Facilitators and Barriers to Implementing Mobile Mental Health Interventions: Qualitative Study of the Consolidated Framework for Implementation Research in Pediatric Oncology Providers

Introduction

Approximately 80,000 adolescents and young adults (AYAs) are diagnosed with cancer annually in the United States [1]. Cancer disrupts key developmental milestones for AYAs, such as identity formation, transitioning to independence, relationship development, and achieving educational and vocational goals [2,3]. AYAs living with cancer and AYA cancer survivors (hereafter referred to as AYA cancer survivors) also have unique psychosocial needs that may be unmet throughout their cancer experience [4,5]. Consequently, AYA cancer survivors experience a myriad of mental health issues, including anxiety, depression, and distress, which often persist into long-term survivorship [6,7].

Evidence-based interventions for AYA cancer survivors are a new and emerging area of research [8-11]. The number of AYA-specific psychosocial intervention trials conducted annually has increased significantly from 2 during the years 2007-2014 to 11 during the years 2015-2021 [9]. This growing research activity highlights a commitment to developing evidence-based, developmentally tailored interventions that address the unique needs of this population [12]. As new interventions continue to emerge, leveraging technology has the potential to address barriers that AYAs face in accessing (eg, lack of service availability and financial challenges) and using (eg, time limitations and discomfort in sharing personal problems) face-to-face services [13,14]. As digital natives who have been immersed in technology from a young age, AYAs frequently use digital tools in their daily lives for communication, entertainment, and to access health information and support [15,16]. In recent years, there has been growing research on digital health interventions designed for AYA cancer survivors, with most studies focusing on symptom management, behavior change, self-care, and emotional health [17]. Digital health interventions have demonstrated feasibility and acceptability for AYA cancer survivors [18]. Compared with standard care, digital psychosocial interventions have been found to significantly reduce anxiety and depression and increase resilience in children and AYAs undergoing cancer treatment [10,19]. With 95% of teenagers having access to smartphones [20], mobile health (mHealth) interventions have the potential to expand evidence-based psychosocial care.

On average, it takes 17 years for research evidence to be translated into clinical practice, and only 1 in 5 evidence-based interventions are eventually implemented into routine clinical care [21-23]. While mHealth apps align with AYAs’ digital preferences and hold promise in expanding access to psychosocial care, their integration into clinical care remains a significant research-to-practice gap. Although it is well recognized that successful mHealth implementation depends on the commitment and support of both patients and providers, little prior research has examined provider perspectives [24]. Recent implementation research studies have examined provider perspectives on mHealth tools such as symptom-tracking apps for patients, decision support apps for providers, and cancer triage systems for the general public [25-27]. However, studies have not yet examined provider perspectives on the implementation of psychosocial mHealth interventions. Providers are key stakeholders and partners whose insights are essential for understanding what influences the uptake and implementation of mHealth apps as standard of care within clinical settings.

To address these gaps, we conducted a directed content analysis of qualitative interview data from AYA oncology providers. The primary objectives of our study were to identify facilitators and barriers to implementing new and emerging mHealth apps into routine clinical care from the perspective of AYA oncology providers. Our analysis was guided by the Consolidated Framework for Implementation Research (CFIR), a widely used determinant framework to examine contextual factors influencing implementation effectiveness [28].

Methods

Ethical Considerations

This study was approved by the Seattle Children’s Hospital (SCH) Institutional Review Board (number STUDY00004006). All participants provided written informed e-consent via REDCap (Research Electronic Data Capture; Vanderbilt University). Any and all quotes have been deidentified. Providers received a small token of appreciation (ie, water tumbler) for participation in the study. No other monetary incentives were provided.

Design, Setting, and Participants

Eligible participants were health care providers in the SCH Cancer and Blood Disorders Clinic. All patient-facing providers in the clinic, regardless of role, were eligible to participate in the study. This included psychosocial clinicians (social workers, therapists, and psychologists), oncologists (physicians and fellows), and nurses (nurse practitioners, inpatient, and outpatient) who would potentially integrate mHealth apps into their clinical workflows. Past studies have focused primarily on the perspectives of oncologists. In the interest of inclusivity, we used provider networks across medical and psychosocial staff to try to get equal representation across 6 different patient-facing roles (occupational therapist, pediatric oncologist, psychologist, nurse, social worker, or mental health therapist). We advertised the study on SCH oncology clinic listservs and newsletters and used snowball sampling to identify additional potential participants for recruitment. Participants were queried for convenient appointment times and scheduled for participation in a 1:1 interview. Before the interview, participants completed a brief demographics survey via REDCap.

Research Team

Authors’ backgrounds included health services research (TMP and NL), implementation science (NL), digital health (SJHH, MP, TMP, and EJC), clinical and pediatric psychology (SJHH, MP, TMP, and NL), pediatric oncology (EJC), psycho-oncology research (MP and NL), and qualitative research (KSB and NL). Authors have extensive background and experience in use of the CFIR framework and qualitative methods (SJHH, MP, TMP, KSB, and NL), which informed study design and research objectives. Authors also have extensive clinical experience with AYA cancer survivors (MP, EJC, and NL), which informed the interview guide, codebook development, and clinical implications drawn from the study.

Interviews

Semistructured 1:1 interviews were conducted with 20 health care providers via videoconference and were audio-recorded. This sample size was determined based on qualitative research guidelines recommending group sizes of 15-20 to achieve thematic saturation (no new themes or qualifying patterns) [29,30]. Interviews were deidentified and transcribed verbatim. A clinical psychologist trained in qualitative research methods (NL) conducted interviews from July 2024 to September 2024. Individual interviews were 30-50 minutes in length (average 40 minutes) and focused on facilitators and barriers to incorporating emerging mHealth apps as psychosocial standard of care. The interview script is shown in Multimedia Appendix 1. The interviewer memoed during the data collection process. Interviews were conducted at a single cross-sectional timepoint. We adapted a standardized CFIR interview guide for mHealth and the population of AYAs with cancer focused on the CFIR Innovation, Outer Setting, and Inner Setting domains [28].

Qualitative Analysis

We followed the SRQR (Standards for Reporting Qualitative Research) guidelines for complete reporting of qualitative research data collection and analysis procedures [31]. The completed SRQR checklist is provided in Multimedia Appendix 2. The overall analysis approach was a directed content analysis aimed at identifying facilitators and barriers to implementation of mHealth apps in routine care. We used a standardized CFIR codebook with CFIR domains and constructs, and CFIR definitions. We then developed coding definitions that were specific to mHealth innovations and the population of AYAs with cancer. Using this initial codebook, SJHH, MP, and NL independently coded the first 3 interview transcripts to determine relevance, validity, and fit of each relevant CFIR domain and construct. Over consecutive coding meetings, the analysis team (SJHH, MP, and NL) independently coded transcripts in sets of 3 and iteratively discussed and addressed codebook discrepancies and refined definitions until the codebook was finalized.

The final version of the codebook consisted of 16 of the 67 constructs from the CFIR that were most appropriate to digital mental health interventions. Standardized protocol for applying the CFIR framework is to select a small subset of relevant constructs and domains, specifically operationalized for each research project in accordance with the CFIR qualitative study design guidelines [32]. The most relevant 16 constructs operationalized for our study spanned across all 4 CFIR domains: Innovation (Source, Evidence-Base, Relative Advantage, Adaptability, Design, and Cost), Outer Setting (Local Attitudes and Local Conditions), Inner Setting (Information Technology Infrastructure, Work Infrastructure, Relational Connections, Communications, Recipient Centeredness, Compatibility, and Access to Knowledge and Information), and Individuals (Implementation Team Members). A primary coder (SJHH) and secondary coders (MP and NL) independently coded all transcripts, and all coders met twice monthly to discuss discrepancies and reach consensus. Interview transcripts were managed and analyzed using DeDoose software (SocioCultural Research Consultants).

The interviewer and coder (NL) had prior research relationships with some providers who participated in the study. The interview guide was designed so that questions and probes placed equal emphasis on facilitators and barriers, ensuring that participant feedback covered both domains. Research relationships and solicitations for feedback have historically fostered constructive dialogue, including weighing the pros and cons of different approaches. Dissenting opinions were shared openly and were explicitly encouraged. Reflexive memos and extensive audit trails documenting coding decisions were further used to enhance analytic rigor.

Results

Demographics

Participants were 20 oncology providers (mean 39, SD 7.0, range 29-56 years; Table 1). Most of the sample was female (80%) and self-identified as non-Hispanic White (85%). In terms of provider roles, 60% were medical providers and 40% were psychosocial providers. Provider demographics were representative of the overall staff within the clinic. The mean number of years participants had provided clinical care for oncology patients was 10 (SD 9.2, range 1-24) years. On average, participants reported providing care for 20 AYA patients per month (SD 21.0; range 4-100).

CFIR Framework

We identified facilitators and barriers to mHealth implementation by using constructs across CFIR domains (Innovation, Outer Setting, Inner Setting, and Individuals) and indicated whether each CFIR construct was a facilitator or barrier using (+) or (–), respectively. Thematic saturation was reached with our study sample. A visual representation of the interrelationships among the Innovation, Outer Setting, Inner Setting, and Individuals domains is shown in Figure 1. We demarcated each CFIR construct as a facilitator or barrier to provide a holistic synthesis of our findings. All findings were relevant to AYA cancer survivors across the cancer care continuum from diagnosis to long-term survivorship. Exemplar quotes for each CFIR construct are shown in Table 2.

‎

Innovation

Source (+)

Most providers emphasized the importance of mHealth apps being recommended by trusted sources such as colleagues on their clinical team or national medical specialty associations. Providers also highlighted the value of AYA involvement in developing these apps and incorporating their perspectives and feedback, as “people get excited when our own patients are involved in the creation and implementation of [programs and initiatives]” [P14].

Evidence-Base (+)

Many providers valued robust scientific evidence demonstrating the effectiveness of mHealth apps and being able to see whether apps would be most beneficial for AYA cancer survivors. One provider stated, “[It would] feel like a wasted opportunity to be able to introduce a useful tool if it’s not grounded in any good evidence and theory” [P3].

Relative Advantage (+)

Providers reported several advantages to using mHealth apps compared with traditional clinical practice. Providers emphasized the developmental appropriateness of apps and stated that AYA cancer survivors can “engage within the app in a way that may actually speak their language a bit better than an oncologist sitting in front of them” and that can “facilitate them being able to express themselves in a different kind of way that is probably a little bit more fitting to them” [P3]. Providers also described how mHealth apps could expand access to psychosocial care for AYA cancer survivors following cancer treatment as they transitioned back to a new normal.

Adaptability (+)

All providers discussed the adaptability of mHealth apps. Given that AYAs have unique needs and interests, “an app that has really good uptake is [going to be] one that is able to address [what] an AYA is most interested in” [P1]. A common suggestion from providers for tailoring mHealth app content to AYAs’ unique psychosocial needs was incorporating social connection features (eg, connecting with AYA peers with cancer).

Design (+)

Providers stated that mHealth apps should be well designed and intuitive to use. One provider stated, “I think it's a great way to get health content...to this population as long as A: the content is sound and B: the delivery is engaging to them. I think those are the two key ingredients” [P9]. Based on their experiences with AYA cancer survivors, providers suggested that content should be presented in an interactive and time-efficient format and be highly personalized to their interests. Some app features of interest include short informational videos, gamification elements (eg, leaderboards and streaks), and mood and symptom monitoring.

Cost (+/–)

Nearly half of the sample stated that they were concerned about cost and exclusively interested in recommending mHealth apps that were available for free. Providers emphasized the importance of affordability, with one provider stating, “I wouldn’t recommend ones that are not free, especially if there were a large charge” [P20].

Outer Setting

Local Attitudes (+)

Providers shared a commitment to meeting AYA cancer survivors’ mental health needs and were receptive to using mHealth apps to support clinical needs. One provider stated, “A lot of providers probably...sense the impact of their patients’ mental health on care and may not have the tools...that may be able to help their patients in this way. So [apps] may actually help them feel sort of empowered to at least provide something that the patient could use in the interim before meeting with a mental health provider” [P4].

Local Conditions (–)

Some providers identified challenges within the current conditions of the clinic that could hinder the implementation of mHealth apps. A common challenge, particularly among nurses and social workers, was the time constraints in their clinical workload. To address this, providers emphasized the need for systemic support of clinic-wide mHealth implementation efforts. One provider stated, “Some of the buy-in would also have to be systemic support...It would have to come through management to say we want to provide this opportunity for anyone that’s interested to have this time carved out of their clinical day to put this work into it” [P19].

Inner Setting

IT Infrastructure (+)

Many providers recommended integrating information from mHealth apps (eg, data that show that the patient is improving or struggling) into electronic medical record systems. One provider stated, “Incorporating the results or critical information from that tool back into the environment and EMR...is a critical element for the clear communication and for longevity of the app” [P2].

Providers also suggested leveraging existing technological tools in patient rooms, such as iPads and TVs, to support adoption. For example, they described how educational videos and alerts are preloaded onto TVs for patients and their families. Similarly, TVs could be used to display tutorials of the functions and features of mHealth apps and reminders to download and use them.

Work Infrastructure (+/–)

Providers highlighted how mHealth apps could help with their limited bandwidth, as “having another tool that doesn’t actually require really additional provider time that we may not have would be super useful” [P3]. However, they emphasized that tasks and responsibilities would need to be integrated or automated into their existing workflow to streamline the adoption of mHealth apps.

Relational Connections (–)

Providers noted operating in silos and a lack of awareness about what other clinical teams do and how they collaborate. One provider described having “a lack of clarity...with roles and scope of practice” and “how “we’re in such extreme silos...[that] there is just kind of this lack of understanding of what other services do in comparison to your own” [P14]. As a result, it has been “challenging to understand who’s doing what and how, where there’s reason for overlap...and best practices of how things are delivered and who is delivering it or who’s responsible for it” [P2].

Communications (+)

Although there was no consensus about the frequency of sharing information about mHealth apps among providers, they mentioned that a combination of face-to-face and written communication methods, such as email, one-pagers, and presence at team meetings, would be helpful. One provider stated, “Those Friday AM talks...have always been really well-received...I just feel more visibility about these things and the work that we’re doing or that is being done to support our patients from the psychosocial mental health perspective...I feel like the providers are craving it and needing it...And I think it would actually create a lot of buy-in” [P7].

Recipient Centeredness (+)

Providers demonstrated shared values around supporting and tailoring mHealth apps to the needs and interests of AYA cancer survivors. A commonly perceived need was social connection. Providers also described how AYAs have previously engaged in gamified activities. For example, one provider shared that AYAs “would get so competitive and see that another teen had gotten a certain number of points and they wanted to try to beat them” [P20]. They suggested that a similar leaderboard feature could be incorporated into mHealth apps to enhance engagement.

Compatibility (+)

Providers stated that mHealth apps could “play a role across [the] treatment continuum,” particularly for AYAs getting off treatment and into survivorship as “mental health concerns persist or can absolutely worsen during that time” [P8]. Providers stated that mHealth apps could be “a tool to use especially in moments of distress, which are common in survivorship” or when patients are “coming for scans or labs and have a really high increase of anxiety or stress” [P19]. AYA cancer survivors receiving active treatment also “have a lot of downtime” and mHealth apps would be “a really great way to...harness some of that when they’re between appointments or waiting after labs and before the provider comes in the clinic” [P5]. Providers also described how mHealth apps could be universally used with all AYA cancer survivors, which would “be less of a strain on...the more limited resources that we have for individual patients” [P8].

Access to Knowledge and Information (+)

Providers expressed a need for guidance on the benefits of mHealth apps and how to promote them to AYA cancer survivors. One provider stated that “very basic education about...what it’s for and why it’s important” would be needed so that providers “feel engaged and invested” [P4].

Individuals: Implementation Team Members (+)

Providers recommended bedside nurses and social workers as the clinical team members who would need to be involved as the most ideal potential clinic partners for the successful implementation of mHealth apps. Providers also stated that buy-in and strong engagement from clinical team members would be needed. One provider stated, “You could have created the best thing ever and it won’t make a difference. What you need is a boots-on-the-ground clinic champion to make an impact at the bedside” [P9].

Discussion

Principal Findings

To our knowledge, this is the first study to examine facilitators and barriers to implementing mHealth apps in routine clinical care from the perspective of AYA oncology providers. Constructs across the CFIR Innovation, Outer Setting, Inner Setting, and Individuals domains were identified as facilitators and barriers to mHealth implementation. Our study addresses a key gap in the design-to-implementation pipeline of emerging mHealth interventions.

In terms of facilitators to mHealth implementation, a central theme was strong alignment with patient needs. Providers were interested in collaborating on the development of mHealth apps and emphasized that they should be co-designed with AYAs. Strong research evidence and user-friendly design were other key facilitators, with providers stressing that mHealth apps needed to be both effective and engaging for AYA cancer survivors. Providers viewed mental health care as a valuable component of care in oncology and felt optimistic about the potential of mHealth apps to improve patient outcomes and advance mental health equity. They also expressed openness to receiving guidance and training on using and recommending mHealth apps to AYA cancer survivors. Providers identified bedside nurses and social workers as key implementation partners and highlighted the importance of strong clinical team engagement and buy-in.

Providers also identified several implementation barriers. Cost and accessibility were common concerns, with providers expressing hesitancy in recommending subscription-based or high-fee apps. Usability was another common barrier—providers emphasized that AYA cancer survivors would not want to use mHealth apps with a slow user interface. In terms of organizational barriers, providers reported that heavy clinical workloads and limited cross-team collaboration would make it challenging to implement mHealth apps. Providers further acknowledged the potential challenges of integrating these practices into existing clinic workflows, often referencing the high task demands and the time burden associated with supporting mHealth app implementation.

Comparison With Prior Work

Our findings align with prior research with health care providers who have found positive attitudes and receptivity toward mHealth apps across various health care conditions and disease management functions. A growing body of research also supports the use of mHealth apps in oncology care, which provide opportunities to address patient needs and help them feel engaged throughout treatment [24]. However, to our knowledge, this is the first qualitative study to examine oncology provider perspectives on mHealth implementation. It is also the first to focus specifically on the AYA age group and addressing psychosocial needs in this population. Previous studies that examined provider perspectives have focused on co-designing mHealth interventions with patients and families. Features such as social connection, symptom tracking, and appointment management have emerged as important considerations for these interventions [33,34]. With regard to facilitators of mHealth adoption, our findings align with prior research that has found that providers emphasize the need for evidence-based content and user-friendly design [35]. Our findings support previous recommendations to involve both patients and providers throughout the app development process, ensuring a balance of user and clinical expertise [36,37]. Beyond intervention design, providers offered insights on the integration of mHealth tools and interventions into clinical workflows through electronic health records and existing electronic systems [38]. Providers also shared suggestions for establishing partnerships with boots-on-the-ground clinical champions for large-scale adoptions.

With regard to barriers, our findings align with existing literature on concerns about data privacy, provider liability, and time burden associated with adopting new digital tools [24,34,39]. Our study extends this work by identifying previously unreported barriers, namely, fragmented communication between clinical teams and unclear ownership of implementation tasks. Our findings also highlight opportunities to integrate both medical and psychosocial provider roles into the implementation process and the use of mHealth tools across the cancer care continuum. This includes using mHealth tools during inpatient or outpatient care and throughout the active treatment and posttreatment phases. Additionally, a qualitative review conducted by Jacob et al [40] found that only 3 out of 50 studies examining clinician adoption of mHealth tools used the CFIR framework. This highlights the added value of our study in applying a rigorous implementation science approach to understanding factors that enable and hinder mHealth implementation. While the studies that used the CFIR framework primarily identified structural and organizational factors influencing implementation [41-43], our findings uniquely underscore the value of examining team-level processes when implementing mHealth tools in complex, multidisciplinary care environments.

Limitations

The findings of our study should be interpreted in the context of several limitations. First, our study was conducted at a single academic hospital, which may limit the generalizability of our results. Providers at this institution may have more experience with or interest in digital health than providers in other clinical settings. Second, our sample lacked demographic diversity, with most participants identifying as female and non-Hispanic White. Although this reflects the provider population in the geographic location in which our study was conducted, it may limit the applicability of our findings to more diverse clinical settings. Third, we intentionally predefined 16 CFIR constructs that were applicable to facilitators and barriers within our settings. With the CFIR having 67 constructs total, future research may explore additional factors that could impact implementation, such as policies, partnerships, or external pressures. Fourth, implementation contexts are dynamic. Our interviews captured the perspectives of providers on mHealth implementation as of 2024.

Conclusions

Our findings synthesized key implementation considerations for incorporating mHealth apps as a standard of care in psychosocial support for AYA cancer survivors. Based on these findings, recommendations for practice include co-designing mHealth apps with both patients and providers to ensure alignment with user preferences and clinical needs. These tools should then be embedded within existing clinical workflows such as electronic health records to reduce provider burden and ensure seamless integration. Improvements in cross-team collaboration and communication can also be achieved by clearly delineating roles and responsibilities across clinical teams. Additionally, concerns around data privacy, provider liability, usability, and cost should be addressed to support the sustained use and implementation of mHealth apps. Implementation strategies should also include provider training on the benefits of mHealth apps and best practices for recommending these tools to AYA cancer survivors.

Our findings have important clinical implications. Strong partnerships among AYA cancer survivors, families, and providers will be essential for developing mHealth tools that are patient-centered and evidence-based [44]. Importantly, providers of all represented clinical roles believed that mHealth tools had clinical utility and wanted to be involved in the innovation development through implementation pipeline. They were also ready and willing to partner with researchers in early adoption efforts and had suggestions on how to do so. Future research should focus not only on evaluating the effectiveness of mHealth apps but also on implementing strategies that integrate into clinical workflows and minimize provider burden [45,46]. Understanding how these strategies impact provider uptake and patient outcomes will be crucial to moving toward universal evidence-based digital health care using innovative digital health technologies.