What digital health technology types are used in mental health prevention and intervention? Review of systematic reviews for systematization of technologies

Abstract Digital health technology has been widely applied to mental health interventions worldwide. Using digital phenotyping to identify an individual’s mental health status has become particularly important. However, many technologies other than digital phenotyping are expected to become more prevalent in the future. The systematization of these technologies is necessary to accurately identify trends in mental health interventions. However, no consensus on the technical classification of digital health technologies for mental health interventions has emerged. Thus, we conducted a review of systematic review articles on the application of digital health technologies in mental health while attempting to systematize the technology using the Delphi method. To identify technologies used in digital phenotyping and other digital technologies, we included 4 systematic review articles that met the inclusion criteria, and an additional 8 review articles, using a snowballing approach, were incorporated into the comprehensive review. Based on the review results, experts from various disciplines participated in the Delphi process and agreed on the following 11 technical categories for mental health interventions: heart rate estimation, exercise or physical activity, sleep estimation, contactless heart rate/pulse wave estimation, voice and emotion analysis, self-care/cognitive behavioral therapy/mindfulness, dietary management, psychological safety, communication robots, avatar/metaverse devices, and brain wave devices. The categories we defined intentionally included technologies that are expected to become widely used in the future. Therefore, we believe these 11 categories are socially implementable and useful for mental health interventions.


A new occupational health strategy incorporating digital health technology is needed
The coronavirus disease pandemic led to widespread teleworking worldwide; however, it impaired the physical and mental health of workers. 1 The World Health Organization (WHO) and International Labour Organization have called for occupational health services to provide ergonomic, mental, and psychosocial support using digital health technologies. 1Therefore, a new occupational health strategy incorporating digital health technologies is required in the global workplace.
In recent years, digital health technologies have evolved rapidly.According to the WHO, digital health is an expansion of the e-Health concept that encompasses a wider range of digital technologies for health, including smart and connected devices, the Internet of Things, advanced computing, big data analytics, and artificial intelligence (AI), including machine learning, and robotics. 2Moreover, digital health interventions, such as healthy worker decision support, tracking of health status using digital devices, and health education and training content in digital form for health professionals have been recommended by the WHO. 3 We expect to observe an increasing number of developing countries adopting digital health technology interventions in the future.
The ergonomic perspective of a systems approach is helpful to develop new occupational health strategies that incorporate digital health technologies. 4,5The ergonomic systems approach is a method for optimizing well-being and performance (such as a comprehensive approach to stakeholders, the organizational environment, and the physical environment), 4,5 and from this perspective it is necessary to understand technological trends, optimize the benefits and risks of new strategies, and implement appropriate strategies in society to develop proactive mental health measures using digital health technologies.
Wienert et al 6 defined digital public health intervention as "an intervention that addresses at least one essential public health function through digital means," which includes preventive behavior change, self-management, active monitoring, treatment, and diagnosis.A review of these digital public intervention studies showed that logs recorded by various sensors on smartphones are used for health outcomes, with a particular focus on mental health outcomes. 7As the ability to utilize digital health technologies in the mental health domain is of global interest, familiarity with technologies that are currently widely used or will be developed in the future is important.
Digital phenotypes are expected to enhance mental health measures 8 as various intervention studies are being conducted.A digital phenotype is defined as "moment-by-moment quantification of the individual-level human phenotype in situ using data from personal digital devices". 8,9Avant-garde works exist globally that utilize physical activity levels, sleep quality, and social behaviors for mental health measures based primarily on personal logs obtained from smartphones.Digital phenotypes can be classified as active or passive data. 8Active data are personal experiences captured through user engagement such as ecological momentary assessments. 8,10Passive data are automatically captured by accelerometers, GPS, and light sensors on smartphones and other devices to log an individual's experience without user engagement. 8,10Conversely, some of the services provided are a mixture of active and passive data 11 ; therefore, accurate classification may be difficult. 12Additionally, the terminologies used in previous studies were mixed, causing heterogeneity. 13Furthermore, in addition to the technology used for digital phenotyping, several other digital health technologies, including human-supported digital interventions, such as cognitive behavioral therapy (CBT) by therapists and others via teleconferences, the internet, and telephone 14,15 are used in addressing mental health problems.Technologies including chatbots, gamification, and extended reality (eg, virtual, augmented, and mixed reality) are expected to be utilized as mental health measures in the future. 16ence, no consensus exists on the classification of technologies that can be used to improve mental health.Moreover, from the viewpoint of occupational health technology developers and service providers, classifying the technology element levels is necessary to understand the availability of technologies applicable to mental health interventions.Therefore, we defined the intervention of digital health technologies in mental health as "healthcare services aimed at primary prevention provided to the general workforce using information and communication technology and digital technology.Digital health technology includes services that use technical algorithms (such as, application software [apps], communication robots, wearable devices, information provision through non-contact sensing devices, selfmonitoring, and real-time feedback), and services that do not use technical algorithms (such as, online counseling that solely relies on internet-based means)." 17s mentioned above, various digital health technology representations include digital phenotypes. 13In other words, scoping and surveying digital health technology using only digital phenotypes results in bias.We also incorporate services that do not use technical algorithms into our definition of digital health technology.Therefore, a review is required that incorporates nondigital phenotypic articles to clarify overall digital health technologies.We conducted a review of digital phenotyping and other technologies to integrate and systematize the technologies that are currently being used or will be used in the mental health field.

Systematic review and Delphi technique process
This study implemented a review to investigate digital health technology trends used in digital phenotyping.Our review of articles on digital phenotyping was conducted in conformity with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement, 18 and 2 databases were searched.First, we filtered PubMed by Review and Systematic Review and searched for articles in the last 5 years using the following terms: "digital phenotyp * " AND "mental health".Further, we searched Google Scholar for the following terms: digital phenotype OR digital phenotyping AND mental health.After the search, following the removal of duplicates, articles were screened by 2 researchers (N.T. and T.E.).We defined the criteria for the inclusion of articles as follows: systematic reviews that collected and analyzed data for digital phenotyping of mental health, articles with content relevant to the technical classification, full articles, and articles published in the English language.The following exclusion criteria were used in selecting articles for screening: articles that were not about the technology used for digital phenotyping, systematic reviews not associated with mental health interventions, gray literature, or review protocols.Articles were selected based on the above selection criteria.
However, the lack of a detailed technical classification system with a consensus on digital phenotypes may have led to a selection bias.To cover digital health technologies related to mental health interventions, we needed technical information on digital phenotypes and technical information without digital phenotypes.Therefore, we adopted the snowballing method, 19 which includes 2 types: forward snowballing, which identifies new papers based on those that cite the paper under study; and backward snowballing, which identifies new papers using a list of references.The backward snowballing method was used to identify new articles for inclusion in the reference list.Although this snowballing procedure continued until no new papers were identified, it was conducted only once.Namely, we utilized a 1level backward snowballing approach and included additional articles.In this review, we attempted to identify the technologies used in digital health.We used the Delphi technique 20,21   the responses until group opinions converge.To systematize the technologies resulting from these reviews, we held discussions with experts in psychology and occupational health (C.Y., H.E., K.I., S.K., and N.K.), physiologists (H.F.), information scientists (Y.K.), labor scientists (K.I.), nurses (M.T.), physical therapists (N.T.), and ergonomists (T.E.) using the Delphi technique.

Systematization of digital health technologies and proposal
The review process is illustrated in Figure 1.We extracted 71 articles from the 2 databases and removed 18 duplicates.The remaining 53 articles were screened after duplicates were removed.We extracted 42 articles based on their titles and abstracts.Furthermore, in a full-text review and applying the inclusion and exclusion criteria, we identified 4 articles that were systematic reviews of digital phenotyping for mental health interventions.
Figure 3 shows the 11 categories of technology areas that exist or are expected to become popular, as derived from the Delphi method.The Delphi process was conducted as follows: N.T. was the administrator, and other experts were assigned as panelists; based on the review results, the administrator prepared a tentative technical classification proposal, the panelists provided input on technical classifications via email, and the administrator revised the proposed technical classification based on the panelists' input.The process was repeated until a convergence occurred of the opinion in the technology area on the issue "What are the digital health technology intervention situations and categories used to measure or monitor mental health?"(Figure 4).Using the Delphi method, the 11 categories we defined were heart rate estimation, exercise/physical activity, sleep estimation, contactless heart rate/pulse wave estimation, voice and emotion analysis, self-care/CBT/mindfulness, dietary management, psychological safety, communication robots, avatar/metaverse, and brain wave devices.
3][24][25][26][27][28][29] In a scoping review of health-related outcomes and digital phenotypes, 55% of articles utilized digital phenotyping for mental health interventions, whereas digital health technologies heavily used physical activity and sleep approaches. 7A large cohort study conducted in Norway reported that those who exercised at least 1 hour per week had a 12% lower incidence of future depression. 30Previous studies reported that people with versus without insomnia have approximately twice the risk of developing depression. 31This finding suggests that daily exercise and sleep monitoring are necessary for preventing mental health issues.Notably, the effectiveness of non-contact methods for measuring the heart rate has been reported in recent years. 32Thus, it seems inevitable that heart rate measurement would independently include contactless heart rate/pulse wave estimation, as well as contact measurement by wearable devices.[26]28,29 Voice and emotion analyses were included in this category because they are expected to be used frequently in the future for mood estimation.
Moreover, self-care/CBT/mindfulness and dietary management are expected to be internet-based 13,28 and app-based 10,22,23,25 mental health interventions.CBT and mindfulness can be conducted by a chatbot in an app or by a therapist online. 15,28,33,34dditionally, as digital health technology interventions are effective for eating disorders, 35 online and app-based dietary approaches may become more extensively used in the future.Furthermore, organizational efforts are important for mental health, 36 and interventions such as psychological safety (eg, crew resource management 37 ) may be offered via apps in the future.In recent years, AI-powered chat bots have made headlines, indicating their potential for use in mental health interventions as well. 24,28,29As AI is increasingly used, mental health interventions for communication robots, such as animal robots and robot therapy, should also be considered. 29Presently, avatar therapy has not shown clear efficacy 38 but may become a useful intervention as AI innovations continue to advance.Although no clear effect has been shown for extended reality, 16,34 we have added the category of avatar or metaverse for the possibility that this may be an approach that holds promise for the future.Additionally, although not yet widely used, brain wave devices cannot be ignored in light of future technological innovations. 16Thus, the categories we defined intentionally included technologies that are expected to become widely used in the future.Therefore, we believe these 11 categories are socially implementable and useful for mental health interventions (Figure 5).

Conclusion
In this study, we tentatively systematized the categories in which digital health technologies contribute to mental health interventions.We believe that, from an occupational health perspective, the use of these technologies should not be limited to individual workers; rather, it should be employer-initiated as part of an organizational intervention.the development of employer-friendly tools is lagging.In the field of occupational health, we must encourage developers to develop tools that employers can use to provide organizational interventions.
We reviewed articles that we deemed appropriate for classification as digital health technologies, including those that focused solely on technology, primary prevention, and treatment and those that are expected to become more widespread.This strategy was beneficial for covering various digital health technologies that had begun to spread rapidly.However, this is not limited to primary prevention articles that describe worker-only interventions.Innovation in digital health technology will continue to evolve faster than expected.As digital health technologies become more prevalent, we expect to see more articles focusing on primary prevention among workers in the occupational health field.Therefore, the systematization made in this study is only a temporary classification to organize the current situation, and the 11 categories should be updated periodically.In the future, it will be necessary to further elaborate on systematization by discussing it with stakeholders surrounding digital health technologies and utilizing the Patient and Public Involvement framework.

Figure 2 .
Figure 2. Visualization of technology terms extracted from the review.The technology term frequencies were counted using Natural Language Processing.

Figure 3 .
Figure 3. Systematization of technology using the Delphi method.

Figure 4 .
Figure 4.The Delphi process used in this study.

Figure 5 .
Figure 5.The 11 technology areas considered useful for mental health interventions.
to identify digital health technologies.The Delphi method involves surveying several people with expertise in an issue and repeating

Table 1 .
Summary of reviewed studies and technologies used for digital health technologies.
Abbreviations: EMA, ecological momentary assessments; ESM, experience sampling methodology; GPS, global positioning system; Ref., number in the References; SMS, short message service.
Meanwhile, it is apparent that the type of technology used by individuals, mainly digital phenotypes (eg, active and passive data), is currently widespread; however, Accelerometer, Actigraphy, Gyroscope, Pedometer, Skin Temperature, GPS, Wi-Fi, Bluetooth, Microphone, Heart rate sensor, Light sensor, Screen activity, Call logs, Email/Text/SMS logs, App usage, Social media logs, Keystroke logs, Self-report questionnaires, EMA/ESM, Facial recognition, Voice recognition, Chatbot, Online therapy, Robot therapy, Virtual reality, etc.