Supported self-management for asthma (written action plans and regular review) is highly effective at improving control and reducing acute attacks [1-3]; however, globally, it is challenging to implement for 334 million people living with asthma [4-7]. One of the challenges is the need to tailor support for people with limited health literacy. Health literacy is defined as the degree to which individuals can obtain, process, and understand the necessary health information needed to make appropriate health decisions [8]. Studies have associated limited health literacy with erroneous health beliefs and poor adherence to self-management activities [9,10]. Malaysia has a high burden of limited health literacy in the general population [11], and asthma control is challenged by a lack of patient education, overreliance on unscheduled visits, and lack of action plan ownership [12-14].

The use of digital technologies for internet-based information is more common in the younger age group than in the middle and older age groups [15]. Malaysia’s multigenerational household culture and strong family orientation have helped younger family members assist older generations in using digital technology to stay connected and find information [16]. The pandemic has further seen widespread adoption of digital technologies by a broader age group of users in diagnosis, prevention, and surveillance [17]. Three-fourth of Malaysians are now smartphone users, with most (60.9%) of them in the lowest income group [18]. Although health-related information-seeking behavior on the internet is greater in those with good health literacy [19], our previous qualitative work among people with limited health literacy suggests that a mobile app is a preferred medium to deliver supported self-management, including a pictogram-based asthma action plan and signposting to reliable asthma information sources.

For the aforementioned reasons, developing asthma self-management tailored to limited health literacy needs is an important context for the web-based design sprint workshop, as various studies have shown that the extensive use of pictograms, images, and prompts was appealing to participants and may improve the understanding of information in mobile apps [20-23]. However, it is essential to involve users early in the design stage, as some of the unique features that people want can be time-consuming and costly to build, and a balance may need to be found between desired features [24] and those with evidence-based recommendations [25,26].

Using a design sprint methodology, we seek to optimize user experience in app development by integrating patients into the 5-stage process of mapping, sketching, designing, developing, and testing [24,27]. We used the health literacy framework [8] to underpin the overall structure of the interventional work. The COVID-19 pandemic and physical distancing requirements meant that we had to conduct our workshop in a web-based format. In this study, we report the outcomes of the workshop deliberations and our experience of conducting a week-long remote 5-stage program attended by patients with asthma and health care professionals (HCPs).

The workshop received ethical approval from the Medical Research and Ethics Committee of the Ministry of Health, Malaysia (ID: NMRR-19-3609-52292) and sponsorship approval by the Academic and Clinical Central Office for Research & Development at the University of Edinburgh (ID: AC20011). Informed consent was obtained from all the participants before the workshop.

We conducted a 5-stage design sprint workshop using a web-based and qualitative research approach. We used the 5-stage design sprint process as a roadmap to develop the intervention. We adapted the methodology and constructed the workshop into five stages: (1) understanding and mapping problems, (2) sketching of solutions, (3) deciding on solutions to problems, (4) developing a prototype, and (5) testing a low-fidelity prototype [27,28]. The process was originally designed in the technology sector by the Google Ventures team for business start-up teams [27]. The involvement of the target population and early testing enhance intervention effectiveness and increase the likelihood of adoption at the implementation stage [24]. Owing to the rapid development and testing stages, this is an ideal concept for a low-resource setting, that is, time and cost [24].

Stakeholder (patient and HCP) discussions provided insights into self-management issues in stage 1, which informed mobile app development in stages 2-4. We recruited patients with asthma and limited health literacy to test the low-fidelity prototype in stage 5 and provided feedback through qualitative interviews. A low-fidelity prototype is a modeled prototype with limited technical functionality [29] that is quick to create and can be easily improved in the light of feedback in the testing stages. The details of each stage are listed in Table 1.

The web-based workshop discussions were conducted on a web conference platform (Microsoft Teams), and brainstorming of the idea was conducted on a web-based board (Miro). While conducting the exercise on the web-based board, the workshop participants remained connected to the web conference site to allow ongoing discussion. The web-based board was superseded by a prototype development (Figma) site in stages 4 and 5, whereas discussions remained on Microsoft Teams. The app developer supported information technology activities. Two weeks before the workshop, one of the researchers contacted each of the participants (patients and HCPs) to assess technical skills such as the ability to log on, use a meeting platform, and logistic issues such as the quality of the internet connection.

The workshop, which took place between June 22 and 26, 2020, was conducted through a secure web-based meeting platform using a virtual whiteboard to facilitate information sharing between the researchers and app developers. Our original plan for a face-to-face workshop was changed to a web-based format to overcome the restriction of the order of movement because of the COVID-19 pandemic; an additional advantage was that it allowed participants from different locations and time zones to participate.

The patients were from 2 urban public primary care clinics in central Malaysia. Asthma is managed in primary care clinics, although chronic and acute care management through the provision of an asthma action plan to support asthma self-management is uncommon. Malaysia has a dual health system, public and private, where the public health system provides the leading service for the population with copayment of Ringgit Malaysia, RM 1 (US $0.23) per visit.

Demographic information was collected from the database of patients who agreed to participate. Stakeholders’ discussions in stage 1 focused on challenges by patients and HCPs around (1) asthma education, (2) asthma self-management, (3) monitoring of symptoms, (4) emotional support or lifestyle advice, (5) social support, and (6) clinic set up (Multimedia Appendix 1).

In stage 5, we tested the prototype with 5 patients to assess its utility and usability using a set of semistructured questions in a concurrent think-aloud manner (Multimedia Appendix 2). During this session, HS, the main interviewer, guided the process. Four other researchers, PYL, SSG, ATC, and JW, observed the interview while the 2 app developers, Aidil Goh and Muhammad Marzuqi, managed the technical aspects of the sessions. The interviews took 1 hour and were conducted in Bahasa Malaysia, the patients’ preferred language.

Qualitative interviews were audio recorded, other web-based discussions and browsing activities were video recorded, and web-based board exercises were captured and archived as described in Table 1. All interviews were transcribed verbatim.

For this qualitative study, we used thematic analysis to obtain rich data from the stakeholders’ discussions in stage 1 and the interviews in stage 5. The texts were analyzed iteratively using a deductive thematic analysis approach, as outlined by Braun et al [32]. The deductive thematic analysis seeks to answer the researcher’s theory or analytical interest within the topic [32]. Phases in the thematic analysis included [32] (1) familiarization with the data by reading and rereading and noting down initial ideas (memoing), (2) duplicate coding (HS and JW) of one interview and comparing decisions to agree on standardizing the coding framework before coding all the transcripts, (3) discussing emerging themes with the research team, (4) reviewing themes with the wider research team and generating a map of the analysis (HP, SSG, PYL, ATC, or IY), (5) defining themes iteratively, and (6) presenting the deductive analysis with a selection of extracts. The data were organized using NVivo 11 (QSR International) qualitative data analysis software (HP and IY).

Five patients (including 3 who attended stage 1) attended stage 5 (prototype testing). Table 3 summarizes the demographics of the patients involved in stage 5.

We conducted a design sprint workshop and employed a web-based format to ensure the safety of researchers and patients during the global pandemic. In this 5-stage workshop, we developed a low-fidelity prototype based on theoretical frameworks and refined it based on patient feedback during the design and testing stages. Patients described the resultant app to influence their ability to self-manage in four ways: offering information, providing accessible asthma action plans, motivating and supporting improved medication adherence, and promoting behavior change through a reward system. Specific usability issues were related to navigation, comprehension, and layout.

The involvement of stakeholders and the multidisciplinary approach at the development stage in the design sprint process are among the strengths of this study, which may increase the chances of the intervention meeting the needs of the target population. The 5-stage design sprint structure allowed the development and testing process to be performed quickly and efficiently at a low cost, which would likely be favorable in low-resource settings. Constructing a low-fidelity prototype offers many advantages in the initial stages of prototype development. It allows a quick gathering of requirements, ideas, and concepts and can be built rapidly [29]. The disadvantage was that the low-fidelity prototype lacked some core functionality (such as navigation features), so that the patients on day 5 gave feedback on a limited version of the app. Nevertheless, the feedback was beneficial and enabled the app to be refined after the workshop to produce a high-fidelity prototype.

The web-based approach connected people in different geographical locations and ensured safety during the global pandemic. We recognized that we do not have any participants aged >50 years, perhaps because this approach may be more appealing to younger age groups, although with limited health literacy, and it may not reflect the feasibility of using web-based methods for older age groups. We overcame limited internet access by providing an internet data voucher; however, we had to exclude those with no access to any digital devices. We provided training on the various platforms to be used in the workshop, thus overcoming the lack of digital skills. These strategies may assure researchers or intervention developers working in countries or settings with high levels of limited health literacy that this web-based methodology is of value.

The pandemic context may have explained the relatively small number of participants recruited during the testing stage. Some participants found it challenging to commit uninterrupted time to a web-based workshop while being locked down at home with their domestic or caring or home-schooling responsibilities. For comparison, face-to-face intervention design workshops have been reported with 14 participants in 5-day workshops [33] and 38 participants over 6 weeks [34].

Patients with asthma and HCPs were recruited for a workshop in March 2020, but this could not occur because of the compulsory lockdown imposed by the local authorities in response to the COVID-19 pandemic. Therefore, we adapted the workshop to the challenges of a web-based format and its potential impact on participants’ research experience [35]. A high-speed internet connection and technical skills in conducting this workshop were vital. To overcome some of these practical issues, we supported all participants with mobile web-based access through an RM 10 (US $2.42) internet data voucher to ensure that the participants would not bear the cost of internet access. Participants’ experience with technologies was around social media, that is Facebook, video call, that is, WhatsApp, and information searching platforms, that is, Google and YouTube. We found a lack of experience with videoconferencing platforms and the software we planned to use during the testing day. Thus, we conducted training sessions for all participants to avoid technical problems during the workshop.

In the context of interviews, the literature suggests that web-based data collection can produce data of similar quality to face-to-face interviews [35]. Although using a web-based platform to interview patients was a new experience for the researchers, we found that the web-based programs eased discussions, and interviews took place quickly and effectively. Compared with traditional qualitative interviews, one advantage was that other researchers could observe the interview sessions on the web-based platform, and they could make concurrent fieldnotes. From the patient’s feedback, although they knew they were being observed by additional researchers, being at their own home helped them forget about being observed and anxious.

In web-based discussions and interviews, dictation software has been used to capture audio data in text format, thereby avoiding transcription errors [36,37]. However, this was not possible in our context because the medium of interaction was the Bahasa Malaysia language, and the extensive use of colloquial language made it impossible to use any dictation software. Conducting research on the web raises concerns about participants’ confidentiality and data security. We ensured that entry to the workshop was password-protected to control access to maintain the participant’s confidentiality securely. The recordings were stored in a secure manner. For example, files from the workshop were encrypted and stored in a secure research data storage facility.

Our app focused on designing a pictorial asthma action plan as a core strand of tailoring supported self-management for people with limited health literacy. Other features were a simple language for symptom assessment, education and information resources, provision of visual and audio medication reminders, and practical behavior change strategies such as a reward system. A clear message from our previous qualitative work was that participants wanted an interactive approach to support an asthma action plan with few words and clear pictures. Our original plan was to provide a paper-based pictorial action plan, but as it is not interactive, we reconsidered potential formats and decided to deliver the pictorial asthma action plan using a mobile app. Using a mobile app was seen as promoting a sense of autonomy to feel empowered in managing asthma.

Although previous reviews have reported a lack of interest in action plans [21], others have reported on the keenness of people to use action plans in mobile apps [38], although none have explored plans tailored for people with limited health literacy or the innovative pictorial representation of actions in a mobile app. In the United States, a study found that action plans were written at the literacy level of sixth-grade (11- to 12-year-olds), which will be a challenge to those without formal education or only receiving primary school education [39]. The same study also found that more graphics within an action plan may be needed to increase the ease of use [39]. In our web-based design workshop, the extensive use of images, icons, and the use of simple language were among the strategies used to overcome the challenge of understanding a written asthma action plan.

We outlined recommendations for practice, policy, and research based on this study in Textbox 1. On a practical level, to help researchers’ concentration throughout the 5 days of the 7-hour workshops, we included frequent breaks and provided high-energy snacks. Each session was either 1 hour long with a 15-minute break or 45 minutes long with a 10-minute break. Committing to time in a workshop alone in front of the monitor can be challenging and mentally draining, so a week before the workshop, each researcher received a supply of high-energy snacks through the post. We also provided colored sticky notepads and permanent markers with similar tip sizes to ensure that all scanned sketches and writings were clear when uploaded on the web-based discussion board.

Working with people with limited health literacy enabled the development of an app that could support them in self-managing their asthma. Specific components included sources of information on asthma, pictorial asthma action plan, simple language, audio-visual prompts, and rewards for supporting adherence to daily therapy and scheduled reviews. Despite practical challenges, a 5-day web-based design workshop proved to be manageable, enabling meaningful engagement from patients and HCPs so that a prototype is now ready for feasibility testing.