The Promise of Digital CBT-I

Empirical support for the effective treatment of insomnia with cognitive behavioral therapy techniques (CBT-I) has grown steadily since the 1960s,¹ and CBT-I is now recognized as a first-line treatment option in insomnia therapeutics.² However, its scientific and clinical acceptance has now presented a challenge: the widespread dissemination and implementation of this effective treatment.³ Of the potential solutions to this problem,^4–7 one of the most promising and least studied is the use of digital, internet-based CBT-I (dCBT-I).⁸ The promise of this alternative to standard face-to-face CBT-I is its broad accessibility, high scalability, and relative cost- and time-effectiveness. In order to fully understand the potential advantages and limitations of dCBT-I scalability, trials comparing dCBT-I to traditional approaches must be undertaken; however, there have been no studies to date comparing digital to individual face-to-face CBT-I.

In this issue of SLEEP, Lancee and colleagues⁹ report outcomes of a randomized controlled non-inferiority trial comparing an online treatment program (dCBT-I) to standard CBT-I (6 weekly, 45 min sessions) and a wait-list control. Participants were 90 (30 per group) individuals with DSM-5 insomnia recruited from a sleep clinic in Amsterdam, the Netherlands. Standard validated efficacy assessment techniques were used, including an online consensus sleep diary completed pretreatment and posttreatment, as well as at 3- and 6-month follow-ups.

Although differences between treatments emerged, dCBT-I and face-to-face CBT-I both significantly improved many aspects of sleep relative to wait list controls, including insomnia severity index scores, sleep efficiency, and total wake time. Where clinical differences between active treatments were seen, face-to-face individual treatment was found to be superior to dCBT-I, a result that is inconsistent with previous findings comparing dCBT-I to group treatment where dCBT-I was found to be non-inferior.¹⁰ The long-term durability of CBT-I is a critical strength, and was evident in the results of Lancee et al. Indeed, posttreatment improvement in insomnia severity was not diminished at any subsequent evaluation. However, compared to dCBT-I, face-to-face treatment produced significantly greater improvement in both insomnia severity and remission rates at 3 and 6 months (Table 4 in Lancee et al.⁹).

Though the advantages of dCBT-I are promising, there are obstacles to be addressed before effective dissemination can be implemented. For example, despite typically large effect sizes for the sleep components listed above, few CBT-I trials, including Lancee have demonstrated significant improvement in total sleep time (TST) following treatment. Considering that the recent consensus regarding 7–8 hours as adequate sleep duration,¹¹ the posttreatment means of 6.4 and 6.6 h for digital and face-to-face CBT-I highlight TST as a critical area for improvement in both approaches. Limited generalizability to the larger population of insomnia patients is another issue. As the study was conducted in a sample of patients from a sleep clinic, it is unclear how results would apply to insomnia patients who have limited access to healthcare beyond primary care, and thus the target population for non face-to-face approaches. This calls for additional research that directly targets the larger population of insomnia patients.

A growing consideration in the dissemination of CBT-I is cost.¹² While the authors presume that the dCBT-I approach is more cost-effective, the opposite may currently be true given that few, if any, insurance carriers provide dCBT-I as a benefit, or allow reimbursement for it. Importantly, the results found by Lancee may serve as an impetus to examine the effectiveness of implementing internet-based approaches in the context of today's payer driven health-care system. The financial viability of dCBT-I rests in part on mainstream acceptance of this treatment modality by insurance companies, employer wellness programs, and/ or patients themselves.

Another factor in the widespread acceptance of any CBT-I modality by healthcare providers and the community at large is patient preference. In the study by Lancee, patients indicated a significantly higher preference for traditional CBT-I (77.8%) compared to dCBT-I (52.2%). Similarly, patient adherence to dCBT-I may also be a challenge, as indicated by the 50% completion rate for dCBT-I found in this study. A stepped-care model using both digital and face-to-face CBT-I has been suggested,¹³ with access to face-to-face treatment available for complex cases and dCBT-I treatment failures. Future studies could help elucidate underlying moderators of this effect to expand the utilization of dCBT-I.

An important missed opportunity in the study by Lancee was the specific assessment of potential adverse aspects associated with CBT-I, such as excessive sleepiness and performance impairments.¹⁴ Such effects are important in the context of any CBT-I treatment modality, but are critical for therapies with sleep restriction, particularly in the context of minimal ongoing supervision by trained sleep specialists.

Behavioral sleep medicine specialists are almost universally invested in increased patient access to CBT-I, with efforts aimed at brief treatments,^5,15,16 dissemination of training to non-specialists,⁷ and several other non face-to-face approaches (e.g., phone, video, telehealth).¹⁷ Another important way of emphasizing the value of CBT-I is demonstrating its potential for not only improving sleep health but also for reducing the impact and/ or risk of the mental and physical disorders with recognized links to insomnia.¹⁸ Lancee et al. show that both dCBT-I and face-to-face treatment resulted in improved anxiety and depressive symptoms, which is consistent with evidence from other dCBT-I^19,20 and face-to face trials.²¹ This positive impact on mental health functioning, if replicated in future studies, will be a tremendous benefit to scalable dCBT-I. If similar effects on functioning are found in larger and longer-term randomized controlled trials of dCBT-I, it will likely facilitate integration of this insomnia treatment modality into standard clinical practice.

The implementation of all types of behavioral sleep medicine remains a great challenge for the field, but the study by Lancee may help to push the boundaries of dCBT-I. If integrated with improved access and appropriate triage to behavioral sleep specialists, dCBT-I shows promise as an efficacious and cost-effective treatment for insomnia.

As a field, we must now work to replicate potential differences between digital and in-person delivery of insomnia treatment, to identify mediators of these differences, and, finally, to determine ways to leverage the advantages of each approach while reducing their respective limitations. While future replications are needed, this study is an important contribution to the growing scientific literature on alternative CBT-I approaches (telephone, telehealth, email and video, etc.). Efforts can also be directed to large-scale population-based studies that attempt to prevent and reduce the mental and physical health-related morbidity of insomnia through scalable implementation of behavioral interventions. Research determining if specific patient populations benefit differentially from various CBT-I interventions is also an important goal, and one that may help evolve both digital and face-to-face approaches.

The study by Lancee et al. is a much-needed step in that direction and the authors should be commended for an excellent study which reveals important strengths and limitations to each of these effective approaches to CBT-I treatment. The burden is now on dCBT-I providers to demonstrate efficacy in large effectiveness trials, before widespread implementation can be realized.²²

CITATION

Drake CL. The promise of digital CBT-I. SLEEP 2016;39(1):13–14.

DISCLOSURE STATEMENT

Dr. Drake has indicated no financial conflicts of interest.