Digital, Internet and App-Delivered CBT-I: Effects, Dropout and Cost

The sources on this entry are meta-analyses and reviews of cognitive behavioural therapy for insomnia delivered over the internet, by smartphone app or in fully automated form, with insomnia severity, sleep-diary measures, dropout and cost as the outcomes. Soh 2020 (meta-analysis, 33 studies, 4719 digital and 4645 control participants) reports that digital CBT-I reduced the Insomnia Severity Index by 5.00 points against control after treatment, 3.99 at short-term follow-up and 3.48 at one year, and that face-to-face CBT-I produced greater improvement than the digital form (3.07 points). Zachariae 2016 (meta-analysis, 11 randomised trials, 1460 participants) reports improvements in insomnia severity, sleep efficiency, sleep quality and other sleep parameters with effect sizes of 0.21 to 1.09, comparable to face-to-face CBT-I, with longer duration and higher clinical support associated with larger effects. Seyffert 2016 (meta-analysis, 15 trials) reports that internet-delivered CBT-I improved sleep efficiency by 7.2% and decreased the Insomnia Severity Index by 4.3 points against control. Gao 2022 (network meta-analysis, 61 trials, 11,571 participants) reports that individual, group, guided self-help, digital-assisted and unguided self-help formats all significantly improved sleep efficiency, total sleep time and insomnia severity against treatment as usual and waitlist, with individual, group and digital-assisted formats appearing to be the more effective options. Huang 2025 (meta-analysis, 16 studies, 2146 participants) reports mobile-health CBT-I effective for adults with insomnia, with longer, more complete, group-format interventions associated with better outcomes. On fully automated delivery, Nazari 2025 (meta-analysis, 49 randomised trials, 20,118 participants) reports a significant reduction in insomnia severity against control (WMD -3.42) with high heterogeneity, and Hwang 2025 (meta-analysis, 29 randomised trials, 9475 participants) reports moderate to large effects on insomnia severity against control but less effect than therapist-assisted CBT-I, concluding that a hybrid model combining therapist support is more beneficial. Linardon 2024 (meta-analysis, 19 randomised trials) reports significant effects favouring stand-alone apps for self-reported insomnia (g = 0.60, number needed to treat 4.8) and sleep disturbances, with dropout higher for apps (13.1%, OR 1.78). On cost, Le 2025 (systematic review, 28 economic evaluations) reports that digital CBT-I generated healthcare and societal cost savings against face-to-face CBT-I or pharmacotherapy, and Natsky 2020 (systematic review, 7 studies) reports that CBT-I appears cost-effective against pharmacotherapy or no treatment, with the limited number of studies warranting caution. In young people, Tsai 2022 (meta-analysis, 4 randomised trials) reports that digital CBT-I significantly improved self-reported sleep quality with a medium-to-large effect but had limited effect on objective sleep quality, and calls its findings preliminary; Lu 2024 (meta-analysis, 13 studies, 5251 participants) reports that digital sleep interventions significantly improved sleep quality in college students and young adults; and Werner-Seidler 2018 (systematic review, 3 studies) reports improvements in sleep efficiency, sleep quality, sleep-onset latency and total sleep time and calls eCBT-I a promising intervention for young people, with more studies needed. Yu 2019 reports that commercial apps were moderately adherent to CBT-I principles. CBT-I is a treatment for insomnia; this entry describes what the research found about a delivery format and does not diagnose insomnia or prescribe treatment.