JMIR Formative Research最新文献

Background: Mobile health tools that use the Experience Sampling Method (ESM) appear to be a promising tool to streamline and improve the treatment of substance use disorders. However, patient involvement in the development of these tools is uncommon, and research on the preferences of people being treated for substance use disorders has been scarce. In the scope of the European Union IMMERSE (Implementing Mobile Mental health Recording Strategy for Europe) consortium, an ESM-based tool for Digital Mobile Mental Health (DMMH) was first codeveloped and later tested in 4 European countries.

Objective: This study aimed to achieve an understanding of preferences for features of DMMH among mental health service users with problem substance use.

Methods: In 4 European countries, service users were recruited for a semistructured qualitative interview, which started with the presentation of a prototype of the DMMH. Data analysis was performed through directed qualitative content analysis.

Results: The analytical sample consisted of 12 (5 female, 6 male, and 1 nonbinary person) participants with problem substance use aged 18-50 years. There were 4 participants from Slovakia, 3 from Belgium, 4 from Germany, and 1 from Scotland. Patient preferences were classified into 7 categories: notifications, questions, user interface, functionality changes, visualizations, sharing data with clinicians, and sharing data with others. The proposed intensive notification schedule was deemed acceptable by service users as long as the questionnaire is short. Participants expressed a preference for open-text response options, Ecological Momentary Interventions, and options for individual customization of several elements of the tool. Data visualization was considered an important aid for communication with clinicians, with whom all participants wanted to share data obtained with DMMH. The possibility of sharing data with other people depended on the quality of the relationship with them.

Conclusions: In the development of ESM-based mobile health tools for people with problem substance use, their preferences for content, functionality, and appearance of the tools should be considered so they match patients' treatment needs.

Background: Online harm affects many people and has been associated with self-harm and suicidal ideation. Although there is an emerging body of evidence that addressing adverse online experiences should be part of the support offered to people who are at risk of self-harm and suicide, there has been little guidance to date on how this support might be provided and how safe conversations can be had on the subject. A UK charity dedicated to offering emotional support to anyone experiencing mental discomfort, having difficulty coping, or being at risk of suicide developed a digital intervention, the Online Safety Hub (the Hub), to address this shortfall.

Objective: The study aimed to evaluate the impact of the Hub on practitioners (people who provide support) and people with lived experiences of suicide and self-harm and to determine what learning environment is best suited to increase and maintain learning in the context of the Hub.

Methods: A sequential explanatory mixed methods evaluation comprised a rapid literature review, data collected from people with lived experience (n=6) and practitioners through an analysis of the Hub's activity data, 2 surveys (survey 1: n=45; survey 2: n=368), interviews (n=9), and focus groups (n=7). Surveys were analyzed for descriptive purposes only, and the interview and focus group analyses comprised coding of data and thematic analysis. The study design was informed by a panel of people with lived experience of online harm resulting in either self-harm and/or suicidal ideation.

Results: Initially, the evaluation found limited uptake of the Hub. Engagement with the Hub was impeded by a lack of clarity on the part of practitioners as to whether they were the intended audience. The evaluation process prompted the charity to design and deliver webinars to facilitate uptake of the Hub. Practitioners who engaged with the Hub via webinars found the content useful and were able to consider incorporating their learning into practice. The webinars offered a more social learning experience than individual engagement with the Hub, providing a community of practice for people with common interests across diverse organizational settings. Opportunities for shared learning and the supportive nature of the community of practice were valued when learning about the sensitive and difficult topic of online harm in relation to self-harm and suicide. The Hub contributed to awareness-raising and shared learning.

Conclusions: Online resources alone may not be sufficient for an intervention to effectively raise awareness and change practice. Social learning facilitated through communities of practice can enhance engagement, uptake, and learning.

Background: The evolution of language models, particularly large language models, has introduced transformative potential for psychological assessment, challenging traditional rating scale methods that have dominated clinical practice for over a century.

Objective: This study aimed to develop and validate an automated assessment paradigm that integrates natural language processing with conventional measurement tools to assess depressive symptoms, exploring its feasibility as a novel approach in psychological evaluation.

Methods: A cohort of 115 participants, including 28 (24.3%) individuals diagnosed with depression, completed the Beck Depression Inventory Fast Screen via a custom ChatGPT interface (BDI-FS-GPT) and the Chinese version of the Patient Health Questionnaire-9 (PHQ-9). Statistical analyses included the Spearman correlation (PHQ-9 vs BDI-FS-GPT scores), Cohen κ (diagnostic agreement), and area under the curve (AUC) evaluation.

Results: Spearman analysis revealed a moderate correlation between PHQ-9 and BDI-FS-GPT scores. The Cohen κ indicated moderate diagnostic agreement between the PHQ-9 and the BDI-FS-GPT (κ=0.43; 76.5% agreement), substantial agreement between the BDI-FS-GPT and the clinical diagnosis (κ=0.72; 88.7% agreement), and moderate agreement between the PHQ-9 and the clinical diagnosis (κ=0.55; 71.4% agreement). The BDI-FS-GPT demonstrated excellent diagnostic accuracy (AUC=0.953) at a cutoff of 3, detecting 89.3% of participants with depression with an 11.5% false-positive rate compared to the PHQ-9 (AUC=0.859) at a cutoff of 5 (sensitivity=71.4%; false-positive rate=13.8%). Participants also reported significantly higher satisfaction with the automated assessment compared to the traditional scale (P=.02).

Conclusions: The automated assessment paradigm framework combines the interactivity and personalization of natural language processing-powered tools with the psychometric rigor of traditional scales, suggesting a preliminary feasibility paradigm for future psychological assessment. Its ability to enhance engagement while maintaining reliability and validity provides encouraging evidence, warranting validation in larger and more diverse studies as large language model technology advances.

International registered report identifier (irrid): RR2-10.1101/2024.07.19.24310543.

Background: Uncontrolled diabetes contributes to serious comorbidities and mortality. Effective self-management can improve outcomes, though barriers such as limited education and support often prevent patients from engaging in such behaviors. Automated texting systems show promise to deliver diabetes self-management education as they are accessible and scalable. Furthermore, customizing these systems may further enhance patient engagement compared to standard, one-size-fits-all approaches. However, such customization is more resource-intensive, and it remains unclear whether the added effort meaningfully enhances diabetes self-management and outcomes.

Objective: This study aimed to describe the development of 2 versions of an automated texting system intervention for diabetes self-management: (1) a standard, education-only intervention (Diabetes Self-Management Support; DSMS) and (2) an interactive, customizable intervention (Diabetes Self-Management Support + Interactive and Customizable Messages; DSMS+).

Methods: Two versions of an automated texting system intervention were developed using a participatory design approach that incorporated input from veterans and expert clinicians. Message content was refined through feedback from a multidisciplinary team, veteran coinvestigators, national surveys, interviews, clinical expert panel reviews, and beta testing. Surveys were mailed to 1000 potential participants, oversampling rural, low-income, minority, and female participants. Respondents rated message relevance and provided preferences for content, timing, and frequency. Interviews provided customization preferences. A clinical expert panel reviewed all messages for safety and appropriateness. Beta testing informed final refinements.

Results: Ninety-two surveys were completed (9.2% response rate). Respondents rated 62% of the messages as personally relevant and 61% confidence-enhancing. Interviews with 23 respondents revealed a preference for 1-2 texts per day, emphasizing topics such as healthy eating and weight management. The clinical expert panel reviewed 536 messages, flagging 81 for revision. Beta testing confirmed feasibility and informed refinements to clarity and timing. The 2 resulting interventions were built in the US Department of Veterans Affairs' automated texting system, Annie.

Conclusions: Two text messaging interventions, DSMS and DSMS+, were developed to support diabetes self-management among US veterans. DSMS delivers standard educational content, while DSMS+ incorporates interactive features and personalization. The subsequent clinical trial will assess whether customization enhances engagement and improves diabetes outcomes, providing insights into the potential of tailored mobile health interventions for chronic disease management.

Background: Methadone is a first-line treatment for opioid use disorder, which is delivered in federally regulated opioid treatment programs (OTPs). Federal policies require directly observed dosing of methadone followed by graduated provision of nonobserved doses to take at home (ie, "take-home" dosing) after demonstrated stability is achieved. Policy changes since the COVID-19 pandemic have greatly expanded take-home dosing. Video directly observed treatment (video DOT) is an approach in which patients submit videos of themselves taking medications, which are asynchronously reviewed to verify adherence.

Objective: In preparation for an implementation trial evaluating the adoption of video DOT in OTP settings, we conducted a rapid needs assessment with multidisciplinary stakeholders to assess acceptability, perceived benefits, and needed support for video DOT to monitor take-home methadone dosing.

Methods: In our rapid needs assessment, we explored perspectives of multidisciplinary stakeholders (N=20) at 3 clinical sites within a single OTP in western Washington state. Trained qualitative researchers took ethnographic field notes during meetings with organizational leadership and in-person site visits with clinical and administrative staff. Field notes were analyzed via a team-based rapid assessment process using coding templates informed by the Consolidated Framework for Implementation Research. Summaries of qualitative data were iteratively reviewed by the study team and further confirmed with site stakeholders.

Results: Stakeholders included leadership (n=6, 30%), medical providers (n=4, 20%), substance use disorder counselors (n=7, 35%), and clinic managers and support staff (n=3, 15%). Stakeholders perceived that video DOT could lessen the barriers patients face, including travel burden (eg, time and cost) and stigma. They also identified that video DOT could have important impacts on early care retention, given expansions of take-home dosing. However, stakeholders anticipated an added burden for clinical staff and emphasized the need for implementation supports that would limit burden, such as additional staff support for video submission review and clear communication pathways when video submissions require additional clinical input.

Conclusions: A rapid needs assessment of OTP sites for a future implementation study suggested that stakeholders saw potential benefits for patients receiving video DOT, but there were concerns that this would add to their work burden. Learnings informed the subsequent tailoring of clinical use cases and implementation supports.

When given a sample of 100 emergency department discharge instructions, Claude Sonnet, a large language model, produced accurate Spanish translations as evaluated by Spanish-speaking physicians and medical interpreters.