Digital Health Interventions for Depression and Anxiety in Low- and Middle-Income Countries: Rapid Scoping Review

Introduction

Background

Mental illnesses are a leading contributor to the global burden of disease [1]. The COVID-19 pandemic has contributed to further deteriorations in population mental health and exposed critical and persisting gaps in mental health care [2,3]. The Global Burden of Disease 2020 study estimated that the COVID-19 pandemic has led to increases of 28% in major depressive disorders and 26% in anxiety disorders [4], 2 of the most common mental disorders [5]. Depression and anxiety disorders are associated with profound disease and socioeconomic burden to individuals and their relatives [6]. Low- and middle-income countries (LMICs; countries with a gross national income per capita of ≤US $14,005 [7]), which bear a larger proportion of the global mental illness burden [8], have been disproportionately impacted due to preexisting mental health care system deficiencies [8,9]. Increased mental health issues have, in turn, placed additional substantial pressures on strained mental health care systems [2]. This is particularly concerning in LMICs, where 75% of people lacked access to mental health care before the COVID-19 pandemic [10]. Persistent and growing barriers to treatment in LMICs include low mental health awareness, stigma that impacts help seeking, a shortage and maldistribution of mental health specialists, and care that is concentrated in specialized and tertiary settings with little to nonexisting care at the community level [11,12].

Worldwide, the pandemic has also impacted the delivery of mental health care, leading to a considerable increase in care delivered through digital health interventions to address the widespread reduction of in-person care imposed by public health constraints. These include interventions delivered via telephone, videoconferencing, messaging, web-based programs, and smartphone or other mobile device apps [13], collectively referred to herein as digital mental health interventions (DMHIs). Many of these DMHIs target depression or anxiety. For example, there has been a considerable increase in smartphone apps since the onset of COVID-19. Estimates indicate that >10,000 mental health–related apps are available [14], with most focused on depression and anxiety [6,15]. Digital interventions for depression and anxiety can help with health promotion and literacy at the population level, prevention by targeting high-risk individuals or subpopulations, and treatment and preventive services for people with mental health conditions [16].

Active components of these DMHIs are often based on evidence-based psychological interventions developed for in-person delivery [17]. While it is critical that the interventions are implemented with high fidelity, there is limited evidence on which considerations are most relevant and what should be included in guidance for fidelity of digital adaptations from evidence-based in-person formats. There is a need to evaluate if and how the theoretical and technical components are applied in the new modus or can be enhanced by drawing on the unique interactive features of a digital format (eg, videos and animations) [18]. There has also been limited attention paid to methods for adaptation and evaluations of adaptations across contexts to ensure cultural appropriateness for relevance; applicability; and, ultimately, effectiveness [19-21], particularly for the adaptation of digital interventions [22]. Cultural adaptation can be defined as “the process of systematically modifying an evidence-based intervention to be congruent with the language, values, beliefs, and context corresponding to clients’ cultural backgrounds” [23].

While DMHIs have been used for 2 decades, their development and uptake has been greatly accelerated by the COVID-19 pandemic and hold promise as a feasible way to address the growing mental health needs of populations amidst the persistent and growing mental health treatment gap [12]. This might have been a “black swan” moment as described by Wind et al [24], acting as a catalyst in the marked shift from in-person mental health care provision to digital health. Challenges that have slowed their implementation into routine care include a lack of acceptance by mental health professionals, who may be concerned about threats to the therapeutic alliance, which they believe can only be established in person [25,26]. Other challenges include digital and language literacy [27]; data privacy and security concerns [28], especially as LMICs have seen an increasing number of cyberattacks [29]; and structural issues such as underdeveloped technological infrastructure [12]. There is a need to ensure that all those who need digital mental health care, especially the traditionally underserved groups such as those who live in rural and remote areas, with low income, and from culturally diverse communities, have the digital literacy and opportunity to engage with DMHIs to reduce the digital divide [2]. Another key concern for implementation is usability, the extent to which a product can be used easily, efficiently, and with satisfaction to achieve specified goals [30]. Usability has been identified as a prerequisite for the success of mobile apps [31]. Conversely, poor usability is a primary reason for low user engagement, limiting clinical utility and resulting in the failed adoption of DMHIs [32,33].

Rationale and Objectives of This Review

Active components of these DMHIs are often adapted from in-person formats [17], most commonly developed and tested in high-income countries [12]. Despite this, there is a paucity of evidence on their fidelity to the original format and limited examination of the cultural adaptation process. Furthermore, although there has been increasing research on DMHIs from LMICs [34,35], where >85% of the global population lives [36], most DMHIs have been developed and studied in high-income countries [12,37], leading to potential biases and limiting generalizability. Knowledge gathered from LMICs will be critical for these countries. Thus, the purpose of this research was to conduct a rapid scoping review of the available evidence on digital health interventions for depression and anxiety, the 2 most common mental disorders, in LMICs.

The specific objectives of this rapid scoping review were to: (1) identify the types of digital health interventions for depression and anxiety being used and studied in LMICs; (2) explore their fidelity, usability, and cultural adaptation to a digital format in LMICs; and (3) examine the challenges to and facilitators of their uptake in LMICs.

This rapid scoping review contributes important evidence on the growing adaptation, use of, and challenges to DMHI applications for mental health in the context of the constantly changing landscape of the COVID-19 pandemic and beyond. While some scoping reviews have focused on digital health interventions for mental health in general [38-40] and others have incorporated a focus on LMICs [11], to our knowledge, this is the first review that focuses on digital health interventions in LMICs specifically for depression and anxiety and includes an in-depth review of their fidelity, usability, and cultural adaptation.

Methods

Data Sources and Search Strategy

Considering the need for a timely review, a rapid scoping review was conducted, and a protocol was not registered. We used the following definition of a rapid scoping review as proposed by Tricco et al [41]: “a type of knowledge synthesis in which components of the systematic review process are simplified or omitted to produce information in a short period of time.” Furthermore, scoping reviews seek to provide a broad mapping overview of the extent, range, and nature of the existing evidence base, focusing on breadth regardless of quality [42-44]. For these reasons, appraisal of the methodological quality of the individual studies was not conducted. This rapid scoping review was conducted following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) and using a search strategy developed in consultation with a liaison librarian at Simon Fraser University. Processes for rapid reviews as outlined by Tricco et al [45] were followed.

Two databases were searched (PubMed and PsycINFO) following common search methodologies for rapid reviews, which advise the search of at least one electronic database (PRISMA [Preferred Reporting Items for Systematic Reviews and Meta-Analyses] standards [46]), with most published rapid reviews using ≥2 [45]. The PubMed and PsycINFO databases were searched for records published between January 2020, when COVID-19 was declared a public health emergency of international concern by the World Health Organization (WHO; declared a pandemic on March 11, 2020), and January 2025. The COVID-19 pandemic accelerated the development of DMHIs, and a rapid review within this time frame was conducted to capture the recent literature that may have incorporated new methods of development and application.

The search strategy was developed across the following three domains: (1) digital health interventions, (2) depression or anxiety, and (3) LMICs. While we were interested in articles related to the COVID-19 pandemic, we made the decision not to include COVID-19 as a search domain as we were interested not only in DMHIs developed to respond specifically to COVID-19–related anxiety and depression but also in DMHIs developed for anxiety and depression in general during the search time frame. Instead, the search included the aforementioned 3 domains and the time frame of the COVID-19 pandemic (2020-2025). A combination of search terms was applied for the automatic search of publications in the aforementioned databases: (“digital” OR “technology” OR “e-health” OR “telemedicine” OR “internet” OR “online”) AND (“depress*” OR “anxi*” OR “mood disorders” OR “anxiety disorders”) AND (“low and middle income countries” OR “developing countries”). Searches included both subject headings and keywords. Search terms for LMICs were taken from the University of Alberta’s Health Sciences Search Filters resource [47]. Multimedia Appendix 1 provides the full search strategy for PubMed, and Multimedia Appendix 2 provides the full search strategy for PsycINFO.

Screening Strategy and Inclusion and Exclusion Criteria

Records identified in the database search process were screened using titles and abstracts by one researcher (the first author, LWC). Records were selected for full-text review if they: (1) reported on quantitative or qualitative evidence related to DMHIs for depression and anxiety; (2) were from LMICs; (3) were written in the English language; (4) were published between January 1, 2020, and January 31, 2025; and (5) had the full text available. In addition to original research articles, other types of articles (eg, protocol papers and scoping and systematic reviews) were included if they provided information on DMHIs being used in LMICs and included depression or anxiety as a primary focus. We included all articles on depression and anxiety regardless of other conditions or specific patient groups (eg, pregnancy or cancer) as long as the primary outcome included either depression or anxiety.

We excluded articles that did not include depression or anxiety as a primary focus. Any articles that presented data from multiple countries (eg, literature reviews) were included if data from LMICs and non-LMICs were disaggregated and reported separately; in these instances, only data related to LMICs were reported in this review. Duplicates were removed across the 2 databases. Records included for full-text review were cross-checked for relevance and inclusion by an additional researcher (senior author JO).

Data Extraction and Synthesis

Data were extracted from the records that were eligible through full-text review and charted into 2 tables according to the following headings: author, year of publication, participant characteristics (study location and study population), intervention characteristics (platform and techniques and functions), and study characteristics (design, mental health conditions, and outcomes) in the first table and discussions on fidelity (either in relation to the original in-person or paper-based format or ensuring adherence by lay health workers to the intervention protocol), usability (both quantitative and qualitative explorations), cultural adaptation (of DMHIs to the local cultural context), and challenges to and facilitators of the interventions’ uptake in LMICs in the second table. Findings across the discussions are presented using numerical counts accompanied by descriptive text. See Multimedia Appendix 3 for the PRISMA-ScR checklist.

Results

Article Selection

The searches of PubMed and PsycINFO were conducted on March 13, 2025. Figure 1 shows the PRISMA flowchart for the full record selection process. A total of 1010 records were returned, and after the removal of 6 (0.59%) duplicates, 1004 (99.41%) were included for screening. After performing a title and abstract review, 20.92% (210/1004) of the articles were included for full-text review based on our inclusion criteria. A total of 79.08% (794/1004) of the articles, which did not meet the inclusion criteria, were excluded from full-text review. Of the 210 articles included for full-text review, 2 (1%) were not retrieved as the full text was not available.

After completing the full-text review of the 208 available articles, a total of 80 (38.5%) records were included in the final analysis. This included 80% (64/80) original research articles, 15% (12/80) protocol papers, and 5% (4/80) literature reviews (1/4, 25% scoping reviews [48] and 3/4, 75% systematic reviews [49-51]). Reasons for exclusion of articles at full-text review that did not meet the inclusion criteria (128/280, 45.7%) were recorded and reported in Multimedia Appendix 4. The most common reasons for exclusion were focus on another topic (eg, app use in general), lack of a DMHI (eg, therapists using video or telephone to deliver regular counseling without an accompanying app or online platform), results from reviews (global focused) combined without data from LMICs being reported separately, focus on mental health in general, and no description of the use of a DMHI. A quality appraisal was not conducted considering the nature of this rapid scoping review, which did not focus on an examination of the effectiveness of the DMHIs, along with the emergent nature of the literature. The examinations of fidelity and usability were reviewed and described independent of whether there was an exploration of intervention effectiveness and, if there was, whether the intervention was effective.

‎

Study Characteristics

Table 1 presents the main characteristics of the studies included in the rapid scoping review. Six main DMHI platforms were identified: (1) mobile app (software app used on a mobile device such as a smartphone or tablet), (2) web-based program (intervention delivered on the web through an internet-connected device), (3) virtual reality (VR; involved the immersion of the user in simulated and controlled computer environments [52]), (4) videoconferencing (intervention delivered via video through an online platform; eg, Google Meet and Zoom), (5) telemedicine (intervention delivered via phone calls or text through an instant messaging app [eg, WhatsApp] that went beyond the usual psychoeducation delivered by therapists by incorporating additional components such as educational videos and relaxation music), and (6) social media–based interventions (delivered through a social media platform; eg, Facebook).

Synthesis of Results

Overview

This section presents a summary of the findings from the 80 articles that reported on fidelity, usability, and cultural adaptation of DMHIs, along with the challenges to and facilitators of their uptake in LMICs. The synthesis was organized by the three research objectives: (1) type of DMHI intervention platform; (2) discussions on their fidelity, usability, and cultural adaptation; and (3) identified challenges to and facilitators of their uptake in LMICs. A summary is presented in Table 2 organized according to the type of DMHI intervention platform used, along with other articles that reported the use of more than one type of DMHI.

Types of Digital Health Interventions for Depression and Anxiety Being Used in LMICs

This search and review of the 80 articles showed that, of the DMHIs identified, the largest category was mobile apps (n=46, 58%). Other DMHIs involved web-based programs (n=10, 12%), VR (n=5, 6%), videoconferencing (n=4, 5%), telemedicine (interventions delivered via an instant messaging app; n=4, 5%), and social media–based interventions (n=1, 1%). A total of 12% (10/80) of the articles reported on delivery of mental health care using more than one platform (eg, delivered via app or the web or reviews that included various DMHIs). DMHIs can be stand-alone self-guided programs or incorporate a hybrid model whereby human assistance, either from a therapist (ie, mental health specialist) or a lay health worker to augment their care, is available with varying levels of support.

Most of the articles (64/80, 80%) reported strengths of DMHIs either on improving symptoms of depression or anxiety or other factors such as high level of usability, acceptability, and satisfaction during usability and pilot-testing. Overall, DMHIs show promise as a way to bridge the mental health treatment gap for depression and anxiety in LMICs.

Fidelity, Usability, and Cultural Adaptation

Fidelity

Despite the accelerated increase in the development and application of DMHIs in recent years, with many being adapted from existing in-person interventions, only 20% (16/80) of the articles reported on fidelity. Of these 16 articles, only 4 (25%) discussed fidelity in relation to the original in-person or paper-based intervention format. One of these articles mentioned adherence to the original intervention structure and format in the adapted digital format [80]. Another article described the process of involving psychologists and psychiatrists to ensure theoretical alignment of app content with psychosocial components [83]. The third article mentioned the development of 1 video per session to deliver the therapeutic ingredients with fidelity and at the right dose [57]. The last of the 4 articles mentioned that the intervention delivery format (mobile app) ensured that key therapeutic elements were delivered with fidelity and at the appropriate dose using the app’s automated features [64].

Most of the articles that discussed fidelity (13/80, 16%) described the process for ensuring fidelity or adherence by lay health workers to the intervention protocol. This included fidelity and audit checks of lay health workers by supervisors and reviews of case notes [55,70-73,79] and use of various tools (eg, fidelity checklist [71,72,79] or the practical robust implementation and sustainability model framework [95]). Regarding those reporting on the use of a tool, 2% (2/80) of the articles provided further information, indicating that 5% of the guidance calls and messages were rated (no additional information was provided on who performed the rating) [71,72]. Another article reported that the structured steps on the app made it easier for lay workers to adhere to fidelity when navigating the sessions [73], whereas another article indicated that the intervention itself, which was used for detection and treatment, could increase consistency in diagnosis and treatment [89]. Finally, one article stated that fidelity to the intervention protocol was demonstrated by lay health workers delivering the intervention adapting easily to the technology (smartphones) due to familiarity with the protocol [57]. Regarding the articles that reported on measuring fidelity of lay health workers to the intervention protocol, overall, it was shown that lay counselors adhered to the intervention delivery components.

Usability

Less than half (29/80, 36%) of the articles described usability in some manner, including through usability testing or qualitative exploration. Most of the articles (23/29, 79%) were related to mobile apps. Of the 29 articles, only 7 (24%) reported the use of measures, either standardized (eg, the System Usability Scale or mHealth App Usability Questionnaire [MAUQ]) or created by the authors for the study (eg, satisfaction survey), to capture usability [73,78,84,95,111,115]. In addition, only 31% (9/29) of the articles described the usability testing process along with findings in detail, with most (7/9, 78%) related to mobile apps [62,64,65,80,95,105,110,111,115] and the rest (2/9, 22%) pertaining to web-based programs [55,84]. In total, 10% (3/29) of the articles were protocol papers that indicated how app development would follow a user-centered iterative design.

Most of the articles that described usability testing outcomes (19/80, 24%) reported high satisfaction. For example, the study by Nurbaeti et al [100] showed that the evaluation of an app for postpartum depression symptoms in Indonesia was mostly positive, with a large percentage of the participants reporting the following: appearance of the app being adequate (92.7%) and the app being easy to download (89.9%), understandable (96.3%), easy to fill in (94.5%), and beneficial (96.3%) and reflecting psychological conditions (90.8%). Another article reporting the protocol for a feasibility cluster randomized controlled trial to be conducted in Nepal and Nigeria indicated that health workers found the prototypes of the app for the adapted WHO electronic Mental Health Gap Action Programme Intervention Guide to be intuitive, appropriate, and feasible for clinical work [112].

Of the 10 articles that reported more than one delivery platform, 4 (40%) reported on usability testing [51,73,78,101]. In total, 75% (3/4) of these articles indicated the evaluation of a usability-related construct using particular measures—a usability questionnaire (MAUQ) [73], satisfaction survey [78], and completion rates with participant feedback [51]. The remaining article mentioned that piloting had been conducted to evaluate the appropriateness of the intervention without describing this in detail [101]. Usability was reported as being high (eg, the MAUQ score was 78.5, and the satisfaction survey showed that 92% of participants were very satisfied).

Cultural Adaptation

Finally, 39% (31/80) of the articles described a process for adapting the DMHIs to the local cultural context. Of these 31 articles, 23 (74%) reported on mobile apps. Various methods were used, including a narrative approach involving local women, a participatory approach, identification of measures that had been validated in the implementation country or working with local partners to enhance cultural validity where needed, and the intervention being led by a local partner. Many of the articles (13/80, 16%) discussed the involvement of a broad range of stakeholders in the adaptation process using a “participatory approach” [67]. For example, one article mentioned how “the vignettes were developed in consultation with local women who had suffered from depression” [64]. Another article described that, to improve cultural fit, “a team of clinical psychologists, psychiatrists, and researchers from the U.S. and India adapted the intervention,” and “modifications were informed by student feedback” [84].

A few articles (7/80, 9%) described the cultural adaptation process in detail. For example, one article described how the VR space was designed to be similar to the environment of the end users (urban refugee youths in Kampala) and presented in their language of choice combined with training materials and content adapted with peer navigators to enhance relevance [88]. Another article detailed the extensive cultural adaption process conducted by the WHO and the Lebanon National Mental Health Programme to develop relevant content (story, local idioms, and illustrations) for the Lebanese population, displaced Syrians, and other groups in Lebanon. The article also included descriptions of piloting for face and content validity, clarifying language, assessing readability, and establishing content appropriateness within the intended context [55].

Most of the articles (18/80, 20%) were vaguer, indicating simply the importance of identifying what would be culturally acceptable, that the app was designed for the specific population, or the use of a guide adapted for the health system of the target country. A total of 2% (2/80) of the articles, which reported on web-based DMHIs, indicated that cultural adaptation was conducted with the intervention or culturally adapted approaches were used, resulting in high cultural relevance and acceptability; however, the process was not described in detail [58,84]. Similarly, 67% (2/3) of the articles that reported more than one delivery platform mentioned cultural adaptation but did not describe the process in detail [71,72]. These 2 articles pertained to the same intervention. Another article mentioned that the narrative was adapted to the local context considering cultural and linguistic nuances among displaced Syrians and that participants could choose the appearance of the avatar used in the program, reflecting the main cultural groups in Lebanon (Lebanese, Syrians, and Palestinians) [64].

Challenges to and Facilitators of Intervention Uptake in LMICs

Challenges to Uptake

A total of 28% (22/80) of the articles reported on challenges. In total, 59% (13/22) of these articles were related to mobile apps, describing challenges that pertained to the technological system (eg, system lags, app navigation challenges, and bugs); user engagement issues (sustained engagement of individuals who are depressed and may have low motivation, high churn, and issues with adherence to treatment), which are linked to poor usability; contextual issues (eg, illiteracy, technological proficiency, accessibility of mobile phones, data limits, battery charging issues, and lack of family involvement); structural barriers (eg, network connectivity issues and resource constraints); and concerns about privacy and confidentiality [48,57,62,64,79,80,89,90,107,111,117,125,127]. Of the 22 articles, 2 (9%) highlighted how many of these challenges contribute to the digital divide, or the division between individuals or populations who have access to and use digital technology and those who do not [64,117]. Other challenges mentioned included that of a practice gap between health care providers and application developers [48] and challenges with the online delivery method where participants complete tasks in their own time [79].

In addition to similar challenges described previously, other challenges reported in relation to web-based programs and those delivered via more than one platform included individuals forgetting their log-in credentials, time constraints due to busy schedule of the participants, retention of lay health workers, sustained engagement, and stigma [55,66,84]. There were also concerns reported about the interventions’ efficacy, effectiveness, and usability [73].

Facilitators of Uptake

In total, 21% (17/80) of the articles reported on facilitators. Of these 17 articles, 12 (71%) were related to mobile apps, describing various facilitators that supported their uptake and scale-up, including widespread mobile phone use, built-in supervision and training features, and an app’s convenience and integration into existing social (eg, schools) and health care systems [53,57,64,67,79,83,89,90,105,111,117,125]. One of these studies suggested a web-based delivery platform rather than an app-based one as the former does not require regular updating and tailored interventions or distinct adaptations that account for the unique needs of the target population [79]. Other articles (3/80, 4%) reported on the importance of orientation and training programs and mobilization of community health volunteers [89,111] and how an understanding of the intervention and its potential benefits will likely support its uptake [90]. In line with some of the concerns that clinicians expressed, described previously, one article reported on the importance of buy-in from frontline health care providers and clinical experts, which may be facilitated by demonstrated clinical utility and efficiency [117]. Other facilitators reported in relation to web-based programs or interventions involving more than one delivery platform included facilitators at the individual level, such as being technologically savvy and literate and having access to devices [55], and incorporating features to boost user engagement [101]. Finally, the importance of cultural appropriateness was emphasized by a number of the articles (11/80, 14%; see, eg, the studies by Alavi and Hirji [58] and Pozuelo et al [105]).

Discussion

Principal Findings

The 80 articles identified in this rapid scoping review reported the use of mobile apps, web-based programs, VR, videoconferencing, telemedicine (interventions delivered via instant messaging apps) platforms, and social media, in addition to instances in which the DMHIs were delivered via more than one platform (eg, delivered via app or the web). To our knowledge, this rapid scoping review is the most comprehensive presentation of articles examining DMHIs for depression and anxiety and their fidelity, usability, and cultural adaptation, along with challenges to and facilitators of their uptake in LMICs.

Fidelity

Active elements of these DMHIs are often modeled on evidence-based psychological interventions developed for in-person delivery [17], offering further promise to provide scalable effective mental health care. As such, it is critical that the interventions are implemented with high fidelity while balancing the need for necessary adaptations to the digital context, including the different motivational needs and preferences of digital users. However, inconsistencies regarding treatment fidelity have been identified as a main contributor to the mixed evidence on the effectiveness of many DMHIs [128]. Similarly, results from this review showed that less than a fifth (16/80, 20%) of the articles reported on fidelity, with heterogeneity in process and presentation format. Most of these articles (13/80, 16%) described the process for ensuring adherence by lay health workers to the intervention protocol rather than for ensuring fidelity of the DMHI to its original in-person or paper-based format. Furthermore, only 1% (1/80) of the articles mentioned specifically adherence of the DMHI to the original intervention structure albeit delivered in an automated digital format [80]. Interestingly, 2% (2/80) of the articles mentioned that the delivery format of the intervention (app) supported fidelity via the nature of its automated and structured format [64,89]. This is consistent with previous research demonstrating that less complex interventions are easier to implement with high fidelity [129]. As more DHMIs are being developed and adapted from in-person or paper-based formats, it is crucial to ensure intervention fidelity to maintain consistency in intervention delivery across platforms while balancing necessary adaptations to suit digital modes of delivery and the local context [130,131].

Usability

Poor usability has been cited as a primary reason for failed adoption of and engagement with digital health apps [31,32], where approximately one-third of users who download health apps stop using them after a short time, often no more than 2 weeks [132]. Hurdles to app usability include being difficult to set up or use or being buggy, not working well, requiring too much time to enter data, or having small screens that display limited information at a time [32,133]. Other challenges include concerns about privacy. An additional consideration for individuals with depression include specific symptoms such as low motivation, concentration difficulties, and behavioral avoidance that may make it more difficult for them to engage with DMHIs, posing further challenges to real-world efficacy [133]. Although usability testing is understood increasingly as an important component of app development and evaluation [134], usability currently remains underexplored in depression app development [135,136]. Findings from this review reflect this as only 36% (29/80) of the articles described usability testing in some manner despite overwhelming evidence highlighting how the efficacy of mobile apps can be greatly enhanced by conducting usability testing to determine factors impacting user adherence and satisfaction [137-139].

Articles that reported on usability outcomes (9/80, 11%) demonstrated mostly positive findings, with above-average usability and satisfaction scores being reported. Certain challenges to usability were also reported, including a confusing registration process and content that could be improved to be more engaging [80], which was a barrier to adherence [55]. Despite the generally positive usability and satisfaction scores reported by the various articles, poor usability has been clearly identified as a barrier to sustained user engagement across various digital health interventions [140,141]. A lack of user engagement with the DMHI regardless of users’ self-reports on satisfaction and utility likely indicates poor usability [142] as usability is a precursor for engagement [143]. This is also reflected in the fact that, despite the large number of medical health apps available, their overall downloads are decreasing [31]. Methods to increase usability include incorporating interactive features (eg, tooltips) [144] and gamification (eg, badges and avatars) [145]. Usability is also intricately linked with acceptability, which includes the intuitive attitudes toward the intervention, use intentions, actual use, and satisfaction after using [146]. Acceptability is intricately linked with prevailing social and cultural norms [147], highlighting further the importance of cultural adaptation.

Cultural Adaptation

Ensuring that interventions are culturally appropriate and relevant is important to their success [148,149]. While depression and anxiety affect people from all cultures, the experience of these illnesses is culturally bound, shaped by local beliefs, practices, and norms. This cultural variation profoundly influences mental illness expression; patterns of coping and help seeking; the process of recovery; adherence to treatment; and relationships between individuals and their networks of family, friends, and health care providers [150,151]. Cultural adaptations to increase relevance for individuals from ethnic minority groups have been shown to have greater efficacy [152] and more favorable health outcomes [153].

However, there has been limited focus on cultural adaptations of DMHIs to ensure relevance and acceptability [19,20]. Most of the 39% (31/80) of the articles that reported on cultural adaptation only touched on the cultural adaptation process undertaken, for example, “clarify[ing] language, and establish[ing] the appropriateness of the developed content within the intended context of use” [101] and “the narrative was adapted to the local context, considering gender, linguistic, and cultural nuances” [71,72]. Only 9% (7/80) of the articles described the cultural adaptation process in detail. While it is encouraging that many of the articles included in this review (31/80, 39%) discussed cultural adaptation in some form, there is a need for researchers and developers to place more emphasis on the cultural adaptation process to ensure that evidence-based interventions are properly adapted to enhance favorable health outcomes.

Promisingly, a number of the articles (13/80, 16%) discussed the involvement of a broad range of stakeholders in the development and cultural adaptation process. Engaging a broad range of stakeholders, including experts, end users, and community members, has been shown to contribute critical insights for development, implementation, and evaluation [153]. Developing DMHIs that address the culture and context of the diverse groups in LMICs by including necessary cultural adaptations while maintaining intervention fidelity is critical to increasing engagement of historically underserved cultural groups [21].

Challenges and Facilitators

A total of 28% (22/80) of the articles reported on challenges, whereas 21% (17/80) reported on facilitators. Unlike the topics of fidelity, usability, and cultural adaptation, where most of the articles discussed mobile apps, articles that discussed challenges and facilitators were spread more evenly across the different delivery platforms, highlighting the consistency of challenges and facilitators experienced with different types of DMHIs. Many of the challenges reported in the articles (eg, contextual issues and structural barriers) were perceived as possibly leading to inequitable digital access and contributing further to the digital divide. These findings are consistent with the literature [12], which indicates that, while DMHIs offer an accessible and promising method for delivering mental health care, a number of challenges remain from the end user and developer side that must be addressed. For example, many of the individuals who use these types of DMHIs may live in rural areas where contextual (eg, being busy with multiple responsibilities such as farming, taking care of family, and side jobs) and structural issues (underdeveloped technological infrastructure) may prevent them from engaging fully with DMHIs. Structural issues and digital health literacy have been widely reported in the literature for a number of years [154,155]. A lack of digital literacy, especially among the older adult population and those from lower socioeconomic brackets, is one of the main challenges in scaling up digital health interventions [155,156]. In addition, DMHIs may not be accessible for subpopulations such as racial and ethnic minority groups due to language barriers [157], emphasizing again the importance of ensuring cultural appropriateness. Finally, for DMHIs that incorporate a hybrid model involving human assistance, in particular lay health workers to augment their care, the articles emphasized the heavy burden on the providers, who often have heavy existing workloads and are largely underpaid or even unpaid [158]. This may pose a challenge for sustainability.

Clinician buy-in was reported as a challenge to the overall implementation of DMHIs into routine care and centered on time management, regulatory structures and licensure, lack of nonverbal evidence, and payment. Similar challenges have been reported in the literature, demonstrating that there has been a reticence from mental health professionals to engage with DMHIs due to the perceived threat to the therapeutic alliance, which some believe can only be established in person [25,26]. Research suggests that there is presently limited theoretical and practical training on novel digital tools and interventions in psychiatry, which influences clinicians’ understanding and perceptions of DMHIs [159]. The importance of in-person contact in psychotherapy has also been widely discussed in the literature [160-162]. However, Newman et al [163] showed that, while clinician-assisted treatments are indeed optimal, interventions with reduced in-person contact and at reduced intensity have also been shown to be effective and cost-effective. This has been researched most heavily in the case of mobile apps [164,165]. For example, 2 systematic reviews of mobile apps (both self-guided and hybrid models) demonstrated their effectiveness in reducing depression symptoms [16,139]. Nevertheless, clinician concerns need to be addressed to enhance their acceptance of and engagement with DMHIs, particularly for those that are integrated into standard care and may require their support. One method may be engaging clinicians more throughout the app development process to provide expert input and also increase their understanding and acceptance of DMHIs [13].

Facilitators of the uptake and scale-up of DMHIs in LMICs included widespread mobile phone use, built-in supervision and training features, their convenience, and integration into existing social and health care systems. For the hybrid models that required lay health workers for ongoing implementation, mobilization of community health volunteers, including through training and professional support, was considered an important facilitator. Training and professional support have been shown to increase lay health workers’ satisfaction and enhance their agency and social status [166]. Ensuring cultural appropriateness to the local context by using tailored adaptations was also emphasized. This can increase the reach of and engagement with DMHIs and increase equity by avoiding the systematic exclusion of cultural groups who have been historically underserved [21]. Furthermore, increased mental health awareness was highlighted as an important facilitator to increase understanding of the DMHI and its potential benefits. In addition to demonstrated clinical effectiveness of DMHIs for depression and anxiety, their use has been shown to increase mental health literacy and awareness [154]. Finally, the importance of buy-in from clinicians and other providers was highlighted. In addition to DMHI integration into standard care, clinician behaviors could support adherence. For example, Paxling et al [167] demonstrated that certain clinician behaviors, such as task reinforcement, task prompting, self-efficacy shaping, and empathetic utterances, were correlated with module completion [168].

Limitations

A limitation of this review is that we only included English-language papers, leading to an unknown number of articles from LMICs published in languages other than English being excluded. However, we believe that this review of 80 articles identified many of the key issues concerning DMHIs in LMICs. A second limitation is that this review focused only on depression and anxiety; examination of DMHIs for other mental health conditions may elucidate additional considerations. However, as depression and anxiety are the 2 most common mental health conditions [5] and many DMHIs are developed for depression and anxiety [6,15,169], the results of this review can inform the further development and delivery of DMHIs for these 2 conditions and provide information on general rather than condition-specific features of DMHIs focusing on other conditions. Finally, research published before the search time frame may have contributed additional findings. Considering the rapid advancement of DMHIs and related research stimulated by the COVID-19 pandemic, we believe that this review of research published during the applied time frame captured the most crucial and timely evidence.

Given the rapidly changing landscape of development and application of DMHIs, further research is warranted to capture emergent changes, including the interventions’ fidelity, usability, and cultural appropriateness and challenges to and facilitators of their uptake in LMICs. Additional future research should also focus on underexplored areas, including an examination of the percentage of DMHIs that have no basis in evidence; the potential for harms in the unrestrained development and delivery of DMHIs; and the threats to privacy and security, such as potential for financial harm through unregulated mobile apps, including those presented as DMHIs. Finally, the accelerated growth of artificial intelligence (AI) and its increasing application in various health care fields in recent years could transform the landscape of digital mental health, including DMHIs [168,170]. While AI represents a potential means to advance mental health care, there will be an ensuing urgent need to examine in detail ethical implications associated with the integration of AI into digital mental health [171].

Conclusions

DMHIs are one of the most promising ways to better meet population mental health needs and respond in culturally appropriate ways to diverse population groups. Despite the opportunities they offer for reducing the severe and persistent mental health treatment gap, there were several key issues identified in this scoping review that warrant further examination. These include further work on the fidelity, user engagement, and cultural adaptation of DMHIs. In addition, various challenges to the uptake of DMHIs in LMICs, including challenges pertaining to the technological system, contextual issues, structural barriers, and concerns about privacy and confidentiality, must be mitigated in the ongoing implementation of DMHIs in LMICs to avoid contributing further to the digital divide. In summary, DMHIs in isolation are not a solution to address the significant mental health care system deficiencies that are present; rather, they may serve as a low-cost and low-barrier treatment option—especially those DMHIs that are appropriately adapted from evidence-based in-person interventions—in low-resource settings where community-based care is severely limited.