This paper characterizes the mental health technology "kits" of individuals managing depression: the specific technologies on their digital devices and physical items in their environments that people turn to as part of their mental health management. We interviewed 28 individuals living across the United States who use bundles of connected tools for both individual and collaborative mental health activities. We contribute to the HCI community by conceptualizing these tool assemblages that people managing depression have constructed over time. We detail categories of tools, describe kit characteristics (intentional, adaptable, available), and present participant ideas for future mental health support technologies. We then discuss what a mental health technology kit perspective means for researchers and designers and describe design principles (building within current toolkits; creating new tools from current self-management strategies; and identifying gaps in people's current kits) to support depression self-management across an evolving set of tools.

Artificial intelligence (AI) supported clinical decision support (CDS) technologies can parse vast quantities of patient data into meaningful insights for healthcare providers. Much work is underway to determine the technical feasibility and the accuracy of AI-driven insights. Much less is known about what insights are considered useful and actionable by healthcare providers, their trust in the insights, and clinical workflow integration challenges. Our research team used a conceptual prototype based on AI-generated treatment insights for type 2 diabetes medications to elicit feedback from 41 U.S.-based clinicians, including primary care and internal medicine physicians, endocrinologists, nurse practitioners, physician assistants, and pharmacists. We contribute to the human-computer interaction (HCI) community by describing decision optimization and design objective tensions between population-level and personalized insights, and patterns of use and trust of AI systems. We also contribute a set of 6 design principles for AI-supported CDS.

Patient work encompasses a challenging set of activities necessary for learning about and managing chronic conditions over time. Many patient-centered health technology interventions focus on supporting types of patient work, such as symptom tracking, medication adherence, and information sharing between patients and providers. However, people may not always follow, or may actively resist, the activities prescribed by their formal patient role. In this paper, we present three case studies about patients with different chronic conditions to critically reflect on the types of patient behavior commonly taken up in health technology design as acts of "noncompliance." Detailing conflicts that emerge when patients are caught between meeting their personal needs and following clinical best practices, we show how everyday life and health system goals are often misaligned in ways that can't be easily reconciled through current design approaches. As a way forward, we argue for alternative ways of understanding the tensions routinely shaping people's healthcare experiences. We introduce the term care frictions as a sensitizing concept useful for helping designers reframe "noncompliant" behaviors as legitimate forms of patient work. Our paper also offers design considerations-both on challenges and generative possibilities-for future CSCW research seeking to support a wider breadth of patient behavior. In this, we call attention to the value of designer and researcher reflexivity in making visible the problematic assumptions in health technology design that can lead to social and emotional patient harms.

Depression is a challenging condition that requires individuals to manage their moods and emotions over time. Within CSCW, there has been an interest in understanding how individuals seek and share support on social media and in online communities. However, less attention has been paid to how collaboration as an aspect of self-management of depression unfolds in people's daily lives. In this paper, we explore the collaborative self-management work of 28 individuals managing depression who live in the United States. Data collection included remote semi-structured interviews with an associated cognitive mapping exercise. Our findings describe who participants turn to for day-to-day collaborative support, how collaborative activities are enacted (across both mood-focused and preventative support practices), and where these often technology-mediated interactions occur across text, phone, video, and picture-based channels. We discuss collaborative self-management in the depression support context, including key characteristics: agency, reciprocity, time, and interaction. We also present a four-step model of how the process occurs over time (awareness, planning, interaction, and reflection). We conclude by discussing how technology ecosystems support individuals' collaborative self-management.

Much human-computer interaction work related to depression focuses on the population level (e.g., studying social media hashtags related to depression) or evaluates prototypes for digital interventions to manage depression. However, little is known about how people living with depression perceive and manage technology use, such as time spent on social media per day. For this study, we interviewed 30 individuals living with depression to explore their technology and social media use. We find that these individuals demonstrated emergent practices related to self-regulation, such as learning to monitor and adjust technology use to improve their emotional, cognitive, and behavioral health. Our findings add a human-centered viewpoint to the relationship between living with depression and technology and social media use. We present design implications of these findings for better empowering individuals with depression to encourage their natural inclinations to self-regulate technology and social media use.

People diagnosed with a chronic health condition have many information needs which healthcare providers, patient groups, and resource designers seek to support. However, as a disease progresses, knowing when, how, and for what purposes patients want to interact with and construct personal meaning from health-related information is still unclear. This paper presents findings regarding the information work of chronic kidney disease patients. We conducted semi-structured interviews with 13 patients and 6 clinicians, and observations at 9 patient group events. We used the stages of the information journey - recognizing need, seeking, interpreting, and using information - to frame our data analysis. We identified two distinct but often overlapping information work phases, 'Learning' and 'Living With' a chronic condition to show how patient information work activities shift over time. We also describe social and individual factors influencing information work, and discuss technology design opportunities including customized education and collaboration tools.

Millions of Americans struggle with depression, a condition characterized by feelings of sadness and motivation loss. To understand how individuals managing depression conceptualize their self-management activities, we conducted visual elicitations and semi-structured interviews with 30 participants managing depression in a large city in the U.S. Midwest. Many depression support tools are focused on the individual user and do not often incorporate social features. However, our analysis showed the key importance of sociality for self-management of depression. We describe how individuals connect with specific others to achieve expected support and how these interactions are mediated through locations and communication channels. We discuss factors influencing participants' sociality including relationship roles and expectations, mood state and communication channels, location and privacy, and culture and society. We broaden our understanding of sociality in CSCW through discussing diffuse sociality (being proximate to others but not interacting directly) as an important activity to support depression self-management.

In emergency response organizations like the fire service, personnel require easy access to reliable, up-to-date safety protocols. Systems for creating and managing Standard Operating Procedures (SOPs) within these command and control organizations are often rigid, inaccessible, and siloed. Open collaboration systems like wikis and social computing tools have the potential to address these limitations, but have not been analyzed for intra-organizational use in emergency services. In response to a request from the Fire Protection Research Foundation (FPRF) we evaluated a high-fidelity open collaboration system prototype, FireCrowd, that was designed to manage SOPs within the U.S. fire service. We use the prototype as a technology probe and apply human-centered design methods in a suburban fire department in the Chicago area. We find that organizational factors would inhibit the adoption of some open collaboration practices and identify points in current practices that offer opportunities for open collaboration in the future.