Design and Development of a Digital Program for Training Non-Specialist Health Workers to Deliver an Evidence-Based Psychological Treatment for Depression in Primary Care in India

Abstract

Background:

Digital technologies hold promise for building capacity of non-specialist health workers towards scaling up depression care in low-resource settings. The purpose of this study was to describe the systematic approach to designing a digital program for training non-specialist health workers to deliver an evidence-based brief psychological treatment for depression, called the Healthy Activity Program, in primary care in rural India.

Methods:

The design and development of the training program involved 5 steps: 1) develop program blueprint; 2) create instructional content; 3) digitize content for a smartphone app; 4) develop a platform for uploading and hosting the digital content; and 5) user-testing and refinements to ensure program functioning. This was followed by field-testing and focus group discussions with non-specialist health workers recruited from primary care facilities in Madhya Pradesh, India, to inform further modifications and improvements to the digital training program.

Results:

Training program development occurred over 12-months, and the final digital training consisted of 16 modules with videos, role-plays, and digital content tailored to the local culture and context. Focus group discussions with 19 non-specialist health workers generated three key action items and modifications to the digital training in response to participant feedback: 1) addressing technical challenges by making the digital content available offline; 2) account for low digital literacy by including a comprehensive orientation session about navigating the smartphone app; and 3) addition of remote coaching to support participants in completing the training.

Conclusions:

This study illustrates a step-wise approach to combine evidence-based content with iterative feedback from stakeholders to develop a digital training program tailored to the context in a low-resource setting. Further research is needed to validate this approach and to evaluate the effectiveness of the final modified digital training program, while considering whether this approach can be adopted and replicated in other settings.

Introduction

Mental disorders are a leading cause of disease burden globally, yet most individuals living with mental illness do not have access to treatment (Patel et al., 2018). In India, about 90% of the nearly 50 million people living with depression do not have access to care, a significant challenge driven largely by shortages of trained mental health service providers (Sagar et al., 2020). Task sharing involves training non-specialist health workers, which includes frontline health workers or community health workers without specialized training in mental health care, in delivering evidence-based brief psychological treatments for common mental disorders at the primary care level (Raviola, Naslund, Smith, & Patel, 2019). Research shows that task sharing is an important strategy for bridging the care gap in lower resource settings, such as in India (Singla et al., 2017). However, there remain significant challenges with training and building capacity of non-specialist health workers towards widespread implementation and scale up of mental health services.

With the emerging use and reach of digital technology globally, there may be new opportunities to make training programs more widely available to non-specialist health workers (Dalberg Global Development Advisors, 2012; Naslund et al., 2017). Research suggests that digital programs for training counsellors and therapists are not only needed but also crucial (Anthony, 2015). Digital training for psychotherapy has been shown to be highly effective, as it can offer opportunities for the learner to work closely with other learners, and engage in direct communication and self-disclosure that are not always possible in typical face-to-face training situations (van Deurzen, Blackmore, & Tantam, 2006). Growing evidence shows that digital training is a potentially effective and sustainable method to train and build capacity of non-specialist health workers for delivering treatments for various conditions (Funes, Hausman, Rastegar, & Bhatia, 2012; Mishra et al., 2019; Winters, Langer, & Geniets, 2018). In a report summarizing over 140 projects using mobile technologies among non-specialist health workers in low-resource settings, one key function of technology was to support training and education (Agarwal, Rosenblum, Goldschmidt, Carras, & Labrique, 2016). Mobile learning could also reduce costs associated with travel, arranging classroom space, and requirements for the time and availability of instructors (Sissine et al., 2014), while giving learners freedom to acquire new skills at their own pace (Sissine et al., 2014). In the context of task sharing, a recent study from rural Pakistan showed that a digital tool could support traditional face-to-face training of non-specialist health workers to deliver a brief psychological treatment for perinatal depression (Rahman et al., 2019; Zafar et al., 2016).

Despite the promise of using digital technology to support training for non-specialist health workers, there remain gaps in understanding how to effectively develop a digital training program for this cadre of frontline health workers in low-resource settings. Specifically, there has been limited emphasis in the literature on the systematic approach for developing a digital training program delivered entirely remotely in a rural area of a low-resource setting, as well as the steps required to adapt evidence-based manuals for a brief psychological treatment for depression into digital content tailored to the local context, culture and language. Therefore, the purpose of this study was to employ a systematic approach to the design and development of a digital program for training non-specialist health workers to deliver an evidence-based brief psychological treatment for depression, called the Healthy Activity Program (HAP), in primary care settings in Madhya Pradesh, India. We describe the step-wise process of adapting the evidence-based content for digital instruction, optimizing the use of a smartphone app for hosting the training content, and ensuring the feasibility and acceptability of the program through iterative user testing and feedback.

Methods

Setting

In this study, development of the digital training program was conducted in Sehore district of Madhya Pradesh, India. Madhya Pradesh is one of the largest states of India, yet it is also one of the lowest-resourced states with a predominantly rural population (Census Organization of India, 2011; Menon, Deolalikar, & Bhaskar, 2008; National Health Mission, 2018; Suryanarayana, Agrawal, & Prabhu, 2016). Importantly, there is poor access to mental health services in the state, making it a key priority for the health system (Kokane et al., 2019; Shidhaye, Lyngdoh, Murhar, Samudre, & Krafft, 2017). In this study, the target cadre of non-specialist health workers for the design and development of the digital training program are community health workers, called ASHAs, deployed within India’s National Health Mission. ASHAs (Accredited Social Health Activists) are all women and are selected by community-level governance mechanisms, and trained and supported by the health system. ASHAs typically serve a population of 1000, primarily in rural areas, with key responsibilities about maternal and child health services. The ASHA acts as a bridge between the community and the health system in India (Scott, George, & Ved, 2019; Ved et al., 2019; Waskel et al., 2014). To support the development of the digital training program, we worked with ASHAs and ASHA Facilitators (ASHAs who are promoted to provide support to groups of ASHAs) recruited from community health centers in Sehore district.

Research Design

All study procedures were approved by institutional review boards at Sangath, India, and Harvard Medical School, USA. Additional Health Ministry Screening Committee (HMSC) approval was obtained from its secretariat at the Indian Council of Medical Research (ICMR). We employed an iterative approach to the design of the digital training program, informed by the ADDIE (Analysis, Design, Develop, Implement and Evaluation) framework, a recognized instructional design model that offers flexibility for the development of digital learning content (Kruse, 2002). Our approach was also guided by prior digital intervention development projects related to mental health (Gonsalves et al., 2019), and the broader instructional design literature (Torre, Daley, Sebastian, & Elnicki, 2006), with consideration of the target learner, learning environment, approaches for cultural and contextual adaptation, and strategies to support program engagement (Yardley, Morrison, Bradbury, & Muller, 2015).

Throughout the development process, we referred to the core content and principles of the Healthy Activity Program (HAP), an evidence-based brief psychological treatment for depression (Patel et al., 2017). The HAP manuals, covering general counselling skills as well as specific treatment skills, formed the source content for the digital training program. During our program development, it was essential to ensure that we maintained fidelity to this original HAP treatment content and that we did not deviate from proven strategies for successfully treating depression in primary care. We selected HAP because it is an effective treatment for depression based on behavioral activation (Dimidjian, Barrera Jr, Martell, Muñoz, & Lewinsohn, 2011), and that is recommended as part of the World Health Organization mhGAP program (World Health Organization, 2017). HAP is delivered over six to eight sessions, covering core content related to psychoeducation, behavioral assessment, activity monitoring, activity structuring and scheduling, problem-solving, and activation of social networks. HAP is designed to be delivered by lay health counsellors and was originally developed and evaluated in Goa, India (Chowdhary et al., 2016), and has demonstrated effectiveness and cost-effectiveness (Patel et al., 2017), as well as sustained benefits for treating adult depression in primary care (Weobong et al., 2017). As outlined in Figure 1, and described in the sections below, we followed a step-wise approach to support the training program development.

Figure 1.

Overview of the Digital Training Program Development

Step 1: Create Blueprint of Training Program

The design process started by creating a ‘blueprint’ of our digital training program to ensure alignment between the learning objectives and the core content of the HAP treatment manuals. The blueprint was developed to specify the knowledge and skills required to effectively deliver HAP and served as a way to validate the evaluation content for the program as well as a document to guide the development of the course content and learning experiences (Coderre, Woloschuk, & Mclaughlin, 2009). The manuals cover general counselling skills and specific skills required for treating depression and are available open-source from the Sangath website [http://www.sangath.in/]. Three members of our research team reviewed these manuals to create one comprehensive blueprint to serve as the architecture for the digital training program. Table 1 summarizes the sections of the blueprint that were converted into modules with accompanying digital content. To ensure that the blueprint adhered to the core content in HAP, we had multiple rounds of reviews from experts, including counsellors with experience delivering HAP (n=2), researchers and clinicians with expertise in depression care (n=4), and members of our research group involved in the initial conceptualization and development of HAP (SD and VP).

Table 1.

Digital Training Program “Blueprint”

Module Number	Topic	Source Materials	Learning Objectives In this module, we will learn:	Video lecture	Role-Play Video	PowerPoint lecture video	Summary video
1	Understanding Depression	HAP Manual; Module 1; Page 6	• What is Depression? • What are the symptoms of Depression? • How to screen for symptoms of Depression	5	-	1	1
2	An Introduction to Counselling	CR Manual; Module 1; Page 6	• What is meant by counseling? • What is the difference between counseling and a friendly chat?	3	-	-
3	An Effective Counselling Relationship: • The meaning of an effective counseling relationship • The key skills for an effective counseling relationship • The different styles of counseling	CR Manual; Module 2; Page 9	• What is meant by an effective counseling relationship? • What are the key skills for developing an effective counseling relationship? • What are the different styles of counseling?	3	3	-	1
4	Creating the Right Conditions for Getting Started: • Preparing yourself for the session • Greeting the patient and introducing yourself • Talking about confidentiality • Conducting home visits • Using the telephone in counseling	CR Manual; Module 3; Page 17	• How to prepare ourselves for the counseling session • How to greet the patient and introduce ourselves • How to talk about confidentiality	3	-	2	1
5	Managing Patients’ Crises: • Managing suicide risk • Managing personal crises • Managing patients who are bereaved (i.e., someone close to them has died) • Managing domestic violence	CR Manual; Module 4; Page 25	• How to manage suicide risk • How to conduct a suicide risk assessment • When to refer suicidal patients • How to apply effective skills and strategies for counseling a suicidal patient • How to manage personal crises • What are the goals of crisis counseling? • How to assess a patient in a crisis • What are the steps in crisis counseling? • How to help patients in bereavement • How to manage domestic violence	5	2	2	1
6	Involving Significant Others (SO) in Counselling	CR Manual; Module 5; Page 42	• Why do we need to involve an SO in counseling? • What are the situations in which we will need to involve an SO in counseling? • How to involve an SO in counseling • What are the necessary precautions while involving an SO in counseling?	1	1	-	1
7	Becoming a More Effective Counsellor: • Keeping contact with patients • Maintaining standards – documentation and record keeping • Supervision • When to refer and to whom • Maintaining boundaries • Taking care of ourselves	CR Manual; Module 6; Page 49	• How to keep contact with patients • How to maintain standards • What is supervision? • When to refer and to whom • How to maintain boundaries	2	-	6	1
8	Style of a Healthy Activity Program Counsellor	HAP Manual; Module 3; Page 15	• What are the key aspects of the counsellor’s style when delivering the Healthy Activity Program?	1	-	-	1
9	Healthy Activity Program: Phase-wise Guide	HAP Manual; Module 2; Page 12 & HAP Manual; Module 4; Page 19	• What is the Healthy Activity Program? • What are the phases of the Healthy Activity Program? • Moving through Healthy Activity Program Phases	-	-	2	1
10	Phase 1: Getting Started (1–2 sessions)	HAP Manual; Module 4; Page 20	• How to deliver each session of the Healthy Activity Program in a step-by-step manner • Engaging and establishing effective relationship • Helping patients understand the Healthy Activity Program • Eliciting commitment for counseling	6	2	-	1
11	Phase 2: Learning together, getting active and solving problems (3–6 sessions) Part 1: Learning Together	HAP Manual; Module 4; Part 1 Page 31–36	Part 1: • Identifying activation targets	2	2	1	1
12	Part 2: Getting active and Solving problems	HAP Manual; Module 4; Page 37 to 50	Part 2: • Strengthening the understanding of Healthy Activity Program and applying it to daily life • Encouraging activation • Identifying barriers to activation and learning how to overcome these • Helping patients solve (or cope with) life problems	6	2	-	1
13	Phase 3: Ending Well	HAP Manual; Module 4; Page 51	• Help patient review the HAP model in general and specific actions that support patient’s mood • Help patient identify possible challenging future situations • Help patients make a plan to deal with such situations using the skills they have learnt	2	-	-	1
14	Useful Strategies for Specific Problems	HAP Manual; Module 5; Page 55	• Ways to deal with 5 common problems that can make it more difficult to get active and solve problems: • Thinking too much • Feeling anxious or tense • Problems with people close to the patient • Difficulties with sleep • Using tobacco	-	1	4	1
15	Telephone Sessions	HAP Manual; Module 6; Page 67	• How to do telephone sessions to deliver the Healthy Activity Program	-	-	1	-
16	Role of Medication	HAP Manual; Module 7; Page 70	• How medicines work in Depression • When medicines are used in Depression • Our role when a patient is taking medicines for Depression	-	-	1	-
Conclusion	Thank student for completing the program	New material
Appendices
	Appendix1: Dealing with difficult situations	HAP Manual; Appendix 1; Page 73
	Appendix2: Glossary of Difficult Words	HAP & CR Manuals
	Appendix3: Suggested further reading	HAP & CR Manuals

Step 2: Develop Training Program Content

The next step involved adapting written content from the HAP manuals into content that aligned with the structure from the blueprint, and that could be converted into a digital format accessible from a smartphone app. To support this process, we reviewed existing training manuals available for ASHAs from the National Health Mission to consider the format and level of written content currently in practice. This allowed us to better understand how to adapt and tailor the HAP curriculum content specifically for this cadre of non-specialist health worker. We also considered the broader digital education literature, and the need to account for the educational qualifications and prior experiences using technology for accessing training programs among the target non-specialist health workers (Chaiyachati et al., 2013). Drawing from our prior formative research activities and focus group discussions with ASHAs, we found that participants expressed a preference for video-based content and use of visuals, as this could more effectively reflect realistic scenarios representative of the health system that are both engaging and relatable (Muke et al., 2019).

Content from the HAP manuals was divided into the 16 modules outlined in the blueprint, with each module further broken down into short segments covering key materials. We worked with a scriptwriter to develop scripts for the short video segments. The scripts were developed in English (the original language of the HAP manuals) under the guidance of the training program lead (AK), an experienced psychologist and HAP Master Trainer (refers to having certification in order to offer training to health workers to deliver HAP), and primarily included dialogues, montages and directions for shooting specifications. Two experienced HAP counsellors and one of the HAP developers (SD) reviewed the scripts and focused on ensuring the fidelity of the content and adherence to the core principles of HAP treatment. The English scripts were finalized by incorporating expert feedback and comments and were translated into Hindi (the language of the local setting where the training program will be delivered). Our team ensured substantial consideration to safeguard the contextual and cultural fit of the content when contextualizing the Hindi scripts. This process involved consultation with 2 experienced counsellors and 2 community health workers to ensure that the language was simplified through the removal of jargon or complex terms and that the content was appropriate for the local language and context in Madhya Pradesh, using words that are typical for the region. The last step involved having a HAP expert with certification as a Master Trainer review the final Hindi scripts, ensuring that the core content from the English scripts was captured in the translations and that the adaptations and use of local words for describing depression were appropriate for the study setting. This process of creating English scripts and accompanying instructional content followed by translation and expert review required approximately eight months to complete.

Step 3: Digitize Training Program Content

Digitization involved contracting vendors to develop video-based content. Our team worked with the vendors to compile a list of materials, locations, and the number of actors in preparation for the video shoot. The videos were filmed in several locations, including community health centers and other locations in the community where the ASHAs work, to create a context that would be relatable to the target audience of health workers. Additional videos were filmed in offices in Bhopal made to look like typical clinical settings. The videos consisted of lectures delivered by HAP experts, and the use of actors who could memorize and deliver the lines. In role-play videos, actors were selected to act as mock patients receiving care, or as counsellors to demonstrate how to deliver elements of HAP based on different scenarios. We completed the video shooting over two rounds, first requiring 11 days with one vendor, and then over 8 days with another vendor. The training program lead (AK) was present during all the days to guide and supervise video production, and to ensure that the filming adhered to the scripts covering HAP. Two vendors were selected to expedite the process because, following the video shoot, multiple rounds of editing were required to finalize the videos for uploading to the digital platform. The vendors worked together to ensure that the video content followed a similar style. The training program development team was also present throughout the editing process and reviewed each iteration of edits to ensure the correct use of subtitles, icons and images, montages, placement of bullet points and text, sound quality for spoken content and narration, and use of music and other sound effects. During the editing phase, PowerPoint lectures with voiceover, and graphics were added to supplement video-based content.

Step 4: Develop a Learning Management System (LMS) and Upload Content

Concurrently with Step 3, described above, we developed a Learning Management System (LMS) for hosting the digital training content using the Moodle platform. In prior formative research conducted by our team, we learned that the Moodle platform appeared both feasible and acceptable for use among non-specialist health workers (Muke et al., 2019). Importantly, we selected Moodle because it is an open-source LMS that is used in numerous countries worldwide and offers many specific features to support learning and engagement with educational content (Bulaeva, Vaganova, Koldina, Lapshova, & Khizhnyi, 2017; Paragina, Paragina, Jipa, Savu, & Dumitrescu, 2011). These include content that can be accessed from a smartphone app, learner performance tracking and metrics, customizable interface to multiple languages, and user-friendly dashboard. We contracted an external IT agency for supporting the development and customizations to the LMS. Two members of our team worked closely with the IT agency for designing, developing and continuously testing the features and functionality of the LMS. As the digital training content was finalized from Step 3, we began to upload this content onto the LMS platform. This involved continuously testing the content within our team to ensure that it could be accessed through the smartphone app. As part of this process, we also created short engagement quizzes and assessments that could be integrated within the platform to support learners with acquiring the program content. This process involved working closely with the IT agency, informed modifications to the layout of the platform, including improvements to the appearance of the landing page and easy navigation of the course content.

After uploading the entire digital training program to the LMS platform, and as a precautionary step before user testing, members of our team downloaded the LMS app to a smartphone and tested the various functions including logging in and logging out, accessing the help menu, navigating the interface and course dashboard, accessing the course content, and checking the general performance of the app. Simultaneously, we evaluated the administrative functionality of the LMS from the perspective of the course facilitator. This included testing features for monitoring the progress of learners as they complete the course content, managing access to course activities, and generating performance reports for participants enrolled in the program. This internal testing process was aimed at identifying and addressing as many technical challenges as possible ahead of user testing with non-specialist health workers.

Step 5. User Testing of the Digital Training Program

For initial user testing, we invited 12 individuals from outside our research team who had no prior exposure to the digital platform or training content. This group consisted of 5 males and 7 females and comprised a mix of students, health workers, and others, ranging in age from 18 to 40, with education between 8–12 grade. The goal was to assemble a group of community members who approximated the age and education of the target group of non-specialist health workers to try out the digital platform and provide feedback related to technical challenges and accessing the program content on the mobile app. The user testing took place over four days in our research office in Bhopal. Participants were offered a token incentive for their time. During the user-testing, each participant was provided with a smartphone, a brief orientation to the program, and was instructed to access the program content and to record any issues or difficulties navigating the program content or accessing the different features in the mobile app. Our research team was available to provide support as needed and to record any observations with using the program. This was followed by a semi-structured feedback session to collect insights from participants to inform modifications and improvements to the digital content. The changes included improving some of the reading materials, identifying videos that appeared to be out of sequence, engagement quizzes where the response options were not functioning as intended, and images or graphics that were blurry. Further, we used this user testing as an opportunity to address technical challenges with the digital platform, such as slow download speed for accessing the videos, difficulty with logging into the app, and overall slow performance of the app. Feedback collected during the user-testing was incorporated and the modified version of the LMS was ready to pilot test with a group of non-specialist health workers.

Field Testing and Focus Group Discussions

Following the program development, we conducted a pilot study as an opportunity to field-test the digital training program with a cohort of non-specialist health workers. Results from the randomized pilot study are reported elsewhere (Muke et al., 2020). Here, we report on focus group discussions with ASHAs and ASHA Facilitators who used the digital training smartphone app, and specifically on their recommendations for improving the digital platform. We recruited non-specialist health workers who were aged ≥18 years and had a minimum education level of 8th Standard to ensure sufficient literacy level necessary to access the smartphone app and read the training program materials. These health workers were recruited from primary care facilities in Sehore district, Madhya Pradesh, and had access to the digital training program over four weeks, before being invited to join focus group discussions. A member of our training program development team (RS) joined the focus group discussions to record observations from participants. Our goal was to capture insights from participants to inform modifications to the digital platform. We describe the key action items based on participants’ feedback and resulting changes and modifications that we made to the digital training program.

Results

Digital Training Program

The training program development was completed over 12 months from September 2018 to October 2019. The final digital training consisted of 16 modules, as outlined in Table 1. These modules contained a total of 85 short videos with 39 video lectures, 13 role-play videos, 20 PowerPoint lectures with voiceover, and 13 summary videos providing a recap of the content displayed in a given module. The videos ranged in duration from about 1 minute to 13 minutes, though the average duration was approximately 3–5 minutes. We specifically kept the videos in short segments based on user feedback because it is easier to stay engaged with shorter videos when accessing the content from a smartphone app. In addition to the video content, we included a total of 92 engagement quizzes and 192 knowledge assessment questions embedded throughout the modules. The total duration of the training was about 8 hours of content, but estimated to require about 48 hours to complete accounting for the time to fully read the material and respond to the questions. This duration was matched to the face-to-face training program that is completed over a full 6-day intensive residential course.

As illustrated in Figure 2, the central components of the digital training program were the instructional videos, consisting of expert lectures covering the core content of HAP, and the Moodle LMS smartphone app to house the digital content. The lectures included a variety of digital elements to support learning, such as visual graphics, text in bullet points, figures, tables, narration, footage from local community settings and health facilities, and example interactions between patients and health care providers in the community or at the clinic. Role-play videos were used to demonstrate the use of specific counselling skills, and interaction between patients and counsellors during delivery of HAP sessions. To supplement the lecture and role-play videos, we created PowerPoint lectures with voiceover delivered by a HAP expert, with minimal text and use of graphics and images tailored to the local context (see Figure 3). We also used a variety of engagement and assessment activities, such as multiple-choice questions, drag and drop response options, and true/false questions covering the HAP content to accompany the instructional videos and to support participant engagement in the program. The final digital training is accessible from a smartphone app, and a course administrator can manage the course, track participant progress, and send notifications.

Figure 2.

Overview of sample content from the digital training program

Figure 3.

Sample images adapted from existing training materials for ASHAs in the health system

Focus Group Discussions

Out of 26 non-specialist health workers who were provided with a smartphone for accessing the digital training app, 19 participated in focus group discussions (n=8 in one focus group, and n=11 in a second focus group). These participants ranged in age from 24 to 49 years, with 11 (58%) having education above 8^th standard, and 13 (68%) having prior experience using a smartphone.

In general, participants expressed positive feedback about the digital training, indicating that they found the content easy to understand and relatable and that the program content was useful for understanding the importance of counselling and how to talk to patients, and helpful for understanding mental illness symptoms and breaking stigma. With regards to the smartphone app, participants appreciated the ability to access the training content according to their schedule, liked the layout of the app, and found the interactive questions and activities useful and interesting. Participants also described key challenges with using the smartphone app. The following three broad challenges emerged:

1. Technical challenges due to poor connectivity: the primary technical challenge was poor network connectivity. The participants unanimously expressed frustration with the slow download speed of the videos and accessing the training content. This is largely due to low network coverage in many of the villages where the health workers live and work.

2. Learning to use the smartphone app: for several participants, this was their first time using a smartphone, which contributed to some confusion and anxiety. For instance, some participants indicated that they accidentally deleted the app, could not view messages sent through the mobile app from our team or had to ask their children, neighbours, or family members to help them figure out how to access the program.

3. Need for additional support to finish the training: throughout the training, participants frequently contacted our research team by phone with questions about technical challenges. Participants expressed their appreciation with being able to seek additional support, though they mentioned that more active support from our team could act as a motivating factor. The mobile app had a help feature built-in, though we learned that participants did not use this feature and preferred access to telephone support. Participants also indicated that being able to access the research team for support with using the app outside of the 10 am-5 pm office hours window would be helpful, as this was the time they were typically working, and therefore were not using the training app at that time. Lastly, participants expressed interest in having a refresher orientation training partway through the course for troubleshooting any possible issues with the digital platform.

Modifications to the Digital Training Program Informed by Participant Feedback

The challenges highlighted by ASHAs and ASHA Facilitators were interrelated, and indicated the complexity of real-world issues that could hinder completion of a digital training program in a low-resource setting. We used participants’ feedback to inform several immediate and practical modifications to the digital training program with the goal of addressing these challenges and improving the overall usability of the program in preparation for a subsequent fully powered effectiveness trial. Specifically, we undertook the following changes:

Our first major modification to the digital training program was to ensure that the content could be accessed through the mobile app without requiring an internet connection. Participants overwhelmingly described challenges related to accessing the training program content on the app due to poor connectivity. We worked with the video production team to substantially compress the file size of each video, and we worked with the IT contractor to ensure that the digital program could operate smoothly from our server, to enable full downloading of the course package onto the smartphone so that the training content could be accessed offline as well. Through this process, we learned that it was also necessary to enhance our organization’s server capacity to host the LMS and to run the digital training program course smoothly.

Our next set of changes was in response to participants’ limited prior experience using smartphone devices, where we made changes to our pre-training orientation session by including a more comprehensive overview of the digital technology. Specifically, we created a 4-hour group orientation session offering a detailed overview of how to operate the smartphone (including basics such as charging the phone, protecting the phone, and logging in and out of the digital training app), how to navigate the smartphone app, and how access the digital content. The orientation session includes a brief PowerPoint presentation and is structured in a workshop format where participants have the opportunity to try out the smartphones and ask research assistants any questions as they may arise. The goal is to make sure by the end of the orientation that participants feel confident using the smartphone to access and complete the digital training program on their own.

Lastly, to provide additional support to participants while completing the training, we used there feedback to inform the design of a protocol for remote coaching support to sustain participants’ engagement in learning the content and successfully completing the digital training. We determined that the coach could be an experienced HAP counsellor with certification as a Master Trainer, and who has also completed the digital training. The coaching protocol involves once weekly telephone coaching sessions lasting about 30 minutes that follow a structured plan over the 4-week training duration. Ahead of each weekly call, the coach can review performance metrics collected from the LMS dashboard, such as progress through the modules, so that they can ask participants about how they are doing with the training, whether they have experienced any challenges, or have questions about the platform or program content. The coach will offer encouragement, and set goals with participants. The coach will also moderate a WhatsApp group to share common questions that come up during the individual coaching sessions, post encouraging messages, reinforce challenging concepts from the content, and facilitate peer learning between participants. We will evaluate the potential benefits of adding this remote coaching component to the digital training as part of a large-scale randomized controlled trial.

Discussion

We describe our step-wise approach to designing and developing a digital program for training non-specialist health workers to deliver an evidence-based brief psychological treatment for depression in primary care. The final product of this work is a fully functioning digital training program accessible from a smartphone app with content adapted and tailored to the local context in Sehore, a rural district in Madhya Pradesh, India. One of the strengths of this study was our effort to systematically engage key stakeholders including clinicians with expertise in psychological treatments and HAP, technology vendors and IT designers, and importantly, the target group of non-specialist health workers. Throughout this design process, we balanced two key priorities: first, the need to retain fidelity to the evidence-based HAP treatment for depression; and second, the importance of accounting for the perspectives of the target user group of non-specialist health workers in designing the platform.

By closely adhering to the HAP manuals during program development, we were able to ensure that the training program reflected the core principles of behavioral activation, a widely used and proven approach for treating depression (Dimidjian et al., 2011). The digital platform acts as a way to both replicate the features of conventional Face-to-Face training through the use of role-play videos (i.e., to emulate the use of role-plays activities among participants in traditional classroom-based instruction), and also expand on existing training methods by having videos of clinical scenarios to clearly demonstrate HAP delivery and the necessary skills for responding to patients’ needs, followed by embedded knowledge assessment questions. The ability to create clinical illustrations of the treatment being delivered has been highlighted previously as an important advantage with using digital approaches for therapist training (Fairburn & Cooper, 2011). As the ultimate goal of the training program is to develop the skills and competencies of health workers, this further reinforces the importance of encompassing the essential features of the evidence-based practice to be delivered within the training program materials (Puspitasari, Kanter, Murphy, Crowe, & Koerner, 2013; Rosen, Ruzek, & Karlin, 2017). Therefore, an important strength with our approach was the multiple rounds of revision by experienced clinicians and HAP experts.

Another important consideration in successful digital program design is the involvement of the target user group, as this can contribute to better usability and engagement while ensuring the content is relevant and relatable (Yardley et al., 2015). In this study, the involvement of non-specialist health workers was particularly important for supporting adaptations to the evidence-based HAP content to fit the local context in Sehore district, Madhya Pradesh. Importantly, the use of digital platforms offers unique opportunities to accommodate user preferences and tailor instructional content for facilitating education and training for community health workers (Winters et al., 2018). For instance, this involved simplifying the language to meet the needs of non-specialist health workers with lower education levels while using local terms for describing mental health problems; creating videos with footage showing relevant primary care and community-based settings; and adding culturally appropriate images and graphics. A challenge with existing research in low-resource settings has been the limited engagement of health workers and communities in developing and testing digital interventions, which is potentially a major driver for the low uptake and implementation of these programs (Long, Pariyo, & Kallander, 2018). Therefore, our approach aligns with the growing emphasis on the need to engage stakeholders early on in the intervention development process (Naslund, Shidhaye, & Patel, 2019), and builds on our initial efforts showing interest among this target group of non-specialist health workers in using a digital training program for learning about depression care (Muke et al., 2019).

Our study also highlights the benefits of field-testing any digital program before a large-scale evaluation of effectiveness. After designing the training program, uploading content to the LMS, and addressing preliminary technical challenges through internal user testing, it was only through field-testing with non-specialist health workers were we able to identify several challenges. Participants’ feedback directly informed immediate action items and modifications to the program, the first of which was the need to ensure that the program content could be entirely accessible in an offline format to account for the poor wireless connectivity in the region. Unreliable network connections are a persistent challenge in using digital technology in rural low-resource settings (Labrique et al., 2018), and have been similarly described as a barrier to overcome as part of the design of a digital training program in rural Pakistan where the content was made available offline from tablets (Rahman et al., 2019).

Our addition of a comprehensive orientation session at the start of the training program was necessary to overcome challenges related to limited prior experience using smartphones among the target non-specialist health workers. These considerations are essential for informing specific modifications to the program content, and design of instructional materials (Chaiyachati et al., 2013), and have been highlighted as being especially important for individuals who might not be proficient or familiar with accessing the content in a digital training format (Mastellos et al., 2018). While low digital literacy has also been identified as a barrier to the adoption of digital interventions among frontline health workers, this is likely to change in the coming years as interventions increasingly rely on smartphone-delivered programs for health worker training and support (Winters et al., 2019). Lastly, the development of a remote coaching protocol holds promise for helping learners’ progress through the digital training program. Our approach aligns with prior studies that have used e-mentoring to support participant engagement and completion in online education (Friedman & Friedman, 2013; Homitz & Berge, 2008). Furthermore, a recent study used structured weekly phone calls to support an online program for training therapists to deliver psychological treatments for eating disorders (Fairburn, Allen, Bailey-Straebler, O’Connor, & Cooper, 2017).

Several limitations with our study methods and approach should be considered. Firstly, the sample size was small, and the convenience sampling approach may have introduced selection bias. For instance, the non-specialist health workers who contributed to the intervention development and field testing may have been highly interested in using digital technology for accessing a training program or participating in research and therefore may not be representative of the broader workforce. Second, the research team influenced the development of the training materials and digital content, and may have introduced personal biases during the design of the program. To minimize potential researcher biases, we employed regular consultation and content reviews with external experts, and also made sure to follow the blueprint based on the original program manuals. Third, the field-testing and pilot evaluation of the digital training program and follow up focus group discussions aimed primarily at identifying and fixing technical challenges with the mobile app, digital content, or use of the smartphone devices. We have reported additional pilot findings elsewhere describing the preliminary effectiveness of the digital training program on competency outcomes (Muke et al., 2020), though we have not conducted in depth thematic analysis of qualitative findings as our goal was to collect participant feedback to inform immediate modifications to the digital training program. In depth qualitative exploration of participants’ perspectives about use of the digital training program as part of their routine work activities would be an important future research direction, as this could shed light on the potential barriers and facilitators to the adoption and implementation of digital training for frontline health workers in low-resource settings.

Fourth, it is important to note that the final digital training program that was modified based on participant feedback has not yet been validated through field-testing. Therefore, additional changes may be necessary to address other technical challenges that were not identified in this study, or to further modify the content and platform to promote user engagement. Fifth, while the rigorous process for designing and developing the training program represents an important strength in this study because considerable effort went into customizing the content and program structure to the local culture and context, this also represents a potential limitation. For instance, it may be difficult to scale this program to other communities in India that are not Hindi speaking or where additional considerations may be needed for tailoring the content or language to the local context. As part of ongoing efforts to implement evidence-based treatments for depression in primary care settings in rural India, it will be important to determine whether the training program described in this study can meet the needs in other settings, as well as whether the overall approach to program design can be adopted and replicated for developing other brief psychological treatments for delivery in diverse low-resource settings globally.

Lastly, this study can only be considered the first step towards training and developing clinical skills and competencies of non-specialist health workers to deliver HAP. Future studies are needed to determine whether the training program can achieve its purpose by developing competencies of non-specialist health workers to deliver HAP, something that will be addressed in a large-scale effectiveness study (ClinicalTrials.gov Identifier: NCT04157816). Furthermore, while rigorously designed, another limitation to consider is that the training content only provides opportunities to read about delivering HAP, watch experts through videos and interactive activities, and observe role-play videos. Successful training must be supplemented with direct observation of care delivery in real-world settings, and opportunities to practice these skills and receive feedback from peers and more experienced providers (Fairburn & Cooper, 2011; Kemp, Petersen, Bhana, & Rao, 2019).

Conclusions

The systematic approach described in this study can potentially be replicated in other settings globally and could offer a template for guiding the design and development of digital training programs aimed at building workforce capacity and scaling up brief psychological treatments. Our study offers an important contribution to the broader literature on the use of digital technology for health worker training and health systems strengthening (Long et al., 2018), and adds to the dearth of research on the use of technology for specifically supporting community health workers to deliver evidence-based depression care in low-resource settings (Naslund, Shidhaye, et al., 2019; Winters et al., 2019). The comprehensive program development process described here combined the need to draw from an established scientific evidence-base in depression care with the critical importance of accounting for the perspectives, interests, and context of the target user group (Naslund, Gonsalves, et al., 2019). This approach resulted in a digital training program that is consistent with the core HAP manual and behavioral activation principles and that is appropriate for use in rural India.