Understanding Engagement With Forensic Smartphone Apps: The Service Design Engagement Model

Abstract

Justice services have begun to integrate the use of mobile applications into treatment, support, and rehabilitative programs for forensic clients. One such application that been adopted to support forensic clients is “eRecovery”: a smartphone application that provides clients recovering from a substance addiction with support for managing relapse. In this article, we report on evaluation findings from a trial of eRecovery in an Australian Community Justice Centre, and reflect on several issues relating to fostering and sustaining client engagement with similar applications within forensic and justice settings. We propose the Service Design Engagement Model to organize, visualize, and describe the stages and factors important to adoption, appropriation, and on-going routine use of the software by forensic clients. The model recognizes the role of contextual and environmental factors in supporting users through the early stages of engagement, and the importance of user agency in longer-term engagement with therapeutic apps.

Keywords

eHealth digital interventions mobile applications user engagement

Introduction

The health and mental health sectors have made substantial progress in translating treatment and support interventions into digital and mobile digital channels to support practitioners and consumers (Istepanian et al., 2004; Luxtonet al., 2011). Justice services have been slower to see the potential for digital technologies to support rehabilitation, and where these developments are evident it is often at the intersection between justice and health or mental health services (see Farabee et al., 2016; Kip et al., 2020). Examples of digitally-based rehabilitative interventions that specifically target justice-involved persons include drug relapse prevention for drug court clients (Johnson et al., 2016), drug treatment and recovery for prisoners (Chaple et al., 2016; Elison et al., 2016), intimate partner violence programs in prisons and probation settings (Morris & Bans, 2018), and forensic mental health programs (Kip et al., 2018).

There are several potential advantages that accrue from providing justice clients with rehabilitative interventions via digital channels, including improved accessibility and lower delivery costs, the capacity to tailor interventions to meet the individual needs of users, the remote collection of app-generated and self-report data, and the ability to combine functions into customizable applications that are available to service users when and where they need them. However, there are also some important challenges that must be negotiated before digital platforms for delivering rehabilitative services can be regarded as a viable and effective adjunct to conventional face-to-face approaches. An important limiting factor that has been recognized in the health and mental health sectors is the extent to which digitally-delivered interventions are able to engage users in services or activities, and to support user adherence and compliance over a sufficiently long period to yield meaningful therapeutic outcomes (Baumel et al., 2019).

This article analyzes data on a trial of a smartphone-based drug relapse support app (eRecovery) at a community court in Melbourne, Australia, examining patterns of user engagement with the app as well as steps taken to develop and sustain user engagement. Our article focuses on how insights from Human Computer Interaction (HCI) studies can help justice and health providers better understand and respond to the challenges associated with maintaining client engagement with digital behavior change applications (Murphy et al., 2020). Specifically, we examine the techno-social conditions that must be met for apps such as eRecovery to generate beneficial outcomes for users. HCI studies provides tools for understanding how the design of eRecovery and similar apps informs whether they are adopted by justice clients and ultimately come to use the app routinely and to positive ends. As HCI emphasizes, this is not a technologically-determined process. In understanding the uptake and use of such apps by justice clients, we need to address how such apps fulfil the varying needs of clients.

Engagement: A Pervasive Problem for Digital Interventions

While services and interventions that use a digital platform can be easy and cheap to distribute to potential users, achieving and sustaining engagement with them is much more difficult. Speaking specifically about eHealth applications, Alkhaldi et al. (2017) note that “engagement with digital health interventions (DHIs) may be regarded as a prerequisite for the intervention to achieve positive health or behavior change outcomes.” While the challenges to fostering engagement with digital health interventions have been examined in some depth, the specific challenges associated with fostering engagement with digital justice-oriented mental health and behavior change applications have not received the same level of research attention. eHealth-style applications are often introduced into justice settings in an attempt to increase engagement with and utilization of treatment programs (see Clarke et al., 2015). Yet as Kip et al. (2020, p. 32) note, increased engagement is only a “potential benefit[. . .]” and research evaluating the effectiveness of such applications remains somewhat scant. Relatively few studies have examined: (a) whether justice-involved individuals find these applications engaging (see Chaple et al., 2014; Sygel et al., 2014); (b) the extent to which justice-involved individuals want to use them (Rantanen et al., 2021); or (c) how such applications can be made more engaging for justice-involved clientele (Morris & Knight, 2018; Teng et al., 2019)

Mobile apps have generally low retention rates—a 2018 industry report estimated that for health and fitness apps, retention rates (i.e., where users return to the app at least once in a 30-day period) were 8.5% after 1 week and 4% after 1 month (Statista, 2021). Studies of health and fitness apps show that around half of all downloaded apps are never used, and 90-day retention rates rarely exceed 30% (Birnbaum et al., 2015). Take-up and engagement rates for mental health apps are similarly modest (Baumel et al., 2019; Fleming et al., 2018; Torous et al., 2020). While there is little data on engagement and retention rates for digital interventions that specifically target justice populations (see Johnson et al., 2016), program drop-out rates for conventional forms of forensic interventions are known to be a significant barrier to effective rehabilitation and treatment (Brunner et al., 2019; Lappan et al., 2020), and it seems likely that user attrition for mobile apps is also a significant problem.

The problem of user engagement with digital apps has given rise to a wide variety of theories and models. In their systematic review of engagement in HCI, Doherty and Doherty (2018) identified 372 theoretical frameworks that have been proposed, reflecting an equally wide variety of definitions and HCI contexts (online, personal computer, mobile device, gaming systems, etc.) and domains (physical, cognitive, and affective) that are covered by this term. To bring order to this diverse field, Doherty and Doherty distinguish between actor-focused micro theories that are primarily concerned with engagement as a state of experience, and structure-focused macro theories that are concerned with the higher-level socio-structural and temporal aspects of engagement (Doherty & Doherty, 2018; O’Brien & Toms, 2008).

The eRecovery App

eRecovery is a version of the Alcohol-Comprehensive Health Enhancement Support System (A-CHESS), a smartphone-based relapse-prevention system developed by the Center for Health Enhancement Systems Studies at the University of Wisconsin (Gustafson et al., 2011) and commercialized by CHESS Health. eRecovery was designed to assist people with substance use disorders in the recovery phase following treatment, and specifically to manage the problem of relapse. The app has its basis in two forms of theory: theory about the relapse process in the form of Marlatt’s cognitive-social learning model (Larimer et al., 1999), and theory about motivation to act or change in the form of Self-Determination Theory (Deci & Ryan, 2008). The app aims to support people in managing the problem of relapse by: creating competence in disease management (through the provision of information, notifications of high-risk situations, and strategies to manage distress); building relatedness with others (through social media, sharing of recovery stories, and check-ins), and creating a sense of autonomy in the recovery process (by tracking progress, and responding appropriately to lapses or impending lapses).

eRecovery is a multi-function program available to users through a smartphone or tablet. It includes a client-facing app (Connections) and web-based clinician dashboard (Companion). Users can customize the app with their treatment plans, relapse and lapse triggers, intervention strategies, motivational drivers, and services and people to contact when a crisis arises (see Figure 1). Features include: discussion groups; appointment and medication reminders; GPS-enabled warnings of high-risk locations; progress tracking though weekly surveys; “Beacon button” access to a 24/7 helpline and other resources when urgent support is needed; treatment planning and goal setting; and audio, video and reading content. Through the Companion app and dashboard, clinicians can organize appointments, message clients, and send medication reminders. The app provides clinicians with dynamic updates on risk and protective factors, providing greater insight into clients’ progress.

Figure 1.

eRecovery’s features and functions.

A-CHESS is registered with the US Substance Abuse and Mental Health Services Administration’s National Registry of Evidence-based Programs and Practices as a scientifically established behavioral health intervention for AOD addiction. The app has been shown to be effective in reducing heavy drinking and enhancing long-term abstinence following treatment for alcohol abuse (Gustafson et al., 2013; McTavish et al., 2012), as well as improving post-treatment service utilization, and outpatient treatment use and retention (Glass et al., 2017; Johnston et al., 2019). While A-CHESS has been trialed at a US drug court (Johnson et al., 2016), there is only limited evidence about its use by justice-involved populations. A Spanish language version of A-CHESS has been shown to be effective in reducing drug relapse (Muroff et al., 2019) but to date there are no published outcome studies of the eRecovery version of the app.

The eRecovery Trial at the Neighborhood Justice Centre

The aim of the eRecovery trial at the Neighborhood Justice Centre (a community court in Victoria, Australia) was to assess the app’s potential for use in problem-solving and therapeutic justice programs in the Magistrates’ Court and Children’s Court jurisdictions. Persons enrolled in the trial were charged with an offence and were either receiving clinical support from the client services team at the NJC while awaiting a hearing, or had been sentenced to a community corrections order that was supervised at the NJC. Participants’ use of eRecovery in the trial was voluntary and was not a bail condition or court-mandated. The trial commenced in February 2019 and 28 participants were recruited to the end of 2019. Court activity slowed greatly in the year after March 2020 due to a series of community lockdowns in response to the COVID-19 epidemic, and only seven new trial participants were recruited in the next 12 months. This article reports on data for the 36 trial participants recruited to March 2021, of whom 20 currently had the eRecovery software enabled.

Trial recruitment proceeded by one of two pathways. Clients referred to the clinicians working at the NJC were offered the option of participating in the eRecovery trial as part of their clinical treatment. In the second pathway, Community Corrections workers offered eRecovery to persons with an order supervised at the NJC as a supplement to standard case management and supervision. Trial participants could be provided with a smartphone to support the app and could also be provided with monthly phone credit. As of April 2021, there were five provided smartphone among the 20 currently enabled participants.

As participation in the trial was voluntary, we were particularly concerned with understanding what would motivate people to agree to take part in the trial, as well as what perceived and experienced benefits and disadvantages of eRecovery would influence their decision to continue as users. Unlike all previous studies of A-CHESS except the drug court trial reported by Johnson et al. (2016), this trial involved justice-involved users. There were several corollaries to this: trial participants were engaged in drug or alcohol treatment as part of a bail or community corrections order. In many cases their drug use was intimately related to their offending, and they typically experienced long-term and severe social and economic disadvantage. A core attribute of digital health and mental health apps is that they are “user driven”—users have a great deal of control over how, when, and where they use an app. This contrasts with conventional correctional treatment and program models where clients are substantially coerced participants with only limited control over their participation. Thus, the trial represented an opportunity to examine how engagement with a digital app works with a population characterized by a range of motivational and environmental challenges.

The Service Design Engagement Model

Any theory of engagement with eRecovery needs to address the central attributes of the app’s design and intended uses. These include: its multi-functional nature; the behavior change model embedded in the app; the justice context under which the app is used, and its relationship to the face-to-face supervision that users receive. A key attribute of the eRecovery app is that it is intended to be used over a period of weeks or months to achieve its intended goals. Some of these attributes are at odds with assumptions in micro engagement theories, including that user engagement results from intrinsic motivations for using a system rather than from directed or non-voluntary use, and that user engagement entails an individual losing awareness of the outside world in their focus on a specific task rather than carrying out one task episodically and in conjunction with a range of other tasks (O’Brien & Toms, 2008). In recognition of this, we have adopted Borghouts et al.’s (2021) definition of user engagement (derived from a systematic review of digital mental health interventions) as “a user’s uptake and sustained interactions with a digital intervention” (p. 2). A corollary of this is that any evaluation of user engagement needs to include an understanding of system use over time by its target users (Klasnja et al., 2011).

An influential approach that examines user engagement as a process comprised of multiple stages is O’Brien and Toms’s (2008) process model of user engagement. This model identifies four stages of engagement: (1) a point of engagement; (2) a period of sustained engagement; (3) disengagement; and (4) reengagement. At the point of engagement, an individual becomes engaged in a technology-facilitated task. In the period of sustained engagement, an individual’s attention and interest remain invested in the interaction with the technology. Disengagement occurs when internal and/or external factors cause the user of a technology to cease engaging with it. Finally, reengagement may occur when a “user” returns to a technology after a period of disengagement. Importantly, this process model of engagement does not preclude us from considering attributes that engagement with a technology generates in users—an important qualifier embedded in O’Brien and Toms’s (2008, p. 941) own definition of user engagement as “a category of user experience characterized by attributes of challenge, positive affect, endurability, aesthetic and sensory appeal, attention, feedback, variety/novelty, interactivity, and perceived user control.”

Our model, which we term the Service Design Engagement Model (SDEM), moves beyond examining how such technologies can be used by forensic clients, to examining (a) why users adopt the technology, (b) how they use the technology when they have appropriated it, and (c) why they become routine users of the technology. Our use of the term “service design” acknowledges the contributions of Don Norman and Nicola Morelli in developing the principles of user-centered design (Morelli, 2002; Norman, 1988). The Service Design Engagement Model is structured around three stages that reflect the different types of interaction between the user and the technology as they are introduced to, explore and become regular users of eRecovery. The stages are: adoption, where the technology is designed but as yet unused; appropriation, where the technology is in the hands of the client and being learned and tested; and finally routine use, where the technology is regularly used (see Figure 2). At each of these stages it is proposed that the user evaluates the technology with independent mechanisms and using different criteria. If the user is satisfied in the current stage, engagement, and use proceeds to the next stage. Otherwise, the technology is rejected and use ceases.

Figure 2.

The service design engagement model.

In this article, we describe how the SDEM describes different short-, medium-, and long-term patterns of user engagement with eRecovery, and the motivators and barriers that were associated with each stage of engagement.

Methodology

To examine this issue, a convergent parallel mixed methods approach was utilized (Creswell & Plano Clark, 2011). Specifically, our methodology paired qualitative interviews with eRecovery trial participants with a quantitative analysis of transaction data produced by trial participants using the app (see Han, 2011). Interviews were focused on participants’ experiences of, views, and perceptions of the app. Interviews were conducted in person and by video-link and were recorded. In addition, potential users who declined to participate in the trial were asked for their reasons for not taking up the offer. Qualitative interview data were analyzed in line with a hybrid inductive-deductive thematic analysis approach (Fereday & Muir-Cochrane, 2006). The deductive component of our analysis drew upon the concepts of imagined affordances (Nagy & Neff, 2015) and perceived affordances (Norman, 1988) to examine the features of the eRecovery app trial participants imagined they would find useful (imagined affordances), and the forms of action they perceived the app to offer after engaging with it (perceived affordances).

Transaction data was used to examine patterns of app use, including the duration and frequency with which they used the app, daily check-in records, and “red pin” events (episodes of lapse or relapse). In addition, data from the Kessler Psychological Distress Scale (K10; Kessler et al., 2002) and the Patient Health Questionnaire (PHQ9; Kroenke et al., 2001) surveys conducted at enrolment and at intervals thereafter were also extracted for analysis. The project was approved by the Victorian Department of Justice and Community Safety’s Justice Human Research Ethics Committee (JHREC).

Participants

The trial participants comprised 21 men and 15 women, with a mean age of 36 years (minimum age 24, maximum 56 years). About 19 participants were recruited by the NJC Clients Services team, and 17 were recruited by Community Corrections staff members. In addition, refusal data was collected from a further 11 persons. The most commonly reported primary substance used by trial participants was methamphetamine (around half the sample), followed by alcohol (around one-quarter of the sample), and heroin (one-fifth of the sample), although poly-drug use was common (reported by one in three participants). Over half of the sample was identified as having concurrent mental health problems (most commonly anxiety and depression) and three participants were identified as having an acquired brain injury. Nearly half (48.5%) of participants scored in the “moderately severe” or “severe” range for depression on the PHQ9, and 30% scored in the “severe mental disorder” range for the K10. The sample showed high rates of social and economic disadvantage: two-third reported that they were unemployed or on disability benefits, and one-third reported various degrees of housing insecurity (homelessness, transitional housing, or boarding house accommodation).

Findings

User activity patterns for eRecovery

The three-stage SDEM engagement model reflects a primary feature of the way that trial participants used eRecovery, with some rejecting the app almost immediately, some using it for periods of weeks or even a few months, and some becoming long-term users. One in five trial participants (19%) used the app for less than a week, and in three cases for only 1 day. The same proportion (19%) used the app for at least a week but ceased use within the first 30 days, and a further 1 in 5 (22%) used the app for more than a month up to 3 months. Nearly 4 in every 10 trial participants established longer-term patterns of use, with the two longest-term participants continuing to use the app regularly nearly 2 years after their initial enrolment in the trial. Participant use rates were highest in the earliest stages of engagement when participants were exploring the app functions. For those participants who remained engaged after the first week, on average the app was accessed 2.6 times per day (SD = 2.1). For longer term users (more than 90 days in the trial), average use rates declined to1.9 app activities per day (SD = 1.3).

None of the variables we examined predicted whether a participant would progress to routine use of eRecovery. Routine use was not related to gender (χ² = 0.47, p = .36), age (engaged clients mean = 33 years, not engaged = 36 years, F = 1.1, p = .29), or income category (χ² = 1.78, p = .41), and was unrelated to the participant’s initial level of anxiety/depression (engaged clients mean PHQ9 = 12.4, not engaged = 11.7, F = 0.7, p = .8).

Adoption

Technology as designed

The adoption stage of the SDEM model addresses the “barriers” and “motivators” that a user or prospective user must navigate when they first encounter eRecovery. Adoption barriers are most clearly evident when a participant declines to join the trial or rejects the app within the first few days or weeks of commencing.

“Access” barriers are evidenced in physical access to a smartphone or plan credit, access to identity documents required to register a SIM card, and sufficient literacy and knowledge of technology to participate. It should be noted that access barriers were not consistent across participants, emerged at different stages of the trial and in some cases were alleviated. For some participants, homelessness and feeling unable to keep a smartphone secure were identified as access barriers whilst others engaged in sustained participation despite homelessness. Some participants lost access to the mobile device part way through the trial due to imprisonment whilst others gained access through NJC loaning participants a phone for the period of the trial along with vouchers to compensate for phone data usage.

“Value” barriers arise where clients decline to participate in the trial because they do not believe it would make any difference to their behavior, reportedly did not have time to use it and or were unable to see any value in engaging in activities outside their mandatory correctional order requirements. “Trust” was a barrier, particularly associated with concerns that using the app would mean that other people would have access to personal information, or that the optional GPS-enabled functionality was a covert means of tracking or surveillance:

I felt like I’m signing up for this, but this phone is going to track me everywhere I go, you know, and everything like that. Like everything that’s on this phone is going to be reported to the authorities or whatever. Yeah. That’s just me. I just felt like it was a tracking device [laughs]. But then I got used to it.

Like the other adoption barriers and motivators discussed here, C’s concerns speak to Nagy and Neff’s (2015) concept of “imagined affordances”: actions that designers or users expect a technology to enable. Imagined affordances, Nagy and Neff (2015, p. 5) explain, “emerge between users” perceptions, attitudes, and expectations; between the materiality and functionality of technologies; and between the intentions and perceptions of designers’.

Adoption motivators include features of the app that prospective users imagine or recognize will have immediate benefits for them. Among eRecovery trial participants, these included the ability to enter personal reminders for medication, to schedule appointments with clinicians or workers in a personal calendar, and to communicate directly with the user’s clinician or worker through the direct messaging function:

GP:

Yeah, it keeps me one step ahead of appointments with the way I set the reminder to be a day early so I know “oh yeah, I’ve got that appointment tomorrow”, and I can double-check if I’ve double booked myself

“Community contact” was reported as an interesting feature, although clients questioned whether ongoing interaction with peers would be a positive or negative influence. Finally, the “personalized content” capability of the app was a source of motivation. This included items that could be entered as personal motivators and reminders for recovery (e.g. pictures of family, pets, or a video recording willing themselves to succeed).

Appropriation

Technology in hand

Once a client had overcome or reconciled adoption barriers and identified enough motivators to participate in the eRecovery trial, imagined affordances were superseded by perceived affordances: forms of action that individuals perceive a technology to enable (Norman, 1988). Factors influencing clients during the first weeks and months of using the app can be grouped into “supports” and “hindrances.” Hindrances interfere to a greater or lesser extent with a user achieving a particular goal and detract from the app’s perceived affordances. Conversely, supports assist a user to achieve his or her goals and enhance the app’s perceived affordances. The user’s goals might be instrumental (making an appointment, sending a message or simply reading or listening to some content) or more complex (dealing with an episode of potential relapse).

“Access” hindrances identified by clients who signed up to use the app included problems in maintaining access to the app (e.g., difficulty charging the phone), finding app content to be unsuitable or inaccessible (with, e.g., low literacy inhibiting their understanding of features), and not receiving feedback for comments or queries submitted using the group messaging function. Conversely, supports involved app functions that were found to be useful or rewarding. For example, clinician appointments and personal medication reminders were reported as very useful “organizational tools” by clients, a finding also validated by usage data. Interestingly, the usefulness of this functionality for one client extended to providing a framework to structure their day, starting with reminders and prompts received in the morning.

Supports can also arise from improved “Communication with clinician/worker.” This is mainly attributed to the results from the self-report surveys delivered through the app, results of which are shared with clinicians or workers. These enable clinicians to monitor how the client has been progressing since the last appointment and focus face-to-face time in appointments more effectively on therapy rather than assessment. “Personalized content” was reported to be beneficial as exampled by motivational entries accessed in difficult moments. An additional supporting factor was identified as “Mobility,” enabling clients to access features of the application at any time and location.

Routine use

Technology in use

The final stage of the model considers the technology in routine use, where engagement is driven by deeper “reinforcers” that are products of the user’s progress through the recovery process. In understanding the routine and ongoing use of apps such as eRecovery, we need to address the imbrication of human and technological “agencies” (Leonardi, 2011), where specific human intensions intersect with technologies to produce particular outcomes (Wood, 2021). While it is well documented that digital technologies face high attrition rates (Baumel et al., 2019; Torous et al., 2020) it is also the case that a cohort of users will continue to routinely use the technology over a longer period—in the case of eRecovery, nearly 4 in 10 trial participants progressed to long-term use (more than 3 months). Ultimately, most eRecovery users cease their engagement, although this is complicated by the context of use as an adjunct to clinical support or order supervision. In this case, a conscious decision is made to cease using Connections as part of the recovery process. To understand why this is the case within justice and forensic settings, we need to understand the factors underpinning the sustained use of a technology, as well as the factors contributing to both productive, and unproductive, user attrition.

In examining routine use among eRecovery users, we identified three key factors that contributed to users’ productive engagement with the app, and that were all associated with long-term use. These were, (1) that eRecovery helped them create positive routines, (2) that they found eRecovery’s potential for self-quantification empowering, and (3) that they found that eRecovery helped them feel connected to other users, and society in general.

Positive Routines

For one participant, eRecovery’s medication reminder nudged them toward productive routines that helped them avoid relapsing. As M details:

Well mainly I use it for the medications, my medication reminder, and that prompts me to do my daily survey as well. Those two things in the morning kind of just set me off and since I’m such a routine person, it just triggers my day and makes it easier for me to follow through from there so. . .

Nudges, as Thaler and Sunstein (2009, p.6) explain, are “any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any option or significantly changing their economic incentive.” Thaler and Sunstein’s (2009) nudge model is informed by a rich literature in the cognitive psychology of decision making that has refuted rational choice models of decision making, offering models that account for heuristics, cognitive shortcuts, and the often unreflective nature of individual decision-making.

Technology-facilitated nudges can, as Caraban et al. (2019) explain, take a number of forms, several of which are employed by eRecovery. Firstly, eRecovery makes use of a form of nudge that falls under Caraban et al.’s (2019, p. 5) category of confront nudges. Nudges that confront “attempt to pause an unwanted action by instilling doubt” and in doing so tapping into regret aversion bias (Caraban et al. 2019, p. 5). One key confront bias is what Caraban et al. (2019, p. 5) term reminding of the consequences: nudges that utilize the availability heuristic to prompt an individual to reflect on the potential consequences of an action. eRecovery employs just such a reminding of the consequences (confront) nudge through notifying users when they are approaching areas that the user has designated as high risk for them.

Further, eRecovery makes extensive use of what Caraban et al. (2019, p. 8) term reinforce nudges: nudges that “attempt to reinforce behaviors through increasing their presence in individuals’ thinking.” As M’s account demonstrates, such reinforce nudges occurred when the eRecovery medication reminder “notification” appeared on their phone outside of the app itself:

Interviewer:

With the um, . . . medication reminder . . . do you get that when you open the app, or does it just come up, a notification, come up on your phone?

It kinda comes up as a notification, just a little dot that comes up there and so

Interviewer:

Ah ok, and then that prompts you to open the app

But in saying that, I’m one of those people where if I see a dot I need to get rid of it [laughs]

This form of reinforce nudge represents what Caraban et al. (2019) term a just-in-time prompt: time-appropriate reminders to engage in a particular behavior. Such just-in-time prompts are employed in eRecovery to remind users to take their medications, attend therapeutic and justice-related meetings, and undertake the daily survey.

Empowering Through Self-Tracking

eRecovery affords users regular feedback on their progress. Participants are asked to do both weekly and monthly surveys that track their recovery. For at least one participant in the eRecovery trial, the self-tracking enabled by this regular feedback was a key facet of their productive engagement with the app:

. . . I know I was quite happy for you guys to monitor it to see if I was going down a slippery slope, or it gives you guys a chance to monitor me as well as me monitor myself. Which I really thought was fantastic

Interviewer:

So a tool for you to monitor yourself?

Exactly, and just to calibrate where I’m at mentally, like if I’m going up and down, cause when I combine that with seeing [mental health nurse] once a week, I get her to see how I am compared to last week. How’s my mood compared to last week? With it able to give me that information, like what I’m seeing on the graph, that information is really, really helpful.

Monitoring responses to treatment is known to significantly improve mental health and substance user treatment outcomes (Goodman et al., 2013). Further, Attig and Franke (2020), found that tracking motivation was closely linked to continued engagement with personal fitness and health apps. Yet as our findings indicate, prompts to self-track can act as a negative or a positive reinforcer for engagement. Participants who dropped out of the trial within the first month rarely completed any of the weekly surveys. One user who disengaged in the first month reported that he found the surveys “overwhelming,” and the tracking surveys may also be a contributing factor for users who disengaged because of concerns about privacy. In contrast, users who stayed engaged for more than a month continued to complete the self-tracking (PHQ-9) surveys on average twice every month, and there was no reduction in the frequency of survey completion with longer engagement periods.

Feeling Connected

Another affordance central to eRecovery is the connections it provides to clinicians, peers, and other members of a nominated support group via message boards and discussion groups. There were many references in the participant interviews to the importance attached to feelings of connectedness. Importantly, this sense of connection didn’t only arise from the explicitly “social” affordances in eRecovery but were also identified as arising from the prompts provided by the daily check-in, and appointment and medication reminders:

Interviewer:

Yeah, and um, how important is this phone, now that you’ve got it in your day to day life?

Life-changing mate. I can’t tell you, it’s been like, it makes me feel like I’m a part of a community again. And just like I said, getting a message in the morning saying, how are you feeling today? Did you get out of bed today? Um, those little things, things like that, you don’t realize until you’ve got them, coming every day, how inspirational and effective they really are. It’s like, aw yeah, I need to, I should get out of bed today. Aw look I haven’t got out of bed today, let’s get out of bed today.

You know what, I actually like it because it . . . because, you know, the daily check-In thing that always pops up. And it reminds me that, like, I’m part of the normal . . . the normal world like, you know, society.

Interviewer:

Yeah.

And so even if I if I feel, like, really isolated and lost like I’ll go into that and that’s like connection to Corrections and like . . . How do I explain it? . . . like. I just feel like I’m connected to . . . like I’m doing the right thing by using it, you know? Yeah. I feel like it connects me to the real world.

It is well established that social media can generate a sense of social connectedness independent of direct social interactions arising from family, employment, or recreational activity. Yet, in offering this, therapeutic apps may raise privacy and confidentiality-related concerns, especially for people with mental health conditions (Hilty & Mucic, 2016)

Conversely, there are factors that will contribute to counterproductive engagement (Smith et al., 2017) leading to eventually abandoning the software or being barred from use. Such counterproductive engagement can take the form of clients repurposing the affordances of forensic technologies to pursue harmful ends, such as abusing another client. It may also include psychological distress caused by engaging with the intended content of the app (see Wood, 2022). An example of this would be distress resulting from a client receiving feedback from the self-tracking survey that showed a significant deterioration in well-being.

Discussion and Conclusions

The SDEM provides a framework to organize and describe the stages of engagement with an app designed to support recovery from AOD. Long-term engagement with the app ultimately derives from the productive routines, sense of self-control and feelings of connectedness that accompany regular use of eRecovery’s affordances. But to get there, users need to negotiate the first two stages of adoption and appropriation. If the motivators and supports in these two stages of engagement are more powerful than the barriers and hindrances, then the user can move to the routine use stage. An important feature of the SDEM model is, therefore, that it emphasizes the user’s agency in engaging with eRecovery. Establishing positive routines, self-tracking, and feeling connected all depend on the user exercising choice and control in the way they use the app. A second is that the SDEM model acknowledges the role of contextual and environmental factors in supporting users through the early stages of engagement. Some of the greatest challenges in the eRecovery trial were about addressing the environmental barriers to regular and reliable access to the app, especially those associated with the high levels of social disadvantage and often chaotic lifestyles of trial participants. These barriers in turn reflect aspects of digital inequality (Reisdorf & Rikard, 2018) where people are cut-off from routine contact with the range of digital channels and services that form an increasingly important part of normal daily life. The NJC made use of material incentives (phones and phone credit) in combination with detailed user briefings to support users in the early weeks of engagement. However, given that digital inequality is likely to be present in many justice-involved groups, successful engagement strategies may need to be more proactive and structured. Sugie (2018) outlines a variety of strategies to assist disadvantaged groups to use smartphones including training sessions in smartphone use, encouraging the use of voice to text translation, and providing assistance to set up email and other accounts. These factors—and the problem of engagement more broadly—highlight how important it to avoid approaching digital technologies in corrections through a lens of technological solutionism: the belief that wholly technological “solutions” can be produces to solve long-standing social problems (Morozov, 2013). Addiction won’t be “fixed” by an app, as the most ardent technological solutionist might believe. For this reason, we want to again emphasize that mobile applications should represent an adjunct to, rather than a replacement of, traditional therapeutic interventions.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors (Stuart Ross, Mark A. Wood, Diana Johns, John Murphy) received Court Services Victoria (CSV) funding in January 2019 to undertake research assessing the trial of the e-Recovery App at the Neighbourhood Justice Centre (NJC) in Collingwood, Victoria, Australia.

ORCID iDs

Mark A. Wood

Diana Johns

References

Alkhaldi

Modrow

Hamilton

Pal

Ross

Murray

(2017). Promoting engagement with a digital health intervention (HeLP-Diabetes) using email and text message prompts: mixed-methods study. Interactive journal of medical research, 6(2), e6952.

Attig

Franke

(2020). Abandonment of personal quantification: A review and empirical study investigating reasons for wearable activity tracking attrition. Computers in Human Behavior, 102, 223–237. https://doi.org/10.1016/j.chb.2019.08.025

Baumel

Muench

Edan

Kane

J. M.

(2019). Objective user engagement with mental health apps: Systematic search and panel-based usage analysis. Journal of Medical Internet Research, 21(9), e14567. https://doi.org/10.2196/14567

Birnbaum

Lewis

D. M.

Rosen

Ranney

M. L.

(2015). Patient engagement and the design of digital health. Academic Emergency Medicine, 22(6), 754–756. https://doi.org/10.1111/acem.12692

Borghouts

Eikey

Mark

De Leon

Schueller

S. M.

Schneider

Stadnick

Zheng

Mukamel

Sorkin

D. H.

(2021). Barriers to and facilitators of user engagement with digital mental health interventions: Systematic review. Journal of Medical Internet Research, 23(3), e24387. https://doi.org/10.2196/24387

Brunner

Neumann

Yoon

Rettenberger

Stück

Briken

(2019). Determinants of dropout from correctional offender treatment. Frontiers in Psychiatry, 10, 142. https://doi.org/10.3389/fpsyt.2019.00142

Caraban

Karapanos

Gonçalves

Campos

(2019). 23 ways to nudge: A review of technology-mediated nudging in human-computer interaction [Paper presentation]. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems Glasgow, Scotland.

Chaple

Sacks

McKendrick

Marsch

L. A.

Belenko

Leukefeld

Prendergast

French

(2014). Feasibility of a computerized intervention for offenders with substance use disorders: A research note. Journal of Experimental Criminology, 10(1), 105–127.

Chaple

Sacks

McKendrick

Marsch

L. A.

Belenko

Leukefeld

Prendergast

French

(2016). A comparative study of the therapeutic education system for incarcerated substance-abusing offenders. The Prison Journal, 96(3), 485–508. https://doi.org/10.1177/0032885516636858

10.

Clarke

Jinks

McMurran

(2015). Readiness enhancement management strategies (REMS): A proof of concept and evaluation feasibility study of staff training to improve service engagement by people with personality difficulties. Journal of Psychiatric and Mental Health Nursing, 22(7), 502–509.

11.

Creswell

J. W.

Plano Clark

V. L.

(2011). Designing and conducting mixed methods research. SAGE.

12.

Deci

E. L.

Ryan

R. M.

(2008). Self-determination theory: A macrotheory of human motivation, development, and health. Canadian Psychology, 49(3), 182–185. https://doi.org/10.1037/a0012801

13.

Doherty

(2018). Engagement in HCI: Conception, theory and measurement. ACM Computing Surveys, 51(5), 99. https://doi.org/10.1145/3234149

14.

Elison

Weston

Davies

Dugdale

Ward

(2016). Findings from mixed-methods feasibility and effectiveness evaluations of the “breaking free online” treatment and recovery programme for substance misuse in prisons. Drugs: Education, Prevention and Policy, 23(2), 176–185.

15.

Farabee

Calhoun

Veliz

(2016). An experimental comparison of telepsychiatry and conventional psychiatry for parolees. Psychiatric Services, 67(5), 562–565.

16.

Fereday

Muir-Cochrane

(2006). Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International Journal of Qualitative Methods, 5(1), 80–92.

17.

Fleming

Bavin

Lucassen

Stasiak

Hopkins

Merry

(2018). Beyond the trial: Systematic review of real-world uptake and engagement with digital self-help interventions for depression, low mood, or anxiety. Journal of Medical Internet Research, 20(6), e199. https://doi.org/10.2196/jmir.9275

18.

Glass

J. E.

McKay

J. R.

Gustafson

D. H.

Kornfield

Rathouz

P. J.

McTavish

F. M.

Atwood

A. K.

Isham

Quanbeck

Shah

(2017). Treatment seeking as a mechanism of change in a randomized controlled trial of a mobile health intervention to support recovery from alcohol use disorders. Journal of Substance Abuse Treatment, 77, 57–66. https://doi.org/10.1016/j.jsat.2017.03.011

19.

Goodman

J. D.

McKay

J. R.

DePhillipis

(2013). Progress monitoring in mental health and addiction treatment: A means of improving care. Professional Psychology: Research and Practice, 44(4), 231–246. https://doi.org/10.1037/a0032605

20.

Gustafson

D. H.

Boyle

M. G.

Shaw

B. R.

Isham

McTavish

Richards

Schubert

Levy

Johnson

(2011). An E-health solution for people with alcohol problems. Alcohol Research and Health, 33(4), 327–337.

21.

Gustafson

D. H.

McTavish

F. M.

Chih

M.-Y.

Atwood

A. K.

Johnson

R. A.

Boyle

M. G.

Levy

Driscoll

Chisholm

S. M.

Dillenburg

Isham

Shah

(2013). A smartphone application for alcoholism recovery: A randomized controlled trial. JAMA Psychiatry, 71(5), 566–572.

22.

Han

J. Y.

(2011). Transaction logfile analysis in health communication research: Challenges and opportunities. Patient Education and Counseling, 82(3), 307–312. https://doi.org/10.1016/j.pec.2010.12.018

23.

Hilty

D. M.

Mucic

(2016). Technology, health, and contemporary practice: How does e-mental health fit it and what does it offer? In Mucic

Hilty

D. M.

(Eds.), e-Mental health. Springer.

24.

Istepanian

R. S. H.

Jovanov

Zhang

Y. T.

(2004). Guest editorial introduction to the special section on M-health: Beyond seamless mobility and global wireless health-care connectivity. IEEE Transactions on Information Technology in Biomedicine, 8(4), 405–414. https://doi.org/10.1109/TITB.2004.840019

25.

Johnson

Richards

Chih

M.-Y.

Moon

T. J.

Curtis

Gustafson

D. H.

(2016). A pilot test of a mobile app for drug court participants. Substance Abuse: Research and Treatment, 10(1), 1–7. https://doi.org/10.4137/SART.S33390

26.

Johnston

D. C.

Mathews

W. D.

Maus

Gustafson

D. H.

(2019). Using smartphones to improve treatment retention among impoverished substance-using appalachian women: A naturalistic study. Substance Abuse: Research and Treatment, 13(1), 1–10. https://doi.org/10.1177/1178221819861377

27.

Kessler

R. C.

Andrews

Colpe

L. J.

Hiripi

Mroczek

D. K.

Normand

S. L. T.

Walters

E. E.

Zaslavsky

A. M.

(2002). Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychological Medicine, 32(6), 959–976.

28.

Kip

Bouman

Y. H. A.

Kelders

S. M.

van Gemert-Pijnen

Y. J. E. W. C.

(2018). eHealth in treatment of offenders in forensic mental health: A review of the current state. Frontiers in Psychiatry, 9, 42. https://doi.org/10.3389/fpsyt.2018.00042

29.

Kip

Oberschmidt

Bierbooms

J. J.

(2020). Ehealth technology in forensic mental healthcare: Recommendations for achieving benefits and overcoming barriers. International Journal of Forensic Mental Health, 20(1), 31–47.

30.

Klasnja

Consolvo

Pratt

(2011). How to evaluate technologies for health behavior change in HCI research [Paper presentation].ACM CHI Conference on Human Factors in Computing System, Mat 7–12 2011, Vancouver, BC, Canada.

31.

Kroenke

Spitzer

R. L.

Williams

J. B. W.

(2001). The PHQ-9 validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613.

32.

Lappan

S. N.

Brown

A. W.

Hendricks

P. S.

(2020). Dropout rates of in-person psychosocial substance use disorder treatments: A systematic review and meta-analysis. Addiction, 115(2), 201–217. https://doi.org/10.1111/add.14793

33.

Larimer

M. E.

Palmer

R. S.

Marlatt

G. A.

(1999). Relapse prevention: An overview of marlatt’s cognitive-behavioral model. Alcohol Research and Health, 23(2), 151–160.

34.

Leonardi

P. M.

(2011). When flexible routines meet flexible technologies: Affordance, constraint, and the imbrication of human and material agencies. Management Information Systems Quarterly, 35(1), 147–167.

35.

Luxton

D. D.

McCann

R. A.

Bush

N. E.

Mishkind

M. E.

Reger

G. M.

(2011). mHealth for mental health: Integrating smartphone technology in behavioral healthcare. Professional Psychology: Research and Practice, 42(6), 505–512.

36.

McTavish

F. M.

Chih

M.-Y.

Shah

Gustafson

D. H.

(2012). How patients recovering from alcoholism use a smartphone intervention. Journal of Dual Diagnosis, 8(4), 294–304. https://doi.org/10.1080/15504263.2012.723312

37.

Morelli

(2002). Designing product/service systems: A methodological exploration. Design Issues, 18(3), 3–17.

38.

Morozov

(2013). To save everything, click here: The folly of technological solutionism. Public Affairs.

39.

Morris

Bans

M. K.

(2018). Developing digitally enabled interventions for prison and probation settings: A review. Journal of Forensic Practice, 20(2), 134–140. https://doi.org/10.1108/jfp-08-2017-0030

40.

Morris

Knight

(2018). Co-producing digitally-enabled courses that promote desistance in prison and probation settings. Journal of Criminological Research, Policy and Practice, 4(4), 269–279.

41.

Muroff

Robinson

Chassler

López

L. M.

Lundgren

Guauque

Dargon-Hart

Stewart

Dejesus

Johnson

Pe-Romashko

(2019). An outcome study of the CASA-CHESS smartphone relapse prevention tool for Latinx Spanish-speakers with substance use disorders. Substance Use & Misuse, 54(9), 1438–1449.

42.

Murphy

Densley

Ross

(2020). Understanding uptake to support mobile service design - Towards a practical model to assess the uptake of a mobile application supporting clients with drug and alcohol addiction [Paper presentation]. Proceedings of ServDes.2020 Tensions, Paradoxes and Plurality Conference Melbourne, Australia.

43.

Nagy

Neff

(2015). Imagined affordance: Reconstructing a keyword for communication theory. Social Media+ Society, 1(2), 1–9. https://doi.org/2056305115603385

44.

Norman

(1988). The design of everyday things. Basic Books.

45.

O’Brien

H. L.

Toms

E. G.

(2008). What is user engagement? A conceptual framework for defining user engagement with technology. Journal of the American Society for Information Science and Technology, 59(6), 938–955. https://doi.org/10.1002/asi.20801

46.

Rantanen

Järveläinen

Leppälahti

(2021). Prisoners as users of digital health care and social welfare services: A finnish attitude survey. International Journal of Environmental Research and Public Health, 18(11), 5528.

47.

Reisdorf

B. C.

Rikard

(2018). Digital rehabilitation: A model of reentry into the digital age. American Behavioral Scientist, 62(9), 1273-1290.

48.

Smith

Ploderer

Wadley

Webber

Borland

(2017). Trajectories of engagement and disengagement with a story-based smoking cessation app [Conference session]. ACM CHI Conference on Human Factors in Computing Systems, ACM Press, New York, NY (pp. 3045–3056). https://doi.org/10.1145/3025453.3026054

49.

Statista. (2021). Retention rate on day 1 and day 30 of mobile app installs worldwide as of August 2020, by category. Author. https://www.statista.com/statistics/259329/ios-and-android-app-user-retention-rate/

50.

Sugie

N. F.

(2018). Utilizing smartphones to study disadvantaged and hard-to-reach groups. Sociological Methods & Research, 47(3), 458–491.

51.

Sygel

Kristiansson

Furberg

Fors

(2014). Reactions on display/intimate partner violence (ROD/IPV)—A study of a new interactive computer simulation program for the treatment of men convicted of intimate partner violence. International Journal of Forensic Mental Health, 13(4), 369–380.

52.

Teng

M. Q.

Hodge

Gordon

(2019, May). Participatory design of a virtual reality-based reentry training with a women’s prison [Conference session]. Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1–8).

53.

Thaler

R. H.

Sunstein

C. R.

(2009). Nudge: Improving decisions about health, wealth and happiness. Penguin Australia.

54.

Torous

J. B.

Lipschitz

Firth

(2020). Dropout rates in clinical trials of smartphone apps for depressive symptoms: A systematic review and meta-analysis. Journal of Affective Disorders, 263, 413–419. https://doi.org/10.1016/j.jad.2019.11.167

55.

Wood

M. A.

(2022). Mapping technology-harm relations: From ambient harms to zemiosis. Crime, Media, Culture, 18(4), 509–526. https://doi.org/10.1177/17416590211037384

56.

Wood

M. A.

(2021). Rethinking how technologies harm. The British Journal of Criminology, 61(3), 627–647.