Journal of Sleep Disorders: Treatment and CareISSN: 2325-9639

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Research Article, J Sleep Disor Treat Care Vol: 5 Issue: 4

Implementing Cognitive Behavioral Therapy for Insomnia- The Patient Perspective

Megan R Crawford1,2*, Simon D Kyle3, Delwyn J Bartlett2,4, Ron R Grunstein2,4 and Colin A Espie2,3
1Department of Behavioral Sciences, Rush University Medical Center, Chicago, IL, USA
2NHMRC Centre for the Integrated Research and Understanding of Sleep (CIRUS) Sydney Medical School, University of Sydney, Sydney, Australia
3Nuffield Department of Clinical Neurosciences/Sleep & Circadian Neuroscience Institute, University of Oxford, UK
4Woolcock Institute of Medical Research, Sydney, Australia
Corresponding author : Megan R. Crawford
Department of Behavioral Sciences, Rush University Medical Center, 710 S Paulina, Chicago, Il, 60612
Tel: 312-942-0549
E-mail: [email protected]
Received: December 02, 2015 Accepted: March 15, 2016 Published: March 20, 2016
Citation: Crawford MR, Kyle SD, Bartlett DJ, Grunstein RR, Espie CA (2016) Implementing Cognitive Behavioral Therapy for Insomnia- The Patient Perspective. J Sleep Disor: Treat Care 5:4. doi: 10.4172/2325-9639.1000180


Objective: Despite strong evidence for the efficacy of cognitive behavioral therapy for insomnia (CBT-I), patient experience of therapy implementation remains unexplored. This study sought to deliver a comprehensive account from the patient perspective and provide a unique insight into treatment adherence.

Methods: Semi-structured interviews were conducted with individuals with insomnia (n = 11, female = 8), who had completed a cognitive behavioral therapy program.

Results: Using thematic analysis, three themes emerged depicting the experience of implementing CBT-I components: ‘making sense of CBT-I’, ‘ongoing evaluation of components’ and ‘obstacles to implementation’. Each theme was associated with three subthemes that provided a more nuanced and interpretative account of how individuals implement cognitive behavioral strategies.

Conclusions: Implementing cognitive behavioral strategies is far from simple and adherence to these components is a dynamic process. Future research avenues are discussed that might further advance our understanding of adherence to this treatment.

Keywords: Cognitive behavioral therapy; Insomnia; Sleep


Cognitive behavioral therapy; Insomnia; Sleep


In recent years the adherence literature has emphasized a transition toward a more patient-focused stance on adherence, emphasizing that patients actively choose whether or not to implement treatments, rather than simply following the practitioner’s advice [1]. Our field could benefit from adopting this transition to more patient-focused research; letting patients describe first hand, what it is like implementing these treatment components. Qualitative studies asking patients about their treatment experience have revealed that adherence to treatments for chronic conditions can be challenging and complex [2]; however, these barriers to implementation are often outweighed by perceived treatment benefits [2-5]. These studies have identified positive patient-practitioner interactions [3,6,7], social support [3-5,7-9], illness- or treatment-specific factors, including dosing timings, side effects and social stigma [7,10,11] as influential in shaping adherence. Studies using qualitative methodologies in areas such as exercise, chronic heart failure, antiretroviral therapy and general medication taking have contributed to the understanding of patient behavior, and have been systematically synthesized, with the goal of translating findings into future research and clinical practice [9,12-14].
Cognitive behavioral therapy for insomnia (CBT-I) is a multicomponent therapy that combines single or multiple behavioral and cognitive strategies to target dysfunctional thoughts and behaviors that contribute to the maintenance of insomnia. CBT-I typically includes sleep hygiene/education, relaxation strategies, stimulus control, sleep restriction, and cognitive restructuring components. These strategies are associated with significant behavioral change and this is often dependent on a substantial cognitive shift on the patient’s behalf. For example, during stimulus control, individuals are asked to remove all waking activities (e.g., watching TV, reading) from the bedroom, and to limit the time in bed awake by removing oneself from the bedroom if awake for more than 15-20 minutes. With the efficacy of CBT-I established, the focus has shifted to treatment implementation, with patient adherence being one of those areas. Adherence to CBT-I is multifaceted, which makes it challenging to present a consensus statement on the extent of non-adherence [15]. Most studies have reported on adherence to the behavioral components derived from sleep diaries [16-21], actigraphy [22], or questionnaires [21,23,24]. Predominantly associated with lower adherence are psychological factors such as lower self-efficacy [16,25], depressive or dysthymic symptoms [21,24,26], decreased readiness or motivation for change [17,25] and better sleep quality and increased sleepiness [27]. Less conclusive have been factors such as age, social support, medication use and fewer perceived barriers to treatment [25,28].
To the authors’ knowledge only three studies have used qualitative, exploratory frameworks to probe the unique experiences related to implementing CBT-I strategies [29-31]. Kyle, Morgan, Spiegelhalder, and Espie [31] conducted a systematic investigation of patient’s experiences of implementing a sleep restriction program, one of the behavioral components of CBT-I. During sleep restriction therapy (SRT), individuals are instructed to reduce their time in bed and avoid daytime napping, which will increase their homeostatic drive for sleep and lead to improved sleep at night. Sleep windows are then gradually increased based on the extent of improvement in nocturnal sleep. Kyle et al. were able to map the patient’s experience of implementing SRT strategies using validated self-report measures, qualitative audio diaries (recorded twice a day) and semi-structured interviews. Adjustment issues and difficulties implementing the program were frequently recorded. Particularly problematic was the impact of sleep restriction on daytime functioning and dealing with an increase in daytime sleepiness. Individuals were also concerned with boredom or loneliness induced by the extended wake period. Certain coping strategies were employed to counteract these effects (e.g., engage in wakefulness promoting activities). Relapsing back to old behaviors that were aimed at compensating for lost sleep (e.g., napping) led to increased sleep onset latency, and patients who witnessed the negative effect of non-adherence felt encouraged to re-engage in behavior change. This study highlighted the complexity and effort associated with the implementation of sleep restriction. Although this study was useful in contributing to the understanding of the potential mechanistic principles, SRT is often delivered as one component in a CBT-I package, and administering it alongside other cognitive strategies and behavioral components, such as sleep hygiene, relaxation therapy, stimulus control might significantly alter the experience of treatment. Adherence might be facilitated by the ability to ‘pick and choose’ components, or the combination of all components may exacerbate the negative effects of individual behavioral components.
Currently, only two studies have investigated the patient experience of completing multi-component CBT-I from a qualitative perspective [29,30]. After completing a 7-week CBT-I program, 43 participants with chronic insomnia completed open-response feedback forms with questions regarding their treatment experience [30]. Written responses were subsequently coded into themes; half of the participants reported the completion of CBT-I in a group format and sharing the insomnia experience with others was most helpful. Unfortunately, this study fell short of providing an in-depth rich, account of patient experience. Ellis, Deary and Troxel [29] interviewed 21 individuals who had completed CBT-I and explored the influence of the partner on implementation of CBT-I principles. Using quantitative content analysis, positive statements that related to partner’s encouragement to implement techniques, and negative statements that reflected the partners’ discouragement were identified and counted. An increased number of positive statements about the partner’s involvement were related to a greater reduction in insomnia severity post treatment. The authors hypothesized that greater partner alliance facilitated the implementation of challenging techniques, which then led to a superior treatment effect.
We know from the above-mentioned qualitative studies that CBT-I components are not easily implemented, and which strategies can be utilized to facilitate adherence. It is important to not only validate these previous studies, but also to expand on them with an indepth qualitative analysis of multi-component CBT-I. Furthermore, quotes from real patients might be used as a clinical tool, providing prospective patients with a glimpse of what others have experienced. The aim of this study is thus, to adopt an in-depth qualitative approach to understand the patient experience of CBT-I, with a particular focus on adherence.


Participants were recruited at two separate sites: University of Glasgow Sleep Centre (UGSC), Glasgow, UK and the Woolcock Institute of Medical Research, Sydney, Australia. Individuals were identified if they had completed a research study, which included manualized CBT-I, and had consented to being contacted for future research studies. The same CBT-I manual was used at each site; having been adapted from protocols published elsewhere [32-34]. The CBT-I encompassed 5 group sessions covering sleep education, behavioral and cognitive-restructuring strategies and was delivered by a graduate student trained in CBT-I and supervised by a clinical psychologist. This recruitment method was employed to reduce the variability in treatment delivery, and thus this study included only those who underwent manualized CBT-I.
Once a person was identified, they were contacted to complete a screening interview to determine their eligibility. Individuals had to be fluent English speakers, aged between 18-65 years, and had to have completed CBT-I within the past 24 months. Eligible participants also had to have a diagnosis, prior to CBT-I initiation, of primary insomnia according to the research diagnostic criteria35. Those with a diagnosis of co-morbid psychiatric or unstable medical disorders were excluded unless there was a temporal and causal separation from their primary sleep concern [35]. Suspected presence of a co-morbid sleep disorder (e.g., sleep apnea, restless legs syndrome, parasomnias) was a further exclusion criterion. Individuals who were using other forms of treatment for their insomnia (e.g., pharmacotherapy) were not excluded. Twelve eligible individuals were identified and approached. One potential participant was unable to attend the interview because of travel distance to the research center, resulting in a final sample size of 11.
Ethical approval was obtained from the NHS Greater Glasgow and Clyde and Sydney Local Health District Ethics Review (RPAH Zone) committees to conduct this study at two separate sites (University of Glasgow Sleep Centre, Glasgow (Ref: 09/S0701/12), UK and Woolcock Institute of Medical Research, Sydney, Australia, (X09-0304 & HREC/09/RPAH/510).
Once participants were contacted and written consent was obtained, individuals were invited to the research centers for a faceto- face interview. The lead researcher (MRC) conducted all interviews during 2009-2010, and questions and prompts were pre-determined. This enabled the interviewer to focus on the responses, permitting each individual to speak freely and with as little guidance from the interviewer as possible. These types of semi-structured formats enable the exploration of the individual’s experience, thoughts and beliefs in rich detail [36]. Table 1 highlights the topics covered in the interviews. Duration of the interviews was between 40-90 minutes (Table 1).
Table 1: Four topics covered as part of interview, and sample questions
Data Preparation and Analysis
The lead researcher transcribed recorded interviews verbatim, and the data were analyzed using thematic analysis [37]. This flexible technique, unrestricted by pre-existing theoretical frameworks has been used extensively, often without explicit reference to its use [37]. An inductive, descriptive approach to analysis was chosen to allow for preliminary exploration of the poorly understood topic of CBT-I adherence. In line with Braun and Clarke’s guidelines for thematic analysis, each interview was read and re-read, familiarizing oneself with the data set. Subsequently, passages were coded at the basic, descriptive level, and these codes were then collated into themes. The clustering of codes and identification of themes was refined in an iterative process; data extracts were re-read to ensure the themes were internally consistent and captured the verbatim evidence accurately. Generated themes were then compared within and across transcripts and were subsequently organized into a meaningful and credible account of the phenomenon using sufficient verbatim evidence and telling the story in the words of the individual. The “keyness” of data required to be counted as a theme was made according to researcher judgment and was not defined by quantifiable measures. Themes were identified and refined through discussions with secondary reviewers (CAE/SDK/ DB) throughout different stages of the analysis.
Participant’s names and additional identifying information were changed to ensure anonymity. Small changes were made in the final report to facilitate legibility of the verbal extracts. Brief vocal utterances (e.g., ‘ehm’ ‘oh’) were omitted and semantic repetitions (e.g., my partner, you know my partner) or brief tangents from the topic, were also transcribed, but replaced in with ‘[…]’ in the final write up. Short pauses in the interview were transcribed as ‘…’ and incorrect grammar indicated with [sic]. Quotes are presented within the text forming a narrative account of the individuals’ experience of adhering to CBT-I.


Patient characteristics
The group of 11 participants was relatively homogeneous. The majority was female (73%), and well educated. Participants had completed their CBT-I program between 1-4 months prior to the interview. For the ten participants who provided this information, the mean years of formal education was 16 (SD=3 yrs.). All individuals gave details about their ethnicity; the largest proportion described themselves as White (British/Scottish, Australian or Canadian, n=10, 91%) and followed by Asian (n=1, 9%). Individuals reported having problems sleeping for on average 10 years (SD=8). Half were married or had a live-in Partner (n=6), whilst others were single, divorced or separated (n=5). Table 2 presents the characteristics of each participant (Table 2).
Table 2: Participant characteristics including alias, gender (female/male), age, education, marital status, duration of insomnia and CBT-I group (1-4). Participants with the same number completed CBT-I in the same group.
Qualitative Results
Three themes emerged from the thematic analysis of the interviews: “Making sense of CBT-I”, “Ongoing evaluation of components” and “Obstacles to implementation”. In the following, each theme is presented with the respective subthemes and patient quotes embedded in the text to feature the experience of patient adherence to CBT-I.
Theme 1: Making sense of CBT-I. Individuals described their initial impressions of treatment. Very few had concrete conceptualization of what CBT-I was, and individuals were willing to “give it a go” because CBT-I presented itself as the last resort. CBT-I was also conceptualized in reference to the meaning of their insomnia.
What is CBT-I really? At initiation of CBT-I, it was uncommon for participants to have a tangible understanding of treatment. Very few had further investigated this type of treatment and were unsure of the actual content. Apart from identifying CBT-I as “talking therapy”, individuals described entering treatment with a “blank canvas”. Interestingly, what motivated individuals to participate in this treatment was the hope for improved sleep. The simple distinction from the familiar pharmacological options seemed to be an additional influence for taking part in something individuals had little knowledge about:
I didn’t really know what to expect I think it was just going in there almost, with a sort of blank canvas, open minded, willing to try and take on… I wouldn’t have been willing to take on any medication as such, […] I just don’t like taking medication basically.’ (Ben, l. 205- 210)
CBT-I as the last resort. In the interviews it emerged that, for the majority, CBT-I was their last hope for improvement; having tried various other treatment options (pharmacotherapy, hypnotherapy and herbal remedies) individuals were willing to engage in this treatment that was, as Karen describes, the “last port of call”. Individuals described the emotional impact of sleeplessness, feeling “depressed” and “fed up”, out of options, which resulted in a sense of emotional “relief”, when the treatment was suggested:
‘When I got the chance of this, I thought well, I can’t go on like this. Not sleeping or sleeping for a week, and I find sometimes [sighs], it really gets you down, you feel quite tired and depressed; because you are not sleeping.’ (Belinda, l. 54-58)
CBT-I was seen as the solution, both in terms the pathway to improved sleep, and in terms of providing an acceptable alternative to pharmacotherapy —a route they were less comfortable in pursuing. The combination of feeling helpless, and CBT-I taking shape as an option that would not produce any side effects, without risk of dependency, proved motivation for engagement in this therapy:
‘You know that everything was worth a try, I mean obviously I hadn’t gone the sort of medication route, […] which is something I could have discussed with my GP, if I had wanted to. So on some level, I have always avoided that, so […] doing therapy for it, seemed like a good idea.’ (Abigail, l. 46-49)
The meaning of insomnia. Making sense of CBT-I occurred in light of how these individuals conceptualized their insomnia. Although many described the frustration, helplessness and hopelessness associated with having insomnia and were ready to embark on the changes outlined in the treatment, some participants described an initial disbelief in the potential effectiveness of CBT-I, based on how they viewed their insomnia. For some CBT-I was seen as too simple to deal with their complex, chronic problem.
Rachel: I didn’t think it would work.
Interviewer: Why was that?
Rachel: When you first called I thought ‘oh this might be a waste of time’ [laughs]
Interviewer: What was the reason for that?
Rachel: I don’t know, maybe because I have had this problem for so long, then I just thought it was the way I slept, or lack of sleep. I just thought it was me. And I guess you get to a point where you just accept what you have and just get on with it.
(Rachel, l. 130-140)
Others perceived CBT-I as ideal because it promised “long-term benefits”. Others again felt sleep should come naturally, and does not deserve treatment. Interestingly though, when the attempt was made to implement these techniques, individuals realized how difficult some of the components actually were. Despite the overall sense that insomnia is a debilitating disorder, for a small number of interviewees, insomnia was associated with secondary gain. Insomnia became part of one’s identity; sleeplessness during the night provided opportunities to catch up on tasks that were not accomplished during the day or was used as an excuse for irritability or low mood. This led to initial “trepidation” or resistance towards CBT-I.
It’s shedding those things that make you different, and having the badge of being an insomniac, it’s like “oh no, shedding that one too”.’ (Karen, l. 121-123)
Theme 2: Ongoing evaluation of components. It emerged in the interviews that individuals did not simply comply with what was outlined in the program, but experienced a process of evaluation of the components on which they based their subsequent adherence behavior. Each aspect of CBT-I was continuously evaluated in terms of efficacy: individuals used what worked; what worked for them; and what worked for others.
Evaluation against past, current, future self. Once initial impressions had been established and the treatment accepted, there was a more detailed evaluation of the individual treatment components in respect to the image they had of themselves. Treatment components were deemed valid when they were assimilated with individuals’ past, current or future self-image.
Newly imposed behaviors and sleep routines were compared to how they used to sleep, or seen as a mechanism to re-establish the old sleep patterns:
‘That’s what I used to do, so maybe, if I start trying that again, and it’s [sic] really reaffirms the “yes, I have been on the wrong track and there is a way back”.’ (Sally, l. 460-462)
Participants also continuously assessed whether or not there was an alignment with their current habits and preferences. Abigail, who described herself as not having a “whole lot of negative thoughts”, did not find using the cognitive strategies very helpful. Similarly, Mike who disliked “routine imposed upon [him]” sometimes struggled to stick with the sleep restriction program. For Belinda, Abigail and Madeline, there was congruence between consistent rise times and the perception of their “self” as someone who “doesn’t lie in bed any length of time” “ha[s] never been someone to snooze the alarm” or “would try anything” respectively. Others again identified their present circadian preference (“a night-person”) or insomnia subtype (“problem staying asleep”) when evaluating the appropriateness of treatment components such as sleep restriction and stimulus control.
In addition to their image of their past and current self, participants also described their future self and the goals they aimed to achieve in terms of their sleep. The prospect of reaching these goals provided added motivation to adhere to the treatment components. For some it was also the image of what they could become, if they failed to initiate change that provided a push to continue:
‘With older people, like my mum is in a nursing home at the moment, and when I get there, it doesn’t matter what time I go there, they are all asleep, a lot of them are napping, I thought “crikey, if I fall into that category now, what hope have I got when I am that age” I don’t want to get into that.’ (Polly, l. 397-400)
Using what works. Of components that were deemed appropriate for themselves, individuals used what proved effective in improving their sleep patterns on an ongoing trial and error basis. Some individuals identified the specific components that worked for them. Individuals described a cost-benefit evaluation when deciding to follow recommendation. In the majority of cases, like for Madeline, the effectiveness seemed to outweigh any difficulties individuals were having and provided a motivator for continued use.
‘I don’t want to get up and get out of bed, when you don’t have to and you know you are going to go back there shortly, so I would rather not, but I mean it does work, you get up and you go have a little wander and then you go back.’ (Madeline, l. 778-781)
Seeing the effects of non-adherence provided a form of validation for the efficacy of components and this increased their motivation to be more rigorous with the treatment routine.
‘I woke up at about half past 5, I could have got up then quite easily and went, “no I will just sneak some more sleep in” […], then I felt pretty, not as alert all day, just a bit sort of lethargic, and then that night, I just had so much trouble getting to sleep.’ (Karen, l. 511-515)
‘Saturday night, come bedtime I wasn’t tired, […] but that was my bedtime, so I kind of footed about and went to bed, which was a big mistake, because I didn’t sleep all night.’ (Belinda, l. 440-445)
Comparing self to others. Lastly, participants also evaluated components after comparing their behavior with others, either others in the treatment group, or reports of previous success stories. These comparisons also gave hope and reassurance that the improvements would be seen, if they continued to adhere to components. Whilst comparing themselves to others, individuals also recognized dysfunctional behavior or thought processes they were also trying to change. As both Sally and Karen describe, this process helped them internalize the underlying treatment rationale to provide validation of the treatment components:
‘It’s like me giving the advice that I don’t take myself really, […] as if I am trying to convince myself that these things will work and then by explaining to someone else, it helps me get it straight in my head really.’ (Sally, l. 216-220)
‘I do really like the group thing and hearing how other people are going, and hearing myself in other people, who aren’t going so well, and recognizing parts of me, […] “they are still hanging on to that bit, I am too”, […] “if it is not working for them, why do I think it is going to work for me?”.’ (Karen, l. 868-872)
Theme 3: Obstacles to implementation. Throughout the CBT-I program, individuals were faced with certain obstacles that made implementation of the treatment components challenging. Participants described obstacles such as trying to implement components at the wrong time, holding on to certain beliefs and the presence of unwanted consequences.
Is it the right time? Timing plays a large role in behavior change. Adherence might be difficult on specific days (e.g., weekends, vacation), or at a certain time in life (e.g., readiness for change):
‘I think I probably had the advantage, that because I am not currently working, it was probably easier for me to organize my time to fit in to the program.’ (Jonathan, l. 80-83).
Unfortunate timing might then result in practical or motivational obstacles for implementation. Karen describes the reason she found implementing sleep restriction on weekends difficult, was because it is ‘traditional’ to stay in bed longer. Other members of the household, who continued a normal routine, were especially impacted during these times. Some described the timing of CBT-I initiation in terms of their current lifestyle; having a relatively free schedule facilitated implementing components, whereas a hectic life-style produced obstacles for adherence. Individuals also acknowledged that importance of timing in reference to their internal motivation and whether they were ready for change, as Polly denoted, one of the biggest things was “that we are willing to change”(l. 936) Some individuals described at the start being “resistive” “having trepidation” or thinking the treatment would not work.
Beliefs about sleep. Beliefs about sleep requirements provided a further obstacle for implementation. One belief that prevented many from implementing the behavioral components was aspiration to obtain more sleep, or the magical ‘8 hours’.
‘I couldn’t sit up till 11.30 anyway and get up when I have to get up, I find that I just needed that sleep, […] when you say you only need 8 hours a night sleep, I must need 8 hours a night sleep.’ (Caroline, l. 423-426)
Others were less keen to implement stimulus control strategies that seemed to promote wakefulness and thus were counter-intuitive. In the following quote, it appears Belinda is retaining the belief that sleep needs to be controlled rather than just letting it happen.
‘If you don’t fall asleep you know sometimes, I find too, if I do get up and I sit, I become more wide awake and that kind of defeats ... you can be sitting there for a long time before you’re feeling tired.’ (Belinda, l. 253-255)
Interestingly, individuals explain how “getting their head” round the treatment rationale, facilitated the rebuttal of these dysfunctional beliefs.
"I really need to get my head round this as well, if I don’t feel tired when I should be going to bed, then I just don’t go” (Belinda, l. 443-445)
For Karen it was not solely about internalizing the rationale, but also internalizing sleep and accepting sleep as part of herself, which made the insomnia less of a foe:
‘Well, is it [sleep] something to be conquered? That is the thing, it’s just, I don’t conquer being awake […] but for me sleep actually has to be so normal, that is part of who I am and part of what I do, rather than this almost external entity that has to be gotten better of. […] So to normalize sleep, makes it less of a foe, less of something to be afraid of, yeah I think that works for me.’ (Karen, l. 384-393)
Negative/Unwanted Consequences. Individuals described difficulties implementing the behavioral components —especially sleep restriction and stimulus control— when these produced unwanted or even unexpected consequences. One of the most frequently noted adverse effects was an increase in sleepiness:
‘It was really slow [in the] mornings, then a period of feeling fine and then early evenings starting to get really tired and having to stick with it and stay up and stay out of bed and so it was challenging.’ (Mike, l. 154-157)
Despite their willingness to adhere, they found it difficult to ward off these feelings that emerged during treatment. The presence of unwanted practical consequences inhibited implementation of treatment components. Individuals felt that wakeful activities associated with sleep restriction and stimulus control impacted on daily routines and especially on others in the household. In order to overcome these side effects, participants described the need to develop certain coping strategies. For many participants implementing change was more difficult than expected and side effects like tiredness or boredom had to be counteracted by these strategies. For example staying active, changing habits like sitting on a firm chair rather than the couch, or finding activities that would encourage rising in the morning, facilitated adherence. Likewise, individuals had to develop strategies to prevent the more practical unwanted side effects that disrupted others in the household:
‘How to set the alarm and finding something that wouldn’t wake the whole house up —just an iPod under the pillow with a low volume— it woke me up, but not my partner’ (Karen, l. 258-260)


This study aimed to explore, through semi-structured interviews, the patient experience of implementing CBT-I. Three superordinate themes were extracted from the analysis: “making sense of CBT-I”, “ongoing evaluation of components” and “obstacles to implementation”. Figure 1 provides an overview of the themes and their respective subthemes. Making sense of CBT-I at the global level is core to the process of implementing the components, as indicated by the framed box in Figure 1. Establishing the meaning of this treatment is fundamental and with this in place, individuals begin to focus their concerns on the individual components. This incorporates both the evaluation of each component against several criteria and also how certain obstacles may hinder implementation. These themes, “ongoing evaluation” and “obstacles”, might inform each other (as indicated by the dashed arrow in Figure 1). For example, beliefs about sleep as an obstacle might cloud the patient’s evaluation of a component’s effectiveness. Likewise, something valued as effective might facilitate the continued implementation despite marked obstacles. These relationships were not explicitly examined in this study and remain to be investigated in further quantitative and qualitative research (Figure 1).
Figure 1: Themes and Subthemes and their Interrelations.
Our findings have potential implications for the field of sleep medicine and research, which are discussed in detail below.
Increasing awareness, availability and acceptability of CBT for insomnia
The data highlight the problem of availability and accessibility of CBT-I for individuals with insomnia. The importance of increasing availability of CBT-I has been raised elsewhere [38] and there is evidence to show that lack of awareness can act as a barrier to seeking treatment for insomnia [39,40]. Certainly recent studies have explored novel ways of improving accessibility, whether that is through online technology, [41] or increased training of health professionals able to administer CBT-I [32,42,43].
A benefit of increased awareness might be that individuals recognize CBT-I to be a challenging, complex and valid treatment that is not simply “talking” therapy and based on principles rooted in physiological and psychological models. Whether any stigma attached to psychotherapy plays a role in help seeking in this population remains to be established. Indeed, in our study some individuals had initial trepidation when presented with CBT-I. Treatment acceptability was evaluated against the patient’s perception of insomnia. There is a large body of evidence highlighting the relationship between illness perceptions and health behavior [44]. Whether these types of perceptions are related to CBT-I adherence is relatively under-investigated. Interestingly, there is some evidence that the perception of the illness may have an impact on either treatment seeking behavior [40,45], or outcome [27]. Patients’ treatment and illness perceptions prior to initiating behavior change should be considered at both the research and clinical level.
Encouraging individualized, patient-centered care
Indiscriminant compliance with treatment recommendations was not present in this sample, and it was evident that many of the participants continuously evaluated the treatment components on different levels. A recently published review of quantitative and qualitative studies on the patient perceptions of treatments for insomnia also reported that patient experiences might change through the input of new information throughout treatment [46]. This conveys not only how dynamic the process of adherence might be, but that for these participants treatment implementation was not merely about complying/not complying. They engaged in an active decision making process whether to adhere or not to adhere. In this sample, the decision to implement recommendations was highly individualized. If patients are choosing the components that overlap with the image of their self and their sleeping patterns [47], there might be value in investigating treatment-matching strategies [48-50]. It might be futile to search for the most acceptable and most often used components of CBT-I [51-53]. Rather, in the therapeutic encounter, patients need to explore and select components they believe will be most effective for them [54].
Increased focus on (non-) adherence to CBT-I
The data presented here has potential implications for how we treat non-adherence, both at a clinical and empirical level. It is not well documented how to address non-adherence in the clinical setting. In this sample, after identifying effective treatment components, individuals still struggled to implement them, hindered by certain obstacles. This chimes well with the mounting evidence that CBT-I is not a simple treatment, easily implemented without the possibility of adverse effects or consideration for the individual’s life circumstances, motivation and beliefs about sleep. The acute stages of a sleep restriction program have been linked to adverse effects, particularly daytime sleepiness and impaired attention [31,55,56]. It might be useful to discuss the obstacles in the clinical setting and if necessary introduce behavioral countermeasures as touched on here and described in other studies [31].
Non-adherence itself might be used as a clinical tool. It allows the individual to recall the experience of poor sleep and highlights how their behavior may perpetuate the insomnia problem. To prevent individuals from abandoning treatment, preparation for relapse needs to be an integral part of CBT-I [33]. A technique that could be integrated is motivational interviewing/ enhancement therapy (MI/ME), which acknowledges relapse as an important part of change. This approach has been proven useful for problem behaviors like substance abuse [57]. Moreover, some studies have provided preliminary evidence for the effectiveness of MI/ME in improving Continuous Positive Airway (CPAP) use [58-61].
Inability to implement treatment recommendations might also be a sign for the clinician that the patient is not ready for change. Currently, we know little about optimal timing for implementing CBT-I. Indeed, there is a presumption that all individuals are ready for change once treatment is sought. When faced with the behavioral components and the prospect of reducing time in bed, however, motivation might be altered for some individuals. Readiness for change has been extensively studied within the framework of the transtheoretical model. This theory suggests that behavior change can only occur when the individual is at a stage ready for change [62-64]. This theory has been helpful in understanding and predicting CPAP adherence, [65,66] and might be worthwhile pursing in the context of CBT-I adherence.
The treating clinician might also consider dysfunctional beliefs about sleep as another source of patient non-adherence and identify patients that have not been able to make a cognitive shift. Research has indicated that dysfunctional beliefs are subject to change across CBT-I; certain cognitions —such as needing 8 hours of sleep— were associated with outcome [67,68] and most recently adherence [69]. If, for example, the dysfunctional belief of needing 8 hours of sleep is still heavily endorsed, removing oneself from bed in accordance with sleep restriction and stimulus control will seem counter-intuitive and will be challenging [31]. A successful restructuring of individual’s thought processes prior to implementation of the behavioral components might be an interesting avenue to pursue [69]. Unfortunately, to date there has been no systematic investigation of the order of cognitive and behavioral elements of CBT-I. Most packages described in the literature will provide the behavioral prior to cognitive components. However, it might be useful, at least for some individuals, to offer the cognitive components first. Matching individuals to a specific treatment based on their readiness for change has been successful in other areas [70].
Finally, the data has highlighted some aspects of patient adherence that have not been scientifically evaluated. The first pertains to the impact of treatment delivery (group vs. individual). Participants in our study reported evaluating treatment components and their behavior against that of other’s in the group. Bandura’s social cognitive model states that people form knowledge about their environment from observing and modeling others’ behavior [71]. To our knowledge no study has investigated the impact of group vs. individual treatment delivery on adherence. Group administration of CBT-I is well established and leads to clinically meaningful improvements in sleep outcome variables [72-74]. Despite the consensus that group therapy might increase adherence to treatment components [72,74,75], there is only one qualitative study reporting how individuals valued meeting others like them during group CBT-I [30]. Kyle, Espie and Morgan [76] reported that individuals with untreated insomnia felt isolated and poorly understood. These findings suggest that groups may have important advantages for some patients. Similarly, the incorporation of previous treatment completers as a source of motivation for current patients has not yet been considered. One study reported that a CBT program, which included a short 15-minute video of current CPAP users displaying both treatment success and failure, was associated with an increase in average nightly machine use of 2.9 hours compared to treatment as usual [77].
The second area that has been neglected is the adherence pattern or daily variability. To date, self-report scales that have been used to measure adherence to the behavioral components have not included separate items for weekends and weekdays [21,28]. There is little consideration as to whether weekends or holidays are more difficult times to adhere to the sleep scheduling practices. The social norm of ‘sleeping in’ or additional social activities on weekends and holidays might make adhering to “out-of-bed times” difficult [31]. It seems prudent for the field to consider quantitative measurement of both weekday and weekend adherence.
Limitations and Future Directions
The aim of this study was to describe the experience of this particular sample and to generate hypotheses for further quantitative and qualitative testing. There is a need for further qualitative and quantitative studies to explore the themes highlighted here. The primary concern of qualitative studies is not to produce data that is generalizable to other populations, but rather provide in-depth insight into the phenomenon studied. A benefit of qualitative studies is to identify possible areas that can be further examined with other qualitative or quantitative studies.
It is important, to evaluate our study in light of several limitations. One limitation of this study is it sampled participants from both the UK and Australia. Samples in qualitative studies are generally comprised of a homogenous group. Recruiting participants from only one cultural background might have generated different qualitative data; on the other hand, themes emerged in this qualitative investigation, are reflective of experiences across both settings, relatively uninhibited by effects of different cultures and health systems. There were no apparent differences in themes emerging across these two sites.
One further limitation is the largely female, well-educated and young sample of individuals with primary insomnia, selected from previous research studies. Although this is quite representative of research samples within the insomnia literature, it will be imperative to investigate experiences of a real-world clinical sample that, in addition to insomnia, are likely to be dealing with a co-morbid sleep disorder like sleep apnea, or other physical (e.g., pain) or psychological conditions (e.g., depression). Furthermore, the experiences of a largely motivated, health literate research participant might differ from the journey of someone who is presenting to their practitioner in their first treatment-seeking attempt. Interestingly though even this motivated group of research participants, who largely responded to treatment, found CBT-I challenging.
This qualitative investigation relied largely on the recall of experiences in these individuals. All participants were interviewed between 1-4 months after completing the treatment protocol, thus recall effects might have biased retrospective accounts of the experience of implementing components. Additionally, the retrospective nature of the study resulted in the exclusion of treatment dropouts; this might have led to the omission of important experiences related to the early phases of CBT-I. Future studies might want to follow individuals through the entire CBT-I process to obtain a more detailed account of individuals’ experience in real time. Techniques such as audio diaries, with which individuals can record their experiences on a daily basis, have been successfully used in a previous study involving an insomnia population [31] and might provide a valuable resource for this endeavor.
For reasons of feasibility, the interviewer in this study was also the therapist or shadowed the therapist in all CBT-I groups. This is a potential bias; however this is not uncommon for qualitative treatment investigations. Refinement of the themes occurred through discussions amongst authors (MRC, SDK, CAE, DJB). Having multiple coders to reduce rater bias does not necessarily align with the philosophy of qualitative analysis. It is acknowledged in thematic analysis particularly, that coding is perceived as an active and iterative process, and the analysis is shaped by the coder [78,79].
Lastly, this study did not include polysomnography, structured interviews or medical exams to corroborate the individuals’ self-report of non-existing sleep, psychiatric and medical conditions.


This present study is a systematic exploration of the patient’s experience of implementing multicomponent cognitive behavioral therapy for insomnia and adds to our knowledge of the treatment of insomnia. The study has validated some of the themes emerging in previous qualitative studies: the implementation of some of the behavioral components is far from simple. However, this study has also provided some novel themes to be further explored, for example how adherence can be a dynamic process. In this study individuals continuously evaluated CBT-I and the components against certain criteria. This started as early as treatment initiation, and continued throughout treatment. Patients engaged in various thought processes in determining which treatment components to implement and did not simply “comply”. The study has also highlighted some novel issues about the therapeutic delivery and context that could be further investigated, for example group vs. individual therapy and the order in which components are provided. Further research is warranted to explore these avenues, and to further understand treatment experiences and factors that influence behavioral adherence to CBT-I.


Funding for this study was provided by a Medical Research Council UK postgraduate scholarship awarded to the University of Glasgow and the first author (MRC).

Conflict of Interest

Dr. Espie is Clinical & Scientific Director of Sleepio Ltd, and a shareholder in the company but receives no income from Sleepio. He has conducted paid consultancy for Boots UK and Novartis and paid speaking engagements for UCB and Boots UK. Dr. Kyle has previously consulted for Sleepio LTd. None of these activities have a direct bearing on this study. The remaining authors have no conflicts of interests to disclose.


Track Your Manuscript