JMIR Formative Research

Background: Mental health issues are prevalent among men, yet help-seeking remains low due to stigma, masculinity norms and access barriers. Digital mental health (DMH) screening questionnaires offer opportunities for early detection, but their uptake among men is limited. Objective: This study explored the barriers and facilitators influencing mens willingness to use DMH screening questionnaires, with the aim of informing user-centered design that supports early detection and engagement. Methods: This interpretive qualitative study was conducted through semi-structured interviews with 17 purposively sampled Estonian men (aged 20-54) in a highly digitalized context until data saturation was reached. Thematic analysis followed a mixed deductive-inductive approach: deductive codes were derived from theoretical frameworks (Technology Acceptance Model, Health Belief Model, User-Centered Design, Behavioral Design), while inductive themes emerged from participants responses across the three research questions, including their evaluations of four screening questionnaire (PHQ-2, PHQ-9, EEK-2, WHO-5). Results: Key barriers included data privacy fears, distrust of digital solutions, lengthy questionnaires, and poor user experience (UX). Facilitators were anonymity, institutional trust, short (5-10 min) questionnaires, mobile-optimized design, personalized feedback, and clear next steps. As main contribution, four archetypes were identified: Skeptic, Self-Manager, Explorer, and Situational Seeker. They reflected distinct patterns across privacy concerns, institutional trust, user experience preferences, and help-seeking orientations. Skeptics were characterized by low institutional trust, high concern about data misuse, and a preference for anonymous, low-friction interactions, often delaying help-seeking. In contrast, Self-Managers emphasized autonomy, transparency, and evidence-based support, engaging in structured self-monitoring and purposeful help-seeking. Explorers showed openness to experimentation and engagement, particularly when supported by intuitive, interactive, and visually clear UX, while data sharing depended on perceived value. Situational Seekers demonstrated episodic engagement patterns, where trust, data-sharing, and help-seeking were highly context-dependent, preferring fast, low-effort interactions when needed. Conclusions: Mens uptake of DMH screening questionnaires is influenced by a combination of social, psychological, and usability factors. Effective design should integrate anonymity, institutional credibility, and user-centered features to support engagement and early mental health detection. Personalized, actionable feedback with transparency, user control, and clear next-step guidance emerged as key drivers of sustained engagement, while poor usability and lack of meaningful feedback led to disengagement. Importantly, the proposed archetypes capture how these factors co-occur in dynamic, context-dependent user profiles, offering a more actionable alternative to one-size-fits-all and demographic approaches for designing DMH questionnaires tailored to male users.

Background: Concerns about "AI psychosis" have swirled in the media since ChatGPT's release, but few systematic analyses exist. We therefore conducted an electronic health record (EHR) analysis to identify the frequency, clinical characteristics, and quality of AI interactions in patients experiencing psychosis treated in a medical center. Methods: AI keywords (e.g., ChatGPT, AI) were used to search Vanderbilt University Medical Center's EHR from 12/1/2022-4/1/2026. Records were discarded if they were not AI-related or if the primary diagnosis did not include psychosis. Three raters read notes to determine if a patient was experiencing AI psychosis and classified the interactions using 4 a-priori categories (Catalyst, Amplifier, Co-Author, Object) formulated to explain how AI-related negative outcomes emerge. Findings: 73 patients met our criteria. 28 patients were rated as experiencing AI psychosis, 17 had neutral interactions, and 28 expressed delusional content related to AI without documented evidence of conversational AI use. ChatGPT was the matching keyword for 53.6% patients experiencing AI psychosis. The majority of AI psychosis cases were documented after ChatGPT's "4o" model was released in May 2024. Notably, the AI Psychosis group had significantly more patients experiencing a first psychotic episode (60.7%) compared to the other two groups. Amplifier was the most common (64.3%) qualitative rating in the AI Psychosis group. Interpretation: "AI psychosis" is an infrequent but real phenomenon observed in clinical practice. Most affected patients were experiencing their first psychotic episode and presented with AI psychosis following the release of the more sycophantic GPT-4o. Among the affected patients, AI most often exacerbated an existing condition by reinforcing distorted ideas.

Background. Complex cases in specialized pediatric care require consistent adherence to evidence-based clinical pathways and protocols to ensure safe, high-quality, and equitable care. Currently, clinical pathways and supporting documentation are frequently distributed across multiple platforms, leading to fragmentation. Human-centered design principles can guide the development of healthcare technologies that minimize cognitive load and support rapid, efficient access to relevant information in clinical settings. The purpose of this study is to design and evaluate perceived usability of a pediatric cardiac center digital guideline management system that is embedded within the electronic health record leveraging human-centered design. Methods. This study used a mixed-methods usability evaluation to assess a digital guideline management system prototype embedded into clinical workflow. Through human-centered design principles, the prototype provides a centralized digital document library that organizes cardiac-specific clinical pathways, guidelines, procedures, and related resources. A small but diverse sample, encompassing a wide variety of roles and clinical areas within the pediatric cardiac center, was recruited to evaluate the perceived usability of the prototype. Usability was evaluated by stakeholders using the validated System Usability Scale (SUS) with additional optional questions to understand perceptions of the information architecture and clinical value. Results. Preliminary usability testing showed a mean SUS composite score of 76.5, indicating above average usability. Questions related to the complexity of the system and user confidence received high scores across participants. Lower scores were observed for questions related to usage frequency and ability to learn the system very quickly. Conclusion. Leveraging human-centered design when building a digital guideline management system embedded within clinical workflow revealed positive perception from participants. By centralizing access to clinical resources, this prototype can reduce current-state fragmentation. Further evaluation of larger samples is needed to develop a list of future recommendations.

Background: Head and neck cancer (HNC) patients experience clinically significant anxiety and depression in 65-85% of cases during active treatment. Current supportive care lacks personalized, real-time non-pharmacological interventions. Skitii is a novel HRV-adaptive music therapy system that uses continuous RMSSD (root mean square of successive differences) monitoring via a Polar H10 chest sensor to select music in real-time, targeting parasympathetic recovery (RMSSD >=30ms). Methods: This is a prospective, open-label, randomized controlled trial (1:1 allocation) at Yenepoya Medical College Hospital, Mangalore, India. Adults aged 18-75 years with confirmed head and neck cancer (any subsite, Stage I-IV) undergoing radiotherapy and/or chemotherapy with baseline distress (HADS >=8 or NCCN Distress Thermometer >=4) will be enrolled. Participants are randomized to Skitii adaptive music therapy (20-minute sessions, 3 times daily, 3 weeks) or static music therapy control. Skitii uses a two-phase algorithm: Phase 1 (0-2.5 minutes) uses heart rate as a stress proxy for immediate music selection; Phase 2 (2.5-20 minutes) uses RMSSD to adapt music every 2.5 minutes when physiological state changes by >=20%. Primary endpoints are HADS-Anxiety score and resting RMSSD at Week 3. Sample size is 70 (35 per arm), powered at 80% to detect a 2.5-point HADS difference (SD=3.8, alpha=0.05, 15% dropout). Analysis is ANCOVA, intent-to-treat. Discussion: This is the first randomized controlled trial evaluating RMSSD-based adaptive music therapy in cancer patients. The active control design isolates the effect of the adaptive algorithm from music exposure alone. If positive, results will support a scalable, cost-effective supportive care intervention with objective physiological monitoring, and provide the clinical evidence base for CDSCO Class B medical device approval for Skitii in India, with future CE Mark and FDA applications planned. Trial Registration: Clinical Trials Registry - India CTRI CTRI/2025/11/116732

Objectives: Older adults with comorbidities and polypharmacy have disproportionately high risk of hospitalization as well as readmission from adverse drug events (ADEs), of which 28%-71% are preventable (pADEs). This paper introduces an LLM application, CommunicADE, designed to support risk-mitigation of pADE-related readmission for the aforementioned population. We aim to evaluate CommunicADE's technical performance with OpenAI's HealthBench criteria: accuracy, completeness, communication quality, context awareness, and instruction following. Materials and Methods: Our technical validation study used an LLM (KimiK2.5) to simulate interviews between CommunicADE and nine high-fidelity synthetic patients hospitalized and at increased risk for pADE-related readmission (65+ years, comorbidities, 5+ medications). Some pADE risk mechanisms clues were visible to CommunicADE in patient H&Ps, but most mechanisms were solely discoverable in interviews. Two pharmacists evaluated CommunicADE's interview questions and EHR notes with HealthBench-informed variables. Analyzes used descriptive statistics. Results: For 35 mechanisms across 9 patients (avg=3.89 mechanisms/patient), CommunicADE's precision and recall were 0.92 and 0.63, respectively. Hallucinations were absent. Coherence and person-centeredness scored 4.28 and 4.44 on a 5-point scale (5=highest). On average, communication was at a 5th grade level and objective for 78% of patients. Most patient-reported quotes included in notes (92%) supported detected mechanisms. CommunicADE followed all instructions regarding interview length and patient approvals. Discussion: CommunicADE's strongest performance was in accuracy (precision, hallucinations), communication quality (coherence, readability), context awareness (person-centeredness). Completeness (recall) and instruction following (objectivity, pADE mechanism/quote alignment) show room for improvement. Conclusion: Findings suggest technical readiness for a feasibility pilot with real-world patients, and key areas for performance improvement.

BackgroundSmartphone addiction is an emerging public health concern among university students in sub-Saharan Africa. Limited data exist on its prevalence and associated factors in Uganda. ObjectiveThis study aimed to determine the prevalence of smartphone addiction and its associated sociodemographic and economic factors among students at Islamic University in Uganda (IUIU). MethodsA cross-sectional study was conducted among 287 undergraduate students at IUIU Kampala campus. Data were collected using a structured self-administered questionnaire incorporating the Smartphone Addiction Scale Short Version (SAS-SV). Bivariate and multivariate analyses were performed using modified Poisson regression. ResultsThe prevalence of smartphone addiction was 76.7% (95% CI: 71.4 to 81.2). Female students were 1.16 times more likely to be addicted than male students (APR: 1.16; 95% CI: 1.04 to 1.32). Students who spent more time on smartphones than on academic revision were 1.33 times more likely to be addicted (95% CI: 1.11 to 1.61). Those using smartphones for five or more hours daily were 1.32 times more likely to be addicted (95% CI: 1.02 to 1.48). ConclusionSmartphone addiction is highly prevalent at IUIU. Female gender and prolonged daily screen time are significant independent predictors. Targeted digital wellness programmes and institutional policy interventions are urgently needed.

Youth in the United States are experiencing growing mental health challenges, yet many face barriers to accessing timely, affordable, and stigma-free support. At the same time, artificial intelligence (AI) chatbots have become widely available and are increasingly being used by young people as tools for information seeking, coping, and self-reflection. This brief report explores how youth are engaging with AI chatbots for mental health support, drawing from qualitative interviews conducted in New Jersey. Nine semi-structured interviews were completed with participants ages 19-22. Thematic analysis revealed five core themes: (1) generational change, peer communication, and humor as coping and normalization tools; (2) internal and external barriers to self-recognition and help-seeking; (3) AI chatbots as a safe and accessible first step; (4) AI chatbots as a tool for filling information gaps; and (5) limits of AI chatbots and the preference for human connection. These findings indicate that young people see AI chatbots as private, judgment-free starting points for exploring their emotions and seeking early support. However, they also recognize that these tools cannot replace human connection or professional care. For public health, this presents both challenges and opportunities in utilizing the accessibility of AI chatbots while ensuring ethical design, cultural responsiveness, and protections that safeguard youth privacy and equity.

Purpose: Obstructive sleep apnea (OSA) is a treatable chronic condition associated with significant health and societal consequences. Primary care providers (PCPs) often manage OSA with support from sleep specialists but face challenges navigating a complex system of care. By developing a Journey Map, we sought to identify factors influencing primary care OSA management and the associated PCPs' perspectives and emotions. Methods: Twenty-one Calgary-based PCPs were interviewed as part of a study evaluating a primary care management pathway for OSA. We used conventional content analysis, utilizing inductive coding to define journey phases and deductive coding via the Theoretical Domains Framework (TDF) to identify barriers and enablers. These were then mapped onto journey phases for OSA management to create a Journey Map. Results: The Journey Map included five phases of OSA care. PCPs described feeling neutral during the Learning phase and expressed neutral to positive emotions during the Patient Encounter and Diagnosing OSA phases. In contrast, the Initial Treatment and Ongoing Management phases were associated with neutral to negative emotional experiences. Barriers included limited OSA-related training and education, unclear roles among provider groups, and low patient engagement. Enablers included accessible knowledge resources, a shared key role in OSA screening, and availability of sleep testing. Opportunities to enhance primary care OSA management were identified at each step. Conclusion: This study identified several behavioural factors influencing PCP decision-making across the OSA care continuum. The Journey Map illustrates how high diagnostic confidence of PCPs shifts to escalating challenges and negative sentiment during treatment and long-term management of OSA. Keywords: obstructive sleep apnea; primary health care; health service delivery; process assessments; attitude of health personnel

Perioperative patient monitoring requires clinicians to integrate multiple physiological data streams under time pressure and frequent interruptions. Conventional monitors predominantly present vital signs as separate numerical values and waveforms, which must be sequentially interpreted and mentally integrated, imposing substantial cognitive demands. Audible alarms are intended to enhance safety but contribute to alarm fatigue and increased workload. Time spent outside predefined safe ranges for key physiological variables and excessive alarm burden are associated with adverse outcomes, motivating approaches that support earlier detection and improved situation awareness without increasing cognitive load. The Philips Visual Patient Avatar is an avatar-based visualisation technology displayed on the patient monitor that supports clinicians' situation awareness by integrating multiple vital signs and sensor states into a single animated virtual patient, while retaining conventional numerical displays. Although laboratory, simulation and qualitative studies suggest benefits of avatar-based monitoring, its impact on objective monitoring outcomes has not been systematically quantified.

Cardiac rehabilitation is critical for secondary prevention, yet long-term adherence remains low. We present CUOREMA, a new personalized mobile health system integrating self-monitoring diaries, wearable data, virtual coaching, and reinforcement learning-enhanced adaptive interventions to support lifestyle change during and after outpatient cardiac rehabilitation. In a six-month two-center feasibility study (N = 53, Switzerland and France), we evaluated usability, engagement patterns, and preliminary health-related outcomes. Attrition was high: only 19\% of participants used the app on more than 100 days, and questionnaire response rates declined from 55\% at baseline to 13\% at six months. Despite these limitations, exploratory data-driven analysis revealed three distinct engagement clusters (dropout, sporadic, and consistent), which were further supported by matching patterns in app component usage, medication diary adoption, and smartwatch wearing time. Engagement clusters were not associated with demographic factors; instead, psychological themes of patients' personal goals suggested that intrinsic motivation characterized sustained users, whereas extrinsic motivation predominated among early dropouts. User experience was rated positively, and validated questionnaire scores showed no deterioration over time. One center demonstrated a statistically significant improvement in 6-minute walking test performance, though the study was not powered to detect clinical outcomes and selective dropout cannot be ruled out. These findings highlight engagement variability as a central challenge in digital cardiac rehabilitation and suggest that tailoring interventions to individual motivational profiles may improve long-term adherence.

Background: Statins are key to preventing atherosclerotic cardiovascular disease and lowering low-density lipoprotein cholesterol and cardiovascular events. However, skepticism regarding their safety and value persists and is increasingly influenced by social media. TikTok has emerged as a major source of health information, but its content varies in quality and accuracy. This study evaluated the quality, attitudes, misinformation, and engagement of statin-related content on TikTok. Methods: Public TikTok videos were collected using predefined search terms and coded by creator type, thematic content, and overall attitude. Video quality was assessed using the DISCERN instrument, the Patient Education Materials Assessment Tool for Audiovisual Materials, and the Global Quality Score. False or misleading claims were independently reviewed by two cardiology fellows. Associations between engagement and quality were also examined. Results: Of 1,349 screened videos, 258 met inclusion criteria. Most were educational (91.0%), with non-physician healthcare providers (34.5%) as the largest creator group. Risks or negative effects were discussed more often than benefits (63.2% vs 42.2%), and 39.5% contained at least one false or misleading claim, most often from complementary and alternative medicine providers and wellness promoters. Quality differed by creator type across all instruments, with physician-created content scoring highest. Video popularity showed minimal association with informational quality. Conclusion: Statin-related TikTok content frequently emphasizes harms, often contains misinformation, and varies substantially in quality by creator type. Greater involvement of healthcare professionals on social media may help improve digital health literacy and counter misleading information about statin therapy.

While wearable activity trackers (WATs) are widely used in the present day, with device ownership increasing, some individuals subsequently discontinue device use. Existing research primarily examines the initiation and maintenance of device use, with less focus on device discontinuation. Examining this phenomenon can provide valuable insight into human-computer interactions and habit reversal. Therefore, the current study examined the perceived emotional and behavioural impact of WAT discontinuation, alongside reasons for this action in former WAT users. Fifteen former WAT users (9 female, aged 23 to 56 years) who reported either full or partial device discontinuation were interviewed. Three themes and nine sub-themes were identified which detailed the impacts of device discontinuation. Participants reported a mindset shift around ones body image, exercise performance and exercise motivation. Device discontinuation removed numerical feedback provision which led to participants gaining bodily intuition and a sense of freedom. However, discontinuation also resulted in short-term negative emotions including frustration around the loss of external praise and envy in current WAT users. Current findings hold important implications around digital safety from user perspective, highlighting the need for guidance around healthy WAT use and vulnerable user profiles. More broadly, findings also raise the need for physical activity promotion whilst protecting individuals well-being.

Background: Large language models (LLMs) are increasingly used in telehealth, but their safety in antibiotic prescribing remains uncertain, particularly in the presence of patient misinformation. Methods: A cross-sectional analytical study evaluated 5,000 responses from five chatbot models using 1,000 primary-care vignettes of mild infections. Guideline adherence, overprescribing, misinformation effects, and safety behaviors were assessed. Inappropriate prescriptions were classified using the WHO AWaRe framework. Results: Overall, 76.2% of responses were guideline-concordant, while 6.6% showed unprompted overprescribing and 17.2% were influenced by misinformation. Some models were more vulnerable to misinformation than others. Although most responses correctly noted that antibiotics do not treat viral infections, fewer advised consulting a doctor, and warnings against self-medication were rare. Many inappropriate prescriptions involved broad-spectrum antibiotics. Conclusion: LLMs show potential in telehealth but remain prone to misinformation and inappropriate prescribing. Stronger guideline integration and clinical oversight are necessary to ensure safe use. Keywords: antimicrobial stewardship; large language models; telehealth; antibiotic prescribing; misinformation; clinical safety

Background: Poor glycemic control, a contributor to the development of diabetes related complications among patients with diabetes mellitus, remains a public health challenge in low- and middle-income countries. In Uganda, limited evidence exists on predictors of poor glycemic control among diabetic patients attending peri-urban primary healthcare facilities. The study assessed predictors of poor glycemic control among adults attending the diabetic clinic at Kasangati Health Centre IV in Wakiso district. Methods: We conducted cross-sectional study among 283 diabetic patients attending Kasangati Health Centre IV between March and April 2025. Data were collected using interviewer-administered structured questionnaires and data abstraction tools. Poor glycemic control was defined as glycated hemoglobin (HbA1c) levels [≥]7%. Modified Poisson regression with robust standard errors was used to determine factors associated with poor glycemic control. Adjusted prevalence ratios (aPRs) with 95% confidence intervals (CIs) were reported. Results: Overall, 67.8% of the participants had poor glycemic control. Poor glycemic control was significantly associated with older age, low income status (aPR: 1.4, 95%CI: 1.24-1.58), use of multiple anti-diabetic medications, non-adherence to regular follow-up (aPR: 1.5, 95%CI: 1.33-1.65), medication side effects (aPR: 1.2, 95%CI: 1.01-1.32), physical inactivity (aPR: 1.1, 95%CI: 1.05-1.21), non-adherence to recommended dietary plans (aPR: 1.1, 95%CI: 1.02-1.22), perceived treatment regimen complexity (aPR: 1.2, 95%CI: 1.12-1.34), stress (aPR: 1.1, 95%CI: 1.08-1.20), lack of peer support groups (aPR: 1.2, 95%CI: 1.08-1.23), and high costs of accessing care (aPR: 1.2, 95%CI: 1.17-1.33). Conclusion: Almost two-thirds of the diabetic patients suffered from poor glycemic control which was determined by various socio-economic, behavioral, clinical and health system factors. Enhancing adherence counseling, encouraging healthy lifestyles, adopting age-based supportive healthcare approaches, better psychosocial support and reduction of cost barriers in accessing diabetic healthcare could improve the glycemic status of diabetic patients in peri-urban primary healthcare settings.

Background Wearable-based studies have largely examined activity and sleep using static summaries or single time windows, potentially missing how chronic patterns and recent behavioral changes jointly relate to depressive symptom severity. We evaluated whether combining long-term habitual behavior with short-term dynamics improves characterization of moderate-to-severe depressive symptoms. Methods We analyzed Fitbit data from All of Us participants with Patient Health Questionnaire-9 (PHQ-9) assessments, defining moderate-to-severe symptoms as PHQ-9 [≥] 10 (N=248). Logistic regression evaluated long-term measures (past-year step count and awake time after sleep onset) and short-term dynamics (30-day step decline and 30-day sleep duration variability), adjusting for demographics. Performance was assessed via repeated stratified 10-fold cross-validation. Results Thirty percent of participants (n = 74) had moderate-to-severe depressive symptoms. Higher long-term step count was associated with lower odds of elevated symptoms (OR = 0.75 per 1,000 steps/day), greater awake time after sleep onset with higher odds (OR = 1.27 per 1%), a 30-day step decline with higher odds (OR = 2.70), and greater 30-day sleep variability with higher odds (OR = 1.07 per percentage point). Short-term dynamics provided complementary information beyond long-term measures alone. The combined model achieved the highest discrimination (area under the curve [AUC] = 0.80 vs. 0.73 demographics-only), though findings should be interpreted as exploratory given the modest sample size. Limitations The sample was modest in size (N = 248), PHQ-9 reflects symptom severity rather than clinical diagnosis, causal inference is not possible given the cross-sectional outcome assessment, and Fitbit users may not represent broader populations. Conclusions Long-term behavioral patterns and short-term changes in activity and sleep were associated with depressive symptom severity, supporting wearable-derived measures as potential adjunctive markers in mental health research.

Objective: Several endotypes contribute to the development of Obstructive Sleep Apnea (OSA). However, efforts to measure these endotypes have been challenging. In this paper, we propose a new method that overcomes some of these challenges. Methods: To test the feasibility of this new method, data from the Sleep Heart Health Study (SHHS) were analyzed and two oxygen-based endotypes were identified and plotted on a graphical model: the steady-state SpO2 and the SpO2 arousal threshold. The first is the oxygen saturation that would occur during sleep if there were no arousals, and it is a measure of upper airway collapsibility (a more collapsible airway produces a lower SpO2). The latter is the oxygen saturation that triggers arousals. These endotypes were validated by assessing their ability to detect positional and state-related changes in airway collapsibility and arousal threshold. Results: The study showed that it was feasible to measure oxygen-based endotypes in 95% of SHHS participants. As expected, steady-state SpO2 was lower during supine vs. non-supine sleep, as well as during REM vs. NREM sleep. Also, the SpO2 arousal threshold was similar between supine and non-supine sleep. However, SpO2 arousal threshold was not lower in REM sleep vs. NREM sleep. Therefore, in 3 of the 4 conditions, the oxygen-based endotypes moved in the expected direction due to positional or sleep state changes. Conclusion: Although further validation experiments are required, this study indicates that OSA endotyping using the pulse oximetry signal is feasible. The oxygen-based endotypes could be used to aid therapeutic decision making.

Migraine detection and sentiment analysis in healthcare have become increasingly important, particularly with the rise of social media platforms like Twitter, where users often share their personal health experiences. This study presents MASHA (Multi-Agent System for Healthcare Sentiment Analysis), an artificial intelligence (AI)-driven framework that integrates multiple machine learning (ML) models for sentiment analysis of Arabic tweets related to migraines. The system leverages a multi-agent architecture to handle tasks such as data acquisition, pre-processing, model training and real-time decision-making. Key ML models, including Support Vector Machines (SVM), Naive Bayes (NB) and Logistic Regression (LR), are integrated using ensemble techniques, leading to improved classification performance. Experiments conducted on a dataset of Arabic tweets demonstrate that MASHA outperforms traditional methods, achieving an accuracy of 90.0% and an F1-score of 89.46%. Moreover, the system's scalability and flexibility make it suitable for real-time public health monitoring, offering valuable insights into patient experiences and public sentiment regarding healthcare services. MASHA's adaptability suggests its potential application for analysing other healthcare-related conditions, reinforcing the system's scalability and broader relevance. Future work will focus on incorporating deep learning (DL) models and expanding the dataset with content from additional social media platform.

Objective: Stigmatizing language in the electronic health record (EHR) has been associated with adverse patient experience in substance use disorder care, including opioid use disorder (OUD). This study evaluated a privacy-preserving, locally-deployed large language model as a method to detect stigmatizing language documentation in OUD patients with patient-directed discharge (PDD). Methods: A retrospective cohort study of 477 inpatient admissions from the MIMIC-IV database with a diagnosis of opioid use disorder were classified using a locally deployed Gemma-4-31b-it-bf16 model and predefined 140 term lexicon to identify stigmatizing language in clinical documentation. Results: Analysis of clinical documentation showed stigmatizing language was present in 84.1% (190/226) in the PDD cohort vs 62.2% (156/251) in the non-PDD cohort, with an unadjusted odds ratio of 3.21 (95% CI 2.07-4.98; p < 0.0001). After adjustment for age, sex, insurance status, marital status, and race, PDD discharge remained an independent predictor of stigmatizing documentation (aOR 2.24, 95% CI 1.40-3.59; p < 0.0001). Further analysis of stigma intensity showed higher stigmatizing markers in the PDD cohort vs the non-PDD cohort (2.85 {+/-} 2.39 vs 2.02 {+/-} 2.44; p < 0.0001). Discussion and Conclusion: Stigmatizing language is detected with increased frequency and prevalence in clinical documentation of OUD patients that initiate PDD compared to those that adhere to standard discharge processes. A locally deployed large language model (LLM) offers a scalable, privacy-preserving method to audit clinical documentation for stigmatizing language.

Ambient AI documentation systems rely on automatic speech recognition to transcribe patient-provider conversations before generating clinical notes. However, little empirical evidence exists on how these systems perform in mixed-language clinical encounters. We conducted a mixed-method heuristic evaluation of an ambient AI documentation tool using 24 reenacted primary care conversations involving Spanish-English and Mandarin-English code-switching. Quantitative analyses measured mixed error rate (MER) and code-switching detection. Overall MER was low, with a median of 4% and less variation in Spanish-English conversations, and 9% in Mandarin-English conversations, but with outliers reaching 67%. The system generally detected language switches reliably, although deletions occurred frequently in Mandarin-English transcripts at switch points. Qualitative analysis revealed transcription errors related to phonetic similarity, automatic language translation, clinical terminology recognition, and language-specific challenges. These findings highlight considerations for improving ambient AI clinical documentation systems to support multilingual providers in delivering care for linguistically diverse populations.

Background: Generalized anxiety disorder (GAD) is associated with substantial psychological burden, autonomic dysregulation, and limitations of existing pharmacological and psychotherapeutic treatments. Transcutaneous auricular vagus nerve stimulation (taVNS) has emerged as a promising non-invasive neuromodulation approach, but evidence regarding home-based application in GAD remains limited. Objective: To evaluate the feasibility, safety, and preliminary clinical and physiological outcomes of a home-based taVNS intervention in adults with psychologist-confirmed moderate-to-severe GAD. Methods: In this prospective single-arm feasibility study, 48 participants initiated a 4-week home-based taVNS intervention consisting of two daily stimulation sessions performed five days per week. Clinical assessments were conducted at baseline, Week 2, Week 4, and follow-up visits at Weeks 6 and 8. Ambulatory electrocardiographic monitoring was performed before treatment initiation, at Week 2, and at the end of treatment to assess heart rate variability (HRV) using the root mean square of successive differences (RMSSD). Primary outcomes included feasibility, safety, adherence, and change in clinician-rated anxiety severity (HAM-A). Results: Thirty-four participants completed the study and were included in the primary analyses. HAM-A scores decreased significantly from baseline to Week 4 ([EMD] -6.9, 95% CI -10.4 to -3.4, p = 0.001), with partial maintenance during follow-up. Improvements were also observed in Beck Anxiety Inventory scores, whereas changes in GAD-7, perceived stress, depressive symptoms, and sleep quality were not statistically significant. RMSSD increased significantly from baseline to Week 4 (EMD 6.7 ms, 95% CI 2.1-11.3, p = 0.009). Greater increases in RMSSD were associated with larger reductions in HAM-A (R^2 = 0.18, p = 0.031) and BAI scores (R^2 = 0.21, p = 0.019). No serious adverse events occurred. Mean adherence was 79.8%, and 73.5% of participants completed at least 70% of prescribed stimulation sessions. Conclusions: Home-based taVNS was feasible and generally well tolerated in adults with moderate-to-severe GAD. Preliminary improvements in clinician-rated anxiety severity and autonomic physiological measures were observed; however, the single-arm design precludes causal inference. These findings support further evaluation of home-based taVNS in adequately powered randomized sham-controlled trials.