eHealth-Based Psychosocial Interventions for Adults With Insomnia

Background: Worldwide, insomnia remains a highly prevalent public health problem. eHealth presents a novel opportunity to deliver effective, accessible, and affordable insomnia treatments on a population-wide scale. However, there is no quantitative integration of evidence regarding the effectiveness of eHealth-based psychosocial interventions on insomnia. Objective: We aimed to evaluate the effectiveness of eHealth-based psychosocial interventions for insomnia and investigate the influence of specific study characteristics and intervention features on these effects. Methods: We searched PubMed, Embase, Web of Science, PsycINFO, and the Cochrane Central Register of Controlled Trials from database inception to February 16, 2021, for publications investigating eHealth-based psychosocial interventions targeting insomnia and updated the search of PubMed to December 6, 2021. We also screened gray literature for unpublished data. Eligible studies were randomized controlled trials of eHealth-based psychosocial interventions targeting adults with insomnia. Random-effects meta-analysis models were used to assess primary and secondary outcomes. Primary outcomes were insomnia severity and sleep quality. Meta-analyses were performed by pooling the effects of eHealth-based psychosocial interventions on insomnia compared with inactive and in-person conditions. We performed subgroup analyses and metaregressions to explore specific factors that affected the effectiveness. Secondary outcomes included sleep diary parameters and mental health–related outcomes.


Background
Insomnia is a common complaint in primary care and a prevalent public health problem [1].The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, defines insomnia as dissatisfaction with sleep quantity or quality, characterized by difficulty initiating or maintaining sleep for 3 or more days per week for a minimum duration of 3 months, accompanied by considerable distress and functional impairments (eg, intellectual-, behavioral-, social-, occupational-, and mood-related impairment) [2].Approximately 25% of adults experience unsatisfactory sleep, and approximately 6% to 10% of adults meet the diagnostic criteria for insomnia [1].In low-income settings in Africa and Asia, the prevalence of sleep problems can reach up to 40% [3].Insomnia is often persistent and debilitating, increasing the risk of other physical or mental illnesses or exacerbating existing medical or psychiatric disorders [4,5].Thus, insomnia carries a heavy individual and societal burden, including the burden on the health care system, which hinders societal development and leads to socioeconomic losses [6,7].
Although a range of pharmacologic treatments and psychosocial therapies exist for insomnia, the treatment of insomnia remains a major challenge [8].Previous research has demonstrated that benzodiazepines and benzodiazepine receptor agonists have a short-term efficacy on insomnia, whereas long-term use is usually associated with potential side effects, including memory dysfunction, somatic symptoms, drug dependence, and interactions [9].Cognitive behavioral therapy (CBT) for insomnia is effective with long-lasting effects when compared with medications, and is recommended as the first-line treatment for insomnia [10].However, because of the limited number of trained therapists, high costs, and time-intensive nature of in-person CBT for insomnia (CBT-I), millions of patients still do not have access to this effective treatment to improve their sleep outcomes [11].Hence, there is an urgent need for inexpensive, innovative delivery modalities of CBT or novel treatment options to be effective and accessible for the larger population at a lower cost [12,13].
eHealth is increasingly being developed and implemented for the delivery of remote, timely, high-quality, and limited-contact care [14].eHealth can be defined as "health services and information delivered or enhanced through the Internet and related technologies" [15].In a broader sense, it can encompass a range of services or systems that facilitate health care practice through the use of information and communication technologies, including electronic health records, e-prescriptions, digital interventions, telemedicine, and mobile health [16].Furthermore, eHealth is increasingly being applied to the prevention and treatment of several mental illnesses, including but not limited to smoking cessation, anxiety, depression, and suicidal ideation [17][18][19][20].In these eHealth programs, information about illness, treatment, self-management strategies, health status tracking, support, and feedback are delivered via the internet and related technologies [21][22][23][24].Their results consistently show eHealth to have high accessibility, interactivity, and effectiveness with limited cost.In addition, a meta-analysis revealed the effects of internet-delivered CBT to be equivalent to those of face-to-face CBT for psychiatric and somatic disorders [25].
Over the past decade, a number of telemedicine interventions, smartphone apps, and websites have been created to help users develop good sleep habits and improve sleep quality through sleep monitoring, sleep hygiene, CBT-I, or mindfulness meditation [26][27][28][29].Systematic reviews suggest that internet-based CBT-I has medium to large effects on sleep outcomes among youth and adults [30,31].Meta-analyses indicate that digitally delivered CBT and telemedicine-based CBT are noninferior to face-to-face CBT [32,33]; however, only a small number of randomized controlled trials (RCTs) directly comparing 2 treatments were available to be included to pool the effects (n=4 and n=2).A recent systematic review of mobile phone sleep interventions demonstrated the effectiveness of mobile health technologies for improving sleep [34].However, these reviews covered a single psychosocial intervention (eg, CBT-I) or particular eHealth modality (eg, mobile devices).We found no systematic review or meta-analysis summarizing and comparing the effects of multiple eHealth-based psychosocial interventions for insomnia.In addition, more novel trials with robust study designs and large sample sizes have been published in the past 5 years that were not included in the previous reviews [12,13,26,[35][36][37][38][39].Hence, an updated systematic review is warranted to examine the effectiveness of eHealth as a treatment for adults with insomnia.

Objectives
The primary purpose of this systematic review was to summarize the evidence of the effectiveness of eHealth-based psychosocial interventions on insomnia symptoms, including insomnia severity and sleep quality.We investigated the effectiveness of eHealth-based psychosocial interventions compared with inactive controls and in-person comparators.Subgroup and metaregression analyses were performed to identify whether and to what extent population and intervention characteristics were related to treatment effectiveness on insomnia symptoms.As secondary outcomes, we evaluated the effects of eHealth-based psychosocial interventions on improving sleep diary parameters and mental health-related outcomes, for instance, sleep efficiency, maladaptive beliefs about sleep, and fatigue and depression symptoms.

Methods
This systematic review and meta-analysis were conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) reporting guidelines [40].This study was registered in PROSPERO (CRD42021233241).

Search Strategy
A systematic search was performed in the following databases-PubMed, Embase, Web of Science, PsycINFO, and the Cochrane Central Register of Controlled Trials-from inception to February 16, 2021, with PubMed searched up to December 6, 2021.Our search strategy combined index terms and text words associated with insomnia, eHealth, and intervention (the full list of search terms is provided in Multimedia Appendix 1).We also searched the gray literature, including dissertations, clinical trial registries, and conference proceedings, for unpublished studies.Furthermore, we manually scanned the references of relevant studies and reviews to identify any additional studies of relevance.

Eligibility Criteria and Study Selection
The search strategy and selection criteria were developed using the Population, Intervention, Comparison, Outcomes, and Study Design framework [41].Studies were eligible if they assessed eHealth-based psychosocial interventions with the primary aim of improving insomnia symptoms; detailed information on inclusion and exclusion criteria is provided in Textbox 1.After identifying and removing duplicates, 2 reviewers (WD and JW) performed the screening, with disagreements resolved through discussion with a third reviewer (RK).
Textbox 1. Inclusion and exclusion criteria for this study.

•
Inclusion criteria

•
Participants: participants were adults with Diagnostic and Statistical Manual of Mental Disorders-or International Classification of Sleep Disorders-diagnosed insomnia or self-reported insomnia complaints.
• Intervention: "eHealth-based psychosocial interventions" included psychosocial treatments that were delivered via computers (email or websites), mobile phones (apps or SMS text messages and phone calls), telemedicine, digital games, and tablets or related technologies [1].Blended interventions for insomnia, which combine little face-to-face care with the intervention primarily via eHealth channels [14], were also included if the eHealth component constituted ≥75% of the intervention sessions or the core of the intervention was eHealth-based [30].
• Comparison: we only included studies with "in-person" controls or "inactive" controls.
• "In-person" meant face-to-face psychotherapy, for instance, cognitive behavioral therapy.

•
Control groups classified as "inactive" were those in which participants were put on a waiting list or placed in a placebo group, for instance, usual care or sleep hygiene [4,5].
• Outcome measures: we included studies that had at least one insomnia severity-or sleep quality-related outcome measure.
• Study design: we included randomized controlled trials.

•
We excluded studies that included participants with sleep disorders other than insomnia (eg, obstructive sleep apnea or narcolepsy), with specific medical conditions (eg, epilepsy, chronic pain, cancer, or recent surgery), and belonging to certain special populations (eg, overnight shift workers and pregnant or puerperium women).

•
We further excluded trials using smart devices solely to record sleep data or physical therapy.
• Trials without a control group or noninferiority trials comparing 2 eHealth interventions were excluded.computer to log into a website, dashboard, or video chat room), or mixed mode (eg, an intervention delivered using both computer-assisted and phone-delivered components or an intervention that only stated the use of the internet or website).We also classified the functions of eHealth as informing, instructing, displaying, guiding, reminding, and communicating [42].For studies with overlapping data sets, we used the most recent study with relevant outcome measure data.If the data were abstracted or unclear, we contacted the corresponding author by email for clarification.If the author did not respond after 2 contact attempts, we excluded the study.Disagreements between reviewers regarding data abstraction were resolved through discussion with a third reviewer.Interrater reliability between the 2 reviewers, assessed using the Cohen κ, indicated acceptable agreement (Cohen κ=0.86) [43].

Data Analysis
A random-effects model was used in the meta-analysis, which allows for subtle differences across studies because of variability in sampling or treatment [44].The Hedges g was used to estimate the effect size by pooling the mean difference of the continuous measures between the intervention and the control condition, which is the unbiased standard mean difference [45,46].For multi-arm studies including 2 eHealth-based psychosocial interventions, we included each pairwise comparison separately by evenly dividing the shared control condition among the comparisons [47].For studies comparing an eHealth-based psychosocial intervention with both in-person and inactive control conditions, corresponding comparisons were used in the meta-analyses for the different conditions.We presented pooled results using forest plots.Between-study heterogeneity was estimated using the Cochran Q and I 2 statistics.Heterogeneity was interpreted according to the following thresholds: low (0%-40%), moderate (30%-60%), substantial (50%-90%), and considerable (75%-100%).We conducted influence analysis on all studies via the "leave-one-study-out meta-analysis" method recalculating the pooled effect sizes after removing each study to detect potential outliers [48].Publication bias was examined graphically using funnel plots.The degree of asymmetry was tested using the Egger regression test of the intercept with a 1-tailed significance level of α=.05 applied to the primary and secondary outcome analyses [49].Furthermore, the trim-and-fill analysis by Duval and Tweedie [50] was applied to adjust the effect size for missing studies.All statistical analyses were conducted using Stata (version 14.1; StataCorp) [51].

Quality Assessment
The Cochrane risk-of-bias tool for randomized trials was used to assess the methodological quality of the RCTs [52].This tool examines 5 domains of trial design: the risk of bias in the randomization process, deviations from the intended interventions(effect of assignment to intervention), missing outcome data, measurement of the outcome, and selection of the reported result, ranking each domain as high, low, or with some concerns for risk of bias.Studies with a high risk of bias in at least one domain were rated as having a high overall risk of bias.In total, 2 reviewers (WD and JW) evaluated the risk of bias for each domain independently, with good interrater agreement (κ=0.70).Disagreements between reviewers were resolved through discussion with a third reviewer (RK).
The delivery mode of the eHealth interventions ranged from mixed-mode interventions (20/37, 54%) to computer-assisted interventions (13/37, 35%) and phone-delivered interventions (5/37, 14%).Sleep Healthy Using the Internet [82] and Sleepio ([83]; Big Health Ltd) were frequently used eHealth programs, each appearing in 16% (6/37) of the studies.i-Sleep [84] was also used quite frequently.Regarding the guidance modality, 46% (17/37) of the included studies reported that the eHealth intervention was instructed by a trained human therapist, for example, under the guidance of an expert clinician or trained coach in vivo.A total of 19% (7/37) of the studies reported that the participants were guided by an animated therapist, with the remainder (15/37, 41%) reporting no guidance.Regarding feedback, the eHealth interventions in 86% (32/37) of the studies provided tailored feedback, including feedback on the web using real-time user data such as personal summary statistics, progress scores, or automated individual advice.In contrast, a small proportion of the studies (7/37, 19%) did not provide tailored feedback.Notably, there were 5% (2/37) of the studies comprising 2 separate eHealth intervention arms, one with tailored feedback and the other without tailored feedback [13,59]; both concluded that tailored feedback could enhance the efficacy.In addition, most of the eHealth interventions in the included studies (34/37, 92%) reminded or encouraged participants to stay involved via email, SMS text message, or phone call, whereas only a minority (5/37, 14%) did not mention the use of any reminder or encouragement; 5% (2/37) of the studies included 2 eHealth arms, where only 1 arm set reminders [13,59] (see Table 2 for the intervention characteristics).The duration of the interventions ranged from 2 to 12 weeks, with a mean of 7.05 (SD 2.24) weeks, and the average number of treatment sessions was 6.2 (SD 1.0) ranging from 2 to 8 sessions, with durations of 20 to 60 minutes per session.Questionnaires and sleep diaries were used to evaluate the effectiveness of the eHealth interventions in improving insomnia symptoms, sleep status, or mental health.Detailed information on the outcome indicators and assessments can be found in Table 3.On average, the percentage of participants in the eHealth intervention groups that completed the postintervention assessment was 74.6% (SD 18.8%).The follow-up duration ranged from 3 weeks to 12 months, with a mean of 5.6 (SD 3.1) months.Multimedia Appendix 2 [12,13,26,[35][36][37][38][39] includes a summary of the full details of the included studies and information on process outcomes.

Effects of eHealth-Based Psychosocial Interventions on Primary Outcomes in Comparison With In-Person CBT
The results of the eHealth-based psychosocial interventions versus in-person CBT are shown in Figure 3 [26,55,67,78].We found that in-person CBT showed greater improvement in insomnia severity (4/37, 11% of the studies; Hedges g=0.41, 95% CI −0.02 to 0.85; P=.06;I 2 =65%) compared with eCBT; however, the assumption of noninferiority of eHealth interventions compared with in-person CBT was not rejected (P=.06).In terms of improving sleep quality, in-person interventions had a significantly superior performance (3/37, 8% of the studies; Hedges g=0.56, 95% CI 0.24-0.88;P<.001;I 2 =9%).Heterogeneity was low to moderate across the studies.

Exploratory Subgroup Analyses and Metaregression Analyses
The results of the exploratory subgroup analyses conducted using only the studies comparing eHealth-based psychosocial interventions with inactive controls are shown in Table 4.   Exploratory subgroup analyses pivoting on intervention features showed that "guidance by trained therapists" and "tailored feedback" were moderators that caused significant differences in effects.Specifically, eHealth-based psychosocial interventions providing "guidance by trained therapists" were more effective in reducing insomnia severity ( The difference between the subgroups was the marginal significance level (Q=3.27;P=.07).However, there were no statistically significant associations between therapeutic approach, eHealth delivery mode, or provision of reminders and effect sizes.
The results of the metaregression analyses are presented in Table 5.The baseline insomnia severity and intervention duration (in weeks and sessions) had moderating effects on the study effects on insomnia severity.Higher baseline insomnia severity was associated with larger effect sizes (b=−0.11;P=.004.In addition, a longer intervention duration was associated with larger effect sizes (in weeks: b=−0.09 and P=.01 in sessions: b=−0.20 and P=.03).See the plots of the metaregression analyses in Multimedia Appendix 5.However, no statistically significant associations were observed among sleep medication, number XSL • FO RenderX of intervention components, number of eHealth functions, and the effects on insomnia severity or between these moderators and the effects on sleep quality.

Effects of eHealth-Based Psychosocial Interventions on Secondary Outcomes
All the pooled effect sizes related to sleep parameters and mental health-related outcomes remained statistically significant, as shown in Figure 4

Quality Assessment
The results from the Cochrane risk-of-bias tool for randomized trials assessment are shown in Figure 5 [12,13,26,[35][36][37][38][39].Of the studies, 54% (20/37) were rated as having a high overall risk of bias, 41% (15/37) were rated as having some concerns regarding the overall bias, and 5% (2/37) had a low overall risk of bias.The most frequent risk factor identified was deviations from intended interventions, which was most often owing to inadequate blinding of participants or caregivers, adherence problems, or lack of appropriate analysis; 49% (18/37) of the studies were evaluated as having a high risk of bias in this domain, with 43% (16/37) of the studies evaluated as having some concerns.Owing to a lack of information on concealment of the allocation sequence, 57% (21/37) of the included studies had some concerns regarding the randomization process.A total of 95% (35/37) and 68% (25/37) of the studies were evaluated as having a low risk of bias for missing outcome data and measurement of the outcome, respectively.None of the included studies was evaluated as having a risk of bias regarding selectively reported results.[12,13,26,[35][36][37][38][39].

Principal Findings
This systematic review identified 37 RCTs that reported data on 13,227 individuals from 12 countries.eHealth-based psychosocial interventions were delivered via a website, computer, smartphone, telephone, or mixed mode, in most cases (32/37, 86%) based on CBT-I.Questionnaires and sleep diaries were used to evaluate the effectiveness.Our findings are outlined in Textbox 2.

•
We found that eHealth-based psychosocial interventions yielded a large reduction in insomnia severity (Hedges g=−1.01) and a moderate improvement in sleep quality (Hedges g=−0.58) as compared with inactive controls.Despite the heterogeneity between studies on primary outcomes, no outlier studies were identified through influence analysis, funnel plots, and Egger tests, indicating no significant publication bias.
• There was no significant difference in the reduction of insomnia severity when comparing eHealth interventions and in-person cognitive behavioral therapy (CBT).However, in-person CBT was shown to be more effective in improving sleep quality.
• Guidance from trained therapists and tailored feedback were associated with larger treatment effects on insomnia symptoms.All subgroup analyses on primary outcomes favored eHealth interventions and indicated at least moderate effect sizes (Hedges g=−0.33 to −1.21).

•
Higher baseline insomnia severity and longer intervention duration were associated with a larger reduction in insomnia severity.In addition, our review showed a larger improvement in insomnia severity in the clinical samples and samples with higher baseline insomnia severity.Possible explanations for this include clinical samples with more severe insomnia symptoms having more space for improvement.In addition, user requirements and motivation may influence attention engagement and, thus, treatment effectiveness.The observed effects suggest that eHealth-based psychosocial interventions for insomnia could be applicable to a broad range of populations and are well suited to be integrated into a stepped-care approach [88,89].
With regard to eHealth intervention features, our results imply that professional guidance and tailored feedback are associated with greater effect sizes.Thus, this could be crucial in facilitating the effectiveness of eHealth-based psychosocial interventions as the involvement of trained therapist support and individualized advice seem to promote effectiveness.This further supports the view that blended interventions integrating therapeutic support in face-to-face treatment with the cost-effectiveness of eHealth could be a way to increase treatment effectiveness while saving time and reducing costs [90].Contrary to expectations, eCBT and other psychosocial interventions were found to be roughly equivalent in effectiveness, although caution is needed as only 4 non-CBT eHealth studies were included in this review.Alternative therapeutic techniques such as acceptance and commitment therapy, problem-solving therapy, psychodynamic therapy, mindfulness therapy, and interpersonal psychotherapy are also worth exploring in conjunction with eHealth to promote sleep and well-being.Considering the structured nature of CBT, it might not be beneficial for people with complex needs, and it fails to address deeper causes or the possible underlying causes of mental illness, such as childhood experience, family history, or relations [91].In addition, the typical treatment period for CBT is 6 to 20 weeks, which, to some extent, requires people to adhere to a long period.Indeed, a more thorough investigation of the effectiveness of different therapeutic techniques implemented in eHealth for the treatment of insomnia is necessary.Furthermore, a longer intervention duration was found to be associated with a larger reduction in insomnia severity; similar results were reported in a previous meta-analysis on CBT-I [31].Finally, no significant effects on insomnia were found for moderators such as delivery mode, reminder settings, sleep medication use, number of intervention components, and eHealth functions.Additional studies are needed to determine optimal intervention characteristics, including the number of treatment sessions and intervention components.

RenderX
We further found that eHealth-based psychosocial interventions effectively improved sleep efficiency and TST and reduced SOL, WASO, and NWAK.These findings are in line with those of previous studies [31,86].There is also evidence supporting a mixed effect on mental health-related outcomes, adjusting people's maladaptive beliefs about sleep and alleviating fatigue, anxiety, and depression symptoms.Thus, eHealth-based psychosocial interventions for insomnia could promote mental health and prevent the exacerbation of comorbid medical and psychiatric conditions.Previous studies have demonstrated the effects of eHealth interventions for cancer survivors on improving sleep and reducing fear of recurrence, depression, and anxiety [92,93].These low-cost and convenient insomnia treatments can be widely disseminated, along with support for eHealth-illiterate populations, among people with mental or physical disorders accompanied by insomnia [94].
This meta-analysis focused on a wide range of sleep and mental health outcome measures to assess the impact of the intervention in a holistic manner.Our findings highlight several directions for future research.Given the importance of user engagement and therapist support for treatment effectiveness and adherence, future research could pay more attention to increasing user engagement and interaction in the design of eHealth-based psychosocial interventions for insomnia.This could be achieved by developing blended, appealing, and adaptive interventions as well as making eHealth interventions accessible to those with a lower eHealth literacy [22].Furthermore, research is needed to directly compare eHealth-based psychosocial interventions in different delivery modes in noninferiority trials while assessing the cost-effectiveness, treatment credibility, satisfaction, and therapeutic alliance.This would help tremendously in the optimization of eHealth-based psychosocial interventions.In addition, given that most trials to date have been implemented in high-income countries with little cultural diversity, eHealth-based psychosocial interventions for insomnia should also be investigated in more low-and middle-income countries to increase the accessibility of eHealth.

Limitations
Some limitations should be noted.First, this review only focused on adults with insomnia; further research is needed to evaluate the effects of eHealth-based psychosocial interventions in specific populations, including children, adolescents or employees of specific sectors.Second, although the random-effects model aimed to account for between-study heterogeneity statistically, our analyses still indicated significant between-study differences.Variability in the control condition was particularly identified as a source of heterogeneity.However, the heterogeneity in this study might be multivariate, which may be caused by different outcome measures or confounding bias.As not all studies used the same outcomes, the pooled effects for insomnia severity and sleep quality and the secondary outcomes were based on different numbers of studies and, to some extent, using different outcomes.The difference in outcome measures might more or less influence the treatment effects.Third, our meta-analyses focused on the immediate intervention effects.Owing to a certain level of dropout, few studies assessed insomnia symptoms for >6 months, and the included studies set different follow-up times.
Future studies could include long-term follow-ups at standardized lags to observe the prolonged effects on sleep.Fourth, this review depicted participants' completion of various questionnaires and assessments.However, because of the limited number of studies that reported the same outcome measure on treatment adherence, we failed to summarize how participants implemented treatment recommendations and how therapists followed treatment protocols, which are important factors in treatment adherence [95].The impact of treatment adherence on eHealth-based psychosocial interventions is worth investigating, and future RCTs should use standardized methods to comprehensively assess treatment adherence and its relationship with treatment effects.

Conclusions
In

Figure 1 .
Figure 1.PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart of study selection.

c ICSD- 3 :
International Classification of Sleep Disorders-Third Edition.d ISI: Insomnia Severity Index.e IG: intervention group.f CG: control group.g eCBT-I: eHealth-based cognitive behavioral therapy for insomnia.h CBT: cognitive behavioral therapy.i NS: not specified in the study.
j ePST: eHealth-based problem-solving treatment.k DSM-5: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition.l SHUTi: Sleep Healthy Using the Internet.m eCBT: eHealth-based CBT.n TAU: treatment as usual.o SCI: Sleep Condition Indicator.p QIDS-SR: Quick Inventory of Depressive Symptomatology.q DSM-4: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition.r CBT-I: CBT for insomnia.s PSQI: Pittsburgh Sleep Quality Index.t mHealth: mobile health.u PHQ-9 Patient Health Questionnaire-9.
b CG: control group.c Italicized P values are significant.d Differences between subgroups are shown in the next row.e CBT: cognitive behavioral therapy.f eCBT: eHealth-based cognitive behavioral therapy.

Figure 4 .
Figure 4. Effects of eHealth-based psychosocial interventions on outcome measures.DBAS: Dysfunctional Beliefs and Attitudes about Sleep; NWAK: number of nocturnal awakenings; SOL: sleep onset latency; TST: total sleep time; WASO: wake after sleep onset.
a The studies by Espie et al

Table 3 .
Summary of outcome indicators and assessment.

Table 4 .
Exploratory subgroup analyses of effects on the primary outcomes (N=37).
a IG: intervention group.

Table 5 .
Metaregression analyses of effects on the primary outcomes (N=37).
a Italicized P values are significant.

•
With respect to secondary outcomes, eHealth interventions for insomnia had significantly small to moderate positive effects (Hedges g=−0.21 to −0.56) on sleep parameters and mental health-related outcomes.
conclusion, this review provides an up-to-date and comprehensive overview and quantitative integration of current research on the effectiveness of eHealth-based psychosocial interventions for insomnia.eHealth-based psychosocial interventions have the potential to reduce both insomnia symptoms and other mental health-related outcomes.Our findings suggest that, as a less costly intervention, eHealth-based psychosocial interventions should be disseminated widely and integrated into a stepped-care model.Professional guidance and tailored feedback should accompany eHealth interventions to improve effectiveness.Blended care integrating face-to-face care with eHealth may further improve effectiveness and benefit a more diverse population with insomnia complaints.Further investigations of intervention components and blended interventions are needed to better understand the effectiveness of different intervention components, especially as pertaining to people of low socioeconomic status or low eHealth literacy.