ABSTRACT
BACKGROUND: Before the COVID-19 pandemic, adolescents with type 1 diabetes (T1D) had already experienced far greater rates of psychological distress than their peers. With the pandemic further challenging mental health and increasing the barriers to maintaining optimal diabetes self-management, it is vital that this population has access to remotely deliverable, evidence-based interventions to improve psychological and diabetes outcomes. Chatbots, defined as digital conversational agents, offer these unique advantages, as well as the ability to engage in empathetic and personalized conversations 24-7. Building on previous work developing a self-compassion program for adolescents with T1D, a self-compassion chatbot (COMPASS) was developed for adolescents with T1D to address these concerns. However, the acceptability and potential clinical usability of a chatbot to deliver self-compassion coping tools to adolescents with T1D remained unknown. OBJECTIVE: This qualitative study was designed to evaluate the acceptability and potential clinical utility of COMPASS among adolescents aged 12 to 16 years with T1D and diabetes health care professionals. METHODS: Potential adolescent participants were recruited from previous participant lists, and on the web and in-clinic study flyers, whereas health care professionals were recruited via clinic emails and from diabetes research special interest groups. Qualitative Zoom (Zoom Video Communications, Inc) interviews exploring views on COMPASS were conducted with 19 adolescents (in 4 focus groups) and 11 diabetes health care professionals (in 2 focus groups and 6 individual interviews) from March 2022 to April 2022. Transcripts were analyzed using directed content analysis to examine the features and content of greatest importance to both groups. RESULTS: Adolescents were broadly representative of the youth population living with T1D in Aotearoa (11/19, 58% female; 13/19, 68% Aotearoa New Zealand European; and 2/19, 11% Maori). Health care professionals represented a range of disciplines, including diabetes nurse specialists (3/11, 27%), health psychologists (3/11, 27%), dieticians (3/11, 27%), and endocrinologists (2/11, 18%). The findings offer insight into what adolescents with T1D and their health care professionals see as the shared advantages of COMPASS and desired future additions, such as personalization (mentioned by all 19 adolescents), self-management support (mentioned by 13/19, 68% of adolescents), clinical utility (mentioned by all 11 health care professionals), and breadth and flexibility of tools (mentioned by 10/11, 91% of health care professionals). CONCLUSIONS: Early data suggest that COMPASS is acceptable, is relevant to common difficulties, and has clinical utility during the COVID-19 pandemic. However, shared desired features among both groups, including problem-solving and integration with diabetes technology to support self-management; creating a safe peer-to-peer sense of community; and broadening the representation of cultures, lived experience stories, and diabetes challenges, could further improve the potential of the chatbot. On the basis of these findings, COMPASS is currently being improved to be tested in a feasibility study.
ABSTRACT
OBJECTIVES: Symptom misattribution is a central process in the nocebo effect but it is not accurately assessed in current side effect measures. We have developed a new measure, the Side Effect Attribution Scale (SEAS), which examines the degree to which people believe their symptoms are treatment side effects. METHODS: The SEAS was tested in three New Zealand studies: a vaccination sample (n = 225), patients with gout or rheumatoid arthritis (n = 102), and patients switching to a generic medicine (n = 69). The internal reliability of the scale was examined using Cronbach's alpha. To assess validity, the Side Effect Attribution Total Score and Side Effect Attribution Binary Score were related to a number of psychological measures associated with side effect reporting. RESULTS: The scale showed good internal reliability across the three studies, with Cronbach alphas ranging from 0.840 to 0.943. Analysis of the effect sizes showed that the Attribution Total Score was generally more strongly associated with nocebo responding than Attribution Binary Score. Participants had greater Side Effect Attribution Total Scores if they had higher expectations for vaccination side effects (r = 0.18, p = .028), more worry about future vaccine effects (r = 0.16, p = .046), a higher perceived sensitivity to medicines (r = 0.50, p < .001), greater anxiety (r = 0.25, p = .016), greater intentional non-adherence (r = 0.30, p = .003), greater medicine information seeking (r = 0.26, p = .010), lower trust in pharmaceutical agencies (r = -0.29, p = .026), and lower medicine efficacy beliefs (r = -0.46, p < .001). CONCLUSIONS: The SEAS provides a more nuanced assessment of symptom attribution beliefs. It appears to be more sensitive measure than just a side effect total, as it is associated with a greater number of relevant psychological variables. Future research should examine the scale in other populations and settings.
