Efficient Type 1 Diabetes (T1D) management necessitates comprehensive tracking of various factors that influence blood sugar levels. However, tracking health data for children with T1D poses unique challenges, as it requires the active involvement of both children and their parents. This study aims to uncover the benefits, challenges, and strategies associated with collaborative tracking for children (ages 6-12) with T1D and their parents. Over a three-week data collection probe study with 22 child-parent pairs, we found that collaborative tracking, characterized by the shared responsibility of tracking management and data provision, yielded positive outcomes for both children and their parents. Drawing from these findings, we delineate four distinct tracking approaches: child-independent, child-led, parent-led, and parent-independent. Our study offers insights for designing health technologies that empower both children and parents in learning and encourage the sharing of different perspectives through collaborative tracking.

OBJECTIVES: To understand healthcare providers' experiences of using GlucoGuide, a mockup tool that integrates visual data analysis with algorithmic insights to support clinicians' use of patientgenerated data from Type 1 diabetes devices. MATERIALS AND METHODS: This qualitative study was conducted in three phases. In Phase 1, 11 clinicians reviewed data using commercial diabetes platforms in a think-aloud data walkthrough activity followed by semistructured interviews. In Phase 2, GlucoGuide was developed. In Phase 3, the same clinicians reviewed data using GlucoGuide in a think-aloud activity followed by semistructured interviews. Inductive thematic analysis was used to analyze transcripts of Phase 1 and Phase 3 think-aloud activity and interview. RESULTS: 3 high level tasks, 8 sub-tasks, and 4 challenges were identified in Phase 1. In Phase 2, 3 requirements for GlucoGuide were identified. Phase 3 results suggested that clinicians found GlucoGuide easier to use and experienced a lower cognitive burden as compared to the commercial diabetes data reports that were used in Phase 1. Additionally, GlucoGuide addressed the challenges experienced in Phase 1. DISCUSSION: The study suggests that the knowledge of analytical tasks and task-specific visualization strategies in implementing features of data interfaces can result in tools that lower the perceived burden of engaging with data. Additionally, supporting clinicians in contextualizing algorithmic insights by visual analysis of relevant data can positively influence clinicians' willingness to leverage algorithmic support. CONCLUSION: Task-aligned tools that combine multiple data-driven approaches, such as visualization strategies and algorithmic insights, can improve clinicians' experience in reviewing device data.

In time-series forecasting, future target values may be affected by both intrinsic and extrinsic effects. When forecasting blood glucose, for example, intrinsic effects can be inferred from the history of the target signal alone (i.e. blood glucose), but accurately modeling the impact of extrinsic effects requires auxiliary signals, like the amount of carbohydrates ingested. Standard forecasting techniques often assume that extrinsic and intrinsic effects vary at similar rates. However, when auxiliary signals are generated at a much lower frequency than the target variable (e.g., blood glucose measurements are made every 5 minutes, while meals occur once every few hours), even well-known extrinsic effects (e.g., carbohydrates increase blood glucose) may prove difficult to learn. To better utilize these sparse but informative variables (SIVs), we introduce a novel encoder/decoder forecasting approach that accurately learns the per-timepoint effect of the SIV, by (i) isolating it from intrinsic effects and (ii) restricting its learned effect based on domain knowledge. On a simulated dataset pertaining to the task of blood glucose forecasting, when the SIV is accurately recorded our approach outperforms baseline approaches in terms of rMSE (13.07 [95% CI: 11.77,14.16] vs. 14.14 [12.69,15.27]). In the presence of a corrupted SIV, the proposed approach can still result in lower error compared to the baseline but the advantage is reduced as noise increases. By isolating their effects and incorporating domain knowledge, our approach makes it possible to better utilize SIVs in forecasting.

Children with Type 1 Diabetes (T1D) face many challenges with keeping their blood glucose levels within a healthy range because they cannot manage their illness by themselves. To prevent children’s blood glucose from becoming too high or too low, parents apply different strategies to avoid risky situations. To understand how parents of children with T1D manage these risks, we conducted semi-structured interviews with children with T1D (ages 6-12) and their parents (N=41). We identified four types of strategies used by parents (i.e., educated guessing game, contingency planning, experimentation, and reaching out for help) that can be categorized according to two dimensions: 1) the cause of risk (known or unknown) and 2) the occurrence of risk (predictable or unpredictable). Based on our findings, we provide design implications for collaborative health technologies that support parents in better planning for contingencies and identifying unknown causes of risks together with their children.

Engaging with multiple streams of personal health data to inform self-care of chronic health conditions remains a challenge. Existing informatics tools provide limited support for patients to make data actionable. To design better tools, we conducted two studies with Type 1 diabetes patients and their clinicians. In the first study, we observed data review sessions between patients and clinicians to articulate the tasks involved in assessing different types of data from diabetes devices to make care decisions. Drawing upon these tasks, we designed novel data interfaces called episode-driven data narratives and performed a task-driven evaluation. We found that as compared to the commercially available diabetes data reports, episode-driven data narratives improved engagement and decision-making with data. We discuss implications for designing data interfaces to support interaction with multidimensional health data to inform self-care.

Although child participation is required for successful Type 1 Diabetes (T1D) management, it is challenging because the child’s young age and immaturity make it difficult to perform self-care. Thus, parental caregivers are expected to be heavily involved in their child’s everyday illness management. Our study aims to investigate how children and parents collaborate to manage T1D and examine how the children become more independent in their self-management through the support of their parents. Through semi-structured interviews with children with T1D and their parents (N=41), our study showed that children’s knowledge of illness management and motivation for self-care were crucial for their transition towards independence. Based on these two factors, we identified four types of children’s collaboration (i.e., dependent, resistant, eager, and independent) and parents’ strategies for supporting their children’s independence. We suggest design implications for technologies to support collaborative care by improving children’s transition to independent illness management.

Tools for self-care of chronic conditions often do not fit the contexts in which self-care happens because the influence of context on self-care practices is unclear. We conducted a diary study with 15 adolescents with Type 1 Diabetes and their caregivers to understand how context affects self-care. We observed different contextual settings, which we call contextual frames, in which diabetes self-management varied depending on certain factors - physical activity, food, emotional state, insulin, people, and attitudes. The relative prevalence of these factors across contextual frames impacts self-care necessitating different types of support. We show that contextual frames, as phenomenological abstractions of context, can help designers of context-aware systems systematically explore and model the relation of context with behavior and with technology supporting behavior. Lastly, considering contextual frames as sensitizing concepts, we provide design direction for using context in technology design.

Human–computer interaction research on personal informatics in health care has focused on systems that aim to support patient empowerment and enable better health outcomes with data monitoring and tracking. Through examining the lived experience of personal data used to manage chronic illness, we show how such technology design is also the site of radical dependencies, collaborative care arrangements, and wider sociopolitical concerns tied to new forms of technical labor and shifts in medical expertise. Drawing from ethnographic research with open source, do-it-yourself collectives engaged in opening up corporate-controlled type 1 diabetes devices and data, we propose the analytical lens of lived data. Lived data emphasize data as an integral way of living, enacted through a multiplicity of things, relations, and practices, from bodies and needles, social media support groups, and legal processes to writing code, making visualizations, and hacking devices. Building on critical and feminist scholarship of human–machine relations, we articulate the work that goes into producing and living personal data, the physical and emotional costs of data tracking, and the consequences of do-it-yourself as a form of individual empowerment in health and wellness.

Objective: Sharing personal information about type 1 diabetes (T1D) can help adolescents obtain social support, enhance social learning, and improve self-care. Diabetes technologies, online communities, and health interventions increasingly feature data-sharing components. This study examines factors underlying adolescents' willingness to share personal T1D information with peers. Materials and Methods: Participants were 134 adolescents (12-17 years of age; 56% female) who completed an online survey regarding experiences helping others with T1D, perceived social resources, beliefs about the value of sharing information and helping others, and willingness to share T1D information. Hemoglobin A1c values were obtained from medical records. Results: Adolescents were more willing to share how they accomplished T1D tasks than how often they completed them, and least willing to share glucose control status. In multivariate analyses, sharing/helping beliefs (β = 0.26, P < .01) and glucose control (HbA1c value; β = -0.26, P < .01) were related to greater willingness to share personal health information. Glucose control moderated relationships such that adolescents with worse A1c values had stronger relationships between sharing/helping beliefs and willingness to share (β = 0.18, P < .05) but weaker relationships between helping experience and willingness to share (β = -0.22, P = .07). Discussion: Many adolescents with T1D are willing to share personal health information, particularly if they have better diabetes health status and a stronger belief in the benefits of sharing. Conclusion: Social learning and social media components may improve intervention participation, engagement, and outcomes by boosting adolescents' beliefs about the benefits of sharing information and helping others.

Designing systems to support the social context of personal data is a topic of importance in CSCW, particularly in the area of health and wellness. The relational complexities and psychological consequences of living with health data, however, are still emerging. Drawing on a 12+ month ethnography and corroborating survey data, we detail the experiences of parents using Nightscout--an open source, DIY system for remotely monitoring blood glucose data-with their children who have type one diabetes. Managing diabetes with Nightscout is a deeply relational and (at times) contested activity for parent-caregivers, whose practices reveal the tensions and vulnerabilities of caregiving work enacted through data. As engagement with personal data becomes an increasingly powerful way people experience life, our findings call for alternative data narratives that reflect a multiplicity of emotional concerns and social arrangements. We propose the analytic lens of caring-through-data as a way forward.

Making effective use of patient-generated data (PGD) is challenging for both patients and providers. Designing systems to support collaborative and individual use of PGD is a topic of importance in CSCW, considering the limitations of informatics tools. To inform better system design, we conducted a study including focus groups, observations and interviews with patients and providers to understand how PGD is interpreted and used. We found that while PGD is useful for identifying and solving disease-related problems, the following differences in patient-provider perceptions challenge its effective use - different perceptions about what is a problem, selecting what kinds of problems to focus on, and using different data representations. Drawing on these insights, we reflect on two specific conceptualizations of disease management behavior (sensemaking and problem-solving) as they relate to data specific activities of patients and providers and provide design suggestions for tools to support collaborative and individual use of PGD.