Information Studies Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/2780
Browse
3 results
Search Results
Item Show and Tell: Exploring how audio narratives can complement visualizations of stroke survivors’ personal health data(2023) Shettigar, Aishwarya; Choe, Eun Kyoung; Library & Information Services; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Wearable technology in healthcare could give individuals awareness and independence in rehabilitation. In this qualitative work, I investigate how using speech-based, audio narrative summaries alongside graphical visualizations affect users’ understanding of their personal data. I conducted this work in the context of stroke recovery, where stroke survivors experiencing hemiparesis can monitor their physical progress using a wearable ring sensor. Using a co-design approach, I engaged with stroke survivors and their caregivers to elicit recommendations for multimodal (speech/visual) feedback of the wearable ring data. Reflexive thematic analysis of the sessions showed that multimodal feedback can potentially lend therapeutic support for stroke survivors. Audio narratives helped to reinforce the visual feedback, and positively framed narrative content that was reflective, motivational, and suggestive was able to support stroke survivors as they navigate their independent recovery journeys.Item “IT’S TOO EXHAUSTING GOOGLING 50 THINGS!”: RECOMMENDATIONS FOR THE LOW-FIDELITY DESIGN OF A CROWDSOURCED HEALTH INFORMATION SYSTEM WITH LOCAL HEALTH-RELATED RESOURCES FOR INDIVIDUALS WHO HAVE CHRONIC HEALTH CONDITIONS(2020) Jindal, Gagan; St. Jean, Beth; Library & Information Services; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Individuals who have chronic health conditions often encounter considerable barriers when trying to find out about local resources in their communities (e.g. libraries, senior centers, fitness classes, nutrition services, faith-based services, support groups, etc.) that can help them better manage their health. In this dissertation, I outline a series of three studies investigating the acceptability and optimal content and design of an online health information system to streamline this information-seeking process with a crowdsourced repository of information of local health resources for this population. I initially conducted 15 in-depth semi-structured interviews to assess the strategies used, and the challenges faced, by these individuals in their attempts to identify these types of local resources in their communities (Chapter 2). The evidence from this first study suggested the potential for the uptake of a novel online health information system that will rely on users to crowdsource and maintain an up-to-date repository of information on relevant local health resources. Based on the results of my first study, I conducted a second study using a card-sorting method to determine the system functions and features, as well as the types of information, individuals who have chronic health conditions felt they would need in this type of system to find a useful local resource and then determine if that local resource would be useful for them (Chapter 3). Based on the results of this card-sorting study, I developed a series of low-fidelity wireframes representing the system features and functions and types of content my study 2 participants wished to see in the proposed crowdsourced health information system (CHIS). I then further refined these low-fidelity wireframes drawing on the findings from my third study in which I garnered direct feedback on the initial wireframes from individuals who have chronic health conditions in a series of participatory design sessions, enabling me to finalize the design recommendations for the proposed CHIS (Chapter 4). Finally, I conclude (Chapter 5) with an overview of the overarching contribution of this research, illuminating a crucial unmet information need and proposing an actionable strategy to better meet this need. I also propose opportunities for future research to further improve the uptake of the proposed CHIS.Item CONTRADICTIONS AND OPPORTUNITIES IN MOBILE CARE MANAGEMENT (“mCare”): AN OBSERVATIONAL ANALYTIC COHORT STUDY(2019) Crowley, Patrick Kenyon; St. Jean, Beth; Butler, Brian; Library & Information Services; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Chronic diseases such as diabetes are among the most widespread, expensive, and preventable of all health problems, accounting for approximately 86 percent of the United States’ $2.7 trillion annual health care expenditures. In the face of such staggering numbers, it is surprising that our current approach to chronic disease care management has remained largely unchanged for decades, where the care team evaluates the patient and related data infrequently and episodically. However, mobile care management (mCare) information system use is growing, whereby individuals with chronic medical conditions such as diabetes are taught to monitor and manage their disease through the use of a mobile application for tracking, education and feedback, along with monitoring of vital signs with “connected” medical devices, and the support of a remote health coach. These mCare systems offer promise, but many unanswered questions exist surrounding their effects on the health and healthcare of the users, and how user individual differences may influence these effects. Informed primarily by the mobile health systems and health behavior literatures, this study provided a deeper understanding of the effects of an mCare platform on health outcomes and health services utilization of chronic disease patients, principally those with diabetes mellitus, and the effects of a user’s social support on these outcomes. This study analyzed administrative claims, device readings, app usage, demographic and social determinant data of 163 diabetic mCare users from a 21-week observation period from mCare initiation, along with a well-matched control group of diabetic non-users, and a supplemental cohort of 127 non-diabetic mCare users with other chronic medical conditions. mCare had a significant positive effect on users’ adherence to physician’s office visits, suggesting greater continuity of care, chronic care management, and a possible reduction in inpatient use (1.2 fewer encounters over 5 months, on average). The findings show that mCare had a significant beneficial effect, on average, towards the cardiovascular health of the users as measured by the change in their diastolic blood pressure (- 2.8 mmHg, - 3.3%) and systolic blood pressure (- 6.7 mmHg, - 4.9%) in the five-month observational period, which is a primary therapeutic target for diabetes care and clinically important. Furthermore, linear mixed models of cardiovascular outcomes uncovered how those mCare users with a moderate degree of social support are likely to achieve greater benefit in from mCare on average relative to those with very high or very low social support in their lives. This additional impact equated to on average a 2.4 mmHg drop (2.9%) in diastolic blood pressure and a 3.9 mmHg (3.1%) drop in systolic blood pressure over the five-month observational period, which is clinically significant. These results provide evidence to support a more precisely tailored future healthcare paradigm beyond the current one-size-fits-all archetype. A primary goal of mCare is triaging emergency department use where appropriate; however, this study found that this did not happen in a significant manner in the treatment group compared to the control group. Furthermore, the study identified specific medical problems where improved mCare design is needed, including processes to prevent hyperglycemia, hypoglycemia and exacerbations of hypertension and pulmonary issues (such as asthma and chronic obstructive pulmonary disease), and a need to assess pain more effectively to foster more appropriate healthcare utilization. Additional training for health coaches, as well as training and development of machine intelligence algorithms to better triage patient problems to appropriate sites of care, are productive directions for future research. mCare designers should seek to better gauge the severity of pain, and develop new sensor technologies to assess emergent issues, especially abdominal pain. mCare vendors should also seek to refine their processes to better manage glucose and respiratory issues to avoid exacerbations, and predict exacerbations earlier to intervene.