A Consumer Behavior Perspective of Adopting Mobile Contact Tracing Apps in a Public Health Crisis: Lessons from ABTraceTogether for COVID-19 Pandemic

Abstract

Responses to the COVID-19 pandemic included m-Health innovations, such as contact tracing and exposure notification applications to track virus exposure. Such apps were released by over 45 international governments throughout 2020, becoming the first m-Health innovation with such widescale deployment. Most regions relied on voluntary adoption, and many failed to receive a critical mass of users. Some of these apps can track and share user’s locations, social contacts, and health information, which sparked concerns and misperceptions about the privacy and security of user data. It is important to understand consumer behavior and adoption challenges based on people’s perceptions of benefits, barriers, and risks. To investigate this, we sent an online questionnaire to over 600 participants with open-ended questions asking about their experience with one such app, ABTraceTogether. This chapter covers qualitative findings regarding device and application-level issues participants identified as barriers to their adoption and continued usage of the app, which are accessibility, battery life, downloading challenges, device memory, network connectivity and costs, operating system compatibility, performance issues, and usability. Insight on consumer behavior gained from this study can guide m-Health design and promotion to aid future health crises and personal wellbeing.

1. Introduction

In response to the tremendous human and economic costs of the COVID-19 pandemic, many governments around the world sought ways to curtail the spread of the disease. The pandemic spurred innovation and changes to consumer behavior in dramatic ways. One such change is the widescale adoption of m-Health applications. m-Health refers to mobile applications that promote an individual’s physical and mental wellbeing. The benefits of m-Health to consumers include their “portability (anywhere), immediacy (any time) and convenience (easy access)” ([1], p. 245). m-Health applications are used, for example, to help track the location of dementia patients [2], to allow people with diabetes to monitor their diet [3], for people with disabilities or environmental allergies to avoid barriers or health risks [4], and other purposes. Within health crises contexts, prior to the COVID-19 application, mobile applications were used in limited capacities for tracking influenza [5] and ebola [6], but the scale of use of contact tracing apps was unprecedented prior to COVID-19. With over 45 countries launching a contact tracing app [7], this is the most widescale deployment of m-Health technology.

It can thus be seen that the COVID-19 pandemic providing a test ground for widescale adoption of an m-Health application. In a post-pandemic world, lessons learned from the challenges and responses to this disease can be useful to understand consumer behavior regarding the challenges surrounding adoption and use of m-Health applications, as well as any digital media that collect data about consumers’ location, social contacts, and personal data, such as health status. It is worthwhile to determine what limited the uptake of contact tracing applications to ensure that these issues are fully understood and can be addressed not only for future crises management, such as pandemic or public health emergencies, but also for a wide range of individual health and wellness issues. Using a consumer behavior research perspective to study these issues is important as the individual adoption of a non-mandatory information technology application can be regarded as a decision to purchase a product or service based on their value (i.e., the difference between anticipated benefits and costs) perceived by potential consumers.

Despite the urging of government officials and marketing campaigns, the adoption rates of contact tracing often failed to receive critical mass levels in most regions. Although studies have been conducted on the adoption challenges of contact tracing apps (as covered in the literature review) most of these studies utilized methods that limited the ability of participants to identify their own issues and provide their experiences and beliefs in their own words. Insight gained from qualitative participant data is valuable as it can provide a more complete picture of the overall factors and context influencing user adoption.

The use of contact tracing apps for COVID-19 was voluntary in most jurisdictions. One major drawback of optional usage is that it can be difficult to obtain a critical mass of users. The apps only work if the people users come in contact with are also running the same app at the same time, hence a critical mass of users is needed for the app to be effective. Yet, adoption rates for contact tracing worldwide have been lower than governments hoped for, with one study finding a 22.9% adoption rate in 21 countries [8]. High adoption rates are difficult to accomplish, even in countries with populations with widespread smartphone ownership, given the difficulty of overcoming data security, surveillance, and privacy concerns [9], as well as informing and marketing to the public about the existence of the app, and convincing them to download it and use it continuously. It is therefore crucial to gain an understanding of adoption concerns.

Our study focuses on the adoption and usage challenges surrounding one contact tracing app, ABTraceTogether [10]. The app was launched for iPhone and Android smartphones by the province of Alberta, Canada in April 2020, many months before Canada launched its national contact tracing application. ABTraceTogether uses a proximity-based approach via Bluetooth and decentralized data storage. The app anonymously encrypts and records on a user’s device all other app users that a person comes in contact with for 15 minutes or longer within a two-meter radius. Infected app users are asked to share their log of anonymous contacts with public health officials so that the contacts can be informed of their exposure risk. ABTraceTogether was the first contact tracing app to be launched in North America [11]. Alberta is a province in western Canada with a population of 4 million centered in the urban centers of Calgary and Edmonton. The app was built through a partnership between the provincial government and the private sector and was based on the open-source code provided by Singapore. Despite government and media marketing efforts for the app, within several weeks after launching the app, only about 5% of Albertans had downloaded it [12]. As of January 2021, it had been downloaded by 7.5% of the Alberta population, and it had led to the detection of only 32 COVID-19 cases [13].

Several weeks after becoming available to the populace of Alberta, we launched a questionnaire to approximately 300 users and 300 non-users of the application to understand their motivations for either using or not using ABTraceTogether. From the users of the app, we wanted to learn their motivations and what worked for them, as well as any installation or usage barriers or concerns. From the non-users, we wanted to learn what factors discouraged or prevented them from using the app or caused them to stop using it. This paper reports on the qualitative findings from our study regarding application and device related issues. These findings reveal a range of issues that both encouraged and stymied the adoption and continued usage of the app. Findings from our study detail the microcosm of issues at play that affect whether consumers will sufficiently adopt a technological innovation. Despite the possible health benefits and social value of m-Health, user barriers and concerns must be considered to help achieve widespread adoption and continued usage. It is important for application developers, promoters, and sponsors to address adoption and usage challenges that can result in insufficient uptake or discontinued usage.

2. Background

Contact tracing is an established practice involving public health workers manually contacting people who were exposed to an infected individual [14]. They inform people of their potential exposure and encourage them to get tested and to quarantine, thus limiting the spread of an infectious disease. This manual approach is limited, however, as it requires extensive skilled labour to do the contact tracing interviews and it relies on infected people remembering where they have been and whom they have been in contact with [15]. Expedient contact tracing is crucial with COVID-19 as half or more of infections occur prior to any symptoms appearing [16]. Yet the vast number of COVID-19 infections made it very difficult for public health workers to keep up with manual contract tracing [16]. As a result, many governments decided to use technology-based solutions, such as smartphone applications, to assist with contact tracing. In March 2020, Singapore was the first government to launch a contact tracing app for COVID-19 [17]. Contact tracing apps have two main benefits. First, they record all contacts an individual has with other app users so there is not a reliance on people’s limited memory [15]. Second, they scale up to vast multitudes of users and provide results faster than human-only processes [16].

These apps use various methods for tracking possible virus exposures, but most use either Global Positioning System or wireless signals (e.g., Bluetooth). Bluetooth-based applications are known as proximity-based apps as they do not track the user’s location but rather the device’s proximity (generally two meters) to other devices with the app installed after a designated number of minutes [14]. The Bluetooth approach became the favored approach in Western nations [14] as it ensures the privacy of users and their contacts by logging other nearby app users anonymously. With contact tracing applications apps, such as ABTraceTogether, if a user becomes infected, they are requested to report this information to the app [10]. The app either sends an anonymous alert to those who have come in contact with that individual (as with “exposure notification apps”) or alerts a public health agency who manually conducts contact tracing using the supplied contact data (as with “contact tracing apps”). For the purposes of this article, the term contact tracing app will be used to encompass the various types of similar apps.

The proximity-based approach has limitations as it can generate incorrect or incomplete results, such as requiring a longer time for devices to be in proximity than is needed for virus transmission. As well, Bluetooth signals can travel through walls even though viruses cannot [18]. Another important distinction between contact tracing apps is whether the app uses a centralized or decentralized architecture and data storage model [19]. With centralized applications, the user’s application data is stored on a shared, external server (such as from the app sponsor), which entails greater data privacy concerns. Decentralized applications store user’s data only on the user’s device, thus limiting risks of third-party surveillance or hacking of private data.

From the first launch of COVID-19 contact tracing apps, discourse in popular media and among pundits often centered around privacy and surveillance concerns. Despite some skepticism, contact tracing apps have made a meaningful contribution to limiting the spread of COVID-19 [20]. For instance, one study found that even a 1% increase in uptake of a COVID-19 contact tracing app can reduce cases from between 0.8% to 2.3% [21]. Abueg et al. [22] determined via mathematical modeling that with 15% of a population using a contact tracing app, infections and deaths could be reduced by approximately 8% and 6%. Despite the potential of such apps, adoption rates were often not at the desired level, thus research into consumer behavior aspects of this technology is helpful.

3. Literature review

The main factors found to affect whether a person chooses to use a contact tracing app include usability issues and poor app design [8, 23]. Walrave et al. [24] found that a person’s perception of how well they are able to use mobile applications affects their decision whether or not to use a contact tracing app. A study by Jansen-Kosterink et al. [25] found that attitudes towards technology, privacy concerns, fear of the disease, and age were prime factors driving adoption. A study by Redmiles and Hargittai [26] also found that technical familiarity as well as a person’s risk of health problems were main factors. A survey by Guillon and Kergall [27] found that adoption of France’s contact tracing app was influenced by people’s trust of the government (the app sponsor) and their perceived risk of illness. A study comparing the use of contact tracing apps in five countries, Scotland, Cyprus, Iceland, Ireland, and South Africa, found eight factors present in each jurisdiction that affected adoption contact tracing apps. These factors include, “perceptions of data collection and management, sense of community, communications and misinformation, accessibility and inclusion, trust in public/private institutions, policy and governance, response infrastructure and digital capability” [28]. An Australian survey found that “effort expectancy, perceived value of information disclosure and social influence are critical for adopting contact tracing apps” [29].

Different nations varied in the types of adoption issues for local populations. A study by Hassandoust et al. [30] in the United States found that privacy issues as well as perceived health risk were the main reasons why users decided to use a contact tracing app, while in Germany, offering a financial incentive was found to be effective [31]. Fox et al. [32] found that users specifically are motivated by the health benefits of using the app. Kokkoris and Kamleitner [33] and O’Callaghan et al. [34] found that prosocial beliefs, that is a desire to help the public or one’s community, were leading factors motivating people to use contact tracing apps. A Canadian survey of Alberta doctors found only 27 percent recommended that their patients use the ABTraceTogether app due to reasons such as security and privacy concerns, distrust of the government, and a belief that recommending the app to patients was not their role [35].

Although our study focuses on one government-sponsored application, there are other approaches, such as apps offered by private companies (e.g., Google & Apple), academic institutions (e.g., MIT’s Pact https://pact.mit.edu), and non-profit NGOs (e.g., COVID Watch https://www.covid-watch.org). Indeed, such a multifaceted approach and a lack of standardization has been criticized as an adoption limitation [36].

The literature on adoption and usage challenges of contact tracing apps revealed that studies have largely utilized quantitative instruments, specifically surveys and mathematical modeling. Few studies have used qualitative methods to enable the public to share their issues and concerns more openly for and against using such an app. These approaches limit the ability of actual users of the apps to identify their own issues and provide their experiences and beliefs in their own words. In discussing the “thorny problems of COVID-19 contact tracing apps”, Osman et al. [37] call for a holistic understanding of the issues. Our study offers a more holistic approach than prior studies, providing insight from the public to gives a fuller picture of the overall factors and context.

4. Method

This study was conducted via an anonymous online questionnaire sent to over 600 participants in the summer of 2020. This article focuses on one dimension of our study (see [38, 39] for additional information on our study), specifically the qualitative data gained from three open-ended questions asked to participants. Qualitative data was gathered via a questionnaire that asked participants to respond to open-ended questions that solicited their thoughts and feelings on this issue in an unstructured manner [40]. The questions asked of users and non-users pertained to their concerns and challenges downloading or using the app as well an open question to comment about any dimensions of their experience with the app. This method has the advantage of enabling participants to share their views in an environment free of real or perceived judgment, enabling participants to openly share their thoughts on what proved to be a controversial issue. Participants were drawn from two groups 1) self-identified users of contact tracing app, ABTraceTogether and 2) self-identified non-users of the app to provide a wider spectrum of views. Participation was open to anyone who lived in Alberta, used a smartphone, was 18 or older, and was able to provide informed consent. An external company hosted the questionnaire and for participant recruitment. A small financial incentive was offered to participants. We received participation from 309 users and 306 non-users.

We received a rich depth and variety of responses. The qualitative data from all users was analyzed by the researchers using data analysis software using categories established from the literature and as emerged from the data using Dey’s suggested methodology [41]. All responses from participants were imported into the software, reviewed for a match to existing codes or for new patterns, and was coded as applicable. Subsequently, Dey’s methods of splicing, splitting, linking, and connecting categories were used by the researchers to develop a codebook. This codebook with sample coding was shared with fellow researchers to validate the coding before an additional, final round of focused coding was conducted. The final stage entailed analyzing patterns and establishing connections to established literature to arrive at overall conceptual findings.

This study was approved by the ethics review board of a North American university and followed all established procedures for conducting ethical research. As a qualitative study, the findings presented are not attempting to rank issues or find correlations, but to identify issues and offer insight and context on pertinent factors.

5. Findings & discussion

The qualitative findings covered a diverse array of issues including technological, geo-political, and personal values and beliefs. Our findings can be grouped into two top-level categories, specifically issues at an application and device level and issues at a personal level. These issues are grouped together as the interface or programming structure of the application itself, ABTraceTogether, impacts the individual’s device. For example, the need for the application to be constantly running and active on a user’s device, which drains user battery, is both an application level and device level issue.

This paper covers participant concerns at an application and device level as [39]addresses participants’ personal level issues, so they will only be briefly highlighted here. Personal-level issues center around individual’s personal beliefs and values as they pertain to their willingness to use ABTraceTogether and general beliefs about the pandemic. Despite efforts by officials to assuage participants that their privacy was protected, and their location was never being tracked with ABTraceTogether, participants were nonetheless concerned with the privacy of their data. Other personal issues include one’s level of altruism or pro-social beliefs, sense of personal agency (i.e., that one’s use of the app can make a difference), confidence in the efficacy of the app, personal concern for safety or the safety of loved ones, social pressure, and level of trust in the app sponsor (i.e., the government). Beyond the scope of app developers, issues about the pandemic generally, such as belief in the reality of the pandemic (i.e., that it is a hoax) and the presence of misinformation, was raised by participants as having a detrimental effect on their app adoption and usage.

The remainder of this paper focuses on participants’ barriers and concerns at an application and device level. It should be noted that some participants expressed positive feelings and user experiences with ABTraceTogether and indicated their downloading and usage of the app was seamless. This paper predominantly reports on the barriers and concerns of participants as these are areas that should be addressed to aid adoption and usage.

Our findings include both factors that are genuine (actual problems with the app) as well as participant perceptions that are not based on fact (they are based on faulty understanding or misinformation). For example, there were numerous participants claims that the app was tracking them via GPS. Despite assurances that this were not the case from the app developers, from the government, and from the news – these false beliefs were nonetheless raised a barrier. We also uncovered misunderstandings at a smaller scale, such as that the app would quickly drain one’s battery. Whether actual or perceived problems, these issues are nonetheless crucial as they indicate areas that app sponsors and developers need to address, if not in the app design and functionality itself, then in the educational, support, and promotional material. A final concern regarding some participants’ responses is that at times they appear to be accounts [42], that is, excuses to justify their unwillingness to use the app. Some fundamental barriers identified by participants could be addressed through minimal effort or a small sacrifice, such as removing other apps to make room for the app or contacting tech support. Whether real or perceived concerns, the responses raised by participants should all be considered so as to design and launch apps that are as free of user barriers and concerns as possible, and thus will increase the likelihood of an app being seamlessly downloaded, installed, and continually used.

Participants raised various issues relating to the application and its impact on their smartphone device. Based on our findings from participants’ responses to the open-ended questions, we have grouped their application and device level issues into the following domains, ordered alphabetically: 1) Accessibility, 2) Battery, 3) Download, 4) Memory, 5) Network, 6) Operating system, 7) Performance, and 8) Usability. There is an additional level of issues within these domains. To view these issues and their groupings, we included a diagram below, see Figure 1.

Figure 1.

Diagram of the application and device related issues with ABTraceTogether.

The following paragraphs cover the main dimension of participants’ concerns within these areas. Sample participants’ quotations were selected for inclusion based on their representativeness for addressing salient dimensions of a given issue or for their degree of insight offered.

5.1 Accessibility

Accessibility refers to barriers in downloading or use of the application based on the device owner’s abilities (such as physical or cognitive disabilities), language abilities, or financial constraints (i.e., ability to afford a device). There were only a few comments regarding accessibility based on language or abilities as barriers, but for the people who noted this, it was a profound barrier. One participant was concerned about the cost of running the app and suggested that they “make it cost-effective for all people [with] low income or disabled.” Another was concerned that the app was too hard to use and thought that the makers should “make it easier for the uneducated to understand.” Other participants were concerned about language issues and suggested that there should be “more language options”. Another participant thought that, “The app wouldn’t translate well between other diverse Canadians.”

The bulk of participant accessibility concerns related to financial barriers, both in terms of the necessity to have a newer Android or Apple smartphone and the perceived costs of network connectivity. Some users noted that their phones were too old for the app. For example, one participant expressed frustration that the app would not function on their Blackberry device, noting “I have a refurbished Blackberry with a pre-paid plan of only calling and texting but no data or Wi-Fi. I cannot afford to upgrade my phone or be on a plan where I receive a bill. I feel like this is a program for the majority of society but leaves out those who do not have access to this technology due to income or social standing.” These comments reflect the inequities in availability of contact tracing apps for those who cannot afford newer smartphones. Governments need to look at other solutions to reach all citizens. For instance, Singapore distributed Bluetooth-enabled tokens to citizens who do not have access to smartphones, which can also track and store anonymous contact data [43].

Some people who owned Bluetooth enabled smartphones but had financial concerns about data usage decided not to use the app based on false beliefs and misunderstandings about how the application worked. In particular, a number of participants thought they needed a data plan to use the app, and that it would incur high data costs which would make it too expensive to use. They were unaware that the app uses Bluetooth technology and does not require a constant internet connection or a data plan. The app does need Wi-Fi to initially download the app, for updates, and to share the data if requested by government health officials, but this can be accessed through publicly available free Wi-Fi connections. One participant concerned about data costs stated that they chose not to use the app because, “I have a cell phone that I cannot afford to get a provider for, thus financial hardships makes me decide between many more life needs.” Another participant noted that they thought there would be high data costs and said, “Do not make me feel discriminated against or slighted because I am not able to access this program.” Another user concerned about high data costs noted that, “The people who are at risk for Covid are those in low socioeconomic status…with lack of access to technology. Therefore this app is inappropriate for this.” Participants expressed valid concerns about the inequity of the app being unavailable to those who cannot afford the latest devices. However, the false belief that the app would incur high data charges served as a barrier to people using the application and also seems to have contributed to ill-will and a sense that the Alberta government was only helping the rich by implementing the contact tracing app. Outreach was needed to educate individuals about how the app works and to address these misunderstandings and concerns.

5.2 Battery

One of the main concerns of participants was the misconception or false belief that using the contact tracing app would drain their mobile phone’s battery, presumably because the app requires Bluetooth to be on constantly. A number of participants noted that they stopped using the app because of the battery drain. One stated, “It was using over 60% of my battery life. I can’t use it” while another noted that, “My iPhone 11 Pro’s battery life is terrible when I enable the app, usually I get 20 hours worth of battery life, but with the app I get around 5.” Another participant said, “I use my phone a lot and can’t miss work calls or emails because of a drained battery.” Another was convinced the app was draining the battery and suggested that the government should, “Make the app drain less battery. This app sucks battery in no time.” Participants also expressed concerns that they thought they could not use the app because their device’s battery was not up to the demands required by the app or that it constantly lacked a sufficient charge. Comments from participants included, “My battery is often low” and “My phone is quite old and doesn’t hold a charge.” Another participant thought that using the app would be “…wasting my phone battery ‘cause the charge doesn’t hold” and another noted that one reason they were not using the app was, “My phone is quite old and does not hold a charge.” Another participant did not want to use the app because, they “…don’t have an opportunity throughout the day to charge my phone.” While the ABTraceTogether website does not address the issue of battery drain, the Singapore website for their almost identical contact tracing app offers an FAQ to concerned users explaining that Bluetooth technology uses almost no battery and provides screen captures showing device battery usage with Bluetooth enabled (See support.tracetogether.gov.sg). Addressing battery misconceptions may have encouraged some participants to use the app.

5.3 Download

Although the ABTraceTogether app has a website with installation instructions and is available for downloading on Google Play and the Apple App Store, some participants indicated that they did not know how to download the app. Other participants expressed a general lack of knowledge of information about the app, such as whether or not it was freely available and questioned, “Is there a cost to use?” Participants also expressed difficulties finding the official app website or reported that the website did not work. One participant said that, “I tried to get on it but it says [the] site does not exist” while another said they could not download the app because “the site doesn’t work right.” One participant could not distinguish between the official app and other related apps, noting, “There are too many fake apps claiming to do the same.” Other participants began the process of downloading the app but noticed poor user reviews and stopped. One stated that, “I read reviews [and] almost everybody said battery drain was a huge issue so I’m not even going to attempted [sic] until some of the bugs are fixed.” Another participant who decided not to use the app wrote that, “When I went to download it, I read the reviews and they were not good, so I didn’t use it after all.” Some participants expressed that it would not work on their devices (presumably due to incompatibility issues) or other installation problems. One participant noted that, “I have already tried to get it but it won’t load on my phone” while another said they could not use the app because they, “Don’t know how to activate it.” The download and installation concerns of participants highlight that messaging surrounding downloads and installation needs to be very clear and easy to find, as some potential app users will decide not to use the app if they encounter any issues at all. As well, these concerns show again how important it is for the app sponsor to clearly address concerns and debunk false information spread in reviews and online.

5.4 Memory

After concerns about battery drain, the leading participant concern was the demands of using the app on the participant’s smartphone memory. Comments about a lack of room on their device for the app in the first place include: “I’m always running out of storage so I’m not sure I’ll have enough to download an app,” “Phone has no room to add apps,” and, “It does take up memory on my phone.” Another memory concern with participants is that by the ABTraceTogether using a decentralized approach, it logs all contact data (in encrypted form) on the user’s device. Participants were concerned they had insufficient device memory for this. One stated that, “I only have a limited amount of memory on my phone and don’t want to use it unnecessarily if most people aren’t using the app.” These concerns about memory usage may be accounts or reflect an unwillingness on the part of some users to make accommodations to use the app, such as by uninstalling other apps or files to make room. However, this is also an area that should be clarified by the app developers.

5.5 Network

Participants expressed a variety of concerns related to the app’s use of Bluetooth and their perception that the app required constant Internet access to function.

Many participants had the misconception that the ABTraceTogether app required a smartphone owner to have a data plan and regular internet connection for the app to work. The following quotes are samples of this false perception, “Poor connection is always a problem,” and, “My cell often does not work in the rural area that I live.” One participant noted that, “I visit places where sometimes there is little service so that may be a problem for connection,” while another stated that, “[I] live in an area with low connectivity by which I’m not able to connect properly.” Another participant thought they would not be able to use the plan because, “I am not connected to the internet on my phone, unless I turn a number of things on.”

Some participants noted that although network connectivity was not an issue for them, they perceived that the app would constantly be connecting to their network and thus incurring costs for data use or overages. Participants noted, “I have a small data package so would not have it on”, “Data cost: I don’t want to end up spending money on more data because this app is running in the foreground of my phone while I’m away from home,” and, “I only have 1 Gb data on a monthly basis. I can barely afford my cell phone plan as I’m retired.” Again, better messaging from the app sponsor and debunking of misinformation could help convince people that data plans and constant internet connectivity was not required to use the app.

ABTraceTogether, along with many contact tracing apps, relies on a user’s Bluetooth network constantly being turned on. This raised concerns with a number of participants. The main concern was related to the possibility of a participant’s device being hacked due to this perceived vulnerability. One participant wrote that, “The app runs on Bluetooth and Bluetooth is an easy hack, so this would open your phone to close quarter hacks like [on the] bus, [at the] store, malls, etc.” Another thought that the app would make “it easier for hackers and my privacy on my phone being compromised.” One participant stated they “don’t like the idea of leaving my Bluetooth on. It could allow access to my data &/or information” while another was concerned about “…government or hackers being able to hack my phone easier because Bluetooth is always open.” One participant raised the concern that it, “Sucks that it forces me to use Bluetooth when I want to use that for other things.” Two participants expressed concern about the issue of “radiation exposure” from constantly having Bluetooth on. One participant noted that they were concerned that “…the radiation coming off our phones [is] only getting worse now with 5 g network.”

Issues such as radiation exposure and data costs should be addressed by app sponsors in the educational and promotional material. Although this section only briefly addresses privacy and security concerns, this was a leading concern of participants both for users and non-users of the app. Although the Alberta government did make efforts to assure the public that privacy and security were addressed, more efforts in this area were needed.

5.6 Operating system

ABTraceTogether only works on newer versions of Android and Apple smartphones. This left people with older devices or other operating systems unable to use the app. One participant commented that, “I have already tried to get it but it won’t load on my phone. Being an older phone.” Some participants noted they could not use the app as they had an older smartphone that did not have Bluetooth. One participant noted that the app developers should “understand it is not compatible with iPhone 4s” while another noted they “have an older android smartphone that is incompatible with [the] app.” One frustrated participant wrote, “Don’t assume everyone has a smart phone that is not an Apple. If you are going to make an app that you WANT everyone to use for the good of all - make sure that it works for all the phones out there. This issue with Apple phones when the app first came out could have been avoided if you developed the app with every phone in mind.” A few participants felt that the app would work only on iPhones such as one who wrote that they did not use the app because, “I don’t have an iPhone.” Whether this belief was based on false information or whether this was an excuse is not known. These operating system issues highlight how crucial it is for app developers to ensure that the contact tracing app works before it is released, and that it works on as wide a range of smartphones as possible, even older versions.

5.7 Performance

Performance issues cover genuine and perceived errors and bugs in the app and its impact on the device. ABTraceTogether had difficulties with an early requirement for iPhone users where it was required to always be on in the foreground. Participants also expressed a range of specific issues as well as vague reports or unrelated problems with the app which could be genuine or could be accounts. Some users expressed vague concerns, such as a fear that the app was “damaging my phone” or “slowing down [my] phone.” A few comments stated that the app caused their device to “crash sometimes” or that, “Every once [in] a while my phone randomly restarts now.” Another participant wrote that “My app freezes and lags a lot. I don’t know the cause but it could be on my end.” Some participants expressed concerns that the app had a “slow startup” while another noted an issue with “heating up my phone on some occasions.”

5.8 Usability

Usability refers to both the ability of a user to use the application as well as the degree of user friendliness of the app and any other user experience issues. Most concerns from participants in this domain pertained to the iPhone foreground issue, followed by concerns that expressed doubt about whether the application was working or not. Other issues expressed were ease of use and a lack of app features.

Many users of the ABTraceTogether app expressed doubts about whether the app was working correctly or even whether it was working at all. One participant wrote, “Luckily, I have no idea if it is even working when I go out. It is active on my phone but I don’t know if it is working or not. When I say ‘luckily’, it is because I’ve hopefully never encountered someone who has later contacted COVID *and* informed AHS [Alberta Health Services].’ Other users also reported that the app “never does anything so [I’m] unsure if [it’s] working” and another stated “I don’t know if it works as I’ve never been notified.” Other users were concerned that they were doing something wrong. One said that they were “not sure I’m using the app right.” An Apple iPhone user was worried that the app did not work correctly after the foreground issues with the app. They wrote, “I heard there were issues with the app and Apple users but never heard if this was fixed or if we should delete, then reload the app.” Other participants suggested modifications to the app so that users are informed that it is working correctly. One wrote that the app “needs to give some type of daily update: covid numbers in area. To show it has purpose daily.” Another suggested that “perhaps having a daily notice of how many phones it connects with each day would give an indication of it working.” As the need for assurance that the app is working properly was expressed by many participants, this is a main participant issue that should be addressed when designing applications.

Many users expressed concerns that the app was not easy to use with many simply commenting that it was not “user friendly” or “too complicated”. One participant stated that, “The app also is somewhat confusing to use, I didn’t fully understand how to use the app nor really knew what it was for a while.” A participant said that they “would like it to be more user friendly,” while another said that they are “not sure how to use it.” Some participants from the non-user group reported that they did not like using any mobile apps as they were generally not comfortable with technology. While it is possible that some reports about the app being difficult to use are accounts raised by individuals who are trying to justify their unwillingness to use the app, some of the concerns were also likely raised by participants who lack technological knowledge and confidence. Issues with ease-of-use highlight that it is crucial for app sponsors to ensure that there is clear and easy to understand messaging about how to download and use the app, as well as extensive support available for those who are less comfortable with technology.

Some participants wanted the app to have more features. One participant complained that the app “does not give real time updates.” Another participant thought the developers should “put a live chat in case you need help” while another suggested that the app should offer a “notification if entering a high Covid area.” Other users thought the app could be improved by having “notifiers when leaving a Wi-Fi zone” as this would be helpful when the app needs to update. Another user suggested that the developers should, “Make it possible to turn the location tracker on and off so people can turn it on when in crowded locations but don’t need it to track them on walks through the woods”. Other suggestions participants offering for features including offering general information about the pandemic, personal healthcare tips, a map of infection hot spots, offering a vaccination clinic finder, and adding gamification elements. These suggestions indicate that making an app more useful might encourage more people to download and use it regularly.

Many participants were concerned that the ABTraceTogether app initially ran in the foreground on iPhones. A number expressed that they wanted the developers to, “Allow [the] app to run in the background.” One participant who decided not to use the app wrote that, “I think it’s great in theory and I fully support using it for contact tracing. If it was better to use with my iPhone I would absolutely use it. Right now I don’t because I’m not going to keep it open all the time.” Another user stated, “I don’t want to have to think about using it, I need it to just work seamlessly in the background.” These user experience concerns again highlight the importance for app developers to get usability right the first time, as potential users may not give an app another chance after encountering problems and deciding not to use it.

6. Conclusion

With ABTraceTogether, we see a fascinating situation for m-Health applications. In response to the increasing casualties of the COVID-19 pandemic, the government of Alberta quickly launched their contact tracing application in an effort to combat the spread of the disease. Our study participants identify various genuine issues with the app as well as concerns based on faulty beliefs or misinformation that combined prevented them from downloading or continually using the application. These were early adopters of an application that was in early stages of deployment, at that time there was an emotional component in responses about this disease as we were just a few months into the pandemic, and participants self-selected for the study.

We can speculate that had the application sponsors spent additional time in development and user testing as well as addressing consumer behavior adoption challenges, it may have improved the adoption and usage rates. The ABTraceTogether developers did attempt to address some user concerns in a subsequent app update. However, this update may have come too late and did not substantially increase the user uptake.

In addition to m-Health’s role in health crises, such as COVID-19, this technology has the potential to aid consumers in maintaining their personal health and wellbeing. Yet, for future m-Health or related technology to be successful, understanding consumer behavior relative to people’s adoption and usage issues will help attain sufficient uptake, continued usage, and the resulting consumer benefits.

Acknowledgments

This research and publication was funded by grants from Athabasca University.