Symposium 5: The Digital Health SIG presents: Human Centered Design: International Applications of a Cross-Cutting Methodology for Adapting and Implementing Evidence-Based Interventions

Human centered design (HCD) is a widely used philosophy and methodology in the field of Human-Computer Interaction that emphasizes the importance of understanding the users for whom the intervention is intended. In an ideal HCD process, users are actively involved throughout an iterative process that includes (1) understanding users’ needs; (2) developing prototypes and obtaining user feedback on them; and (3) refining and improving prototypes until a final design that meets the users’ needs is achieved. Prior research suggests that HCD is vitally important in the development of usable, useful, and engaging digital mental health interventions that are tailored to the needs and preferences of end-users. However, HCD has less often been used for adapting and implementing existing interventions to new populations or cultural contexts. In this symposium we will highlight lessons learned from three studies in which HCD methods were employed to adapt and implement established interventions to a novel context. Dr. Murphy will share her experience using HCD to adapt evidence-based interventions for depression to be delivered digitally to adolescent and young adult cancer survivors in the United States; Dr. Wang will discuss HCD as a strategy for cultural tailoring of an mHealth intervention informed by Dialectical Behavior Therapy for men who have sex with men living with HIV in China; and Dr. Triplett will discuss HCD as a strategy for developing a supervision protocol for lay providers of trauma-focused cognitive-behavioral therapy in Kenya. Presenters will highlight the benefits and limitations of engaging in HCD for the purpose of intervention adaptation and implementation, as well as practical recommendations on combining HCD with other methodological frameworks to guide rigorous research on the adaptation and implementation of evidence-based interventions.

Presentation 1: Human Centered Design to Guide the Preparation of Interventions Within the Multiphase Optimization Strategy (MOST) Framework (Karly M. Murphy, PhD)

Introduction: Intervention development typically consists of designing an intervention with multiple components and then testing whether the intervention, as a whole, has a significant effect in randomized clinical trials (RCT). However, intervention effectiveness and efficiency can be improved by gathering and integrating information about which intervention components contribute to the desired effect and which do not. MOST is a comprehensive framework that can be used to optimize an intervention by systematically identifying the most promising intervention components. MOST includes three phases: 1) preparation, which lays the groundwork for optimization (e.g., identification of a guiding conceptual model, choosing candidate components); 2) optimization, in which a trial (i.e. a factorial experiment) is used to identify the components that meet the optimization criterion; and 3) evaluation, in which the effectiveness of the optimized intervention is confirmed (Collins, et al., 2024). The goal of MOST is to strategically balance intervention effectiveness with affordability, scalability, and efficiency. While the optimization and evaluation phases of MOST have established methodologies recommended to achieve their goals, the preparation phase involves several different tasks (e.g., deriving a conceptual model, identifying and adapting a set of candidate components, conducting pilot work, and identifying the optimization objective) with research-specific flexibility in the methodology used to achieve them. As such, establishing concrete methods to achieve the goals of the preparation phase are a priority for future development of the MOST framework. O’Hara et al. (2022) proposed that the discover, design, build and test framework from the field of human-centered design (HCD) can provide a roadmap to achieve the goals of the preparation phase of MOST. I seek to deepen this line of thinking by focusing on the role HCD methods can play in guiding the adaptation of evidence-based interventions. My NCI-funded study focused on the adaptation of evidence-based interventions for depression to the needs and preferences of adolescent and young adult cancer survivors (AYAs) will serve as a case study of how HCD can successfully guide this aspect the preparation phase of MOST.

Methods: The user centered design process set forth by McCurdie et al. (2012) consists of three iterative steps: (1) concept generation and ideation; (2) prototype design and system development; (3) evaluation. Following the final evaluation phase, the product is deployed. Users are at the center of this process, providing input at the concept generation and ideation step as well as the evaluation step. In our study, we involved users (adolescent and young adult cancer survivors and their providers) in the concept generation and ideation step via co-design workshops. Co-design workshops are a participatory design method that engages key stakeholders in developing content and features that are likely acceptable to and useful for the intended end users. During the workshops, AYAs (N = 14) and providers that serve this population (N = 4) engaged in virtual participatory design activities via Figma such as critiquing an existing digital tool using the “I like, I wish, I wonder” framework, brainstorming responses to “How might we…” questions, and using affinity mapping to organize and prioritize ideas. AYAs also participated in a second workshop focused on obtaining feedback about their experiences engaging in a cognitive-behavioral therapy activity and how it could be adapted for digital delivery in a way that is tailored to AYAs. Data collected during co-design workshops was used to inform development of a mid-fidelity prototype of the tool that was presented to AYAs (N = 14) during semi-structured interviews. Interviews focused on obtaining detailed feedback regarding the overall design of the tool, features meant to promote engagement with the tool, and the content to be included in the tool. Ultimately, this HCD process informed the development of a working prototype of a web-based tool to assist AYAs in depression self-management that consists of five components adapted from evidence-based interventions for depression.

Results: Finding from co-design workshops highlighted AYACS desire for tailored information, brief and engaging content delivery, and a positive and encouraging tone throughout the tool. Semi-structured interviews provided more detailed information about what features might promote user engagement (e.g., communication with other AYACS and providers, education on the value of the tool, and gamification), how content should be tailored to AYACS (e.g., age-appropriate use of multimedia), and ideas for improving the usability and interface of the prototype. HCD provided concrete strategies for the adaptation of existing evidence-based interventions to the needs and preferences of a novel population in the context of the preparation phase of MOST. However, researchers new to HCD methodology may find that it is difficult to (1) determine which activities from the HCD literature are most appropriate for their project; (2) make design decisions in the face of inconsistent feedback from users; and (3) decide when a product is ready to be deployed. Despite these limitations, using HCD ultimately resulted in a digital tool adapted for delivery to AYAs that is expected to not only include components that are likely to be efficacious, but also be usable, useful, and engaging.

Presentation 2: Cultural Tailoring of an mHealth Intervention Using Human-Centered Design Strategies (Liying Wang, PhD)

Introduction: Human-Centered Design (HCD) principles, which prioritize user experience, are increasingly recognized in digital health research for their role in creating products that are not only effective and efficient but also engaging, easy to use, and accessible (Dopp et al., 2019; Lyon & Koerner, 2016). These principles are grounded in a thorough understanding of users through methods such as co-design sessions, usability testing, and iterative prototyping (Courage et al., 2005). HCD principles have been successfully applied across various domains, including healthcare, to enhance service delivery and improve user outcomes (Hekler et al., 2016; Lyon & Lewis, 2016). In the context of mental health interventions, there is a growing emphasis on the importance of cultural responsiveness. Research has shown that culturally adapted interventions are more effective in improving mental health outcomes for Black and ethnically diverse populations compared to standard interventions (Rathod et al., 2018; Arundell et al., 2021). The Ecological Validity Framework (EVF) provides a guide for making these cultural adaptations, addressing factors such as language, the personnel involved, content, and behavior change strategies (Bernal et al., 1995). The formative needs assessment stage of the study identified target users’ pain points that were best conceptualized under the Dialectical Behavior Therapy (DBT) framework of emotions. This study used HCD principles and strategies and cultural adaptation approaches to tailor content and delivery design of a set of DBT skills to our target users.

Methods: We conducted co-design sessions (n=10) and usability testing (n=15) to iteratively refine the intervention. The co-design sessions were web-based, using tools like Mural (LUMA Institute, LLC) and BoardMix (Boyun Technology Limited) for live collaboration. Each session involved two research team members and two participants, with one researcher leading the session while the other took notes and asked follow-up questions as needed. These sessions were informed by needs assessments and focused on two key areas: (1) content design, exploring users’ emotional challenges and coping strategies post-HIV diagnosis through ideating, concept mapping, and priority map, and (2) delivery design, generating ideas for potential interventions or services to support coping. Usability testing utilized the think-aloud method to assess the ease of use of key app features (e.g., medication and mood tracking) and its aesthetic elements (e.g., color palette). In line with the EVF, we implemented cultural adaptations across various dimensions of the intervention, including language, personnel involved in delivery, metaphors, content, concepts, client goals, and behavior change strategies.

Results: The co-design sessions highlighted a disparity in coping strategies available for different emotional intensity levels, with fewer strategies identified for high-intensity emotions (n=19) compared to lower-intensity situations (n=26). The co-design process highlighted that participants were able to flexibly choose among their available set of coping skills, while balancing the feasibility and impact of using such coping strategies. Participants noted that substance use, particularly smoking and drinking, was a common coping mechanism in highly emotional situations. Regardless of the emotional intensity, the most frequently cited strategies included distraction, engaging in conversations with trusted friends or family, problem-solving, and exercise. Many participants mentioned preferring solitude when overwhelmed by emotions. These insights pointed to a significant gap in emotion regulation skills, particularly in handling high emotional arousal. Using the DBT model of emotions, we identified three critical components for the intervention: managing biological changes and emotional experiences, improving verbal and nonverbal emotional expression, and addressing emotional vulnerability, which impacts other aspects of emotional response. In addition, the rich discussion from the co-design session facilitated a deepened cultural understanding of the expressions, behaviors, and process of coping around difficult emotions and informed cultural adaptation of the content through adaptations guided by the EVF framework. For example, participants used a famous Chinese poem by Li Bai when describing their tendency to use substances to cope with emotional distress, which we adopted as a cultural idiom in our content design to connect and engage with our intended users.

The resulting intervention program comprises three main components: individual skills learning via the mobile app, group-based online skills training facilitated by community organization staff, and on-demand phone coaching. The app-based component focuses on teaching emotion regulation and coping skills, tailored to the specific needs of the HIV-positive community. Group-based training offers a supportive environment where participants can share experiences, fostering empathy and mutual support. On-demand phone coaching provides immediate, personalized support, particularly valuable during the early stages following an HIV diagnosis.

Presentation 3: Applying and Adapting Human-Centered Design Methods to Develop Protocols for Using Mobile Phones to Supervise Lay Counselors in Kenya (Noah Triplett, PhD)

Introduction: Given the tremendous gaps in access to mental health care across the globe, scalable and sustainable solutions are needed to increase access to care for the most underserved populations. There has been increased attention on the potential of digital tools, such as mobile phone applications or internet-based treatments, to address the mental health treatment gap by directly targeting clients and patients. Despite the potential of these approaches, they may leave behind key groups who do not have access to cell phones or the ability to use them—thereby risking creating or reinforcing health inequities. In these instances, in-person treatment models, such as task-shifting, may be necessary. Task shifting involves training less specialized providers, including lay providers (e.g., community health workers, teachers) without formal mental health training or experience, to deliver evidence-based interventions under supervision. While evidence continues to support the effectiveness of task-shifting, further research is needed to understand how to scale up and sustain task-shifting, including how to sustain supervision. Digital health tools may play a key role in supporting task-shifting; however, considering the needs and preferences of lay counselors and supervisors while co-designing digital tools for supervision is essential in ensuring digital health equity.

Methods: This study builds from a stepped-wedge cluster-randomized-trial in Western Kenya, wherein teachers and community health volunteers have been trained to provide a culturally-adapted, group-based trauma-focused cognitive behavioral therapy (TF-CBT). Using an iterative and mixed methods approach that used principles and methods from human-centered design, we sought to understand the challenges and opportunities for using mobile phones to supervise lay counselors. Following semi-structured interviews with lay counselors (N=24) and supervisors (N=3), we conducted a co-design session with all participants to co-design implementation guidelines to optimize the use of mobile phones for supervision. Co-design activities included small and large group discussions as well as prototyping and role playing “workflows,” or specific series of actions and steps required to conduct mobile phone supervision. Focus group discussions (N=2) were conducted at the end of the co-design session to gather participant perspectives on the human-centered design process (i.e., reflections on the discussions, prototyping, and role playing).

We evaluated the impact of the co-developed implementation guidelines on the acceptability, feasibility, and usability of mobile phone supervision using a QUAN → qual mixed methods approach for data explanation, wherein qualitative data served to contextualize quantitative results. Counselors that received guidelines (n=30) were compared to a control sequence (n=29) that neither received guidelines nor participated in co-development activities. Qualitative data regarding the human-centered design process were analyzed using thematic analysis.

Results: Lay counselors and supervisors co-developed flexible implementation guidelines to facilitate problem solving around mobile phone supervision while allowing lay counselors to select and adapt potential strategies that met their unique needs and preferences. To minimize resources required and maximize acceptability, guidelines were printed on paper. Guidelines were shared and discussed with newly trained lay counselors in educational outreach visits led by supervisors. Guidelines were associated with significant improvements in acceptability [t(57) = -2.1, p = 0.04] and usability [t(57) = -2.3, p = 0.02] of mobile phone supervision. There was no evidence of a significant difference in feasibility [t(57) = –1.9, p = 0.06]. Qualitative interviews contextualized how guidelines impacted acceptability and feasibility—by setting expectations for mobile phone supervision, emphasizing importance, increasing comfort, and sharing strategies to improve mobile phone supervision.

Lay counselors expressed positive sentiments regarding the human-centered design process. Counselors enjoyed working alongside other counselors and felt it was a valuable experience to learn new strategies from their colleagues. They enjoyed the “collaborative spirit” that emerged as part of the human-centered design process. Counselors felt the human-centered design methods fostered participation by creating opportunities for more people to engage and share their thoughts. They suggested the approach be improved by providing more tangible materials (e.g., hand-outs) and more closely following a schedule of activities.

Discussion (Sarah Fadem, PhD)
There is a need for practical approaches to adapt and implement evidence-based interventions in novel contexts. As seen in the studies presented here, HCD is a practical discipline that offers strategic techniques for assessing the needs of and empowering those most impacted by an intervention to be involved in its implementation.

Across these studies, we see a variety of HCD methods for systematically identifying needs and engaging end users. The tools of HCD capitalize on the experiential expertise of end users in understanding the requirements and limitations of implementation context. This is particularly useful when adapting existing evidence-based interventions to novel cultural contexts and patient populations. The use of prototypes, a core feature of HCD, creates opportunities for commentary, reflection, and externalization of knowledge that may otherwise be difficult to capture in traditional research methods.

Since its origins in the 1950s, HCD has been concerned with achieving context-sensitive solutions. The wealth of research in HCD has yet to be effectively applied to issues of implementation. Examining the unique ways that HCD can facilitate adaptation and implementation can provide advances towards achieving context and cultural-sensitive intervention development and implementation.

Possible Discussion Questions:
What is unique about HCD compared to other methods of tailoring and adapting interventions? What does HCD offer that existing methods do not?
There is a lack of clarity around codesign in implementation science (Metz et al., 2021) Consequently, the involvement of users and reporting of design methods varies significantly. As all presenters describe some aspects of codesign--how can we facilitate true participation in the codesign rather than participants simply acting as data sources, particularly when adapting to different cultural contexts?
How can we adapt the “fail early and often” mentality of design to implementation in healthcare spaces? Should we involve professional designers in academic research? When?
How do we reconcile the model of rigorous clinical research with the rapid, iterative, “fail early and often” methods of HCD?

Chair - Karly Murphy PhD  Student
East Carolina University

Presenter - Liying Wang   Student
Florida State University

Presenter - Noah Triplett   Student
University of Maryland

Discussant - Sarah Fadem PhD  Student
Rutgers Robert Wood Johnson Medical School

Human centered design (HCD) is a widely used philosophy and methodology in the field of Human-Computer Interaction that emphasizes the importance of understanding the users for whom the intervention is intended. In an ideal HCD process, users are actively involved throughout an iterative process that includes (1) understanding users’ needs; (2) developing prototypes and obtaining user feedback on them; and (3) refining and improving prototypes until a final design that meets the users’ needs is achieved. Prior research suggests that HCD is vitally important in the development of usable, useful, and engaging digital mental health interventions that are tailored to the needs and preferences of end-users. However, HCD has less often been used for adapting and implementing existing interventions to new populations or cultural contexts. In this symposium we will highlight lessons learned from three studies in which HCD methods were employed to adapt and implement established interventions to a novel context. Dr. Murphy will share her experience using HCD to adapt evidence-based interventions for depression to be delivered digitally to adolescent and young adult cancer survivors in the United States; Dr. Wang will discuss HCD as a strategy for cultural tailoring of an mHealth intervention informed by Dialectical Behavior Therapy for men who have sex with men living with HIV in China; and Dr. Triplett will discuss HCD as a strategy for developing a supervision protocol for lay providers of trauma-focused cognitive-behavioral therapy in Kenya. Presenters will highlight the benefits and limitations of engaging in HCD for the purpose of intervention adaptation and implementation, as well as practical recommendations on combining HCD with other methodological frameworks to guide rigorous research on the adaptation and implementation of evidence-based interventions.

Presentation 1: Human Centered Design to Guide the Preparation of Interventions Within the Multiphase Optimization Strategy (MOST) Framework (Karly M. Murphy, PhD)

Introduction: Intervention development typically consists of designing an intervention with multiple components and then testing whether the intervention, as a whole, has a significant effect in randomized clinical trials (RCT). However, intervention effectiveness and efficiency can be improved by gathering and integrating information about which intervention components contribute to the desired effect and which do not. MOST is a comprehensive framework that can be used to optimize an intervention by systematically identifying the most promising intervention components. MOST includes three phases: 1) preparation, which lays the groundwork for optimization (e.g., identification of a guiding conceptual model, choosing candidate components); 2) optimization, in which a trial (i.e. a factorial experiment) is used to identify the components that meet the optimization criterion; and 3) evaluation, in which the effectiveness of the optimized intervention is confirmed (Collins, et al., 2024). The goal of MOST is to strategically balance intervention effectiveness with affordability, scalability, and efficiency. While the optimization and evaluation phases of MOST have established methodologies recommended to achieve their goals, the preparation phase involves several different tasks (e.g., deriving a conceptual model, identifying and adapting a set of candidate components, conducting pilot work, and identifying the optimization objective) with research-specific flexibility in the methodology used to achieve them. As such, establishing concrete methods to achieve the goals of the preparation phase are a priority for future development of the MOST framework. O’Hara et al. (2022) proposed that the discover, design, build and test framework from the field of human-centered design (HCD) can provide a roadmap to achieve the goals of the preparation phase of MOST. I seek to deepen this line of thinking by focusing on the role HCD methods can play in guiding the adaptation of evidence-based interventions. My NCI-funded study focused on the adaptation of evidence-based interventions for depression to the needs and preferences of adolescent and young adult cancer survivors (AYAs) will serve as a case study of how HCD can successfully guide this aspect the preparation phase of MOST.

Methods: The user centered design process set forth by McCurdie et al. (2012) consists of three iterative steps: (1) concept generation and ideation; (2) prototype design and system development; (3) evaluation. Following the final evaluation phase, the product is deployed. Users are at the center of this process, providing input at the concept generation and ideation step as well as the evaluation step. In our study, we involved users (adolescent and young adult cancer survivors and their providers) in the concept generation and ideation step via co-design workshops. Co-design workshops are a participatory design method that engages key stakeholders in developing content and features that are likely acceptable to and useful for the intended end users. During the workshops, AYAs (N = 14) and providers that serve this population (N = 4) engaged in virtual participatory design activities via Figma such as critiquing an existing digital tool using the “I like, I wish, I wonder” framework, brainstorming responses to “How might we…” questions, and using affinity mapping to organize and prioritize ideas. AYAs also participated in a second workshop focused on obtaining feedback about their experiences engaging in a cognitive-behavioral therapy activity and how it could be adapted for digital delivery in a way that is tailored to AYAs. Data collected during co-design workshops was used to inform development of a mid-fidelity prototype of the tool that was presented to AYAs (N = 14) during semi-structured interviews. Interviews focused on obtaining detailed feedback regarding the overall design of the tool, features meant to promote engagement with the tool, and the content to be included in the tool. Ultimately, this HCD process informed the development of a working prototype of a web-based tool to assist AYAs in depression self-management that consists of five components adapted from evidence-based interventions for depression.

Results: Finding from co-design workshops highlighted AYACS desire for tailored information, brief and engaging content delivery, and a positive and encouraging tone throughout the tool. Semi-structured interviews provided more detailed information about what features might promote user engagement (e.g., communication with other AYACS and providers, education on the value of the tool, and gamification), how content should be tailored to AYACS (e.g., age-appropriate use of multimedia), and ideas for improving the usability and interface of the prototype. HCD provided concrete strategies for the adaptation of existing evidence-based interventions to the needs and preferences of a novel population in the context of the preparation phase of MOST. However, researchers new to HCD methodology may find that it is difficult to (1) determine which activities from the HCD literature are most appropriate for their project; (2) make design decisions in the face of inconsistent feedback from users; and (3) decide when a product is ready to be deployed. Despite these limitations, using HCD ultimately resulted in a digital tool adapted for delivery to AYAs that is expected to not only include components that are likely to be efficacious, but also be usable, useful, and engaging.

Presentation 2: Cultural Tailoring of an mHealth Intervention Using Human-Centered Design Strategies (Liying Wang, PhD)

Introduction: Human-Centered Design (HCD) principles, which prioritize user experience, are increasingly recognized in digital health research for their role in creating products that are not only effective and efficient but also engaging, easy to use, and accessible (Dopp et al., 2019; Lyon & Koerner, 2016). These principles are grounded in a thorough understanding of users through methods such as co-design sessions, usability testing, and iterative prototyping (Courage et al., 2005). HCD principles have been successfully applied across various domains, including healthcare, to enhance service delivery and improve user outcomes (Hekler et al., 2016; Lyon & Lewis, 2016). In the context of mental health interventions, there is a growing emphasis on the importance of cultural responsiveness. Research has shown that culturally adapted interventions are more effective in improving mental health outcomes for Black and ethnically diverse populations compared to standard interventions (Rathod et al., 2018; Arundell et al., 2021). The Ecological Validity Framework (EVF) provides a guide for making these cultural adaptations, addressing factors such as language, the personnel involved, content, and behavior change strategies (Bernal et al., 1995). The formative needs assessment stage of the study identified target users’ pain points that were best conceptualized under the Dialectical Behavior Therapy (DBT) framework of emotions. This study used HCD principles and strategies and cultural adaptation approaches to tailor content and delivery design of a set of DBT skills to our target users.

Methods: We conducted co-design sessions (n=10) and usability testing (n=15) to iteratively refine the intervention. The co-design sessions were web-based, using tools like Mural (LUMA Institute, LLC) and BoardMix (Boyun Technology Limited) for live collaboration. Each session involved two research team members and two participants, with one researcher leading the session while the other took notes and asked follow-up questions as needed. These sessions were informed by needs assessments and focused on two key areas: (1) content design, exploring users’ emotional challenges and coping strategies post-HIV diagnosis through ideating, concept mapping, and priority map, and (2) delivery design, generating ideas for potential interventions or services to support coping. Usability testing utilized the think-aloud method to assess the ease of use of key app features (e.g., medication and mood tracking) and its aesthetic elements (e.g., color palette). In line with the EVF, we implemented cultural adaptations across various dimensions of the intervention, including language, personnel involved in delivery, metaphors, content, concepts, client goals, and behavior change strategies.

Results: The co-design sessions highlighted a disparity in coping strategies available for different emotional intensity levels, with fewer strategies identified for high-intensity emotions (n=19) compared to lower-intensity situations (n=26). The co-design process highlighted that participants were able to flexibly choose among their available set of coping skills, while balancing the feasibility and impact of using such coping strategies. Participants noted that substance use, particularly smoking and drinking, was a common coping mechanism in highly emotional situations. Regardless of the emotional intensity, the most frequently cited strategies included distraction, engaging in conversations with trusted friends or family, problem-solving, and exercise. Many participants mentioned preferring solitude when overwhelmed by emotions. These insights pointed to a significant gap in emotion regulation skills, particularly in handling high emotional arousal. Using the DBT model of emotions, we identified three critical components for the intervention: managing biological changes and emotional experiences, improving verbal and nonverbal emotional expression, and addressing emotional vulnerability, which impacts other aspects of emotional response. In addition, the rich discussion from the co-design session facilitated a deepened cultural understanding of the expressions, behaviors, and process of coping around difficult emotions and informed cultural adaptation of the content through adaptations guided by the EVF framework. For example, participants used a famous Chinese poem by Li Bai when describing their tendency to use substances to cope with emotional distress, which we adopted as a cultural idiom in our content design to connect and engage with our intended users.

The resulting intervention program comprises three main components: individual skills learning via the mobile app, group-based online skills training facilitated by community organization staff, and on-demand phone coaching. The app-based component focuses on teaching emotion regulation and coping skills, tailored to the specific needs of the HIV-positive community. Group-based training offers a supportive environment where participants can share experiences, fostering empathy and mutual support. On-demand phone coaching provides immediate, personalized support, particularly valuable during the early stages following an HIV diagnosis.

Presentation 3: Applying and Adapting Human-Centered Design Methods to Develop Protocols for Using Mobile Phones to Supervise Lay Counselors in Kenya (Noah Triplett, PhD)

Introduction: Given the tremendous gaps in access to mental health care across the globe, scalable and sustainable solutions are needed to increase access to care for the most underserved populations. There has been increased attention on the potential of digital tools, such as mobile phone applications or internet-based treatments, to address the mental health treatment gap by directly targeting clients and patients. Despite the potential of these approaches, they may leave behind key groups who do not have access to cell phones or the ability to use them—thereby risking creating or reinforcing health inequities. In these instances, in-person treatment models, such as task-shifting, may be necessary. Task shifting involves training less specialized providers, including lay providers (e.g., community health workers, teachers) without formal mental health training or experience, to deliver evidence-based interventions under supervision. While evidence continues to support the effectiveness of task-shifting, further research is needed to understand how to scale up and sustain task-shifting, including how to sustain supervision. Digital health tools may play a key role in supporting task-shifting; however, considering the needs and preferences of lay counselors and supervisors while co-designing digital tools for supervision is essential in ensuring digital health equity.

Methods: This study builds from a stepped-wedge cluster-randomized-trial in Western Kenya, wherein teachers and community health volunteers have been trained to provide a culturally-adapted, group-based trauma-focused cognitive behavioral therapy (TF-CBT). Using an iterative and mixed methods approach that used principles and methods from human-centered design, we sought to understand the challenges and opportunities for using mobile phones to supervise lay counselors. Following semi-structured interviews with lay counselors (N=24) and supervisors (N=3), we conducted a co-design session with all participants to co-design implementation guidelines to optimize the use of mobile phones for supervision. Co-design activities included small and large group discussions as well as prototyping and role playing “workflows,” or specific series of actions and steps required to conduct mobile phone supervision. Focus group discussions (N=2) were conducted at the end of the co-design session to gather participant perspectives on the human-centered design process (i.e., reflections on the discussions, prototyping, and role playing).

We evaluated the impact of the co-developed implementation guidelines on the acceptability, feasibility, and usability of mobile phone supervision using a QUAN → qual mixed methods approach for data explanation, wherein qualitative data served to contextualize quantitative results. Counselors that received guidelines (n=30) were compared to a control sequence (n=29) that neither received guidelines nor participated in co-development activities. Qualitative data regarding the human-centered design process were analyzed using thematic analysis.

Results: Lay counselors and supervisors co-developed flexible implementation guidelines to facilitate problem solving around mobile phone supervision while allowing lay counselors to select and adapt potential strategies that met their unique needs and preferences. To minimize resources required and maximize acceptability, guidelines were printed on paper. Guidelines were shared and discussed with newly trained lay counselors in educational outreach visits led by supervisors. Guidelines were associated with significant improvements in acceptability [t(57) = -2.1, p = 0.04] and usability [t(57) = -2.3, p = 0.02] of mobile phone supervision. There was no evidence of a significant difference in feasibility [t(57) = –1.9, p = 0.06]. Qualitative interviews contextualized how guidelines impacted acceptability and feasibility—by setting expectations for mobile phone supervision, emphasizing importance, increasing comfort, and sharing strategies to improve mobile phone supervision.

Lay counselors expressed positive sentiments regarding the human-centered design process. Counselors enjoyed working alongside other counselors and felt it was a valuable experience to learn new strategies from their colleagues. They enjoyed the “collaborative spirit” that emerged as part of the human-centered design process. Counselors felt the human-centered design methods fostered participation by creating opportunities for more people to engage and share their thoughts. They suggested the approach be improved by providing more tangible materials (e.g., hand-outs) and more closely following a schedule of activities.

Discussion (Sarah Fadem, PhD)
There is a need for practical approaches to adapt and implement evidence-based interventions in novel contexts. As seen in the studies presented here, HCD is a practical discipline that offers strategic techniques for assessing the needs of and empowering those most impacted by an intervention to be involved in its implementation.

Across these studies, we see a variety of HCD methods for systematically identifying needs and engaging end users. The tools of HCD capitalize on the experiential expertise of end users in understanding the requirements and limitations of implementation context. This is particularly useful when adapting existing evidence-based interventions to novel cultural contexts and patient populations. The use of prototypes, a core feature of HCD, creates opportunities for commentary, reflection, and externalization of knowledge that may otherwise be difficult to capture in traditional research methods.

Since its origins in the 1950s, HCD has been concerned with achieving context-sensitive solutions. The wealth of research in HCD has yet to be effectively applied to issues of implementation. Examining the unique ways that HCD can facilitate adaptation and implementation can provide advances towards achieving context and cultural-sensitive intervention development and implementation.

Possible Discussion Questions:
What is unique about HCD compared to other methods of tailoring and adapting interventions? What does HCD offer that existing methods do not?
There is a lack of clarity around codesign in implementation science (Metz et al., 2021) Consequently, the involvement of users and reporting of design methods varies significantly. As all presenters describe some aspects of codesign--how can we facilitate true participation in the codesign rather than participants simply acting as data sources, particularly when adapting to different cultural contexts?
How can we adapt the “fail early and often” mentality of design to implementation in healthcare spaces? Should we involve professional designers in academic research? When?
How do we reconcile the model of rigorous clinical research with the rapid, iterative, “fail early and often” methods of HCD?

Chair - Karly M. Murphy, PhD, PhD,
East Carolina University

Presenter - Liying Wang,
Florida State University

Presenter - Noah S. Triplett,
University of Maryland

Discussant - Sarah Fadem, PhD,
Rutgers Robert Wood Johnson Medical School