Insomnia therapy through a screen works nearly as well as in person

Insomnia remains a highly prevalent public health problem worldwide, and psychological treatments cannot reach everyone who needs them when they depend on a therapist in a room. Delivering these treatments through websites and apps promises effective, accessible and affordable care on a population-wide scale. Individual trials have tested digital insomnia interventions for years, but the evidence had never been quantitatively integrated. That left two open questions: how well does eHealth delivery actually work, and does it sacrifice anything compared with sitting across from a therapist?

The researchers searched five databases from inception to February 16, 2021, updated the PubMed search to December 6, 2021, and screened grey literature for unpublished data. From 19,980 identified records, they included 37 randomised controlled trials covering 13,227 adults with insomnia. They pooled the effects of eHealth-based psychosocial interventions on the two primary outcomes, insomnia severity and sleep quality, against both inactive control conditions and in-person treatment. Subgroup analyses and metaregressions then probed which study characteristics and intervention features shaped the size of the effects.

Against inactive controls, digital interventions produced a large reduction in insomnia severity and a clear improvement in sleep quality, with no evidence of publication bias. Against in-person treatment, digital delivery showed no significant difference on insomnia severity, but face-to-face care held a significant advantage on sleep quality. Effects on severity were larger in clinical samples than subclinical ones, and interventions guided by trained therapists beat those with animated therapists or no guidance on both primary outcomes. People with higher baseline severity, and longer interventions, saw bigger reductions, and every secondary outcome, including sleep diary and mental health measures, also improved significantly.

This is the first quantitative integration of the evidence on digital psychosocial treatment for insomnia, and it lands firmly in favour: clearly effective against no treatment, and broadly comparable to in-person care on the core outcome of severity. The authors read their findings as support for wider dissemination and for stepped-care models, where digital programmes carry much of the load and blended care combines both formats. The remaining questions concern ingredients: which specific intervention components drive effectiveness, and how blended approaches might be tailored for people with low socioeconomic status, limited healthcare access or limited digital literacy.

If insomnia treatment has felt out of reach because of cost, waiting lists or geography, this is solid evidence that a well-built online programme is a real treatment rather than a consolation prize. Two details are worth carrying away: programmes with a trained human therapist behind them worked better than fully automated ones, and people with worse insomnia at the start improved more. In-person care kept an edge on sleep quality, so the comparison is not a complete wash, but digital delivery held its own on the outcome that defines the condition.