Mark Matthews, Gavin Doherty
The World Health Organization estimates that approximately one million people per year commit suicide. Mental disorders such as depression are responsible for more than 90 percent of these deaths. In fact, depression is the leading cause of disability in the developed world, and the human and economic cost of mental illness is reaching crisis proportions. The stigma surrounding mental health issues exacerbates the problem, and many people are unable or reluctant to engage in and access effective treatment.
Technology can help address these key problems of access and engagement, particularly for younger people. Interaction design has an important role in developing innovative and worthwhile applications that support the user in an effective way. Given the scale of the problem, even small changes in the effectiveness of mental health services could have a big impact.
Ethnographies, user observation, focus groups, cultural probes, think-alouds, interviewsthese are just some of the tools designers have come to rely on. In mental health cases, the introduction of not just the technology but also the designer could be detrimental. What happens when you can't talk to the user, when they can't be approached or observed, when they are effectively invisible?
The most significant challenge we have faced working with children and teenagers affected by mental health issues is how to develop useful systems where there is little or no direct access to end users? Software that does not consider users' needs and is difficult to use can present another barrier to treatment. While the ethical restrictions that limit access are in place for valid reasons, they leave us with a need for new methods to address the lack of involvement of the end user.
Over the past five years, our group at Trinity College Dublin, in collaboration with the Mater Hospital Child and Adolescent Mental Health Services, has developed several systems in a range of services that are in clinical use today: "Personal Investigator," a 3-D computer game; "Mobile Mood Diary," a personal diary system; and "My Mobile Story," a multimodal storytelling system. Design and evaluation activities have involved 26 different service providers, including schools, hospitals, charities, and specialist clinics.
The goal of our research has been to develop effective support for young people attending public mental health services. When designing these technologies, we needed to consider the needs of both the adolescent clients and therapists. Adolescence is a critical time when many teenagers can feel vulnerable and isolated. Many mental health problems begin during this period. Not dealing with these problems as they emerge increases the risk of more severe issues in adulthood.
Most teenagers do not receive the support they need, and even those who do can find it difficult to engage with their treatment. As one therapist put it, "Young people are not used to walking in and talking to a stranger about their problems." The challenge for therapists is to involve children and adolescents "in treatment and to work toward a change that the child may not view as necessary or even potentially useful" .
A further problem is that the materials and tools used to engage young people in the therapeutic process tend to be outmoded and can seem irrelevant to many teenagers; according to one young person interviewed "things on paper seem like an assignmentmore likely not to do it."
It can be just as demanding to design for therapists who may be concerned about negative impacts of technology. Technology is rarely used in day-to-day therapeutic work or in therapist training. As a result, some therapists can feel undermined by technology. Any technology that is introduced needs to complement their existing work practice; it must not take too much time, given typical busy schedules.
While little technology has been used in the area to date, there are undoubtedly many opportunities, particularly in the development of systems that are relevant to younger people and at-risk groups such as young men.
From the beginning, it was clear to us that a practical approach was essential. Teenagers and clinics cannot afford special devices or high-end computers. In fact, high-tech devices are more liable to attract the attention of peers. In order to ensure that our work could carry over to actual practice, we designed for mobile phones and basic desktop PCs already in the pockets and homes of young people.
Mobile phones remove the need for special equipment: Users are already familiar with their phone, and using it in public places would not attract extra attention. Most important, as most teenagers carry their phones for most of the day, it could provide them with near-constant support when and where they need it. This was evident in the Mobile Mood Diary (see the accompanying sidebar), which significantly increased the amount of mood entries recorded by young people.
When we were designing systems for teenagers to use between and during therapy sessions, we needed to consider two very different environments and two very different levels of engagement.
Inside Session. While in a therapeutic session, there are few competing demands on the young person's attention, teenage clients are notoriously reticent and difficult to engage in face-to-face discussion about their feelings. This makes it extremely important to approach young people using materials with which they are familiar and comfortable. In this setting, technological interventions may provide more appealing alternatives than face-to-face conversation alone. One therapist remarked, "When children come to the clinic... they frequently refer to computers and technology." Indeed, in most cases we have found they were more comfortable using the technology than the therapist, leading to a degree of role-reversal and empowerment that can be positive for the client.
Within a session, providing some control to the client and bringing a new focus into the situation can help to reduce the level of tension and support a less confrontational interaction between the client and therapist. Systems may also provide clients with ways to express emotions and feelings without losing face or appearing weak. During evaluations of the Personal Investigator computer game, therapists reported that some clients were more comfortable talking with the therapist while facing the computer. In these sessions, the teenagers generally controlled the mouse and through navigation of the 3-D space were able to control the pace and direction of the therapeutic session .
While short-lived novelty factors may be of little use in some areas, in mental health care this can be genuinely useful. According to one therapist, "A good ice breaker makes therapy a little less threatening." In some cases, simply using technology as a medium has been effective. Regarding mobile phones, one therapist commented: "I'm 53 and appear old to most clients. For me to have something that they are comfortable with opened up the conversationthey are comfortable talking about their mobile phones. Clients found it very engaging."
Outside Sessions. In the services we were involved with, teenagers typically attended one-hour therapy sessions once a week, leaving significant time before the next session. There are significant benefits to involving teenagers in therapeutic activities between sessions. However, engaging teens in activities outside of therapy can be difficult, because many things compete for their attentiontelevision, friends, schoolwork, sports, and computers. Consequently, teenagers may have no time for long and demanding tasks, or may not have sufficient personal space from peers or family to engage in them. Because of this, we have focused on designing systems that are readily available, provide discreet access, and allow multiple short interactions. Stigma is strongest outside a session, so it is also important that systems are flexible and allow discreet and secure access.
Our two mobile systems use a model in which teenagers generate content through a series of brief interactions between sessions, perhaps at times of significant mental health events, and then elaborate, reflect on, and discuss this content with a therapist in a clinical setting.
The lack of access to users means there is a risk of not understanding users' needs, their cultural context, the practicalities of their domestic life, and existence outside or even within the therapeutic environment.
The challenge then is to identify techniques that can provide the designer with feedback on the suitability of designs before clinical use. Other researchers have experienced similar difficulties getting access to end users, and in some cases we can adapt techniques from these cases. Holloway has described competing with other departments in her company to get access to users to evaluate help systems . Researchers in "inclusive design" (e.g., technologies to support people with autism) sometimes have to rely on other stakeholders because end users may not be able to provide suitable feedback to designers.
While addressing these issues depended on a number of collaborative design activities with therapists, the use of role play, peer evaluation, and clinical pilots were key in the development process.
Therapist involvement. Therapists have expert, sometimes tacit, knowledge of their clients and can provide vital information about their needs. Close collaboration with at least one therapist throughout all our projects has been essential. Therapists have been involved as design partners from early requirements and idea generation all the way through to prototyping, clinical protocol development, and reviewing systems right before they were used in clinical settings.
Peer-user evaluations. While therapists can help to provide the designer with insights into clinical situations, there is no clear way to model a client's outside world. In order to overcome this, we have used peer usersteenagers who are not suffering from mental health problems. In the early stages, they participate in relatively brief usability trials, and at a later stage in more involved evaluations. We generally recruit participants from inner-city schools and after-school clubs, as this matches the catchment area of the services we deal with and provides peer users who share similar socio-economic backgrounds as the ultimate end users.
Usability evaluations usually involve between five and eight young people and last around 30 minutes. When testing mobile systems, we have the teenagers use their own mobile devices. Running usability trials of initial prototypes with peer users has helped us to identify usability problems and provided feedback on designs through direct observation and dialogue. For us, using peer users to evaluate usability has been a critical solution because there is a significant question as to whether it is ethically sound to evaluate software, which might have significant usability problems, on clients who have come solely for treatment.
We have also used peer trials at a later stage of development to help evaluate the suitability, viability, and appeal of systems in realistic contexts. These trials are generally more involved, lasting between one and two weeks.
One benefit of this type of evaluation is that it can provide evidence, at an advanced stage of design, that the system is engaging and is suitable for clinical use. Peer evaluations of Mobile Mood Diary compared existing methods of recording mood information with a new mobile phone system. A large and statistically significant effect was found for improved compliance, providing a strong case for proceeding to clinical evaluation .
One issue with peer user evaluations is that some applications and content may not be meaningful or appropriate for a more general teenage population. For peer-user evaluations of My Mobile Story, a therapeutic storytelling system based on a series of template stories, we provided peer users with a more generic "Mobile Reporter" template, which they used to create nontherapeutic stories. We thought that the existing formats, such as a "thoughts, feelings and behaviors" template, would not have much meaning for peer users.
Peer evaluations can increase confidence in the value of a system and its appropriateness for clinical evaluation. Therapists are more likely to accept and use a system that has strong supporting evidence.
Role playing. Role playing has been used in the design process to numerous ends including: to generate new concepts, to test design ideas on potential users, and as a technique for designers to help them "imagine better...to empathize better" .
Role playing is particularly suited to use in mental health care. Therapists are extremely comfortable with role plays. Most trainee therapists are required to practice role playing therapeutic sessions in their training.
Mobile phones remove the need for special equipment: Users are already familiar with their phone, and using it in public places would not attract extra attention. Most important, as most teenagers carry their phones for most of the day, it could provide them with near-constant support when and where they need it.
In order to simulate the use of therapeutic systems, we provide therapists with simple role cards, which contain basic character details. For the therapist role, this includes information such as experience of technology, years of experience, and their attitude toward technology. For the client roles, it can include elements such as their name, age, mental health difficulty, current mood, and attitude to technology. The client role cards are based on realistic client scenarios and, where possible, actual client cases.
We have used role playing primarily to evaluate the suitability of our systems to the work practice of therapists. It has also proved useful in providing therapists with a safe environment in which to get to know the technologya setting where they can make mistakes, ask "silly" questions, and rehearse important scenarios like introducing the technology to a teenager. Moreover, the use of role playing has helped to "debug" the clinical system, assisting in identifying practical problems prior to clinical use.
Observing role-plays has provided us with insights into the therapists' existing work practices, and has been invaluable in helping to understand how therapists would instinctively use the tool.
Therapist reviews. Just before clinical pilots of a system, we complete a therapist review process. This involves therapists, from a wide range of theoretical backgrounds, evaluating the system. They are asked to imagine themselves in the place of a client and use the system for at least one week. These reviews provide feedback from a therapeutic perspective on the suitability of the system for teenage clients and the protocols for clinical usage. These reviews are different from early-stage expert reviews with HCI specialists and therapists.
In our experience, therapist reviews and role plays had numerous benefits, including assessment of the therapeutic validity of the system, understanding how the system could be used in practice, and increasing therapists' confidence in using the system, which helps in building up a network of evaluators.
Clinical pilots. While peer evaluations and other approaches have their benefits, some aspects of systems can be evaluated only in clinical settings. For example, privacy is more important to clients than to other teenagers, and the sense of stigma is heightened.
Formal clinical trials in mental health generally take the form of randomized, controlled trials (RCTs). These are extremely resource and time intensive. Formal trials are conducted in more tightly constrained situations than everyday practice, with many exclusions (symptoms too mild or too severe) and minimal comorbidity. Consequently, many findings from mental health studies never find their way into clinical practice.
From an evaluation perspective, there are issues regarding the use of formal clinical trials (e.g., RCT) as a primary source of design information regarding the usefulness of mental health technology. Although formal trials measuring outcomes are important once a final system exists, we decided to evaluate our systems initially with practicing therapists in their day-to-day clinics. While this raises logistical issues, the benefit is that we have a greater chance of generating useful design inputs, which have a degree of real-world validity. During these clinical pilots, therapists act as proxies for the designer in their interaction with clients.
Logging. Therapist feedback from clinical sessions tends to focus on the impact of the technology from a therapeutic perspective. Application logging can provide low-level client interaction information that would otherwise not be possible to capture. It can provide useful information both inside a session on a desktop PC and outside the session, logging client interactions on their mobile phone.
The clinical pilot of Mobile Mood Diary provided an idea of the potential of even relatively sparse logging information. Even basic data, such as the time a mood was recorded and how much time was spent on a task, can provide valuable information. As mobile devices become more powerful, it will be possible to record more detailed interaction information without any perceivable performance impact on the system.
As long as it is done sensitively, with the awareness and consent of the client, the logging of client interactions holds promise for providing the designer with contact-free interaction feedback. For the designer, who is distanced from the clinical setting, such data offers a glimpse into the clinic and the client's world, and encourages a degree of empathy with the client. For example, mood diary data may reflect a client going through a very difficult time.
One key point that has emerged from our work is the need to provide a range of options for clients. In a survey of teenage peers, we found that while 85 percent would prefer to use their mobile phone for diaries, a further 15 percent would prefer alternatives such as paper. Preferences are subject to changeone therapist reported that her "client reported wanting to avoid her phone when she felt in a low mood." Other issues we have encountered include teenagers who lived in remote areas, where there was no access to a mobile network, and teenagers who did not have suitable phones.
When possible we try to provide a range of options for clients to engage in therapeutic activities. Mobile Mood Diary clients can record moods using a paper diary, their mobile phone, or a PC.
Finally, while there are clear benefits to using digital devices over paper, such as increased reliability, the potential for greater privacy, and saving admin time, it is important to consider the substantial legacy of paper-based materials in therapy. My Mobile Story was designed to allow therapists to easily incorporate existing paperwork into the system, thereby helping them to continue to work with familiar materials in a new medium and taking advantage of a large, existing body of therapeutic content.
Designing for mental health care involves designing for engagement, for practicality, for privacy, for user control, and for discretion. Mental illness is a very serious and worsening problem with a substantial social stigma that prevents constructive dialogue about the problem.
Design work should seek solutions that put young people at ease, make them comfortable, strengthen the relationship with the therapist, are trusted and worthy of trust, and help to develop trust in the therapy. We can begin to do this through principles of transparency, client control, and privacy. It is also important to remember that engagement with the therapy itself is important, as is supporting the development of a trusting relationship between the client and therapist. Tools should be designed to give therapists the confidence to use them and to introduce them to clients.
This is just beginning in a field where HCI holds much promise for providing a safe and discreet space for reflection, personal development, and meaningful dialogue. Ultimately, it can deliver broader benefits for supporting the personal management of mental health.
Mark Matthews designs systems to help engage people in therapeutic activities. He has worked in this area for six years on projects ranging from therapeutic computer games to mobile diary systems and online virtual environments. Matthews' research interests include designing for sensitive and restricted environments, mobile design, and the use of play to develop relationships. He has recently submitted his Ph.D.
Gavin Doherty is a permanent faculty member at Trinity College Dublin and has led a number of projects in the area of technology in mental health. His research interests include design methods, language technology, design for mobile devices, and visual decision support in complex domains such as health care and manufacturing. Doherty obtained his D.Phil. in the HCI Group at the University of York.
Mobile Mood Diary is based on an established paper-based mood charting exercise in cognitive behavioral therapy. Therapists often ask clients to complete paper charts but compliance is low, they provide unreliable information, and involve added administrative time. A mobile phone mood diary was developed through a collaborative process with therapists.
A psychologist who used the diary in clinical situations remarked, my "client was particularly thorough in their use of the mood diary and this meant that a comprehensive log was maintained for interventions during sessions. I have worked for 20 years in the field and never got such recording compliance."
Sample mood diary entries include: "I feel terrible. Week nd dizzy. No sleep" "very bad i want 2 die now or run away" "Why do the meetings always end when i feel were building momentum" "Tired but happy and refreshed."
My Mobile Story was designed to make use of the multimedia capabilities of mobile phones. Teenagers can use their phone to record sounds, pictures, videos, and text between sessions and then construct therapeutically meaningful stories with their therapists. Therapists can incorporate existing paper therapeutic exercises into the system as "therapeutic plans."
The system was designed to help the client enter into personally meaningful conversations that deal with important issues and identify strengths, ultimately developing the therapeutic relationship. According to one therapist, My Mobile Story "personalizes the content for therapy sessions". Another therapist who used it with two teenagers reported that "it can serve as a starting point for discussion in subsequent sessions".
John [not his real name] was a 15-year old male from a single-parent family, and considered a potential suicide risk. He agreed to use the Mobile Mood Diary for two weeks, but continued using it for several months. He recorded 85 entries over 100 days, and opened the diary many timespossibly to review and reflect on previous entries. John's parent was finding it very difficult to cope with his son's struggles. The therapists suggested that he use the diary "to see if there was a correspondence between the child's and parent's moods."
The parent recorded 68 mood entries over 75 days. This allowed discussion of how the client and the parent influenced each other, providing new insight to both of them, and was an unforeseen use of the system.
Before meetings with the psychiatrist, the client and therapist would login to the website and print out the chart for each person. This was used to support discussions and decisions regarding risk factors (suicide) and medication.
©2009 ACM 1072-5220/09/1100 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.