Depression is a very common disorder and thus represents a heavy mental health burden. Although estimates vary widely, the lifetime and 12-month prevalence of depression may be close to 10% and 5%, respectively, and these values are higher in high-income countries [1]. Suicidal risk and ideations have a high incidence among patients with depression. Although estimates vary widely, the incidence of suicide among patients with depression may be as high as 15%, while up to 70% of patients with acute depression may experience suicidal ideas [2]. Unfortunately, depression is associated with an elevated risk of early death. This is partly due to suicidal risk but also to the heavy physical health burden of this psychopathology. Indeed, depression is a predictor of many physical diseases, including coronary artery disease, myocardial infarction, stroke, diabetes, and various cancers. Certain physical disorders also tend to be more severe among patients with depression. There is a reciprocal relationship between depression and physical pathologies. Biologically plausible mechanisms include behaviors such as smoking, drinking alcohol, obesity, low compliance with treatment regimens, and a variety of hormonal and immune dysregulations. Unsurprisingly, depression is also associated with lower perceived overall health [1]. Caring for mentally and often physically sick patients is expensive and requires resources. Thus, the health care burden of depression is heavy. The costs of this disease have been increasing in recent decades and amount to hundreds of billions of dollars in the United States [3]. Depression is highly socially burdensome in several other ways. It is associated with termination of education, lower probability of marrying, marital dissatisfaction, negative parenting behaviors, adolescent childbearing, unemployment, work disability, absenteeism, low work performance, lower personal earnings, and lower household income [1]. Approximately half of all costs associated with depression are workplace costs [3].

Although traditional in-person consultations are the usual modality of caring for patients with depression, telemedicine may be well suited to psychiatric assessment and management. This modality of care may also offer many advantages, such as improving access to care in rural areas [4]. In fact, medicine as a whole is becoming increasingly digitalized as competencies required to provide health care are evolving [5]. Other components of depression care include collaborative care and culturally sensitive approaches. Collaborative care is an integrated model in which many health care professionals work together with the patient to better manage the patient’s disease in a synergic manner [6]. Cultural aspects must also be considered, as beliefs regarding depression and mood influence how patients perceive their psychological health [7].

For the scope of this review, telepsychiatry is defined as the use of information and communication technologies such as videoconferencing and telephoning for the diagnosis and management of psychopathologies. Telepsychiatry is thus a synonym of telemedicine for mental health. Psychiatrists, general practitioners, and other physicians can practice telepsychiatry. Other health care professionals, including psychologists, social workers, and nurses, can use telemedicine for mental health as well. Broader or narrower definitions of telepsychiatry could be used by other authors. Various technologies could also be included in other definitions, including artificial intelligence, augmented or virtual reality, mobile apps, and the Internet of Things. Other terms such as telemental health, e-mental health, and connected mental health could also be used to describe related concepts [8].

Telepsychiatry is a broad concept, and it has many applications. This results in heterogeneity in the literature. One key aspect of telepsychiatry is its use by physicians for various aspects of depression care, including assessment, evaluation, diagnosis, management, and follow-up. This review therefore aims to evaluate the extent and nature of the existing literature on telemedicine for the care of depression by physicians. This review also aims to examine the effects and perceptions of virtual care and how it compares to traditional in-person care.

This scoping review was conducted by following the Arksey and O’Malley framework as well as the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines. The five methodological stages were the identification of the research question, the identification of relevant articles, the study selection, the data extraction, and the collating, summarizing, and reporting of the results [9].

As described above, this review’s main objectives are to map the literature on telemedicine in the diagnosis and management of depression by physicians and to examine the effects of telemedicine and how it is perceived. In other words, this review examines the virtual care of patients with depression by physicians. The identification of this research question is closely related to the burdens of depression and the potential of innovative psychiatric care.

A systematic literature search was performed in MEDLINE, Cochrane Database of Systematic Reviews and PsycARTICLES. These databases were chosen based on their relevance and specificity for peer-reviewed articles on medical and psychiatric topics. No search for gray literature was performed because the scope of this review did not extend to articles that had not been peer-reviewed. The list of keywords used was developed through a preliminary search on PubMed, and the keywords were selected based on their sensitivity and specificity for relevant articles. The search terms used for the database searches in this review are listed in Textbox 1. The literature search was performed on October 11, 2020, and it yielded 4782 articles.

Following removal of all 532 duplicates, the remaining 4250 articles were screened based on the eligibility criteria detailed in Textbox 2. Articles were initially screened based on their title and abstract, and full text articles were obtained when more information was needed for screening. The final count of studies included in this review is 28.

An Excel file (Microsoft Corporation) was used to chart the data from all included articles. Throughout the charting process, the extraction grid was iteratively revised to refine its components.

Results regarding the study characteristics for both the included review and nonreview articles are reported below. However, analysis of the outcomes and results of the included articles was limited to nonreview articles because the included reviews had high heterogeneity and their aims tended to differ significantly from the scope of this scoping review. In-depth analysis of outcomes and results was also limited to the scope of this review. Tables were used to report relevant data comprehensively. Figures were produced using Excel.

A total of 28 articles were included in this review. The study selection process is detailed in Figure 1 [10-37].

Although all the articles were published in peer-reviewed journals, an absence of conflicts of interest was clearly and explicitly declared in only a small majority (17/28).

All included papers were published between 1998 and 2020. The great majority had been published since 2010 (24/28), and more than half had been published since 2015 (17/28). The publication years of the papers are illustrated in Figure 2.

A total of 13 studies were conducted using a nonreview methodology, while 15 articles were reviews. Most nonreview studies (9/13) compared a virtual intervention group and a nonvirtual group, and almost all of these comparative studies (8/9) included a randomization process.

Most nonreview studies were conducted in the United States. Locations represented in nonreview articles are displayed in Table 1. Another notable characteristic of the populations of these studies was their size, which ranged from 1 to 309 patients. The average study population size was 135 patients. The intervention in almost all these studies included a psychiatrist (12/13) instead of a primary care provider or as well as one. Collaborative care was part of most interventions (7/13). A nonvirtual comparison group was included in nine studies. Characteristics of the populations, virtual interventions, and nonvirtual comparisons of the nonreview studies are detailed further in Table 2 [10-22].

Of the 13 nonreview papers, 11 included results regarding depression variables, 8 included results with regard to perceptions, and 11 included results regarding other variables. Outcomes and results relevant to the scope of this review for the included nonreview articles are detailed in Table 2 [10-22].

Regarding depression outcomes and variables, all 11 studies resulted in improvement in depression with the use of virtual care, and studies that compared the virtual intervention group with a nonvirtual comparison group either obtained equivalent improvement or better results with virtual care. Studied depression outcomes and variables include depression severity, severity improvement rate, depression-free days (DFDs), response rate, remission rate, treatment aspects such as adherence, and depression interrater reliability for diagnosis.

Perceptions regarding virtual care were favorable to virtual care or equivalent to in-person care according to almost every metric studied. Examined perceptions include patient satisfaction, caregiver satisfaction, psychiatrist satisfaction, working alliance, need for time to adapt, willingness to pay for mental health services, and care preferences.

Other variables were significantly improved with virtual care. Studied variables included quality of life, quality-adjusted life years (QALYs), functioning, appointment keeping, number of visits, number of complementary tests ordered, physical health, and comorbid somatization, phobia, and anxiety.

It was also proven that telepsychiatry can be more cost-effective than in-person psychiatry. Examined costs and related variables include consultation time, consultation cost, cost-effectiveness, per-session costs, physician time cost, nurse care manager cost, system cost, system maintenance costs, cost per DFD or QALY, and post–start-up incremental cost per DFD or QALY gained.

This review aimed to map the literature on use of telepsychiatry to treat depression by physicians as well as to examine the effects and perceptions regarding telepsychiatry and how it compares to traditional in-person care. A total of 28 articles were included [10-37], and all 13 nonreview articles were analyzed further [10-22]. The generalizability of the findings may be modulated by the fact that most studies were recent, but it may be limited for primary care populations and countries other than the United States. Although quality appraisal of the studies was not part of the methodology of this study, bias risk may be modulated by the high proportion of randomized studies, the low proportion of clear and explicit declared absence of conflicts of interest, and the average study size, which may arguably be considered sufficiently small or large. The analyzed articles contained measured outcomes related to depression variables, perceptions, and other variables. All applicable studies resulted in improvement in depression with the use of telepsychiatry, which was always measured to be equivalent or better for virtual care compared with in-person psychiatry. Patient satisfaction and other perceptions were examined, and telepsychiatry again performed at least as well as in-person care according to almost every metric studied. Quality of life, functioning, and similar variables were significantly improved with virtual care. Telepsychiatry also tended to perform better than in-person care. Cultural sensitivity and collaborative care were also part of some studied telemedicine programs.

Additional research should be conducted on themes already covered in the articles included in this review. These themes include depression outcomes, perceptions such as patient satisfaction, cost-efficiency, and other outcomes. However, stakeholders would benefit from larger study populations, more studies conducted outside the United States, and more studies about general practitioners, as well as fewer potential conflicts of interest and more transparency.

Research should also be conducted on aspects of telepsychiatric care of patients with depression that are not covered as extensively in the included articles. These aspects include effects on suicide risk, suicidal ideation incidence, evolution of comorbid diseases, health behaviors, societal productivity, work productivity and absenteeism, personal and household income, marital outcomes, and parenthood outcomes, as well as the well-being of caregivers and relatives. Many perceptions could be studied, including those of patients, physicians, health care professionals, caregivers, and relatives. Studies should also be conducted regarding medical student training in telepsychiatry [5].

Research themes could be explored through various methodologies, including systematic reviews and randomized clinical trials.

The results of this review are subject to limitations. As definitions of telepsychiatry may differ, the scope of this review might also be considered too narrow in various ways, such as its focus on physicians, its exclusion of psychotherapy as the main intervention, its exclusion of patients with anxiety disorders, and other eligibility criteria that may be considered limited or firm. The chosen search terms may have further influenced the identification of relevant articles in this direction or may have hindered the identification of relevant articles such as studies about telephone medicine, asynchronous telemedicine, pediatricians, or special patient populations. The selected databases may have also influenced the results of this review; however, this decision was supported by the characteristics of the selected databases as well as the preliminary search. Arguably, other databases could have been considered. Similarly, not searching the gray literature may have influenced the results; however, it was decided that this review would be limited to peer-reviewed articles. Backward and forward reference searching has not been conducted. Including other professions may have better covered some research themes, such as quality of work life, quality of personal life, health behaviors, and other aspects. Some may also disapprove of the use of the PRISMA-ScR guidelines or the Arksey and O’Malley framework [9]. The choice to include a case study may be criticized, as the study population consists of only 1 patient [15]. Finally, the generalizability of the findings may be limited for certain locations, medical specialties, patient subsets, and other clinical environment characteristics not represented among the included articles.

This review examines the literature on telemedicine in the care of depression by physicians, as well as its related effects and perceptions. More research is recommended to fully understand the current and potential roles of telepsychiatry when used by physicians caring for patients with depression. This research should examine various outcomes, including depression variables such as symptom severity and suicidal risk, perceptions such as stakeholder satisfaction and working alliance, health care outcomes such as cost-effectiveness, and other outcomes such as quality of life and work productivity. Studies should be conducted in various clinical contexts, such as urban or rural primary care, and in developing countries.

This review suggests that telemedicine tends to be at least as effective for depression care compared with in-person care, and it may be more cost-effective. Patient satisfaction tends to be high and perceptions tend to be favorable.

As depression is a highly prevalent and burdensome disease, its toll on patients’ mental and physical health as well as its health care burden can probably be reduced by improving and implementing virtual psychiatric care of depression by physicians.