CBT-I to patients and their referring providers and how to modify CBT-I to increase treatment effectiveness for patients with chronic pain.

Regarding our first research question about which aspects of CBT-I were found to be difficult among patients with chronic pain, we found that patients experienced CBT-I, especially sleep-restriction, as challenging but ultimately beneficial for sleep. The effectiveness reported by patients in this study replicates previous findings demonstrating that if patients are able to follow recommended schedules, they notice rapid improvement in sleep quality and in some cases, quantity.(39, 40) Social support emerged as a key facilitator for adherence to sleep restriction, including encouragement from the CBT-I therapist and involvement of family and friends in helping patients maintain the new sleep schedule. Efforts to integrate social support strategies (i.e., involvement of family, partners, friends) into CBT-I is a promising area for future research. Family involvement strategies are increasingly used in evidence-based psychotherapies to improve adherence and outcomes,(41, 42) but this approach has not been systematically developed and tested for CBT-I.

Unexpectedly, neither patients nor CBT-I providers identified pain as a substantial barrier to treatment attendance, adherence or effectiveness. In general, patients and providers did not express a need to modify standard CBT-I for patients with chronic pain, although some providers did integrate chronic pain treatment strategies into CBT-I (e.g., activity pacing, pain-specific relaxation exercises). Providers noted that pain-specific treatments would be helpful in addition to CBT-I, either as separate or integrated treatment. This aligns with ongoing work to develop hybrid CBT-I and cognitive behavioral therapy for pain (CBT-P) treatments, although more work is needed to establish the effectiveness of these hybrid models on sleep and pain outcomes and to ensure that the effects of CBT-I are not diluted by overly complicated and intensive treatment protocols.(18)

Regarding our second research question regarding changes that patients experience in sleep, pain and overall functioning following CBT-I, we found that patients reported improvement in sleep and a range of functional outcomes (e.g., energy, mood, socialization), but did not report substantial changes in pain intensity and frequency. It has been suggested that sleep disruption is an important contributor to chronic pain(43–45) and that CBT-I improves pain outcomes.(18) In theory, CBT-I targets common mechanisms that worsen both insomnia and pain, including maladaptive coping behaviors such as excessive napping and activity avoidance. Re-regulation of these processes during CBT-I may improve both sleep and pain.(18) Clinical trials have shown modest improvement in pain following CBT-I, with benefits tending to accrue over time.(18) The timeline of the current study (3 months posttreatment to facilitate recall of treatment experiences) may not have allowed enough time for full treatment benefits to emerge, especially in the context of pain improvement. As one CBT-I provider in this study noted, CBT-I may indirectly lead to pain improvements over time since better sleep leads to more activity: “they do more and when they do more it helps with chronic pain (provider 7).” Moreover, it is important to note that patients can experience sustained sleep improvement in the context of ongoing pain. Patients reported a sense of agency over their sleep following CBT-I combined with an acceptance of continuing pain.

Although most patients did not report a lessening of pain severity, patients did report that better sleep enhanced functionality in the context of chronic pain. They described being more active in their day-to-day lives due to more energy and less worry about sleep. This aligns with quantitative work suggesting that CBT-I provides widespread benefits beyond sleep, including improvements in mood and energy.(10, 17, 46) Moreover, it supports previous findings that behavioral sleep interventions lead to more robust improvements in pain functioning than pain severity.(47, 48) Although patients primarily viewed CBT-I as a sleep treatment, they readily acknowledged that treatment benefits extended beyond sleep into other valued domains (e.g., relationships, work, self-care). Efforts to increase the uptake of CBT-I would benefit from educating both referring providers and patients about the high probability of sleep improvement in the face of ongoing pain and emphasizing the extended benefit of this treatment for all patients with insomnia.(34)

Finally, regarding our third research question about which aspects of CBT-I were appreciated among patients with chronic pain, we found that patients and CBT-I providers highly value CBT-I as a treatment option for patients with chronic pain. Participants noted the importance of individualized treatment in which a provider works with patients to provide encouragement, holds them accountable for behavior change, and problem-solves with the patient throughout the course of therapy. Efforts to maximize convenience and increase access to CBT-I has resulted in increased focus on self-management approaches, including books, web-based programs, and electronic applications for smartphones and tablets;(49, 50) incorporating the patient- and provider-valued components of standard CBT-I, including real-time support and feedback, into these alternative delivery approaches would enhance the acceptance and sustainability of these efforts.

This study is limited in that all participants were VA participants in a single metropolitan area, so results may not be applicable to other patients with chronic pain, particularly among women and non-white participants. In addition, patients were interviewed within 3 months of completing CBT-I treatment, so long-term treatment effects were unexplored.

Conclusion

This work complements and extends our understanding of patient experiences during CBT-I among patients with chronic pain, for whom sleep problems are pervasive and long-standing. Continuing efforts are needed to enhance CBT-I adherence for all patients with insomnia, with a focus on challenging treatment components like sleep restriction. Systematically integrating and testing strategies that increase social support during CBT-I is a promising area for future intervention development. Finally, this work highlights the value of CBT-I for patients with chronic pain, both in terms of improved sleep and enhanced quality of life. It is our hope that these patient and provider insights help support efforts to increase the use of nonpharmacological sleep treatments like CBT-I for patients with chronic pain so they can live life more fully and “adventure” more.