There are more people who could benefit from behavioral health services than can be served by the currently existing resources for care [1,2]. Mood and anxiety disorders are highly prevalent in the general population [3,4]. These disorders are disabling to individuals and burdensome to communities, resulting in increased service use, loss of productivity, and poorer outcomes for pre-existing medical conditions [5-7]. Common behavioral health problems are frequently treated in primary care settings because of the relative scarcity of behavioral health specialists in many areas [8,9]. However, primary care providers (PCPs) often lack the training and resources to manage these problems effectively, resulting in overstrained primary care practices and the potential for suboptimal care [8,10]. Novel approaches are needed to improve the scale, delivery, and cost efficiency of behavioral health care.

The collaborative care model (CoCM) aims to meet this vast need [11]. The CoCM is a well-established mode of treatment for common behavioral health disorders in primary care settings. Briefly, in this model, a PCP systematically screens patients for common behavioral health disorders and refers those in need to a behavioral health care manager (BHCM). The BHCM, typically colocated with the PCP, sees patients for an initial assessment and provides time-limited psychotherapy and ongoing evaluation. The BHCM liaises with a consulting psychiatrist who provides treatment recommendations. The psychiatrist may supervise multiple BHCMs at multiple primary care sites, thus significantly extending their reach. The BHCM tracks patient outcomes with regularly administered symptom rating scales (eg, the Patient Health Questionnaire-9 or the Generalized Anxiety Disorder-7 [12,13]) in conjunction with clinical evaluation. The BHCM communicates recommendations to the PCP, who prescribes any necessary medications and remains the clinician of record. This model is effective, widely regarded as the best practice, and has been adopted by many health systems across the world since its introduction in the 1990s [11,14-18].

Pitfalls along every aspect of collaborative care (CC) may contribute to unsuccessful implementation. For example, as the CoCM is a specialized multicomponent service, robust adoption requires provider training and stakeholder buy-in [19]. However, local expertise may be scarce. The colocation of the PCP and BHCM may also be a logistical challenge in some settings. To maximize reach, screening for behavioral health disorders within the primary care population must be systematic [20]; subsequent referrals to the BHCM should be streamlined in the clinical workflow. Monitoring symptoms within a large caseload requires efficient tracking mechanisms. The scope of interventions traditionally available to patients includes brief psychotherapy and limited pharmacotherapy [11]. Even under trial conditions, these options may not meet the needs of some patients [17]. Moreover, there are potential financial challenges to implementation [21].

Digital and mobile health technologies have the potential to support multiple components of CoCM [22]. Technologies such as mobile apps, wearable and ambient sensors, social media, and web-based platforms and devices can improve implementation by supporting provider training, patient screening and referral, monitoring, and treatment. For example, improving live and asynchronous communication technology can link novices with experts who can support the development of a CC program. Technologies can continuously collect interpretable patient data and potentially enable quick, flexible, and targeted care [23]. These resources can act as extenders, decreasing provider workload by automating tasks previously required of clinicians in the CoCM and supporting clinician decision-making by providing actionable clinical information in real time [22,24]. Digital and mobile health technologies can also support patients by providing education and real-time feedback from clinicians, reinforcing concepts learned in therapy, and tracking progress [22]. Emerging research has examined the potential of digital solutions for common behavioral disorders in primary care settings [25-27].

Technology has been recognized as important for the optimal functioning of the CoCM from an early stage. For example, clinicians and researchers recognized the necessity of using caseload registries to manage patient information, track outcomes, and have easy access to information from assessments and follow-up appointments [28-32]. In addition, telephone- and 2-way video-based remote care emerged in the 2000s as a viable method of providing CC [33-36]. Thus, there is extensive precedent and continuous interest in the use of technology to augment CC, which is unsurprising, given the model’s spirit of innovation and dissemination.

This review aims to summarize the current state of research into the ability of digital and mobile health to augment the CoCM, highlight important challenges and limitations, and explore areas for further investigation.

The primary aim of this review is to synthesize the evidence available on digital and mobile health technology in CC settings.

The topic of CC augmented with technology includes a wide variety of potential interventions, both patient-facing and provider-facing, and a wide variety of psychiatric disorders, including depression, anxiety, posttraumatic stress disorder (PTSD), bipolar disorder, and others. Thus, we chose to conduct a scoping review on this broad and emerging topic using the methodology recommended by Arksey and O’Malley [37] and refined by Levac et al [38]. To this end, we conducted six stages of review development: defining the research questions to broadly capture the relevant literature in this emerging field; balancing breadth and feasibility in the search strategy; identifying and selecting studies; extracting and charting data; and synthesizing and reporting the findings, including a forward-looking discussion that could guide future research or quality improvement efforts.

We attempted to answer 3 research questions with this scoping review. First, what digital and mobile health technologies have been studied in CC, and at what levels have they been implemented (patient- vs provider-facing)? Second, what is known about the acceptability and feasibility of digital and mobile health use in the CC context? Third, what, if anything, is known about the impact of these technologies on clinical outcomes in this setting?

We searched 3 web-based databases (Web of Science, MEDLINE, and Embase) during February 2020 with search terms related to CC (sometimes termed integrated care) and various technology interventions, including mobile apps, sensors, social media, and wearable devices. Scans of dark literature from Google Scholar were also completed. Scoping reviews on similar topics were consulted in the search strategy development [39,40]. Additional searches with the same search criteria were conducted in December 2020 and September 2021 to fully capture the status of this rapidly growing field of research. A full list of the search criteria is presented in Multimedia Appendix 1.

Articles were determined eligible for inclusion if they described original research on digital and mobile health to augment the CoCM for the treatment of common behavioral health conditions. Novel technologies such as mobile apps, web-based platforms, ambient or wearable sensors, and social media were included. Technologies that are already well-established in the CoCM and in medicine broadly, such as electronic medical records, caseload registries, or telemedicine, were excluded. Opinion pieces, reviews, books, book chapters, protocols, and commentaries were excluded. Post hoc analyses of the trials included in this review were only included if they helped answer the questions posed by this review; otherwise, they were excluded. Articles written in languages other than English were also excluded. Studies were excluded if they recruited patients from a primary care population or other general medical populations that did not participate in a CoCM, as described previously in this paper. Studies in which substance abuse was the primary diagnosis were excluded.

Owing to the large volume of articles obtained in the initial screening, the articles were divided and screened by 2 independent reviewers (KM and MS). In the initial stage of screening, articles were excluded based on the title and abstract. In the second stage, the remaining papers were excluded based on their full text. At the stage of full-text screening, conflicts were resolved by consensus between the reviewers.

The initial search identified 3817 reports; 661 additional reports were identified in December 2020, and 489 were identified in September 2021. The study selection process is summarized in Figure 1. A total of 20 studies were included in this review. The characteristics of the studies are summarized in Table 1, and more details on individual studies are provided in Multimedia Appendix 2 [15,35,41-58]. The 20 studies included a comparative effectiveness trial of an interactive voice recording (IVR) intervention in a safety net population and a series of post hoc analyses on the same study population [41-44], a randomized trial of an internet support group with an associated subgroup analysis [45,46], 30% (6/20) other randomized trials [35,47-52], a nonrandomized trial [53], 25% (5/20) implementation trials without a comparator [15,54-57], and a qualitative exploratory study [58].

In this scoping review, we investigated the use of digital and mobile health technology in CC settings. This study builds on previous research highlighting the potential of technology in improving behavioral health care [22,23,27].

Our results suggest that the implementation of digital and mobile health technology in CC is currently in its early stages in both clinical research and practice. For example, of the 20 studies, only 1 (5%) study using an app was a large trial [53]. Mobile apps were among the most novel technologies used by studies in this review; no studies using social media, wearable devices, ambient sensors, or other more innovative technologies were identified. Overall, there is more to learn about the use of technologies in this setting.

We believe that digital technology has the potential to support the delivery and scale of the CoCM by mitigating several common challenges to their effective implementation, ranging from provider training, patient screening and referral, monitoring and treatment, and sustainability of the practice. We also believe that the CoCM is especially suited to absorb such changes because of its forward-looking, team-based, and measurement-guided approach. We have used the results of our review to scope future directions for augmenting CC with digital and mobile health technologies focusing on provider training, screening, monitoring, treatment, and sustainability.

Research suggests that the implementation of digital and mobile health technologies in behavioral health care requires the addition of a new team member—a digital health coach—who might help patients and providers navigate digital interventions [78]. This can pose a challenge to the cost-effectiveness of digital technology. The structure of the CoCM has an advantage over standard primary care settings because of the existence of the BHCM, who could potentially act as a digital health navigator, and because of the infrastructure that already exists for collaboration between team members in this model.

There are several challenges in the implementation of digital and mobile interventions in CC. To provide evidence for efficacy, clinical outcomes in technology-augmented care should be superior to the clinical outcomes of traditional CoCM. The most comprehensive study of app-augmented care to date, by Carleton et al [53], found comparable effectiveness for both traditional (46%) and augmented CC (47%), as measured by improvement in Patient Health Questionnaire-9 scores [53]. To compensate for the lack of improvement in clinical outcomes, augmented care can be designed to save the cost of care and provider time. Currently, there is a relative lack of evidence on the effect of augmented care on cost and time, with only 5% (1/20) of the studies in this review addressing this issue [42]. In fact, digital and mobile health technologies may come with an additional cost to the providers as they require investment in the technology itself and in the implementation of its use. From this perspective, relatively low-cost interventions such as SMS text messaging and IVR, such as those implemented by Kroenke et al [62], Wu et al [61], and Adewuya et al [47], may be preferred.

Another common challenge in mobile health technology is the lack of engagement and high attrition rates, as demonstrated in several of the reviewed studies [15,53,56]. In augmented CC, the introduction of a mobile app or other technology for remote monitoring and treatment will require compliance with assessment modules and engagement with psychoeducational materials and chat communication. Our results show difficulties with engagement, especially in technologies that were not automated and required user-initiated actions to work [15,44,51]. In initial implementations, encouraging engagement and preventing attrition can become a function of the BHCM or digital health navigator. Future research should address these challenges through the lens of implementation science.

Finally, implementation of many of the discussed technologies requires addressing challenges with legal, ethical, and privacy concerns about the use of these data at both the patient and provider levels. This is a common challenge in technology, especially in the clinical context of behavioral health. Our review identified early reports of privacy concerns with regards to tracking [49,52]. These concerns might be amplified when additional tracking technologies for remote monitoring and treatment are introduced. Therefore, novel strategies must be implemented with the utmost concern for privacy in a cooperative and transparent manner.

Our review has several limitations. Methodologically, we did not use a librarian as part of the search strategy or calibration of the exclusion and inclusion criteria. Although improving the feasibility of our review, these strategies may have limited the breadth of our search and screening process. In a few years, as more technologies are implemented in CoCM, we expect a systematic review of the literature to be conducted to assess the evidence. In this review, we focused on the treatment of common behavioral health disorders in the CC setting specifically because of the unique approach and structure of this model. Therefore, we excluded studies of digital and mobile health technologies in related settings, such as primary care and substance abuse treatment. Nevertheless, as digital and mobile health technologies are used both in primary care and substance abuse, we drew and built on this research in our discussion.

The use of digital and mobile health technologies in the CoCM is still limited. Digital technology was the most successful when it was integrated into the existing workflow without relying on the patient or provider initiative. The effect of digital and mobile health on clinical outcomes in CoCM remains unclear and requires additional clinical trials. To advance the use of digital and mobile health in CoCM, we have introduced a forward-looking discussion for augmenting CC with a focus on improving access to care, remote patient monitoring, and enhancing treatment.