Users of online insomnia program SHUTi reported better sleep in 25 studies, in a review led by its developers

The treatment that guidelines put first for chronic insomnia is a form of talking therapy, and for most people with insomnia it is hard to get. One answer has been to turn the therapy into software, and a new review gathers the published research on one of the most-studied such programs, an online course called SHUTi.

Across 25 studies and more than 14,000 people, those given the program reported much less severe insomnia afterward, and they improved more than people given online information about sleep, a gap that held for a year. The review was written by researchers who include the program’s developers, every result rests on what participants reported about their own sleep, and about a third of people did not complete a full dose of the course.

Why an online insomnia program was built

The therapy in question is cognitive behavioral therapy for insomnia, or CBT-I, a structured course that changes the habits and thoughts that keep insomnia going. In the United States, the American College of Physicians (ACP) put it at the top of its 2016 guideline, advising CBT-I for all adult patients as the initial treatment for chronic insomnia disorder.

The obstacle is supply. A 2016 trial report pointed to a shortage of trained clinicians to deliver this much needed treatment, and an independent review the same year noted that in-person treatment is often unavailable.

SHUTi, short for Sleep Healthy Using the Internet, was one response, developed at the University of Virginia (UVA). “It’s our program that we began developing almost 20 years ago,” Lee Ritterband, a UVA professor and one of the program’s creators, said in a university interview last year. The therapy works, Ritterband said, but it is not widely accessible because few people are trained to provide it.

In 2016 the program ran for six weeks with no therapist involved, a fully automated, interactive, and tailored web-based program built on the main elements of face-to-face therapy. Ritterband lists its ingredients as sleep restriction, stimulus control and cognitive restructuring, along with sleep hygiene and relapse prevention. The first limits time in bed to roughly the hours a person actually sleeps, so that sleep becomes more solid; the second keeps the bed for sleep alone, so that bed and wakefulness stop going together; the third examines and changes unhelpful beliefs about sleep. Users keep a daily record of their sleep, and the program uses it to set a schedule for each person.

Who reviewed the insomnia program

The review, published on October 7 in the journal Sleep, has Kelly Shaffer as first author and Ritterband as last, with Frances Thorndike and Charles Morin, who also worked on the program’s main trials, among the other authors.

Those authors are not bystanders. When UVA announced the 2016 trial, its statement said the program was available through BeHealth Solutions, which licenses SHUTi and its software platform from the university, and that Ritterband, Thorndike, Gonder-Frederick and Morin held shares in it. Three of those four are authors of the new review, though that disclosure is ten years old and described the position in 2016.

The program has since moved further into commerce. “The software has been licensed and the commercial product, Somryst, has been authorized by the FDA,” Ritterband told the university’s interviewer, referring to the US Food and Drug Administration.

That history does not make the pooled numbers wrong, but it does make the review the developers’ account of their own program’s record.

How the review of the insomnia program worked

The authors, who describe SHUTi as among the most-studied digital versions of the therapy, set out to summarize two decades of work on it. They searched four databases up to October 7, 2025, for original English-language reports of the effects of SHUTi on three measures of sleep, all of them reported by the sleeper.

The first is the Insomnia Severity Index (ISI), a seven-question form about trouble sleeping and its effect on the day. Scores range from 0 to 28, and a trial of the program sets out the clinical cutoffs: up to 7 means no insomnia, 8 to 14 mild symptoms below the threshold for a diagnosis, 15 to 21 moderate insomnia and 22 or more severe insomnia. The other two measures come from sleep diaries, the daily logs participants keep: how long it takes to fall asleep, and how much time is spent awake in the night after first falling asleep.

The search turned up 29 publications reporting unique data from 25 studies, with 14,166 participants in total, which had tested the program among diverse populations from around the world. Ritterband has said that trials in Canada, Norway, Denmark and Australia all found the treatment meaningfully reduced insomnia.

The team averaged how much people given the program changed from before to after, and pooled the 10 randomized trials that had compared the program with online education, trials in which chance decided who got which. In the 2016 trial, for example, the education website offered fixed online information about insomnia that was the same for everyone.

What the review found about the insomnia program

People given the program improved on every measure. From the start of a study to its follow-up points, average scores on the insomnia index fell by 8 points or more, time to fall asleep dropped by 21 minutes or more, and time awake in the night dropped by 27 minutes or more, or 35 minutes or more when waking too early in the morning was counted.

Those are before-and-after changes, and people tend to enroll in an insomnia study when their sleep is at its worst, so some improvement follows whatever they are given. The fairer test is the comparison with the education groups, and there the program still came out ahead. Across the 10 trials with an education group, people using SHUTi reported greater improvements from baseline, the starting measurement, at the end of the course and at 6 and 12 months.

The review puts the size of that gap on a common scale, the standardized mean difference, which divides each difference by how much scores vary so that results from different studies can be compared. By convention, 0.2 counts as small, 0.5 as medium and 0.8 as large. For insomnia severity the difference ran from 0.85 to 1.03 across the time points, which is large, while for time to fall asleep it was 0.32 to 0.41 and for time awake in the night 0.45 to 0.48, small to medium.

So the program’s clearest effect was on how people rated their insomnia overall, with a more modest effect on the minutes recorded in their diaries.

Not everyone stayed the course. The review reports 64% completing a full therapeutic dose of SHUTi, meaning enough of the course to be expected to work, which leaves about one in three who did not get that far. The authors also checked whether the program worked better for some kinds of participant or study than others, and concluded that it worked across medically, psychiatrically, and demographically diverse samples.

What single trials and other reviews of online insomnia therapy show

Individual trials show what the pooled figures mean in points, and the 2016 US trial is the one most often cited. It enrolled 303 adults with chronic insomnia who had referred themselves, with 151 assigned to the program and 152 to online education. A year later, 56.6% of the program group members who supplied data (69 of 122) had scores in the no-insomnia range. That share describes the program group on its own; over the year, the trial’s comparisons of the two groups favored the program.

A large trial in Norway, with Thorndike and Ritterband among its authors, randomized 1,721 adults to a fully automated six-session online course or to online education about sleep. Participants were not told which version was the one under test and had no contact with researchers while the course ran. After nine weeks, average scores on the insomnia index had fallen from 19.2 to 10.4 in the course group and from 19.6 to 15.2 in the education group, a difference between groups of 4.7 points, with a range from 4.1 to 5.4.

A version for older adults was tested last year in 311 participants, ages 55 to 95, who were randomized to three groups: 105 to the program alone, 102 to the program with an added support protocol and 104 to online education. Both groups given the program reported less severe insomnia than the education group for a year.

Reviews by other research groups put the benefit of internet therapy in a similar place. One team that pooled 15 trials of various online programs in 2016 found that internet therapy lowered scores on the insomnia severity index by 4.3 points more than in control groups. A larger analysis of digital programs published this year, covering 73 randomized controlled trials and 14,465 participants, found a non-linear dose-response relationship: the benefit grew with total course time and then leveled off beyond about 300 minutes. The largest effect it estimated was 5.77 points on the index against control.

Set beside those figures, the before-and-after drop of 8 points in the new review is larger than the gain over a comparison group that independent reviews find for digital programs, which is expected, because comparison groups improve too.

What the insomnia program review cannot show

The reviewers are the developers. The review was led from the group that built the program, and authors of it have held a financial interest in the company selling it. Its pooled results are in line with independent reviews of internet therapy in general.

Everything is self-reported. Insomnia is defined by how a person experiences sleep, so self-report is the standard measure, but none of the pooled outcomes was measured by a device. UVA’s 2016 statement made the same point about that trial, saying its measure of effectiveness rested on experiences self-reported by study participants rather than on any measurement of time asleep.

The comparison was with education. The trials show the program beats reading about sleep online, but not how it compares with a therapist, with other apps or with sleeping pills. The 2016 review of 15 trials found no clear differences between internet-delivered and in-person therapy with a trained therapist on the measures it pooled.

A third fell short of a full dose. The 64% figure leaves a large minority who stopped early. In the trial in older adults, 131 people (63%) completed all six SHUTi Cores, the program’s six parts, within the nine-week course period.

Who took part. In the main US trial, participants referred themselves. The trial in older adults listed as a weakness a lack of diversity in ethnicity, race and education, a gap that limits how far its results apply to other people.

English-language reports only. The search was limited to papers in English.

Whether an online program, a therapist or another approach suits a particular person with insomnia is assessed by their doctor.

Twenty years of studies show people given this program report better sleep than people given sleep education pages, on the evidence of the team that made it.