Eva-Maria Skoda

Background: Large language model–based chatbots (LLM-CBs) are increasingly used as mental health support tools. Risks and harms are discussed especially for unsupervised use and for vulnerable groups. A naturalistic characterization of the use of LLM-CBs by patients with mental health disorders is currently lacking. Objective: The goal of this cross-sectional study is to investigate demographic and clinical characteristics of users of LLM-CB for mental health and their associations with use behavior. Methods: Across 6 months, qualitative and quantitative data on LLM-CB use for mental health were collected in a consecutive sample of psychotherapy outpatients in routine treatment with at least 1 diagnosed mental health disorder. Results are reported descriptively and supported by exploratory inferential analyses. Results: Within a total of 812 included outpatients who reported on LLM-CB use for mental health, 245 (30.1%) reported having used LLM-CBs for mental health (once: 69/217, 31.8%; weekly: 118/217, 54.4%; and daily: 30/217, 13.8%). Users were more likely to be younger, female, better educated, students, or unemployed. In total, 61.4% (121/197) have used LLM-CBs in self-reported personal crisis situations, of whom, 67.7% (82/121) found the experience unhelpful. LLM-CB use showed a weak, imprecise association with the number of diagnoses in the included diagnosis groups and was higher for anxiety and eating disorder diagnoses and lower for somatoform disorders. Post hoc multivariable analyses suggest that higher use in eating disorders and lower use in somatoform disorders were attenuated after adjusting for age, while higher use for anxiety disorders persisted. LLM-CBs were generally evaluated positively across various dimensions, especially when self-reported personal crisis situation use was experienced as helpful. Qualitative analysis of an open-ended question on use experience shows LLM-CB use across 4 domains reported in the selected subsample: general psychological assistance, disorder-specific assistance, information assistance, and everyday assistance. Conclusions: This naturalistic characterization of LLM-CB use in mental health outpatients implies that harms and benefits of LLM-CB use are associated with demographic and clinical profiles. The results suggest that asking about LLM-CB use may be a useful consideration in psychosocial assessment, particularly regarding crisis use, symptom-reinforcing use, and replacement of human support.

BACKGROUND Chatbots based on large language models (LLM-CBs) are increasingly used as mental health support tools. Risks and harms are discussed especially for unsupervised use and for vulnerable groups. A naturalistic characterization of the use of LLM-CBs by patients affected by mental health disorders is currently lacking. OBJECTIVE The goal of this cross-sectional study is to investigate demographic and clinical characteristics of users of LLM-CB for mental health, and their associations with use behavior. METHODS Across 6 months, qualitative and quantitative data on LLM-CB use for mental health was collected in a consecutive sample of psychotherapy outpatients in routine treatment with at least one diagnosed mental health disorder. Results are reported descriptively. RESULTS Within a total of 812 outpatients, 245 (30.1%) reported having used LLM-CBs for mental health (32% once, 54% weekly, 14% daily). Users were more likely to be younger, female, better educated, students, or unemployed. 61% have used LLM-CBs in crisis situations of whom 68% found the experience unhelpful. LLM-CB use did not differ across number diagnosed mental health disorders but was higher for anxiety and eating disorder diagnoses and lower for somatoform disorders. Diagnosis of depression and anxiety diagnoses showed a positive association with use frequency and somatoform disorders a negative association. LLM-CBs were generally evaluated positively across various dimensions, especially when crisis situation use was experienced as helpful. Qualitative analysis of an open-ended question on use experience shows LLM-CB use across four domains: general psychological assistance, disorder-specific assistance, information assistance, and everyday assistance. CONCLUSIONS This naturalistic characterization of LLM-CB use in mental health outpatients implies that while demographic factors gate initial use, clinical characteristics influence use behavior and experience, suggesting that harms and benefits of LLM-CB use depend on clinical profiles. The results underline the need for systematic clinical assessments of LLM-CB use for clinical practice and targeted research on diagnosis-specific human-LLM-interaction. Clinical practitioners need to be aware that their patients may use LLM-CBs for mental health which may need to be actively included in the treatment.

Despite the evidence for the effectiveness of neurofeedback (NFB) treatment, standardization is scarce, impairing comparability across studies and conditions. Therefore, the acceptance, satisfaction and feasibility of a standardized NFB manual with 14 sessions focused on sensory motor rhythm (SMR) reward and theta and high-beta reduction were evaluated. Methodology: This monocentric, prospective, interventional pilot study in an outpatient setting followed a mixed methods approach. Quantitative data were collected from patients through standardized questionnaires. Qualitative data were derived from interviews with NFB practitioners (NPs) and were analyzed using Mayring’s method of content analysis. Results: The analysis included 24 patients. The mean patient satisfaction (ZUF-8) was 20.81, reflecting moderate satisfaction. The average perceived system usability was 73.48, representing good acceptance. The majority (95.7%) of patients perceived the treatment to be feasible; 78.3% found it useful and would recommend it. Despite reoccurring technical difficulties, NPs perceived the manual to provide security and confidence during sessions. The clear instructions could partially compensate for the heterogeneity in expertise among NPs. Suggested improvements were intensified training on technical aspects and improved patient education. Conclusions: The developed NFB manual is a promising tool for delivering NFB in psychosomatic outpatient settings. It facilitates the establishment of standards, enhances quality assurance and ensures comparability across studies and treatments.

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Hypothyroidism is a common chronic disease that can substantially impair physical and mental well-being and is associated with lower quality of life, a trend that interventions delivered by mHealth applications could ameliorate. Objective: The objective of this study was to evaluate the acceptance and its influencing predictors of mHealth interventions in females affected by hypothyroidism to improve their quality of life. The focus on females reflects the significantly higher prevalence of hypothyroidism in females and their underrepresentation in many prior studies on technology acceptance and mHealth use. Methods: A survey-based, cross-sectional study, which included N = 318 female patients affected by hypothyroidism (assessed via self-reported diagnosis according to ICD-10 criteria, aged 18 years or older) was conducted online between April 2023 and April 2024 in Germany. Participants were recruited via local and online self-help groups, social media platforms and medical practices using flyers. Sociodemographic, health, and eHealth-related data were assessed. To determine acceptance and its drivers and barriers, an extended version of the Unified Theory of Acceptance and Use of Technology (UTAUT) model was applied. Group comparisons (t-tests, ANOVAs) and multiple hierarchical regression analyses were conducted. Only complete data sets were included in the analysis. Results: Acceptance of mHealth applications was high (M = 4.10, SD = 0.91), with 76.1% (n = 242) of the participants reporting high acceptance, 18.6% (n = 59) reporting moderate acceptance, and only 5.3% (n = 17) reporting low acceptance. Significant predictors of acceptance were Place of Residence: Medium-sized city (β = .34, p = .02) and small town or rural area (β = .28, p = .003), Fatigue (β = .54, p < .001), Internet anxiety (β = -.20, p = .002), and the UTAUT predictors Effort expectancy (EE; β = .37, p < .001), Performance expectancy (PE; β = .32, p < .001), and Social influence (SI; β = .20, p < .001). The extended model explained 56.1% of variance in acceptance. Conclusions: The high level of acceptance of mHealth applications observed among females affected by hypothyroidism indicates that mHealth interventions can provide such patients with valuable support to manage the disease in managing their quality of life. Addressing drivers and barriers of acceptance will be crucial for the successful implementation of mHealth interventions in hypothyroidism management, for example by mHealth developers, clinicians, or policymakers. These include intuitive and accessible design (EE), clear communication of app benefits (PE), and fostering healthcare professional support (SI), while addressing barriers such as Internet anxiety. The study also contributes to advancing gender-sensitive mHealth research by applying the UTAUT model to this patient group.

Purpose: The Minimal Documentation System (MIDOS2) is recommended as a systematic screening tool for assessing symptom burden and patient needs in advanced cancer patients. Given the absence of an optimal weighting of individual symptoms and a corresponding cut-off value, this study aims to determine a threshold based on inpatient’s subjective need for palliative support. Additionally, we investigate the correlation between symptom burden and subjective need for palliative support collected through a patient-reported outcome measure (PROM) with survival duration of less or more than one year. Methods: Inpatients diagnosed with advanced solid cancer completed an electronic PROM, which included the MIDOS2 questionnaire among other tools. Differences in symptom burden were analysed between patients expressing subjective need for palliative support and those with survival of less or more than one year using ANOVA, Mann–Whitney-U Test, logistic regression, Pearson and Spearman correlation tests. Cut-off analyses were performed using a ROC curve. Youden-Index, sensitivity, and specificity measures were used as well. Results: Between April 2020 and March 2021, 265 inpatients were included in the study. Using a ROC curve, the MIDOS2 analysis resulted in an Area under the curve (AUC) of 0.732, a corresponding cut-off value of eight points, a sensitivity of 76.36% and a specificity of 62.98% in assessing the subjective need for palliative support. The MIDOS2, with double weighting of the significant symptoms, showed a cut-off value of 14 points, achieving a sensitivity of 78.18% and a specificity of 72.38%. A total of 55 patients (20.8%) expressed a need for support from the palliative care team. This need was independent of the oncological tumour entity and increased among patients with a survival of less than one year. These patients reported significantly poorer physical (p < 0.001) or mental (p < 0.001) condition. Additionally, they reported higher intensities of pain (p = 0.002), depressive symptoms (p < 0.001), weakness (p < 0.001), anxiety (p < 0.001), and tiredness (p < 0.001). Conclusion: Using the established MIDOS2 cut-off value with an adjusted double weighting in our study, a large proportion of inpatients may be accurately referred to SPC based on their subjective need for palliative support. Additionally, subjective reports of poor general, mental, and physical condition, as well as pain, depressive symptoms, weakness, anxiety, and tiredness, increase the subjective need for palliative support, particularly in patients with a survival prognosis of less than one year.

Background: The eHealth Literacy Scale (eHEALS) is a widely used instrument for measuring eHealth literacy (eHL). However, little is known so far about whether the instrument is valid for the assessment of eHL in persons who are affected by the post–COVID-19 condition. This is particularly important as people with the post–COVID-19 condition are frequently affected by false information from the internet. Objective: The objective of our study was to evaluate the validity and reliability of the German Revised eHealth Literacy Scale (GR-eHEALS) in individuals with the post–COVID-19 condition. Methods: A cross-sectional study was conducted from January to May 2022. The self-assessment survey consisted of the GR-eHEALS, health status– and internet use–related variables, sociodemographic data, and (post)–COVID-19–related medical data. Confirmatory factor analysis (CFA), correlational analyses, and tests of measurement invariance were deployed. Results: In total, 330 participants were included in the statistical analyses. CFA revealed that the 2-factor model reached an excellent model fit (comparative fit index=1.00, Tucker–Lewis index=0.99, root mean square error of approximation=0.036, standardized root mean square residual=0.038). Convergent validity was confirmed by significant positive correlations between eHL and knowledge of internet-based health promotion programs, experience in using these programs, and the duration of private internet use. In addition, a significantly negative relationship of eHL with internet anxiety supported convergent validity. Further, significant relationships of eHL with mental health status and internal health locus of control confirmed the criterion validity of the instrument. However, relationships of eHL with physical health status and quality of life could not be confirmed. The 2-factor model was fully measurement invariant regarding gender. Regarding age and educational level, partial measurement invariance was confirmed. The subscales as well as the overall GR-eHEALS reached good-to-excellent reliability (Cronbach α≥.86). Conclusions: The GR-eHEALS is a reliable and largely valid instrument for assessing eHL in individuals with the post–COVID-19 condition. Measurement invariance regarding gender was fully confirmed and allows the interpretation of group differences. Regarding age and educational level, group differences should be interpreted with caution. Given the high likelihood that individuals with the post–COVID-19 condition will be confronted with misinformation on the Internet, eHL is a core competency that is highly relevant in this context, in both research and clinical practice. Therefore, future research should also explore alternative instruments to capture eHL to overcome shortcomings in the validity of the GR-eHEALS.