Brief digital interventions for adults with emerging mental health concerns: a systematic review reporting on socioemotional outcomes Open Access

Epidemiological data from the past two decades (2001–2022) indicate that roughly half of the global population will live with one or more mental health disorders before they reach 75 years old (McGrath et al., 2023), making it the leading cause for years lived in disability globally (WHO, 2022). Recent global challenges such as the COVID-19 pandemic or displacement due to conflict have added additional strains on individual’s mental health, highlighting mental ill-health as a pervasive yet often inadequately treated and underfunded issue (Santomauro et al., 2021; Yasenok et al., 2025). Recent research suggests that the economic impact of ill-mental health was valued at US$5trillion in 2019 (Arias et al., 2022). These costs are expected to rise to US$6trillion by 2030 – surpassing the costs for cancer, diabetes and chronic respiratory diseases combined (Bloom et al., 2011; The Lancet Global, 2020). Given these costs next to the impact on the well-being of individuals and their families, investing in mental health treatment and improving access to interventions is crucial. Research suggests that every dollar spent on treatment for the most common mental health issues, depression and anxiety, could yield a return of four dollars in improved health and productivity (Chisholm et al., 2016).

Digital mental health interventions (DMHIs) offer technical alternatives to traditional face-to-face support, expanding access by using online or technology-based platforms. They significantly increase access to mental health support services, which is especially important for underreached areas (Goodridge and Marciniuk, 2016). To date, young people have more often been targeted with digital solutions to mental health provision. However, accelerated by the general public’s familiarization with technology brought about by the COVID-19 pandemic and the ubiquity of mobile phones, DMHIs offer a universally accessible method for delivering mental health services for all adult populations and targeted intervention in underreached areas of the population (Anderson-Lewis et al., 2018).

Despite this, DMHIs face challenges, particularly in their uptake among targeted populations and maintaining end-user engagement (Andrews et al., 2018; Fleming et al., 2018). Borghouts and colleagues (2021), identified three main barriers:

1. user characteristics (such as age, gender or the severity of symptoms);

2. user experience with the tool (with positive experiences encouraging engagement and the implementation of the tool); and

3. technical difficulties that hinder engagement.

Thus, active end-user involvement in the development and functionality of DMHIs is crucial (Opie et al., 2025). De Beurs et al. (2017) emphasized the challenges in finding effective ways to ensure end-user engagement in Web-based mental health interventions, noting that while it should be standard practice, it requires careful management of resources, including time and funding.

One key factor in increasing DMHI engagement is to supplement the digital intervention with human support and guidance (Borghouts et al., 2021). However, incorporating a human element in DMHIs results in additional resources and costs, often limiting the extent of guidance provided. While automated reminders, such as short messaging service (SMS), offer a low-cost alternative to maintain engagement, they are less effective if lack of motivation or social connection are barriers to participation (Mohr et al., 2011; Schueller et al., 2017). Furthermore, the level of guidance provided also has an impact on the time end-users are willing to dedicate to a DMHI. Shorter self-guided sessions are generally preferred, but some clients may prefer longer sessions with guidance (Borghouts et al., 2021), meaning a single application might not be suitable for all.

Regarding the methodological underpinning of DMHIs, brief psychotherapy using cognitive behavioral methods have demonstrated effectiveness in addressing mental health concerns, like anxiety and depression (Ruisoto et al., 2022), and lends itself to delivery via internet-based programs (Etzelmueller et al., 2020). In this context, single session thinking (SST) offers a progressive and cost-effective alternative for adults with mental health challenges (Hoyt et al., 2018; Hoyt et al., 2021). SST is predicated on the observation that many clients attend only one therapy session, but during this session can receive sufficient guidance and support to mobilize their own resources to meet the challenge they face (Talmon, 1990). The brevity of one session means long-term commitment is not required, nor is there a waitlist, making them more accessible (Young, 2020; von Doussa et al., 2021). While a single session may suffice for some mental health concerns, such as mild depression, and adjustment related concerns (Kim et al., 2023), it can also serve as a screening gateway to further support (Sung et al., 2020). With the emphasis of maximizing the effectiveness of each encounter, the dynamic between client and therapists shifts toward collaborative elements, such as codesigning the session, transparency in expectations and enhanced mutual directness (Young, 2020). Evidence for digital or telehealth applications of SST is emerging, including in complex contexts, such as family therapy (Hartley et al., 2023).

Previous reviews on adult mental health have primarily focused on the impact of brief or single-session interventions delivered via face-to-face interactions, excluding online/digital-based interventions (Aafjes-van Doorn and Sweeney, 2019; Kim et al., 2023; Schleider et al., 2025). However, meta-analytic evidence demonstrates the effectiveness of DMHIs of various length and guidance in treating depression (Karyotaki et al., 2021; Miguel et al., 2023; Sander et al., 2016), anxiety (Olthuis et al., 2016a), stress (Miguel et al., 2023) and posttraumatic stress disorder (PTSD) (Olthuis et al., 2016b). Evidence indicates that DMHIs with human support are as effective as traditional in-person therapy (Andrews et al., 2018; Cuijpers et al., 2019). A meta-analysis by Leung et al. (2022) found no significant difference in efficacy between clinician-guided interventions and nonclinician guided interventions, while nonclinician guided DMHIs were significantly more effective than self-guided ones or control groups. These findings support the use of DMIHs with a human supported component, whether partially or completely.

However, evidence on the efficacy of interventions for adult populations with mild-to-moderate symptoms (i.e. indicated populations) is only just emerging. Conley et al. (2017) found that programs for indicated populations were more effective than universal programs aimed at those with varying levels of symptom severity among college and graduate students. Human guided DMHIs for indicated populations have been more effective than self-guided DMHIs (Conley et al., 2017). Another review found that DMHIs with human guidance, especially those using cognitive behavioral therapy, provided effective management for mild-to-moderate anxiety and depression (Kay-Lambkin et al., 2019) and that text-based interventions, i.e. e-mails and SMS, provided effective support for these conditions. Collectively, these findings support the use of digital interventions as scalable tools for reducing symptoms of anxiety and depression for indicated populations.

Despite the rapid uptake of brief DMHIs in recent years and some emerging evidence, large research gaps on their effectiveness for adult clients remain. Notably, existing reviews examining guided DMHIs for adults focus on common internalizing (i.e. inwardly directed) mental health difficulties, such as anxiety and depression (Karyotaki et al., 2018; Plessen et al., 2025; Willner et al., 2016). Other mental health outcomes, such as well-being, quality of life and social functioning, however, are rarely or not considered. Similarly, the value of DMHIs in addressing externalizing (i.e. outwardly directed) mental health problems or behaviors, such as aggression, conduct problems, impulsivity also remains unclear (Liverpool et al., 2025; Schleider et al., 2025). Thus, a further enquiry into the influence of DMHIs on wider socioemotional outcomes, generally defined as social and emotional functioning, including internalizing and externalizing symptoms (Verhoef et al., 2018; Zarra-Nezhad et al., 2022), is called for. This is necessary to ascertain mental health challenges these tools are most efficacious for, as well as guide and improve their utilization and development in areas that currently hold less evidence of efficacy.

In addition, earlier reviews on brief, adult-focused DMHIs revolve around specific therapeutic models, e.g. cognitive behavior therapy (O'Toole et al., 2025; Wright et al., 2019). This is in stark contrast to the demonstrated potential of DMHIs adhering to various therapeutic models when addressing socioemotional concerns in young people (Madrid-Cagigal et al., 2025; Opie et al., 2024a). Furthermore, knowledge regarding design features and in-built components (e.g., content, multimedia, tailoring) contributing to socioemotional efficacy of DMHIs is limited. The systematic examination of key digital intervention elements positively impacting socioemotional outcomes is, however, essential for delivering optimal and accessible mental health-care (Opie et al., 2024b).

Thus, the specific aims of the current review are to:

• Identify and synthesize all relevant evidence-based, brief (≤3 hour per contact) DMHIs for adults (≥25 years) that incorporate a degree of guidance by a human support, whether completely or partially.

• Report on the socioemotional outcomes of these interventions.

• Provide a preliminary exploration of potential elements of these DMHIs that impact socioemotional outcomes. This includes investigating whether any support elements (e.g. therapeutic approach, content, tailoring, follow-up support) are particularly effective or ineffective for this intervention type.

This systematic review was conducted in accordance with the Joanna Briggs Institute methodology framework (Aromataris and Munn, 2020). Our reporting adheres to the preferred reporting items for systematic reviews and meta-analyses (Page et al., 2021). To ensure transparency, the review’s methodology and inclusion/exclusion criteria were predefined and prospectively registered in PROSPERO (December 12, 2023; CRD42023485124).

The aims for this review have been developed in close collaboration with Beyond Blue, one of Australia’s leading and widely recognized mental health organizations. The authorship team also included a consumer academic (JO) introducing a consumer voice and emphasizing the importance of respecting and valuing the perspectives of people with lived experience in research. By integrating various expertise, experiences and voices of academics, individuals with lived experience and mental health service providers, this study followed community-based participatory research (Hacker, 2013) and participatory action research (Cornish et al., 2023) approaches.

The population, intervention, comparator, outcome and study design framework (McKenzie et al., 2019) guided inclusion criteria eligibility (see Table 1). The search was limited to full text references published in English between January 1, 2018, and March 8, 2024 to ensure a contemporary examination of the literature, given rapid recent technological advancements, associated technological redundancies and the scarcity of relevant literature before 2018 (Ganapathy et al., 2023; Philippe et al., 2022).

A four-step search strategy was implemented. An initial limited search of PsycINFO was conducted to help identify keywords and index terms based on the results. The second search consisted of a systematic examination of four electronic databases: PsycINFO (Ovid), MEDLINE (Ovid), CINAHL (EBSCO) and Cochrane Central Register of Controlled Trials (Central; via Cochrane Library) using the identified keyword and index terms. The third step involved searching for unpublished literature, including dissertations and theses through ProQuest Dissertations and Theses Global. Two Global trial registries (www.anzctr.org.au and www.clinicaltrials.gov) were searched to identify ongoing or completed but unpublished studies. Using the same keywords, Google results were searched up to the first 20 pages. Finally, the reference lists of all eligible studies and relevant systematic reviews were manually examined to identify additional studies meeting the inclusion criteria. Refer to Supplementary Material 1 for a comprehensive search strategy of published and unpublished databases, including concepts and terms used in all included databases.

All records were imported into Endnote 20 (The EndNote Team, 2021) to eliminate duplicates. Subsequently, the remaining studies were imported into Covidence (Veritas Health Innovation, 2022) for screening at title and abstract level, involving two reviewers (JO, AV, including 25% double screening). Following this, the full text of all remaining studies was screened by two reviewers (JO, AV). At both screening levels, disagreements were resolved through consultation and discussion with the last author (SK). Studies were excluded if necessary information was not reported in text and no contact was made with study authors to retrieve missing data.

Data of each full-text article was charted by one reviewer and subsequently reviewed by the second reviewer (JO, AV), with any disagreements being resolved through discussions. The data was extracted into predefined standardized data extraction forms, aligning with the content outlined in Supplementary Tables A1–A3.

Studies were categorized and clustered under subheadings, consistent with Supplementary Tables A1–A3. The reporting of results was data-driven, privileging socioemotional outcomes documented in at least two studies. This encompassed a range of psychological measures, including depression, anxiety, stress and loneliness, as well as mindfulness, emotional regulation and dysregulation, positive affect and anger. Findings were then subjected to narrative analysis (Popay et al., 2006). Given the small number of studies, qualitative data from mixed-methods studies were analyzed together with the quantitative data to provide a more in-depth understanding of the DMHIs’ effectiveness for each outcome. Where available, effect sizes and their significance were reported. Reporting of elements was data-driven, privileging elements documented in at least two studies. Furthermore, elements were categorized in three sections:

1. established efficacy;

2. poor or yet-established efficacy; and

3. inconsistent or contradictory evidence of efficacy.

While this review reports on socioemotional outcomes, a second review explored user experience outcomes and whether DMHIs were perceived as useful, usable, valuable, credible, accessible, findable and desirable (see Opie et al., 2025).

One reviewer (AV, 25% double assessment by JO) conducted the quality assessment using the Effective Public Health Practice Project’s quality assessment tool for quantitative studies (Effective Public Health Practice Project, 2010) assessing the following six quality assessment domains: 1. selection bias; 2. study design; 3. confounding factors; 4. blinding; 5. data collection methods; 6. withdrawal and drop-outs. Each study was rated with three rating criteria – strong, moderate and weak – for each of these domains. The interrater reliability (IRR) was 82% and disagreements were resolved through conferencing.

After de-duplication, the systematic search of the literature produced 13,963 records. Of these, 13,941 were removed after review of their titles and abstracts (n = 13,786) and full texts (n = 155). The initial screening process for titles and abstracts achieved an IRR of 99% with a kappa statistic (κ) of 0.44 for published works, and an IRR of 99% with a kappa statistic (κ) of 0.62 for unpublished works. The IRR for the full-text double-screening was 98% (κ = 0.68) for published literature and 100% (κ = 1.00) for unpublished literature. Ultimately, 22 primary studies fulfilled all inclusion criteria and were retained for detailed analysis. Each stage of the selection process and reasons for exclusions are shown in Figure 1.

Overall, study quality was weak to moderate (n = 19, 86.36%); of which three were strong quality, 11 were moderate quality, and the remaining of weak quality (n = 8; 36.36%). See Figure 2 for a visual representation of study quality. See Supplementary Material 2 for a tabular representation of study quality and individual ratings for each study, including details on how ratings (strong, moderate, weak) across each domain were assigned.

Included studies predominantly consist of published work (n = 16, 69.6%), three unpublished theses (13.6%), one preprint (4.6%), one trial (4.6%) and one report (4.6%); the majority were quantitative (77.3%, n = 17) while five studies used a mixed method approach. Supplementary Table A1 provides a detailed description of all included studies.

Eight studies included at least one comparison group (36.36%, n = 8), either inactive in the form of a waitlist (Carbone et al., 2021; Krohner, 2022; Rubin et al., 2022; Ziadni et al., 2021) or no-treatment control (Dincer and Inangil, 2021). A combination of active and inactive control groups was used in one study (Rubin et al., 2022) or active control groups (Loveys et al., 2022; McLean et al., 2023). Almost half of the studies originated from the USA (n = 10) followed by Australia and Canada (each, n = 3), New Zealand (n = 2) and Turkey, India, Italy and France (each, n = 1).

Reflecting the diverse methodologies, time points for data collection varied. Half of the studies (n = 11) used a pre- and post-intervention data collection framework. Excluding immediate post-intervention assessments, predominantly short-term outcomes (i.e. ≤4 weeks post) were evaluated (40.90%, n = 9). Longer term outcomes (≥5 weeks post) were only assessed in two studies (Hartley et al., 2023; Ranta et al., 2019) and both short-and longer-term outcomes in one study (4.55%; Ziadni et al., 2021).

Sample sizes ranged from 5 to 3,236 (M = 407.50; SD = 922.60) and mean attrition rate of reporting studies (n = 17/22) was 27.3%. More than two thirds of all participants were female (M_female = 71.54%, n = 18 reporting) and the average age was 41.2 years (range 26.1–65.4). Most studies included individual adult clients (n = 12, 54.55%). The remaining studies also included health-care providers, including nurses and therapists (Arevian et al., 2018; Harris-Lane et al., 2023; Krohner, 2022; MHCC, 2019), nurse clients (9.10%, n = 2; Dincer and Inangil, 2021; Tarquinio et al., 2021), caregivers (Van Orden and Wittink, 2022; Wittenberg et al., 2018), parents (4.55%) (Renauld, 2022) or multiple family members or therapists (Hartley et al., 2023). See Supplementary Material 3 for additional information on the study and sample characteristics (e.g. data collection and recruitment methods, race/ethnicity and sexual orientation).

A total of 22 distinct, brief, at least partially, guided DMHIs for adults were identified (unnamed, 40.9%, n = 9; named, 59.1%, n = 13). Named interventions included: Emotional Freedom Techniques (EFT), Single Session Mindfulness Intervention for Loneliness, Walk-In Together, Beyond Blue Support Service, Therapist over video call Cognitive Behavioral Self-Management (T-CBSM), Flash Technique, Crossroads Children’s Mental Health Center’s Walk-in Clinic, Single-Session Consultation, URGent Eye Movement Desensitization and Reprocessing, Lifeline Text, Empowered Relief, Connect 4 Caregivers and Stepped Care 2.0 E-mental health interventions. Details for intervention characteristics can be found in Supplementary Table A2.

Intervention duration ranged from as brief as 10–15 min (Ganesan et al., 2022) up to 134 min (Tarquinio et al., 2021) with a mean intervention length of 46.67 min (SD= 23.09; n = 12). Most (90.91%, n = 20) offered full human guidance, while one study provided partial human support (4.55%) (Manfield et al., 2021) and another (MHCC, 2019) customized the level of human assistance to match the specific needs of the client. Most delivered guidance in real-time (i.e. synchronous; n = 20; 90.91%), one relied on delayed delivery (i.e. asynchronous; 4.55%) (Williams et al., 2021), and another study provided both for different parts of the intervention (Hawkins et al., 2020).

Interventions included guidance provision by clinical practitioners, including psychiatrists, psychologists, social workers, nurses, or a combination of these (n = 9, 40.9%); students (e.g., Doctoral students, Masters students; n = 6, 27.3%); mental health professionals or students (Ganesan et al., 2022); clinicians or peers (MHCC, 2019); research and academic professionals (Dincer and Inangil, 2021; Manfield et al., 2021; Wittenberg et al., 2018); or trained personnel specialized in various areas: business psychology (Williams et al., 2021) and working with elderly participants (Van Orden and Wittink, 2022).

Fully guided interventions (68.18%, n = 15) included videoconferencing (n = 9), online support (n = 4) or telephone support (n = 3) (Ganesan et al., 2022; Ranta et al., 2019; Wittenberg et al., 2018). Others offered phone and online chat/videoconferencing (Harris-Lane et al., 2023; Hawkins et al., 2020; Renauld, 2022), text messaging (Williams et al., 2021), a webinar (Manfield et al., 2021) and e-mental health tools (online or phone/text) (MHCC, 2019). Interventions were conducted in a participant-selected personal space (n = 17), from home (n = 3) (Krohner, 2022; Renauld, 2022; Tacca, 2022), in a hospital setting (Arevian et al., 2018), or COVID-19 isolation wards (Ganesan et al., 2022).

The majority of studies did not report an underlying theoretical framework or theory of change (72.7%). The remaining studies reported the following theoretical methods: attachment theory (Krohner, 2022), emotion-based models (Krohner, 2022), self-determination theory (Van Orden and Wittink, 2022), self-efficacy theory (Renauld, 2022) and common factors theory (Tacca, 2022). Some studies referred to models of practice, i.e. the cancer family caregiving experience model of stress and coping (Wittenberg et al., 2018), and a recovery-oriented model with a strengths-based approach (MHCC, 2019).

While three studies (13.6%) did not report on the clinical techniques followed, most reported clinical modalities: single-session therapy (n = 2) (Carbone et al., 2021; Hartley et al., 2023), solution-focused therapy (n = 3; (Hawkins et al., 2020; MHCC, 2019; Tacca, 2022), mindfulness-based interventions (Rubin et al., 2022; Ziadni et al., 2021), eye movement desensitization and reprocessing (Manfield et al., 2021; Tarquinio et al., 2021), the miracle question within solution-focused sessions (Sung et al., 2023; Tacca, 2022), cognitive behavioral stress management (Loveys et al., 2022), collaborative problem-solving (Renauld, 2022), dialectical behavior therapy (McLean et al., 2023), cognitive behavioral therapy (Ziadni et al., 2021), collaborative problem-solving (Renauld, 2022), crisis intervention support (Williams et al., 2021), psychoeducation (Carbone et al., 2021), anxiety management (Carbone et al., 2021), EFT (Dincer and Inangil, 2021), psychological first aid model (Ganesan et al., 2022) and psychodynamic interviews (Krohner, 2022). One family-centered approach focusing on family communication and stress was reported (Wittenberg et al., 2018).

Individual intervention delivery was predominant (n = 17, 77.3%), while group delivery accounted for 18.2% of all studies (n = 4) (Dincer and Inangil, 2021; Hartley et al., 2023; Manfield et al., 2021; Ziadni et al., 2021). One (4.5%) intervention offered both individual and group support in a stepped care approach (MHCC, 2019). Among the DMHIs evaluated, only one included multiple family members, while the majority focused on individual perspectives (Hartley et al., 2023). Additional information on the intervention characteristics (e.g., referral sources, and intervention design features such as anonymity and peer support) can be found in the Supplementary Material 4.

Findings were examined in ten socioemotional subcategories, summarized in Table 2 and detailed below. Supplementary Table A3 provides a detailed overview of outcome measures.

Fourteen studies reported on one or more of the following four sub-outcomes related to affect: depression, psychological distress, negative affect and positive affect.

Half of the included studies that assessed depression were RCTs (Krohner, 2022; Rubin et al., 2022; Ziadni et al., 2021), while the other half were within-group designs, measuring pre-to-post changes (Ranta et al., 2019; Sung et al., 2023; Tarquinio et al., 2021). Although there was some variation in the six studies evaluating the efficacy of targeted, brief, guided, single-session DMHIs in the management of depression in adults, overall results suggest a positive effect. Two brief DMHIs (Krohner, 2022; Sung et al., 2023) found non-significant changes (d = 0.02 and d_z= 0.21, respectively) in depression, while two other studies (Rubin et al., 2022; Tarquinio et al., 2021) found a meaningful reduction in depression either 24-hours post completion of a videoconferencing intervention (Tarquinio et al., 2021) (Z = 3.5, p < 0.001) or at one-week follow-up (Rubin et al., 2022: b = −3.01, p not reported) [Tarquinio et al., 2021: Mean difference (MD) = −6.1; p not reported]. Rubin et al. (2022) found this effect only when a compassion DMHI component was included. Ranta et al. (2019) observed a notable decrease in depression rates from 74% at baseline to 42% at 12 months, with a stabilization at 41% at 36 months (p-value not reported), suggesting sustained benefits for their telehealth intervention. Further, Ziadni et al. (2021), while finding no immediate treatment effect, found a notable time effect three-months post completion, when average depression significantly decreased over time for both the intervention group (MD = −3.46) and the waitlist control (MD = −0.52). The more pronounced reduction, however, was observed in the intervention group, especially in the first two months post completion (first month: d = 0.37, second month: d = 0.45).

Four single-arm, pre-post studies found that DMHIs resulted in immediate and longer-term reductions in psychological distress. Hawkins et al. (2020) identified a significant decrease in psychological distress (Mean change = −3.1, t = 10.3) at 28-day follow-up. Manfield et al. (2021) reported substantial reductions in distress after two components of the same intervention (Part #1: Target memory; Part #2: Participants had option to focus on different memory or same target memory from Part #1), with large effect sizes noted immediately after treatment (Hedges’ g = 2.39 and g = 2.49, both p < 0.001) and continued improvement at one-month follow-up (Hedges’ g = 0.46 and g = 0.55, respectively). Williams et al. (2021) observed a significant reduction in psychological distress among help-seekers following a text-based support intervention (MD [95% CI] = 1.90 [1.73, 2.07]). Further, the MHCC (2019) observed a significant gradual improvement in psychological symptoms over 37 days (F = 20.04, p < 0.01, η_p² = 0.24), and a significant linear contrast (F = 58.32, p < 0.01, η_p² = 0.23), indicating a steady improvement in psychological wellbeing over the intervention period. The consistent findings from these studies suggest that DMHIs are effective in reducing psychological distress, with three out of four showing sustained benefits beyond the initial treatment period.

Five studies assessed the impact of DMHIs on negative affect. Two RCTs (Carbone et al., 2021; McLean et al., 2023) assessed negative affect using the positive and negative affect schedule (Terracciano et al., 2003). Carbone et al. (2021) found a significant decrease in negative affect post-intervention compared to the wait-listed group and the original condition group (F = 29.45, d = 0.92, p < 0.001). Similarly, McLean et al. (2023) reported a significant reduction in negative affect over time across all participants (F = 346.24, p < 0.001, η_p² = 0.69); however, there were no significant interaction effects between time and condition (all p > 0.283), indicating that the decrease over time was independent from the intervention. In a three-arm study, Loveys et al. (2022) also reported significantly lower negative affect at two weeks follow-up compared to baseline across all conditions (M = 3.89, SE = 0.44, 95% CI [3.01, 4.78] vs M = 5.79, SE = 0.43, 95%CI [4.92, 6.65]). However, the main effect of condition (i.e. both virtual human [VH-CBSM] and human teletherapist over video call [T-CBSM] compared to a self-guided e-manual [E-CBSM]) was again not significant (F = 0.49, p > 0.05, η_p² = 0.03).

Two single-arm, pre-post studies again yielded mixed results. Tarquinio et al. (2021) reported significant reductions in fears related to safety at work at 24 h post intervention (M = 5.52, Z = 3.40, p < 0.001), with improvements sustained at one-week follow-up (M = 6.11, p < 0.05). In addition, following the intervention, they observed a significant decrease in fear associated with going to work at 24 h post intervention (M = 5.80, Z = 3.53, p < 0.001), with improvements sustained at one-week follow-up (M = 5.40, p < 0.05). However, Sung et al. (2023) identified a significant increase in hopelessness post-intervention (d_z = 0.91, p < 0.001). This finding contrasts with a trend observed in the other studies, where general decreases in negative affect were found over time but unlinked to the intervention.

Assessment of positive affect and optimism resulted in mixed findings. Carbone et al. (2021) found in an RCT no significant group effect on positive affect (d = 0.09), indicating that the intervention did not significantly alter participants’ positive affect. Another RCT by McLean et al. (2023) also found no significant effect of time on positive affect and found no interaction effect between time and condition (effect sizes not reported), indicating that the intervention did not differentially affect positive affect. However, within the “Change” condition (intervention focused on change of emotions), those with higher PTSD symptoms experienced a significant increase in positive affect compared to those with fewer symptoms (b = 0.24, p = 0.044) (McLean et al., 2023).

Of the two pre-post study designs with active control conditions, Loveys et al. (2022) observed a large significant main effect of time on optimism (F = 10.78, p = 0.003, η_p² = 0.26), with participants showing an increase in optimism at a two-week follow-up compared to baseline. However, there was no significant interaction between condition and time (F = 0.22, p = 0.807, η_p² = 0.01), suggesting that the increase in optimism occurred independently from the intervention. Similarly, one study (Tacca, 2022) found an increase in positive mood for both the comparator (increase in score = 1.20) and intervention group (increase in score = 1.00) from pre- to post-session. No significant group difference (F = 0.246; p = 0.623) was found, but again a significant main effect revealed an overall increase in positivity for all study participants (F = 29.824, p < 0.001).

Eight studies reported significant reductions in anxiety levels across various brief interventions and different anxiety assessment instruments. In a two-arm observational study, Ganesan et al. (2022) observed a significant decrease in anxiety after telecounselling within the intervention compared to the control group (η_p² = 0.48). Of the two single-arm pre-post (within-group) studies, Sung et al. (2023) reported a significant reduction in anxiety symptoms at two-week follow-up (d = 0.40, p = 0.04), while Tarquinio et al. (2021) observed a significant decrease in anxiety 24-hours post-intervention (Z = 3.63, p < 0.001). The decrease in anxiety remained stable for up to one-week post-intervention (effect not reported).

Five studies measuring anxiety were RCTs (Carbone et al., 2021; Dincer and Inangil, 2021; Krohner, 2022; Rubin et al., 2022; Ziadni et al., 2021). Dincer and Inangil (2021) reported reductions on state anxiety, with a significant decrease in the intervention (F = 19.13, p < 0.001) but not in the control group (F = 1.00, p > 0.05). Krohner (2022) found a medium sized significant reduction in general anxiety from baseline to six-week follow-up in the intervention group (p < 0.001, d = −0.41). In contrast, no change was found for the waitlist control group (p = 0.638, d = −0.06). A significant interaction between group and time confirmed that changes in anxiety levels differed significantly (F = 3.65, p = 0.029) between intervention and waitlist control.

Krohner (2022) found a moderately sized significant reduction in pain-related anxiety in the treatment group at follow-up (p = 0.002, Cohen’s d = −0.36) but no reduction in the waitlist control (p = 0.840, Cohen’s d = −0.02). Again, this difference between groups over time has been confirmed by a significant interaction of time and condition (F = 6.02, p = 0.008).

Carbone et al. (2021) found that compared to a waitlist control, the online counselling intervention resulted in significant decreases in state anxiety (p < 0.001, d = 0.49) post-intervention. Rubin et al. (2022) found that adding a compassion component to their brief intervention led to meaningful reductions in anxiety within a week (b = −3.79). Finally, Ziadni et al. (2021) found that at three-months follow-up the intervention group reported a significantly greater reduction in anxiety (−4.71, 7.91% reduction) than the waitlist control group (−0.42, 0.7% reduction) with a notably steeper decline (p = 0.009, d = 0.59).

Four studies assessed stress as an outcome. Dincer and Inangil (2021) found in their RCT that participants of the treatment group experienced a significant decrease in stress (Z = 16.58, p < 0.001), but not for the control group (Z = 0.286, p > 0.05). Dincer and Inangil (2021) also identified a significant decline in burnout in the treatment group (Z = 5.25, p < 0.001), but again, not for the control group (Z = 1.405, p > 0.05). Similarly, in a three-arm pre-post study, Loveys et al. (2022) reported reduced stress across various measures, including the visual analogue stress scale (p < 0.001, F = 43.14, η_p² = 0.74), perceived stress scale (PSS; F = 30.43, p < 0.001, η_p² = 0.50), as well as physiological indicators such as electrodermal activity (p = 0.003), and skin temperature (F = 8.49, p < 0.001, η_p² = 0.22). While these generalized stress reductions were significant over time, they did not differ between intervention and control group (p’s > 0.05, PSSη_p² = 0.04, electrodermal activity: η_p² = 0.17). In another RCT, Rubin et al. (2022) found the inclusion of a compassion component in their brief DMHI resulted in a meaningful reduction in perceived stress at one-week follow-up (b = −3.75, p-value not reported) compared to the control group. In a pre-post (within-group) study, Wittenberg et al. (2018) found for three of their four included cohorts (Cohort 1 Diagnosis: Pre M = 5.6, Post M = 2.8; Cohort 2 Treatment: Pre M = 4.2, Post M = 3.6; Cohort 3 Survivorship: Pre M = 2.4, Post M = 1.6) a decrease in stress values. However, this general trend was not found for caregivers in the End-of-Life cohort (Cohort 4, n = 5/20, 25% of total study sample), who reported an increase in stress, with the average score rising from 4.2 out of 10 at baseline to 4.8 at one-month follow-up.

Findings indicate that while DMHIs can be effective in managing stress, the degree and durability of the effect may depend on the type of intervention and/or the measurement tools used.

Brief, at least partially, guided DMHIs targeting loneliness and social connection ensued mixed results across two RCTs (Rubin et al., 2022; Ziadni et al., 2021) and two single-arm pre-post designs (Van Orden and Wittink, 2022; Williams et al., 2021). Williams et al. (2021) found a significant improvement in social connectedness following a text-based intervention (MD [95% CI] = −0.95 [−1.11, −0.79]). Similarly, Ziadni et al. (2021) observed a significant reduction in social isolation over time for participants in both groups, the “Empowered Relief” Zoom platform and the waitlist control (effect size not reported, p = 0.02). When testing the interaction between condition and time a significant reduction (effect size not reported, p = 0.02) in social isolation was only found in the intervention group.

Rubin et al. (2022) found a decrease in loneliness at a two-week follow-up (b = −2.36) but no difference between the intervention and the waitlist control group at the one-week follow-up, indicating no immediate intervention effects on loneliness. Concerningly, Van Orden and Wittink (2022), found an increase in loneliness from pre- to post-intervention with mean scores rising from 49.6 to 53.4. However, as no p-value or effect sizes were reported, the significance of this finding is unclear. Overall, results are mixed as some interventions reduced loneliness and social isolation, while others found no significant change.

Two RCTs (Krohner, 2022; Ziadni et al., 2021) found no significant impact of their interventions on anger. Krohner (2022) found that the intervention only resulted in a nonsignificant reduction in anger over time (p = 0.166, d = −0.16) and no change in the control condition (p = 0.902, d = 0.02); confirmed by a nonsignificant interaction between condition and time (p = 0.151, η_p² = 0.01). Similarly, Ziadni et al. (2021) reported no significant time (MD: intervention = 4.10; control = 0.52) or interaction effects on anger, indicating no significant change over the course of the study (p > 0.05, effect size not reported).

The two studies assessing well-being, an RCT (Carbone et al., 2021) and a single-arm pre-post study (MHCC, 2019), found mixed results. The MHCC (2019) found significant improvement in well-being over 37 days (F = 19.2, p < 0.01, η_p² = 0.23), with a strong linear trend indicating sustained improvement (F = 57.21, p < 0.01, η_p² = 0.23). Similarly, global mental health showed a gradual increase over the same time (F = 21.34, p < 0.01, η_p² = 0.25), again with a significant linear trend (F = 62.30, p < 0.01, η_p² = 0.25). Life function followed a similar pattern, with a steady improvement noted (F = 6.44, p < 0.01, η_p² = 0.09) and a significant linear contrast (F = 7.22, p < 0.01, η_p² = 0.09), suggesting a positive trajectory in participants’ daily functioning as a result of the intervention. However, Carbone et al. (2021) found no significant main effect of their intervention on well-being (F = 0.29; p = 0.593, d = 0.17). These contrasting results indicate that the impact of interventions on well-being may either vary depending on the measures used or the nature of the intervention, indicating a need for further research to understand these differences.

Seven studies suggested an increase in the development of knowledge and skills following DMHI participation. Five studies were single-arm, pre-post studies (Hawkins et al., 2020; Renauld, 2022, Van Orden and Wittink, 2022; Williams et al., 2021; Wittenberg et al., 2018). Renauld’s (2022) qualitative interviews highlighted the effectiveness of virtual SST in enhancing parental understanding and insight into their child’s mental health. Parents not only gained knowledge but also acquired practical strategies to support their child’s needs. Wittenberg et al. (2018) found that caregiver confidence in communication increased following their DMHI, with caregivers’ communication confidence improving with patients (Pre M = 7.7, Post M = 8.6) and health-care providers (Pre M = 7.1–8.3, Post M = 9.2–9.5). A general decrease in confidence was found in their Survivorship and End-of-Life cohorts relating to caregiver communication with patients (Survivorship: Pre M = 9.3, Post M = 9.2; End-of-Life: Pre M = 8.1; Post M = 7.2) and health-care providers (Survivorship: Pre M = 9.0, Post M = 8.0; End-of-Life: Pre M = 7.8; Post M = 8.8). Similarly, Van Orden and Wittink (2022) found a modest increase in participants’ perceived knowledge of social engagement immediately post-intervention [Pre M(SD) = 14.8 (2.59)] and at two-week follow-up [post M(SD) = 15.6 (2.3), effect size not reported].

Hawkins et al. (2020) identified marked improvements in coping abilities following their brief DMHI (all p’s < 0.05; immediately post-intervention (Mean change = 1.7, t = −5.3), three-day follow-up (Mean change = 1.1, t = −3.1) and one-month follow-up (Mean change = 0.9, t = −2.4), signifying a rapid enhancement in participants’ stress management capabilities. Similarly, Williams et al. (2021) found a significant increase in the confidence of participants in their ability to cope post-intervention (MD [95% CI] = −0.87 [−1.04, −0.71]).

In an RCT, Krohner (2022) found post-intervention that most participants (84.6%) reported high ratings on knowledge or skill gain outcomes (≥ 4 on 1–5 scale), like gaining insights (94.7%), understanding their emotions (89.5%) and gaining new symptom knowledge (68%).

In a three-arm pre-post study, Loveys et al. (2022) found a moderate sized increase on the awareness of tension skills (F = 3.06, p = 0.090, η_p² = 0.09) at two-week follow-up across all conditions. However, no significant condition effect was found for perceived awareness of tension skills (F = 0.22, p = 0.802, η_p² = 0.01) or relaxation skills (F = 0.24, p = 0.787, η_p² = 0.02), further supported by the absence of a significant interaction of condition and time on both measures (perceived awareness of tension skills, F = 0.20, p = 0.816, η_p² = 0.01; relaxation skills, F = 1.08, p = 0.353, η_p² = 0.07).

This review provides a preliminary exploration of potential associations and trends between intervention elements and socioemotional outcomes. Results were separated by ten socioemotional outcomes presented above (see Supplementary Table A4). Due to limited data available, associations were reported when at least two studies yielded the same result. Therefore, an identified element may still be effective even if they are associated with no outcome improvement in this review. If mixed results were identified, they were reported under elements common to interventions with inconsistent efficacy. It is important to note that only six of the 22 included studies used RCT designs to assess outcomes (Carbone et al., 2021; Dincer and Inangil, 2021; Krohner, 2022; McLean et al., 2023; Rubin et al., 2022; Ziadni et al., 2021).

Content personalization was associated with increases in socioemotional outcomes in brief, at least partially, guided adult DMHIs. DMHIs incorporating personalized elements more often reported decreases in depression (Rubin et al., 2022; Ziadni et al., 2021), psychological distress (MHCC, 2019; Williams et al., 2021), anxiety (Rubin et al., 2022; Sung et al., 2023), stress (Rubin et al., 2022; Wittenberg et al., 2018), loneliness (Rubin et al., 2022, Van Orden and Wittink, 2022; Williams et al., 2021; Ziadni et al., 2021) and knowledge/skill development (Van Orden and Wittink, 2022; Williams et al., 2021; Wittenberg et al., 2018).

DMHIs including safety/stabilization elements (Ganesan et al., 2022; Tarquinio et al., 2021) and action planning elements (Sung et al., 2023; Ziadni et al., 2021) reported effects on anxiety, while co-designed interventions (MHCC, 2019; Williams et al., 2021) and solution-focused therapy (Hawkins et al., 2020; MHCC, 2019) were features of DMHIs associated with psychological distress. DMHIs including mindfulness were found to reduce depression (Rubin et al., 2022; Ziadni et al., 2021), loneliness (Rubin et al., 2022; Ziadni et al., 2021) and stress (Loveys et al., 2022; Rubin et al., 2022).

DMHIs incorporating various cognitive-behavioral therapy approaches, including dialectical behavior therapy and cognitive-behavioral stress management, reported reductions in negative affect (Loveys et al., 2022; McLean et al., 2023), while also enhancing positive affect, as reported in the same studies. Anonymity was an element of DMHIs that effectively addressed psychological distress, knowledge and skill development, including skills to cope with challenges (Hawkins et al., 2020; Williams et al., 2021). Referral information, particularly follow-up support, has been an element of DMHIs that effectively increased knowledge and skill development, as well as enhanced coping abilities (Hawkins et al., 2020; Krohner, 2022; Williams et al., 2021).

DMHIs offering anonymity were associated with reduced psychological distress (Hawkins et al., 2020; MHCC, 2019) and gains in the ability to cope (Hawkins et al., 2020; Williams et al., 2021), indicating that the option to remain anonymous may benefit adult clients engaging with these interventions. Similar results were found for the limited number of co-designed DMHIs (n = 2), where DMHIs including this element resulted in reduced psychological distress at follow-up (MHCC, 2019; Williams et al., 2021). Few DMHIs included information on contact referral and emphasized follow-up support, but for the majority (60%, n = 3/5) an enhanced knowledge and skill development as well as bolstering coping abilities was found (Hawkins et al., 2020; Krohner, 2022; Williams et al., 2021). This underscores a potential benefit of including reengagement features in mental health interventions.

DMHIs using a standardized (or manualized) approach were associated with poor or unconfirmed efficacy, specifically in managing anger (Krohner, 2022; Ziadni et al., 2021).

DMHIs including goal setting were found to be effective in managing anxiety (Krohner, 2022; Sung et al., 2023), but not depression (Krohner, 2022; Sung et al., 2023). DMHIs including aspects of psychoeducation were found effective for various outcomes, including depression (Rubin et al., 2022; Tarquinio et al., 2021), psychological distress (Manfield et al., 2021; MHCC, 2019), anxiety (Carbone et al., 2021; Ganesan et al., 2022; Rubin et al., 2022; Sung et al., 2023; Tarquinio et al., 2021), stress (Loveys et al., 2022; Rubin et al., 2022) and in enhancing knowledge and skill development (Krohner, 2022, VAOrden and Wittink, 2022). However, these DMHIs did not result in significant changes in positive affect (Carbone et al., 2021; Loveys et al., 2022; McLean et al., 2023) and revealed mixed results for negative affect and loneliness, with some studies reporting high efficacy (Carbone et al., 2021; Loveys et al., 2022; Sung et al., 2023; Tarquinio et al., 2021; Rubin et al., 2022) while others reported poor efficacy (Loveys et al., 2022; McLean et al., 2023, Van Orden and Wittink, 2022).

A narrative synthesis of other intervention characteristics (e.g. guidance type, provider/facilitator, length, theoretical framework, clinical modality/approach, target demographic) and associated socioemotional outcomes can be found in Supplementary Material 5.

This review evaluated the effectiveness of fully or partially guided adult-focused (≥25 years old) brief DMHIs (≤3 h) on ten socioemotional outcomes. This review provides first indications on whether specific elements used in DMHIs were associated with positive socioemotional outcomes in the adult population.

Across the 22 studies included in this review, findings on short-term changes were reported for six of the ten socioemotional outcomes: anxiety, stress, depression, psychological distress, negative affect and knowledge/skill development. This is consistent with prior systematic reviews in adults that reported that in-person brief interventions were associated with short-term improvements in depression and anxiety in indicated (Aafjes-van Doorn and Sweeney, 2019; Kim et al., 2023) and selective populations (Bertuzzi et al., 2021). It further aligns with recent meta-analytic findings on universal, self-guided DMHIs (M_length = 9.28 weeks) targeting the adult parent population (Opie et al., 2023). Given this growing body of evidence and with enhanced accessibility, lower cost and flexibility of use, DMHIs form a promising addition to interventive mental health support strategies (Montagni et al., 2020; Witteveen et al., 2022).

Results revealed a short-term mixed efficacy for loneliness and well-being outcomes for brief DMHI participation. Previous systematic reviews examining digital technologies in adult populations also found inconsistent results for loneliness (Gunnes et al., 2024; Shah et al., 2021) as well as wellbeing (Harrer et al., 2019; Lattie et al., 2019). These mixed findings have previously been attributed to intervention and outcomes heterogeneity (Gunnes et al., 2024; Shah et al., 2021), which is also a likely explanation for the findings of this review.

Furthermore, results reveal that adult-focused, brief, partially guided DMHIs seem to be less effective or ineffective for anger management or the promotion of positive affect when assessed short-term. Prior research suggests that interventions addressing anger might be more effective if the primary intervention strategy incorporates cognitive behavioral elements or relaxation (i.e. affect modification) techniques (Conley et al., 2017), both elements not considered in the current set of DMHIs aiming to reduce anger. Similarly, to address positive affect, prior research has emphasized incorporating positive psychology techniques to directly target changes in positive emotions and achieve meaningful improvement in positive affect (Bolier et al., 2013; Mira et al., 2018), again techniques that the considered DMHIs did not implement. The absence of these previously identified intervention approaches in the set of included DMHIs may explain why the present study failed to find evidence on the effectiveness of adult-focused, brief, partially guided DMHIs on anger or positive affect.

Evidence on the long-term socioemotional effectiveness of brief DMHIs is less clear. The present review identified only two studies (Ranta et al., 2019; Ziadni et al., 2021) that conducted follow-up assessments extending three months or longer, offering preliminary evidence for sustained improvements in depression, anxiety and loneliness post-DMHI participation. Beyond this lack of data examining long-term effects, the reviewed DMHIs uniformly omitted options for client re-engagement, which may highlight another area for potential enhancement. Due to the dynamic nature of mental health, which often varies in response to life’s changes, facilitated and simple re-engagement processes with DMHIs seem a key factor in preventing re-escalation of symptoms (Graham et al., 2020).

Overall, the scarcity of data on DMHIs and long-term assessment of intervention outcomes represents a significant gap in understanding the prolonged impact of the here examined brief adult DMHIs. Together with the often-missing description of re-engagement options makes the assessment of long-term impacts of adult, brief, at least partially guided DMHIs difficult. Similar gaps have previously been noted in systematic reviews and meta-analyses examining DMHIs in diverse populations, including parents of children aged 0–5 years (Opie et al., 2023), youth with emerging mental health symptoms (Opie et al., 2024a) and employees struggling with work-related mental health concerns (Moe-Byrne et al., 2022). Thus, the here presented findings endorse prior calls for increased examination of long-term effects of brief interventions (Bertuzzi et al., 2021; Currie et al., 2022).

For affect-related presentations, the few controlled studies examining longer term effects offered an important indication that affect-related improvements may also occur with the simple passing of time as they have been found in both intervention and nonintervention groups (Ziadni et al., 2021). Thus, future studies should examine the relationship between passage of time itself and its interaction with intervention type in more detail. This will help clarify patterns of fluctuation and the overall trajectory of a presentation, including its progression and potential lasting impacts.

Customizing or personalizing the content of brief DMHIs seems to be a stronger predictor for DMHIs that successfully reduced socioemotional burden than manualized, automated content. Customization has been a key element in DHMIs that successfully reduced depression, anxiety, stress, loneliness and psychological distress, and for fostering the development of knowledge/skills. Tailored content on mental health safety and stabilization and action planning was part of DMHIs that resulted in reduced anxiety. On the contrary, DMHIs including standardized or manualized content were found to be less effective in addressing anger and showed mixed results for various other socioemotional outcomes (i.e. depression, psychological distress, anxiety, loneliness, stress, knowledge/skill development, positive affect and negative affect).

These findings align with those of previous systematic reviews across diverse populations (Opie et al., 2024a; Moe-Byrne et al., 2022), outlining the significance of tailoring DMHI content to meet individual needs. Tailoring a DMHI to the user increases the likelihood that users feel understood, while at the same time mirroring the traditional client-clinician therapeutic relationship more closely, one of the greatest predictors of positive mental health improvement (Valentine et al., 2022). Such evidence critiques the “one-size-fits-all” model, highlighting the critical need for interventions that are flexible and adaptable to the unique needs of individuals, and that at least in some parts contain a human response.

Based on the 22 studies examined in this review, some areas successfully applied and integrated in DMHIs in other populations were underexplored in the here presented adult-focused DMHIs. These aspects will be presented in the following.

In the present review, only two interventions were co-designed by the target population and/or health-care providers. These co-designed interventions were effective in reducing psychological distress (MHCC, 2019; Williams et al., 2021). A lack of co-designed interventions has previously been highlighted by Vial et al. (2022) and seems to be a persistent trend, despite the known benefits of co-design approaches (Bevan Jones et al., 2020; Sanz et al., 2021; Thabrew et al., 2018). None of the included studies, to our knowledge, revised the DMHI based on end-user feedback. This approach has previously been recommended (Bond et al., 2023) to align interventions more closely with user and provider needs.

Referral information was only given in about one third of the DMHIs examined, but the results of these revealed boosted knowledge/skill development. This finding aligns with prior literature wherein connecting clients to other services was perceived as useful in providing hope and a positive way forward (Doran et al., 2021). The value of providing alternative entry points and well-coordinated referral mechanisms has also been observed in mental health and primary care services (Lasser et al., 2021; Valibhoy et al., 2017). Relatedly, clients valued practitioners who coordinated next-steps and are proactive in their provision of care (Doran et al., 2021; Lasser et al., 2021).

Despite its established benefits including privacy and in encouraging open participation, as highlighted in other studies (Borghouts et al., 2021; Lehtimaki et al., 2021), anonymity was offered in only four out of 22 studies. In this review, anonymity was associated with reduced psychological distress and enhanced knowledge and coping skills. It is known that stigma and worries around cybersecurity can prevent adults from engaging with DMHIs (Dworschak et al., 2024; Pywell et al., 2020; Opie et al., 2024b) and that incorporating anonymity into these interventions may address these deterrents (Pywell et al., 2020). Thus, the infrequent provision of anonymity in adult, brief DMHIs may have presented a missed opportunity to enhance both engagement and effectiveness by overcoming known obstacles to participation.

Key strengths of the present review include its comprehensive search strategy, inclusion of both published and unpublished literature, and codeveloped with an industry stakeholder, Beyond Blue, as well as the involvement of a consumer academic with lived experience at each phase of the review process.

The diversity of study populations is limited, with most of the included research being conducted in the US and lacking representation from linguistically and culturally diverse groups. There is scant research in vulnerable populations such as older adults, who are more prone to being excluded from DMHIs, e.g. due to a lack of skill or literacy in the use of technology-delivered interventions (Seifert et al., 2019). In addition, females were overrepresented in the included studies. Parts of the gender imbalance may be attributed to men’s lack of engagement with digital health technologies, reflecting their needs and preferences being overlooked in the design process and a lack of attention being paid to what they want to see in DMHIs (Opozda et al., 2024). Further, as males experience greater societal stigma around help-seeking and more negative experiences with health professionals when they do reach out for help (Macdonald et al., 2022), this may discourage participation in intervention research studies.

Beyond this lack of diversity in adult populations, most studies included in this review assessed intervention efficacy immediately after completion and again one week later, potentially neglecting the long-term effects of these interventions. In addition, there is a pronounced emphasis on the socioemotional outcomes of depression and anxiety, likely reflective of their prevalence, but may result in a gap in examining other socioemotional outcomes (e.g., externalizing outcomes such as aggression and conduct problems). Moreover, the mean duration of these 'brief’ DMHIs exceeded 45 min (M_length=46.67 min), with some taking over 60 min (Hartley et al., 2023; Krohner, 2022; Loveys et al., 2022; MHCC, 2019), while a few exceeded two hours (Tarquinio et al., 2021; Ziadni et al., 2021). This may have had an impact on participants’ engagement as briefer DMHIs might be perceived as less onerous and thus, more engaging. As a final limitation, no study explicitly reported their intervention as being “trauma informed,” highlighting a gap in the current brief, adult DMHI literature.

The search strategy of the present study did not cover all mental health symptoms, with personality disorder symptoms and psychotic disorder symptoms omitted as we were looking at the emerging early signs of ill-mental health (“indicated population”) and these are more pronounced and often require more intense treatment. To reduce heterogeneity, qualitative research was also excluded, which may have resulted in the omission of some data. Further, it is possible that not all unpublished literature was identified through the grey search strategy applied. The restriction to contemporary literature (publications from 2018 to March 2024) may have also resulted in the exclusion of older, possibly relevant data. Finally, the heterogeneity in the present review (e.g., sample/intervention characteristics, measurement instruments and study design) limits the ability to draw conclusions about who might benefit most from such interventions and under what circumstances.

This review highlights several key areas for future research and practical application as outlined in the recommendations below.

1. Content personalization: Aligning the DMHI content and delivery with the service user’s expressed goals/needs of care is important. Further, tailoring interventions to diverse populations and embedding culturally sensitive content both honors the context of the individual and their psychological distress and mitigates potential social disadvantage. Standardized and manualized content warrants further examination considering its mixed results and inefficacy in some areas, for example anger management. Given this, we recommend a cautious and considered approach toward reliance on standardized and manualized content within future DMHIs.

2. Study designs of future research: Using robust controlled study designs (e.g., randomized controlled trials) is crucial for progressing the field’s knowledge. Direct comparative studies between brief and traditional, longer-term therapeutic approaches are essential for a comprehensive understanding of DMHI effectiveness. Longitudinal studies examining outcomes up to one-year post-intervention is also needed. Within this, examining re-engagement and the impact of accruing support over time will also help identify if and when additional support might bolster initial effects. Moreover, using the same instruments and assessment periods across different studies examining the same outcomes would significantly improve the consistency and comparability of research findings.

3. Evaluating promising yet underused DMHI elements:

   • Co-design: Involving a diverse group of stakeholders (e.g., current and former clients, practitioners and client support systems) in developing and revising DMHIs is called for. Beyond contrasting the endorsement of views across stakeholder groups, meaningful, culturally safe co-design processes mitigate against potential tokenism and power imbalances.

   • Anonymity: The impact of anonymity in DMHIs on user engagement and intervention effectiveness remain unclear. Investigating potential challenges and solutions related to anonymity, including cybersecurity and user safety, will be crucial. This will provide a clearer understanding of how anonymity can be leveraged to maximize the benefits of DMHIs for adults facing ill-mental health.

   • Assessing socioemotional efficacy in diverse populations: Previous studies included in the present review predominantly focused on samples that are white, heterosexual, female and from Western locations. To ensure the relevance and applicability of DMHIs across different cultural and demographic contexts, the inclusion of more diverse groups is key. This includes men, First Nations communities, and LGBTQIA+ individuals.

4. The examination of additional outcomes:

   • Externalizing socioemotional outcomes: The scarcity of data on socioemotional externalizing issues indicates a significant research gap. Further, considering that men of all ages typically exhibit higher rates of externalizing behaviors (Beauchaine and Hinshaw, 2015; Smith et al., 2016), and most study participants of this review were female, this gender imbalance raises questions about the generalizability of study findings to males. Further, the examination of a broader array of internalizing socioemotional outcomes (e.g., self-efficacy, quality of life) is required.

5. Should additional support be required:

   • Referrals: Incorporating high-quality referral sources and mechanisms may significantly enhance the capacity of DMHIs to offer a comprehensive care strategy that extends beyond the immediate offerings offered by brief DMHIs.

   • Brief DMHIs as a gateway into longer-term support: Brief DMHIs can also serve as an initial step, providing a gateway to longer-term treatment options or facilitating referrals to other appropriate services. This role positions brief DMHIs as a critical screening point for entry into a broader mental health service system.

   • Developmentally suitable: Adulthood covers a broad age spectrum, making it essential to adapt DMHI content for different life stages. By applying a life course developmental perspective, DMHI content could be customized for various phases of adulthood, such as early adulthood compared to later, more senior years. This customization ensures that DMHIs remain relevant, user friendly, engaging and effective throughout an individual’s adult life.

   • Trauma-informed practices: No identified DMHI in the present review reported their intervention as being “trauma-informed,” although tailoring safety and stabilization protocols were used in some. There is a need for research to align with clinical insights more closely and overtly on trauma-informed practices, as well-documented in victimization, trauma and long-term treatment literature (e.g., Bendall et al., 2020; Ting and McLachlan, 2023).

This review represents an important move toward establishing the evidence base for digitally focused mental health services for adults. It provides a comprehensive examination of brief (partially) guided DMHIs for adults with emerging symptoms. The here presented findings to date have been directly embedded into the revised 2024 Support Services Model of Care of our practice partner Beyond Blue. Collectively, findings offer important practice implications elsewhere and will inform research directions in the development, implementation and evaluation of future adult DMHIs. These are all necessary steps to allow for continuous service improvement and improved mental health preventative support services for adults, now and into the future.

This research was supported by Beyond Blue. The authors would like to thank stakeholders at Beyond Blue for their conceptual and contextual contributions.

All data reported and analyzed in this study are included in this published article and the supplementary materials.

Beyond Blue funded this study. Beyond Blue staff contributed to the conceptualization of the study aims but had no influence on the data selection and analyses nor the results or outcomes of the study.

This review was conducted with funding from Beyond Blue.

An Vuong and Jessica E. Opie both contributed equally as the first authors of this work.

• Supplementary material 1: Complete search strategy for published and unpublished databases

• Supplementary material 2: Tabular presentation of study quality

• Supplementary material 3: Additional study and sample characteristics

• Supplementary material 4: Additional intervention characteristics

• Supplementary material 5: Narrative synthesis of intervention characteristics (e.g. guidance type and provider, length, target demographic) and associated socioemotional outcomes

• Supplementary Table A1: Key characteristics of included studies

• Supplementary Table A2: Intervention characteristics

• Supplementary Table A3: Socioemotional outcomes

• Supplementary Table A4: Common intervention elements and associated socioemotional outcomes