This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mHealth and uHealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
The poor engagement of men with HIV care is attributed to a number of factors: fear of stigma, masculine representations, concerns related to confidentiality, and the time commitment needed to visit public health clinics. Digital technologies are emerging as an approach to support the engagement of men with care.
This study aims to deliver a usable and engaging tablet-based app, called EPIC-HIV 2 (Empowering People through Informed Choices for HIV 2), to support men in making informed decisions about engaging with HIV care in rural KwaZulu Natal, South Africa.
We employed a mixed methods, iterative, and three-phased design that was guided by self-determination theory (SDT), a person-based approach, and human-computer interaction techniques. We reviewed related literature and conducted secondary analyses of existing data to identify barriers and facilitators to linkage to care and inform content development and design principles and used focus group discussions with members of the community advisory board and general community to evaluate a PowerPoint prototype of the app; used observations and guided questions with a convenience sample of potential users from the intervention community to iteratively test and refine a functioning interactive app; and conducted qualitative interviews and satisfaction surveys with actual users to evaluate acceptability.
Phase 1 identified supply- and demand-side barriers to linkage to care. Specifically, clinics were feminized spaces unattractive to men with high social costs of attendance. Men did not feel vulnerable to HIV, preferred traditional medicine, and were afraid of the consequences of being HIV positive. Thus, the app needed to allow men to identify the long-term health benefits to themselves and their families of starting antiretroviral therapy early and remaining on it, and these benefits typically outweigh the social costs of attending and being seen at a clinic. SDT led to content design that emphasized long-term benefits but at the same time supported the need for autonomy, competence, and relatedness and informed decision making. Phase 2 indicated that we needed to use simpler text and more images to help users understand and navigate the app. Phase 3 indicated that the app was acceptable and likely to encourage men to link to care.
We found that iteratively developing the app with potential users using local narratives ensured that EPIC-HIV 2 is usable, engaging, and acceptable. Although the app encouraged men to link to HIV care, it was insufficient as a
Poor engagement of men with HIV care is well described [
Digital technologies such as mobile phones are emerging as one of the approaches to close the gaps across the HIV care cascade [
There are many challenges to the development of digital interventions: most importantly, ensuring that the digital platform is usable and engaging for end users [
As part of a larger, cluster randomized controlled clinical trial to increase home-based HIV testing and linkage to care in men [
The primary aim of EPIC-HIV 2 is to support informed decision making by making explicit decisions that are usually implicit and thus, encourage men to link to care, to initiate ART early, and to introduce the potential benefits of staying on ART. Therefore, it is aimed at shifting the motivation of link to care from unmotivated or a motivated to internally motivated because of the benefits of ART. SDT addresses factors that either facilitate or undermine motivation. According to SDT, human motivation is based on the satisfaction of 3 inherent psychological needs: (1) autonomy, (2) competence, and (3) relatedness. Autonomy requires an individual to make personal decisions and act in a way that corresponds to their identity, belief systems, and values [
In the case of HIV, SDT has been used to understand ART adherence, treatment motivation [
As mentioned earlier, EPIC-HIV is being evaluated in a home-based intervention to test and start (HITS), a cluster randomized controlled trial designed to evaluate the effect of 2 interventions—EPIC-HIV (male-targeted HIV-specific decision support app) and microincentives on increasing home-based HIV testing and linkage to care among men, with the ultimate aim of reducing HIV-related mortality in men and hence HIV incidence in young women in rural South Africa [
This was a mixed methods study using a person-centered approach and human-computer interaction (HCI) techniques. The study was conducted in 3 phases. Phase 1 focused on user requirements and early prototyping. It included identifying barriers and facilitators to HIV linkage to care through literature review and secondary analysis of existing qualitative data from AHRI. We then developed a PowerPoint prototype of the app based on SDT and drawing on established behavior change techniques [
Mixed methods approach showing the 3 phases of the study. EPIC-HIV: Empowering people through informed choices for HIV; HITS: home-based intervention to test and start.
This study was conducted in uMkhanyakude district in KwaZulu-Natal, South Africa. The area is predominantly rural and poor, with high levels of unemployment [
All participants were recruited from the AHRI HIV surveillance community. Each phase employed different recruitment and sampling methods; these are described in detail under each phase.
Ethical approval for the study was granted by the Biomedical Research Ethics Committee of the University of KwaZulu-Natal (BREC 398/16).
Below, we provide an overview of each phase, outlining its main aim, study procedures, and participants.
The aim is to (1) gather the app content requirements and formulate design principles and features that the app should include by identifying modifiable barriers to linkage to care, (2) develop an initial prototype, and (3) evaluate it with potential users for overall impressions, appropriateness, and relevance to our setting.
First, we drew on systematic reviews conducted on barriers and facilitators to testing and linkage to care and searched for additional studies in South Africa or elsewhere in sub-Saharan Africa (
To obtain diverse views, we invited a purposive sample of men aged 18 years and above from the CAB and general community. We sought members representing periurban and rural communities in the surveillance area. In general, we aimed for men of different ages and occupying various positions in the community. For example, we recruited men from the taxi industry, church, liquor business, and formal and informal occupations, those who were unemployed, and those in the traditional health sector. The AHRI public engagement (PE) department assisted with the recruitment of participants. Through PE, the research team explained the study to the CAB, and the members who were willing to participate and provide informed consent were phoned for further engagement and to arrange a suitable time for the FGD. Of the 10 CAB members, 7 were included in the study. To recruit community members, a research assistant in the study approached men meeting the study inclusion criteria and explained the study, and those who were willing to participate provided their telephone numbers to be contacted to arrange a time for the FGD. Of 11 men who were approached, 4 participated in the study. We conducted 2 FGDs, one for the CAB and one for the community members.
The main aim of this phase is to iteratively design and test the app to refine it to be usable and engaging in our setting.
We conducted 3 cycles of usability evaluations to assess participants’ ability to successfully navigate the app, comprehend the educational content, and determine whether they found the app to be engaging and relevant. The first cycle was to establish usability requirements, the second cycle was used to evaluate the design changes made as a result of the first evaluation, and the last cycle was used to refine the understanding of the user requirements. User-centered evaluations were conducted with potential users using observations and guided surveys that consisted of a mix of open and closed questions to assess comprehension of the content and record usability issues. The questions were adjusted to suit the objectives of each cycle of evaluation with 2 questions that remained consistent to benchmark the design changes (
We used pop-up user testing [
The aim of this phase is to evaluate the acceptability and perceived value of the resulting app with actual users (participants who received the EPIC-HIV intervention in the HITS trial).
We conducted satisfaction surveys and individual in-depth interviews. An adapted Client Satisfaction Questionnaire (CSQ-8) [
Empowering People through Informed Choices for HIV 2 satisfaction surveys.
Survey item | Response, n (%) | |
|
||
|
Excellent | 13 (93) |
|
Good | 1 (7) |
|
Fair | 0 (0) |
|
Poor | 0 (0) |
|
||
|
All | 12 (86) |
|
Most | 0 (0) |
|
Only a few | 1 (7) |
|
None | 1 (7) |
|
||
|
Yes, definitely | 14 (100) |
|
Yes, generally | 0 (0) |
|
No, not really | 0 (0) |
|
No, definitely | 0 (0) |
|
||
|
Yes, definitely | 13 (93) |
|
Yes, generally | 1 (7) |
|
No, not really | 0 (0) |
|
No, definitely | 0 (0) |
|
||
|
Yes, definitely | 12 (86) |
|
Yes, I think so | 2 (14) |
|
No, I don't think so | 0 (0) |
|
No, definitely not | 0 (0) |
|
||
|
Yes, definitely | 14 (100) |
|
Yes, I think so | 0 (0) |
|
No, I don't think so | 0 (0) |
|
No, definitely not | 0 (0) |
|
||
|
Very satisfied | 12 (86) |
|
Mostly satisfied | 2 (14) |
|
Mostly dissatisfied | 0 (0) |
|
Quite dissatisfied | 0 (0) |
|
||
|
Excellent | 11 (79) |
|
Good | 2 (14) |
|
Fair | 1 (7) |
|
Poor | 0 (0) |
|
||
|
Very satisfied | 14 (100) |
|
Mostly satisfied | 0 (0) |
|
Mostly dissatisfied | 0 (0) |
|
Quite dissatisfied | 0 (0) |
|
||
|
Yes, definitely | 14 (100) |
|
Yes, I think so | 0 (0) |
|
No, I don't think so | 0 (0) |
|
No, definitely not | 0 (0) |
aEPIC-HIV: Empowering People through Informed Choices for HIV 2.
bARV: antiretroviral.
All men aged 15 years and above who received EPIC-HIV 2 between October 2018 and January 2019, as part of the intervention in the HITS trial, were asked to consent and complete the satisfaction surveys. For the individual in-depth interviews, we purposively selected 14 men from the 28 who had received EPIC-HIV 2 between April 2018 and January 2019 (which meant that they had tested positive for HIV but did not link to care within a month).
Each phase is presented separately with a particular focus on the iterative development.
Phase 1
Identified key learning points to incorporate in the content to persuade men to attend clinic and start antiretroviral therapy
Long-term health benefits of starting antiretroviral therapy early and remaining on it
Phase 2
Men in our setting did not understand the seesaw metaphor
The click worked better than swipe
Two options were better than 3 options
Phase 3
App was usable and acceptable in our setting
It encouraged men to link to HIV care
In the literature review and analyses of the earlier study, we identified barriers and facilitators to HIV testing and linkage to care among men, particularly relevant to our setting (
The identified key barriers and facilitators to HIV testing and linkage to care and the key learning points for EPIC-HIV 2. ART: antiretroviral; EPIC-HIV: Empowering people through informed choices for HIV; HITS: home-based intervention to test and start.
A total of 11 men aged 34 to 66 years participated in 2 focus groups to evaluate the first (paper) prototype. They provided feedback on content and design (organization of the app, appropriateness and appeal of language and images, acceptability and relevance, and overall app impressions). On the basis of the feedback from the FGDs, the team modified the initial conceptual design of the app and created a revised plan for the app features and content.
The final realized product (EPIC-HIV 2) has 3 main modules (sections), and to further support the psychological requirements for autonomy, the user has the option to engage with 1 or all 3 modules (
The final conceptual design (realized in the product) starts with an introduction and instructions on how to use the app, and users are then able to choose between “Jabu’s man to man advice” or “healthy and strong with HIV” modules. The “shield” module was not offered initially because it has more focused material and is expected to appeal to fewer men (and to make the number of choices manageable). After users complete the initial module, they are returned to the menu and are able to choose one of the remaining modules or to end.
PowerPoint prototypes showing examples of the content from each module (presented in English, although all content was delivered in isiZulu in the final app).
Map of the pathways through the 3 main modules of empowering people through informed choices for HIV 2.
The main outcome of phase 1 was an in-depth understanding of user needs relevant to EPIC-HIV 2. We subsequently refined the prototypes using a standard app development toolkit to deliver testable interactive demonstrator prototypes on an Android tablet computer. In this paper, we illustrate the approach to iterative design and usability testing by focusing on the development of the decision support tool (a component within
The decision support tool aimed to help men weigh up and reflect on the costs and benefits of attending a clinic. The user was presented with different statements that either support or delay going to the clinic. To facilitate engagement, we wanted to make this section interactive. Initially, we applied a seesaw metaphor for users to attempt to render usually implicit decisions as explicit and interpret whether they were ready to attend a clinic within 2 weeks. This was done by adding the arguments to one side or the other of the seesaw to illustrate how the arguments stacked up in decision making. For example, if the user agrees that they want to see their children and grandchildren grow up, that would be an argument that weighs in favor of engaging with HIV care, whereas if they agree that they do not wish to be seen at a clinic by one of their neighbors, that would be an argument that weighs against engaging with care. In making design decisions about the clinic decision support tool, we prototyped and tested 2 versions: drag-and-drop with illustrations and a Likert scale without illustrations, both employing the seesaw metaphor. These are illustrated in
Example screenshots of drag-and-drop (statement with an image and illustration on how to drag) versus Likert scale (statement only).
Example of the seesaw screenshots with statements only (from Likert scale version) versus statements with images (from drag-and-drop version).
The main objective of the first evaluation was to refine the user requirements. The key objectives and results of this evaluation are shown in
In total, 12 men participated in this evaluation, which was conducted as outlined earlier. In addition to the questions related to the key objectives outlined in
To test which interaction style participants preferred and found easier to use between a drag-and-drop version (dragging each statement into a bucket labeled
Participants found it difficult to use the drag-and-drop version, particularly more aged participants and participants with lower digital literacy levels
A majority of participants needed to be shown how to drag-and-drop, despite the illustration at the beginning of the statements
In contrast, clicking was intuitive; almost all the participants started clicking when they took the tablet
To determine whether the
Some participants preferred drag-and-drop because it had 2 options (agree or disagree): “drag—I found it easy with the scale I was not completely sure of what was happening. With the drag option there was only two options which made it easy to select” [P6]
Another participant reported that he found the neutral option confusing
To test whether participants preferred each option presented as text only or as text with an illustrative picture
Most participants preferred text with an illustrative image to help them select the statements over the text only with many participants stating that images made it easier to understand the statements: “the image makes it clear what the statement says” [P7]
Some found the images to facilitate engagement: “the images capture your attention” [P6]
To test whether participants were able to correctly interpret a seesaw metaphor, in which each response was stacked up on the end of a seesaw depending on whether it weighed toward or against attending a clinic (correct interpretation of the metaphor was that the end that was more heavily weighted and hence
Participants did not understand the seesaw metaphor
Many expressed confusion on how to interpret the results: “not sure how it should look like on the seesaw. When is it good, when it is up or down?” [P1]
To assess whether it was clear how to agree or disagree with a statement and whether it was clear how their selections related to the outcome of the seesaw
It was not clear for some participants on how to agree or disagree with the statements; 2 participants (2/8) in the Likert scale option selected agree for all the statements
On the basis of the findings, the following design changes and decisions were made (as illustrated in
As the app was intended for one-off use, after seeing that clicking was intuitive and users struggled with the drag-and-drop, we decided to use clicking to select statements.
We limited the options to agree and disagree, as some participants did not find value in the neutral option.
To support users to make informed choices regarding attending a clinic for care, we used text and image as users expressed that images made it easier to understand the statements.
It was judged that the seesaw metaphor was not applicable for this exercise, as many participants struggled to understand what it meant if one side was more heavily weighted than the other. We removed the seesaw, summarized the statements, and explained what it means if they were skewed on one side. Furthermore, we labeled the right side as supporting going to the clinic in 2 weeks and left as delaying going to the clinic.
No further changes were judged necessary to address this question.
Example screenshot showing the design change with the 2 options, agree or disagree.
Example screenshot showing statements only without the seesaw and a clear label showing which side supports going to the clinic and which supports delaying (in English, but in isiZulu in the final app). This is a preliminary version, and the data were not final.
The main objective of the second evaluation is to validate the design changes made, particularly the summary screen shown in
The design of the selection tool was received positively by all participants. All participants were selective on the statements they agreed or disagreed with as per their individual experiences, and they were able to interpret what it means if statements were weighted on one side (
Of 7 participants, 4 preferred the text only in the summary page, validating the first evaluation findings that images were only important at the selection point. In addition, 6 participants reiterated that the key message of this section was about the importance of going to the clinic to start ART, with 1 participant adding that one should go to the clinic early and not wait until they are sick. Moreover, 5 participants said they would prefer the agree or disagree buttons to be color coded, with 4 participants suggesting that agree should be green and disagree, red:
colours help, maybe green and red so that they can see that they have made the right or wrong choice.
One participant stated that different colors would be particularly useful for people who cannot read.
On the basis of the preference for color coding and the association of green with right and red with wrong, we decided to use neutral colors that are not associated with right or wrong to help users make an individual choice as opposed to fearing being right or wrong (
Design changes showing the neutral colors for agree and disagree.
This evaluation was conducted as part of the full Jabu section testing, mainly to assess understanding of the content as well as to check whether users find the presentation of content interesting and to validate the changes made as a result of evaluation 2. A total of 8 men participated in this cycle.
This evaluation did not reveal any usability issues, indicating that the changes made to the designs were effective. Participants were paying more attention before deciding whether they agreed or disagreed, and they listened to the audio while looking at the screen. Overall, participants reported finding the decision support tool engaging and useful. For example, when asked which part of the app they thought would be most relevant to their friends or other men, 3 participants referred to the clinic decision support tool. In addition, 1 participant, when asked if he learned anything new as a result of using the app, mentioned that he learned how to use a tablet, suggesting that there was incidental value (not directly related to the primary aim) in introducing such an app to people.
A total of 15 participants received EPIC-HIV 2 between October 2018 and January 2019, and 14 completed the user satisfaction surveys. Their mean age was 35.6 years (SD 13.38299), and a majority (8/14, 57%) had high school education, and half of them were unemployed. User satisfaction ratings were high (
A total of 14 men (aged 35-53 years) who had tested positive for HIV and had not linked to care within a month of receiving their test results received EPIC-HIV 2 between April 2018 and January 2019 agreed to be interviewed; 11 of them had linked to HIV care at the time of the interview. A majority of participants reported that the app offered relevant information about HIV management, boosted their confidence, and encouraged them to start and stay on treatment. However, this was not enough for all participants to link to HIV care. We report the experiences of men who used EPIC-HIV 2 using the SDT framework: autonomy, competence, and relatedness.
Participants described the app as motivating them to start ART by providing them with information on how to manage HIV to live a healthy life:
Mmm... I can say I have listened to the App from the beginning up to the end. Yes, I have really seen a need...I have seen it as a very important thing. I think that has encouraged me because I have never been to the clinic previously, but I went to clinic after that.
For some, this was a catalyst to link to care (even if they did not act on it immediately):
I was supposed to visit the clinic a long time ago, but I didn’t because of work...I was thinking about that when I listened to EPIC [app]...you must be open as the EPIC was saying that you mustn’t fear that people will laugh at you. I took a decision because...this is life and we are living within a community...I realised that in order to be assisted, I must go to AHRI nurses at the clinic.
However, for some participants, the app alone was insufficient to motivate them to go to the clinic for HIV treatment. Other factors reduced the logistic barriers to linkage that nudged them to link. For some participants, it was the availability of our research nurses at local clinics, with the:
referral letter... that was going to make things easy for me...
For others, such as participant 1, it was experiences such as the death of his best friend that motivated him to link to care after using the app:
Er...I can just say that there is nothing new I have learnt (from EPIC)...I started treatment because I've lived with people who had been using it (ARVs) before I knew my health status. One of my best friends passed away after he defaulted from taking treatment. After I found out that I am HIV positive I was motivated, and I said I will try by all means to take treatment accordingly. So, I can say I have learnt a lot from him about the risk of delaying...
This suggests that for men to exercise autonomy to link to care, in addition to the app, external factors such as clinic operation hours, support, and individual experiences of HIV need to be considered.
Several of the participants reflected on how the positive narratives in the app encouraged them to take charge of their lives and manage HIV to improve their health outcomes and “avoid infecting other people with diseases” by describing “how to take the pills on time” and “comply with treatment.”
Some participants also reported that app messaging lessened their fears (particularly fear of disclosure) and motivated them to adopt positive attitudes to advance their health. The quotation below illustrates how some participants felt about their fears:
...I had that feeling of fear to say, oh my Lord how could I do this thing since I am afraid as it will be revealed? Eh, I also gained advices because they start with allowing you to listen from the messages that say once you know your status, you can get help so it’s up to you whether you do it or not. But I chose to be assisted because I am not giving up on life.
Others also drew attention to the message of EPIC-HIV 2 that it gave them tips on how to discuss disclosure with their sexual partners:
Er...like disclosing your health status to your partner so that she can check for HIV as well for both of you to take treatment, yes, I think it’s good not to hide from your partner:
Overall, for many participants, the app supported their competence to link to care through the positive messaging that gave hope, alleviated their fears, and reinforced the need for them to manage their health effectively to live a healthy life while minimizing the risk of transmitting HIV.
Participants reflected how the narratives of different characters in the app reverberated with their present realities and how the app stories shaped their understanding of the importance of HIV treatment and management for them to live a healthy life:
...They [app characters] have mentioned the fact that once you have defaulted from taking treatment you will be very sick…the characters I have seen on the app were fit. They were sharing their experiences as to how they are taking their ARVs...
They [app characters] were sharing simple information, something that is said even on TV by Ministers that you need not to be bedridden until you go to the clinic.
Furthermore, a participant who had previously defaulted treatment had this to say:
The part I have seen as most essential one for me it’s where the characters were talking about the risk of defaulting from taking treatment...It motivated me to decide to be re-initiated on treatment...It is really necessary...because characters on the Epic were talking about things they had experienced.
In sum, the empathetic nature of the personal testimonies of various characters in the app resonated with participants, and they felt cared for. Some of them described the relevance of the information shared through the app and were able to initiate treatment and took the decision to manage HIV to improve their health outcomes.
We found that it was possible to successfully apply a multiphased iterative development process to create a theoretically informed, interactive, tablet-based app to support men in making informed decisions about engaging with HIV care in a low-income, rural South African setting. Men in our setting who had been missing from the HIV care cascade found the app to be acceptable and reported that the stories in the app resonated with their realities and encouraged them to link to care. Previous work has established that digital interventions are acceptable and effective in improving clinic attendance, ART adherence, and turnaround time from testing to treatment [
Drawing on behavior change theory, such as SDT, ensured that EPIC-HIV 2 supported individual decision making. There is an increasing need to apply evidence-based theories to enhance the development of digital interventions and improve their efficacy [
Using a mixed methods approach to understand usability and acceptability and the potential of the app to encourage men to link to care are strengths of the study. However, the generalizability of study findings outside our area may be limited because of the locally tailored design of the EPIC-HIV 2 app, the sampling method, sample size, and specific study sites. In addition, the possibility of social desirability bias cannot be excluded as some participants might have found it easier to say positive things about the app and not necessarily translate to the actual initiation of ART. Furthermore, we cannot establish what the outcome would have been if we had made different design decisions or applied a different theory of behavior change. The decision to develop a tablet-based app was fitting for the specific context of use, but more widespread use of such an app would probably be achieved by implementing it on mobile phones for independent use; this remains an area for future work.
Using a multidisciplinary approach and drawing on evidence-based theories to develop digital interventions can ensure that the resultant products are acceptable and engage in a wide range of users. Our work aims to pave the way for a greater focus on mixed methods and person-centered approaches to the development of digital interventions for HIV care.
Literature review and secondary analysis.
User evaluation questions and guides.
Self-determination theory app content and design.
Example of navigation pathway in the Jabu module.
Africa Health Research Institute
antiretroviral therapy
antiretroviral
community advisory board
Client Satisfaction Questionnaire
Empowering people through informed choices for HIV
focus group discussion
human-computer interaction
home-based intervention to test and start
public engagement
self-determination theory
The authors would like to acknowledge the study participants for their time. They wish to appreciate Velaphi Gumede, Sithembiso Luthuli, Siyabonga Nxumalo, and Jaco Dreyer for participant recruitment, data collection, and data management support. This study was supported by a Partnership Research Fund from the Engineering and Physical Sciences Research Council Interdisciplinary Research Collaboration Early-warning Sensing Systems for Infectious Diseases (i-sense) EP/K031953/1. The HITS trial is funded by the National Institutes for Health R01AI124389. The AHRI’s Demographic Surveillance Information System and Population Intervention Program is funded by the Wellcome Trust (201433/Z/16/Z) and the South Africa Population Research Infrastructure Network (funded by the South African Department of Science and Technology and hosted by the South African Medical Research Council). The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding bodies.
FT, MS, SW, and JS are the investigators of the HITS trial and developed the EPIC-HIV intervention design and protocol in collaboration with AB and PM. PM, AB, AZ, TM, TZ, SW, MS, OA, and FT were involved in the development of EPIC-HIV. SW, TZ, MS, OA, and JS designed the process evaluation. KD provided statistical support. TM wrote the first draft of the manuscript with support from AB and OA and coordinated the study. All authors contributed to the manuscript. All authors have read and approved the final manuscript.
None declared.