bims-librar 2026-07-12 papers

bims-librar

Biomed News

on Biomedical librarianship

Issue of 2026–07–12
43 papers selected by
Thomas Krichel, Open Library Society

MedNLP-Hub: A knowledgebase platform for biomedical NLP tool discovery in clinical informatics.
eGoT: enhanced graph-of-thoughts for multi-hop knowledge retrieval and hypothesis generation in biomedicine.
PixelDeck: A local-first media library manager for biomedical imaging.
Large language models vs. conventional search in identifying delirium guidelines for a scoping review.
Who Is Responsible When AI Gets Cancer Information Wrong? Implications for Patient Education.
Ongoing Changes in Web Searches for Health-Related Information.
Seeing is Believing? Credibility and Behavioral Intentions Toward Evidence-Based and Misleading Health Information.
Boosting Media Literacy Using Lateral Reading and Online Search Interventions.
Perceived cognitive slowing and online health information seeking among older adults: motivational beliefs as pathways and community IT culture as a buffer.
Online health information seeking preferences among pregnant women: a discrete choice experiment.
The association between digital health information acquisition and health-seeking behavior among university students.
Nutrition information-seeking behavior among patients recovering at home after esophageal cancer surgery: a qualitative study.
From information overload to health information avoidance among Chinese adults aged 50 years and above: the association of self-perceptions of aging and techno-exhaustion.
Correction to "Navigating Knowledge: Effects of State Curiosity on Children's Word Learning and Information Seeking".
Supporting patient understanding of cervical and ovarian cancer: how well does AI perform?
Challenges of patient-facing generative artificial intelligence in hypertension care: A cross-platform evaluation of the quality, readability, and actionability of LLM-Generated patient education materials.
Accuracy and readability analysis of ChatGPT responses to frequently asked questions about olecranon fractures.
Assessing Readability and DISCERN Quality of Osteoporosis Education Materials Generated by ChatGPT and Deepseek for Diverse Health Literacy Levels: A Cross-Sectional Study.
Information quality, readability, and empathy of AI-generated public mental health information: a comparative evaluation of eight large language models.
A Comparative Cross-sectional Study Assessing the Readability of ChatGPT and UpToDate Content on Multiple Endocrine Neoplasia Syndromes.
Artificial intelligence-based patient information in rotator cuff injuries: A cross-sectional comparative study of ChatGPT and DeepSeek models.
Clinical safety, content coverage, and patient-centered language of AI responses to periodontal complaint-based queries: a comparative study of four large language models.
Reliability and readability of adenoid hypertrophy information generated by five publicly accessible LLM chatbots: A default-setting snapshot study.
Evaluation of a Retrieval-Augmented Generation-Based Large Language Model for Evidence-Based Herb and Supplement Information in Cancer Care.
Evaluating the quality, reliability and readability of digital and artificial intelligence resources for adults with cancer who have significant caregiving responsibilities for children.
Comparing the Performance of Conversational Artificial Intelligence Models in Answering Patient Questions About Dental Implants.
Trust and verification in AI-enabled physician chatbots for chronic disease management: evidence from digital health behavior.
Artificial Intelligence-Generated Versus Professional Society Patient Education Materials in Gastroenterology, Surgery, Ophthalmology, and Anesthesiology: A Comparative Analysis of Readability and Health Literacy Metrics.
Evaluation of Online Educational Content on Indocyanine Green Fluorescence for Flap Perfusion: Assessment of Quality, Reliability, and Engagement Metrics.
Readability of Online Chronic Traumatic Encephalopathy (CTE) Information: Health Literacy Implications for Parents Navigating Youth Tackle Football Decisions.
The Online Landscape of Brazilian Butt Lift Information: Readability, Quality, and Traffic Analysis.
Online patient education: results of a content analysis of disempowering language in patient education materials about type 1 diabetes.
Brief health information leaflets in general practice: A cross-sectional survey on usage, common topics and sources in Austria.
Emergency Surgical Cricothyrotomy Videos Online: Reliability and Training Quality Evaluation.
Navigating digital resources in dental education: a cross-sectional comparison of the educational quality of rubber dam videos on Youtube and Instagram.
Educational quality and viewer engagement of Youtube videos on ceramic implants: a cross-sectional study.
Is YouTube an Accurate and/or Comprehensive Source of Information on Monosymptomatic Nocturnal Enuresis?
Content Gaps and Informational Quality of TikTok and Bilibili Videos on Pit and Fissure Sealing in China: A Cross-Sectional Study.
Quality and reliability of cardiac rehabilitation-related short Chinese videos on Douyin and Bilibili: a cross-sectional content analysis.
Quality, content, and reliability of lumbar spondylolisthesis-related videos on short-video platforms: A cross-sectional study of TikTok and Bilibili.
The quality evaluation of information in hospice and palliative care short videos on Douyin (TikTok's Chinese version) and Bilibili: a cross-sectional study.
Quality and reliability of anxiety disorder-related short videos on TikTok and Bilibili: A cross-sectional study.
The Quality and Reliability of Short Videos About Psoriatic Arthritis on Bilibili and TikTok: Cross-Sectional Study.

Int J Med Inform. 2026 Jul 05. pii: S1386-5056(26)00332-1. [Epub ahead of print]220 106592

MedNLP-Hub: A knowledgebase platform for biomedical NLP tool discovery in clinical informatics.

Shumin Ren, Yuxin Zhang, Jinyi Zhang, Hongmei Lang, Hui Zong, Bairong Shen.

   OBJECTIVES: Biomedical natural language processing (NLP) tools are used to extract computable information from electronic health records, biomedical literature, and other health-related texts. However, selecting appropriate NLP tools for clinical informatics use remains difficult because available studies are fragmented across heterogeneous tasks, target variables, modeling strategies, and reporting practices. This study aimed to develop and evaluate MedNLP-Hub, a standardized knowledge-driven resource for multi-constraint discovery and comparative analysis of biomedical NLP tools.
METHODS: Biomedical NLP studies published in PubMed up to March 2026 were systematically collected and curated. Each task-level record was annotated using a standardized metadata schema covering study characteristics, text sources, task types, target variables, terminology resources, modeling approaches, accessibility information, and evaluation metrics. Relevant metadata fields were normalized to support structured filtering and semantic retrieval. MedNLP-Hub was then evaluated using 18 task-oriented retrieval scenarios reflecting practical tool-selection needs in biomedical and clinical informatics. Performance was compared with PubMed and three general-purpose LLMs: ChatGPT 5.2, Gemini 2.5-Flash, and Claude Sonnet 4.6.
RESULTS: MedNLP-Hub contains 676 tools contributing 1,994 task-level instances, organized across 24 standardized metadata dimensions. The resource covers 5 NLP task categories, 23 languages, 4 modeling paradigms, and more than 3,200 normalized target-variable elements. The platform supports structured browsing, field-based filtering, semantic variable-level retrieval, tool comparison and visualization. In benchmark evaluations, MedNLP-Hub identified eligible tools for all predefined scenarios, whereas LLM-based systems and PubMed frequently failed to satisfy all structured constraints simultaneously.
CONCLUSION: MedNLP-Hub provides a standardized and clinically relevant infrastructure for biomedical NLP tool discovery. By combining structured metadata, variable normalization, and semantic retrieval, it supports transparent pre-implementation tool selection, methodological comparison, and reuse of NLP systems in biomedical research and clinical informatics workflows.

Keywords:  Biomedical natural language processing; Clinical informatics; Database; Knowledgebase; Large language models; Tool discovery

DOI:  https://doi.org/10.1016/j.ijmedinf.2026.106592
Bioinformatics. 2026 Jul 01. pii: btag216. [Epub ahead of print]42(Supplement_1):

eGoT: enhanced graph-of-thoughts for multi-hop knowledge retrieval and hypothesis generation in biomedicine.

Nihar Sanda, Benjamin M Gyori, Vito Quaranta, Auroop Ganguly, Ayan Paul.

MOTIVATION: Research on biological mechanisms and disease processes is limited by fragmented findings across unstructured text in publications. Question answering and hypothesis generation that can reason across multiple sources can overcome this limitation. However, Large language models (LLMs) are prone to inaccuracies and lack clear provenance to primary evidence. Retrieval augmented generation approaches that have provenance to the original source of evidence address these shortcomings. However, the response richness is dependent on the retrieval process design. Current approaches often fail to produce responses requiring multi-hop reasoning across multiple domains.
RESULTS: To address this, we propose eGoT, which combines automated knowledge graph construction from biomedical literature with a novel graph-of-thoughts approach to query the knowledge base and construct comprehensive responses to natural language questions. Given a corpus of documents, eGoT first uses an LLM-based pipeline to identify and normalize entities and relationships and constructs graph and vector databases. Given an input question, eGoT performs multi-round LLM-based querying of the databases to construct a response. Benchmarking on datasets like MultiHopRAG, HotpotQA, and Ultradomain demonstrates eGoT's superiority over state-of-the-art retrieval methods, including HopRAG, SireRAG, HiRAG, and HippoRAG. We demonstrate eGoT on two biomedical use cases: (i) generate responses to domain expert-curated questions on small cell lung cancer using 1046 PubMed Central publications, and (ii) demonstrate eGoT's ability to find plausible connections between Lupus and climate factors (UV exposure) that affect disease trajectory.
AVAILABILITY AND IMPLEMENTATION: https://github.com/NNeuralDynamics/eGOT.git.

DOI: https://doi.org/10.1093/bioinformatics/btag216
SoftwareX. 2026 Sep;pii: 102803. [Epub ahead of print]35

PixelDeck: A local-first media library manager for biomedical imaging.

Benjamin L Kidder.

  Biomedical imaging workflows generate large collections of histology, fluorescence microscopy, and high-throughput imaging assets that are difficult to organize, search, and reuse using conventional file systems. We present PixelDeck, an open-source, local-first browser application for management of large image and video libraries on commodity workstations. PixelDeck integrates recursive import, SHA-256-based duplicate detection, metadata extraction, thumbnail generation, and full-text search within a responsive interface. The system uses a transparent SQLite-backed architecture with managed filesystem storage and asynchronous processing. Benchmarks on fresh isolated libraries containing up to 1000 assets showed import throughput exceeding 500 files/min and representative median query latencies in the millisecond range.

Keywords:  Digital pathology; Fluorescence microscopy; High-throughput imaging; Histology imaging; Image management; Scientific software

DOI:  https://doi.org/10.1016/j.softx.2026.102803
BMC Med Res Methodol. 2026 Jul 10.

Large language models vs. conventional search in identifying delirium guidelines for a scoping review.

Mary Faherty, Ronan Madden, Breeda Herlihy, Siobhan Bowman, Alasdair MacLullich, Suzanne Timmons.

   BACKGROUND: A literature search of information sources is a key methodological step for evidence synthesis, including scoping reviews. This study compares the performance of large language models with conventional systematic literature search methods in identifying published delirium clinical practice guidelines.
METHODS: Comparative study using parallel searches. Four large language models (ChatGPT, Google Gemini, Claude and Grok) were queried using structured prompts, with results assessed for precision, sensitivity, and efficiency, versus a conventional systematic literature search using established bibliographic databases and grey literature sources.
RESULTS: Large language models sourced 17 of the 45 guidelines identified through a comparable conventional search, missing 28, but also identified two guidelines not found through the conventional search. The sensitivity of individual large language models ranged from 9% to 29%, using the conventional search as a reference standard. Of the 19 guidelines retrieved through large language models, 63% were deemed high quality, compared with 56% identified through the conventional search. Precision of individual large language models varied from 22% to 60%, far exceeding the precision of the conventional search (1.7%). Large language model-based searches required substantially less execution time compared to conventional methods.
CONCLUSIONS: While large language models have a significantly lower yield in retrieving clinical guidelines compared to the conventional search, their efficiency, along with their capacity to identify some unique articles, suggests a possible role in complementing conventional search methods in scoping reviews.

Keywords:  Artificial intelligence; Clinical practice guidelines; Large language models; Literature search; Precision; Scoping review; Search strategy; Sensitivity

DOI:  https://doi.org/10.1186/s12874-026-02941-x
J Cancer Educ. 2026 Jul 11.

Who Is Responsible When AI Gets Cancer Information Wrong? Implications for Patient Education.

Wanich Suksatan.

  Generative artificial intelligence (AI) tools are increasingly being used by patients seeking cancer-related information, creating new opportunities for accessible and personalized cancer education. Large language models can simplify complex medical concepts, improve access to educational resources, and support patient engagement. However, these benefits are accompanied by growing concerns regarding misinformation, hallucinatory content, outdated recommendations, and the potential for harmful health decisions. As AI-generated information becomes more integrated into cancer education, an important ethical question emerges: who is responsible when AI provides inaccurate cancer information? This commentary examines the shared responsibilities of patients, healthcare professionals, healthcare organizations, and AI developers in ensuring the safe use of AI-generated cancer information. This commentary argues that accountability should not rest with a single stakeholder but instead be viewed as a shared responsibility across the cancer education ecosystem. The commentary further argues that AI literacy should become an essential component of modern cancer education to support informed decision-making and safeguard patient well-being.

Keywords:  Artificial Intelligence; Cancer Education; Health Literacy; Large Language Models; Misinformation; Patient Education

DOI:  https://doi.org/10.1007/s13187-026-02952-8
Pharmacopsychiatry. 2026 Jul 07.

Ongoing Changes in Web Searches for Health-Related Information.

Eric D Achtyes, Tasha Glenn, John R Geddes, Rita Bauer, Michael Bauer, Michael Gitlin, Scott Monteith.

DOI: https://doi.org/10.1055/a-2896-4218
Am J Health Promot. 2026 Jul 10. 8901171261468506

Seeing is Believing? Credibility and Behavioral Intentions Toward Evidence-Based and Misleading Health Information.

Tong Lin, Dylan D Wagner, Daphine Choy, Chelsie Hartman, Abigail Henck, Juliette A Louer, Laine O'Shaughnessy, Megan E Roberts.

  PurposeTo examine how adults respond to real-world health videos that vary in credibility.DesignRandomized, within-subjects online experiment.SettingOnline U.S. survey.Sample179 adults recruited from an online panel.InterventionParticipants viewed one misleading and one evidence-based health video, in randomized order.MeasuresBefore viewing, participants reported trust in traditional and social media. After each video, they rated perceived information credibility (PIC) and perceived source credibility (PSC), relevance, and behavioral intention to adopt the recommended behavior. Demographic data were also collected.AnalysisPaired-samples t-tests and hierarchical linear regressions examined responses to evidence-based vs misleading videos and associations among the variables.ResultsThe evidence-based video was rated higher for relevance (t (178) = -4.93, P < .001), PSC (t (178) = -11.36, P < .001), PIC (t (178) = -10.61, P < .001), and intention (t (178) = 5.73, P < .001). For both videos, PSC and relevance were positively associated with intention. Trust in traditional media was related to higher PSC (b = .43, P < .001) and PIC (b = .35, P < .001) of the evidence-based content, whereas trust in social media was related to higher PSC (b = .46, P < .001) and PIC (b = .33, P = .022) of the misleading content. Higher education was associated with a better ability to distinguish between the PSC and PIC of the two videos (ps < .01).ConclusionsFindings highlight psychological and contextual factors shaping engagement with online health information and the importance of speaker credibility and platform trust in promoting informed health decision-making.

Keywords:  credibility; digital health communication; health communication; health misinformation; health promotion; media trust; social media; trust

DOI:  https://doi.org/10.1177/08901171261468506
Psychol Sci. 2026 Jul 06. 9567976261453813

Boosting Media Literacy Using Lateral Reading and Online Search Interventions.

Lisa Oswald, Anastasia Kozyreva, Stefan M Herzog, Pietro Leonardo Nickl, Ralph Hertwig.

  Empowering people to navigate online information competently is essential to complement systemic content moderation and platform regulation. A nationally representative randomized controlled study among adults in Germany (N = 2,666) compared two media-literacy interventions: a source-focused lateral-reading strategy to help participants distinguish trustworthy from untrustworthy news outlets, and a claim-focused search strategy to help them assess the credibility of specific claims. Both interventions showed small improvements in discernment, but not all effects were statistically distinguishable from zero. At a 2-week follow-up, discernment improved in all groups, and differences between the intervention and control groups were no longer statistically distinguishable. We found no evidence of backfire effects. An exploratory analysis of discernment pretreatment indicated the lowest performance for supporters of populist radical-right parties. Behavioral measures suggested increases in information search in both intervention groups. On average, lateral reading reduced trust in untrustworthy sources, and online search increased trust in trustworthy sources.

Keywords:  boost; experiment; intervention; lateral reading; media literacy; misinformation; online search; web tracking

DOI:  https://doi.org/10.1177/09567976261453813
Front Public Health. 2026 ;14 1875038

Perceived cognitive slowing and online health information seeking among older adults: motivational beliefs as pathways and community IT culture as a buffer.

Zian Fang, Zhenjun Zhao, Xianye Cao.

   Background: Online health information seeking (OHIS) can support self-management and health decision-making among older adults, yet many still face barriers to digital health engagement. Perceived declines in information processing speed may constrain older adults' ability and motivation to seek health information online, but the potential explanatory pathways and contextual conditions remain insufficiently understood. This study examined the associations between perceived information processing speed decline and older adults' OHIS, with a focus on the roles of self-efficacy, outcome expectations, and community IT culture.
Method: A cross-sectional survey was conducted among 295 adults aged 60 years and older. Data were collected using a structured questionnaire and analyzed with partial least squares structural equation modeling to test the proposed research model.
Results: Perceived declines in information processing speed were negatively associated with self-efficacy but were not directly associated with outcome expectations or online health information seeking. Self-efficacy was positively associated with outcome expectations, and outcome expectations were positively associated with online health information seeking. Community IT culture moderated the association between perceived declines in information processing speed and self-efficacy, such that the negative association was weaker among older adults reporting stronger community IT culture. Bootstrapping further showed a significant sequential indirect association linking perceived declines in information processing speed to online health information seeking through self-efficacy and outcome expectations.
Conclusion: Perceived declines in information processing speed were indirectly associated with online health information seeking through motivational beliefs rather than through a direct pathway. A supportive community IT culture may buffer the negative association between perceived processing speed decline and self-efficacy. These findings highlight the importance of strengthening self-efficacy and outcome expectations while fostering supportive community-level digital environments to promote digital health engagement among older adults.

Keywords:  community IT culture; information processing speed; older adults; online health information seeking; outcome expectations; self-efficacy

DOI:  https://doi.org/10.3389/fpubh.2026.1875038
Front Public Health. 2026 ;14 1857953

Online health information seeking preferences among pregnant women: a discrete choice experiment.

Jing Xu, Liyan Xie, Qingqing Wang, Yanxia Zhu, Yuhong Li.

   Objective: This study aims to investigate the online health information seeking preferences of pregnant women to provide guidance for optimizing prenatal care information services.
Methods: From August to September 2025, a discrete choice experiment was conducted among 342 pregnant women recruited by convenience sampling from two tertiary hospitals in Anhui Province, China. All attributes and levels were identified through literature review, qualitative interviews and expert consultations. A D-efficient design was employed to create 18 choice sets, which were randomly divided into two blocks. A mixed logit model was used to analyze choice preferences, based on which we calculated relative importance (RI), willingness to pay (WTP) and performed scenario simulation.
Results: The attribute that participants valued most was the information interaction function (RI = 29.80%), followed by information content (RI = 27.01%), cost per payment (RI = 26.56%), information access method (RI = 8.73%), and information format (RI = 7.89%). Pregnant women preferred online consultation (β = 1.16, p < 0.001), knowledge about fetal growth and development (β = 0.59, p < 0.001), lower costs (β = -0.03, p < 0.001), independent search + personalized recommendation (β = 0.27, p < 0.001), and image + text + video (β = 0.23, p = 0.003). The estimated WTP for online consultation was 42.72 CNY, and this service increased the predicted choice probability by 52.40%.
Conclusion: Tailoring the provision of prenatal care information services to the actual preferences of pregnant women can improve their satisfaction with online prenatal care services and their adherence to health advice.

Keywords:  consumer health information; discrete choice experiment; internet; preferences; pregnant women

DOI:  https://doi.org/10.3389/fpubh.2026.1857953
BMC Med Educ. 2026 Jul 06.

The association between digital health information acquisition and health-seeking behavior among university students.

Ali Göde.

   BACKGROUND: This study aimed to examine the associations of digital health information acquisition (based on Web 1.0 and Web 2.0) and digital verification behavior with health-seeking behavior among university students.
METHODS: This cross-sectional study was conducted between February and March 2026 among 302 university students selected using a convenience sampling method. Data were collected through face-to-face questionnaires administered in classroom settings. The survey consisted of a Personal Information Form, the Digital Health Information Acquisition and Verification Scale, and the Health-Seeking Behavior Scale. Data were analyzed using IBM SPSS Statistics 26 software. Descriptive statistics were used to summarize participant characteristics and study variables. Pearson correlation and regression analyses were conducted to examine the associations among the study variables and their contributions to health-seeking behavior.
RESULTS: The findings indicate that there is a positive and significant relationship between digital health information acquisition and verification levels and health-seeking behavior (r = .422; p < .001). The results of the simple regression analysis revealed that digital health information acquisition and verification was significantly associated with health-seeking behavior (R² = 0.178; p < .001). According to the hierarchical regression results, Web 1.0-based health information acquisition emerged as the strongest predictor of health-seeking behavior. Although Web 2.0-based information acquisition and digital verification contributed significantly to the model, their incremental contributions to explained variance were relatively small (final model R² = 0.206; p < .001).
CONCLUSION: Digital health information acquisition and digital verification behavior were significantly associated with health-seeking behavior among university students. The findings suggest that access to digital health information and the tendency to verify information may be related to health-seeking behavior. However, the explanatory power of the model was modest. In addition, the ability to critically evaluate and verify information may play a supportive role in health-related decision-making in digital environments, although its unique contribution to explaining health-seeking behavior was relatively modest.

Keywords:  Digital Health Information; Digital Verification; Health-Seeking Behavior; University Students; Web 1.0; Web 2.0

DOI:  https://doi.org/10.1186/s12909-026-09886-1
Front Nutr. 2026 ;13 1861145

Nutrition information-seeking behavior among patients recovering at home after esophageal cancer surgery: a qualitative study.

Xinhui Song, Lei Zhong, Huiyan Liao, Muxi Cheng, Qiuling Qin, Tong Wu, Mei Li.

   Objective: Patients after esophageal cancer surgery often have unmet nutritional information needs during home rehabilitation. However, research on how these patients acquire, understand, and apply nutritional information in real-life contexts remains limited. This study aims to explore the nutritional information-seeking behaviors of post-esophagectomy patients during home rehabilitation, in order to provide evidence for the development of continuous and individualized nutritional information support strategies.
Methods: A descriptive qualitative design with purposive sampling was employed in this study. Semi-structured interviews were conducted with 15 patients at a tertiary hospital in Guangdong Province, China, between November 2025 and January 2026. Thematic analysis was performed using NVivo 14.0 software to organize and interpret the data.
Results: Four main themes emerged from the interviews: (1) multidimensional nutritional information needs; (2) Diversity of Information Sources; (3) barriers in the process of nutritional information seeking; and (4) behavioral patterns of nutritional information application.
Conclusion: This study demonstrates that nutritional information seeking among patients recovering at home after esophageal cancer surgery is characterized by symptom-driven and fragmented patterns, with patients experiencing substantial difficulties in accessing, interpreting, and applying nutritional information. These findings highlight the need for continuous, individualized, and family-inclusive nutritional information support to strengthen long-term nutritional self-management during postoperative recovery.

Keywords:  esophageal cancer; information-seeking behavior; nutrition; nutritional management; qualitative research

DOI:  https://doi.org/10.3389/fnut.2026.1861145
BMC Psychol. 2026 Jul 06.

From information overload to health information avoidance among Chinese adults aged 50 years and above: the association of self-perceptions of aging and techno-exhaustion.

Jing An, Kexin Wan, Ziyue Xiang, Jinlong An, Pu Han, Weiwei Zhu, Chen Li.

   BACKGROUND: In the era of digital health, online health consultation platforms have become vital resources. However, the complex information environment is often associated with defensive user behaviors. This study investigates the factors correlated with health information avoidance among middle-aged and older adults within the information overload environment of these platforms.
METHODS: A cross-sectional survey was conducted in China, yielding 249 valid responses from individuals aged 50 and above. Drawing on the Stressor-Strain-Outcome (S-S-O) framework, the study examined the structural relationships between information overload (stressor), psychological strains (privacy concerns, techno-exhaustion, and fear of missing out), and health information avoidance (outcome).
RESULTS: The findings demonstrated that information overload was positively associated with privacy concerns, techno-exhaustion, and fear of missing out. Among these psychological strains, techno-exhaustion showed a direct association with health information avoidance behavior. Furthermore, self-perceptions of aging moderated several relationships within the S-S-O framework, suggesting that aging-related attitudes may shape how environmental stressors relate to psychological strain.
CONCLUSIONS: Information overload environments show an indirect association with health information avoidance behaviors through heightened psychological strain. The results highlight the necessity for online health consultation platforms to prioritize aging-adaptive optimizations and psychological support mechanisms to mitigate user fatigue and promote digital health inclusion for middle-aged and older adults.

Keywords:  Fear of missing out; Health information avoidance; Information overload; Privacy concerns; Self-perceptions of aging; Techno-exhaustion

DOI:  https://doi.org/10.1186/s40359-026-05090-4
Dev Sci. 2026 Sep;29(5): e70252

Correction to "Navigating Knowledge: Effects of State Curiosity on Children's Word Learning and Information Seeking".

DOI: https://doi.org/10.1111/desc.70252
Proc (Bayl Univ Med Cent). 2026 Jul 06. 1-5

Supporting patient understanding of cervical and ovarian cancer: how well does AI perform?

Riley Shin, Ashley Shin, Mia Ramirez, Ioana L Bondre, Diego J Aviles, Elizabeth K Nugent.

   BACKGROUND: As the public increasingly interacts with artificial intelligence (AI) chatbots, we compared the answers from five AI chatbots to standardized questions that patients might ask about ovarian and cervical cancer.
METHODS: ChatGPT 3.5, Google Gemini 2.0, Reddit Answers, Bootcamp, and DeepSeek were queried with 15 frequently asked questions (FAQs) on cervical cancer and 11 on ovarian cancer. In a blinded, randomized survey, each deidentified response was independently assessed by three gynecologic oncologists using a 4-point scale (1, accurate and comprehensive; 2, accurate but inadequate; 3, accurate but outdated or inaccurate; 4, completely inaccurate). Readability (Flesch-Kincaid grade level) and word count were recorded.
RESULTS: For cervical cancer FAQs, ChatGPT 3.5, Google Gemini 2.0, Reddit Answers, Bootcamp, and DeepSeek received scores of 1.4, 1.6, 2.7, 1.5, and 1.5, respectively. For ovarian cancer FAQs, average scores were 1.3, 1.4, 2.5, 1.2, and 1.4, respectively. All AI chatbot responses were written at a reading level above 11th grade, making them generally difficult for the average American to read.
CONCLUSIONS: Although generally rated as accurate and adequate, the answers were frequently off-topic and not generalizable. Healthcare providers should be aware of unintentionally generated misinformation to better counsel patients.

Keywords:  Artificial intelligence; cervical cancer; ovarian cancer; patient education; women’s health

DOI:  https://doi.org/10.1080/08998280.2026.2694930
Digit Health. 2026 Jan-Dec;12:12 20552076261464724

Challenges of patient-facing generative artificial intelligence in hypertension care: A cross-platform evaluation of the quality, readability, and actionability of LLM-Generated patient education materials.

Mengqiu Hu, Zhiqiang Wang, Zhiwen Zhang, Muwei Li.

   Objective: To assess the quality, readability, and actionability of hypertension patient education materials generated by six selected patient-facing large language model (LLM) platforms, and to characterize cross-platform heterogeneity to guide the optimization of AI-generated patient education materials.
Methods: Six publicly accessible or commonly accessible platforms were evaluated using standardized prompts to generate patient education materials. Understandability and actionability were assessed using the Patient Education Materials Assessment Tool for Printable Materials (PEMAT-P). Information quality was evaluated using the expanded Ensuring Quality Information for Patients scale (EQIP-36), and overall quality was rated using the Global Quality Score (GQS). Readability was compared using seven metrics, including the Flesch Reading Ease Score (FRES) and the Flesch-Kincaid Grade Level (FKGL).
Results: Overall educational quality and understandability were generally favorable, but substantial cross-platform heterogeneity was observed. Readability remained challenging, with low FRES values and grade-level indices generally exceeding commonly recommended thresholds for patient education materials. Qwen3-Max-Thinking-Preview achieved the highest PEMAT-P total score (77.00), followed by ChatGPT 5.2-Thinking (72.22). For EQIP-36, Qwen3-Max-Thinking-Preview scored highest (49.28), followed by DeepSeek-R1 (45.48). DeepSeek-R1 generated the most readable materials among the evaluated platforms, with a median FRES of approximately 42.14 and FKGL of approximately 10.10, whereas Qwen3-Max-Thinking-Preview showed lower readability, with a median FRES of approximately 16.82 and FKGL of approximately 14.38. Kimi K2 showed high PEMAT-P understandability (76.90) but low actionability (10.00). Post hoc analyses showed that Qwen3-Max-Thinking-Preview significantly outperformed ERNIE Bot 4.5 Turbo, Doubao, and Kimi K2 on PEMAT-P total score, and outperformed Kimi K2 and ERNIE Bot 4.5 Turbo on EQIP-36. DeepSeek-R1 also outperformed Kimi K2 and ERNIE Bot 4.5 Turbo on EQIP-36. Across content domains, actionability was significantly higher for Daily Care and Prevention than for Basic Understanding of the Disease and Complications, Psychological and Social Aspects.
Conclusions: Generative AI shows promise for hypertension patient education, particularly in improving understandability. However, actionability remains a major limitation of current outputs, highlighting the need for platform-aware optimization strategies that explicitly strengthen step-by-step and action-oriented guidance.

Keywords:  actionability; cross-platform evaluation; generative artificial intelligence; hypertension; large language models; patient education; understandability

DOI:  https://doi.org/10.1177/20552076261464724
J Orthop Surg (Hong Kong). 2026 May-Aug;34(2):34(2): 10225536261469293

Accuracy and readability analysis of ChatGPT responses to frequently asked questions about olecranon fractures.

İbrahim Halil Dadir.

  PurposeThis study aimed to evaluate the reliability, quality, and readability of ChatGPT-generated responses to common patient questions about olecranon fractures.MethodsIn this cross-sectional study, twenty frequently asked questions were identified using Google's "People also ask" feature and submitted to ChatGPT-4o in separate sessions between October 1 and October 10, 2025. Two orthopaedic surgeons independently assessed the responses using the DISCERN instrument and Global Quality Score (GQS). Readability was analysed using Flesch Reading Ease (FRE), Flesch-Kincaid Grade Level (FKGL), Simple Measure of Gobbledygook (SMOG), and Gunning Fog Index (GFI).ResultsMean DISCERN and GQS scores (51.25 ± 4.13 and 4.10 ± 0.58) indicated good quality and reliability. The mean FRE was 62.00 ± 5.12, and mean FKGL, SMOG, and GFI values (9.23, 8.42, 9.37) corresponded to a ninth-grade reading level. DISCERN and GQS were strongly correlated (r = 0.913, p < 0.001) and negatively associated with FKGL.ConclusionChatGPT-4o produced coherent responses of generally good quality and reliability regarding olecranon fractures; however, readability levels exceeded those recommended for patient comprehension. These findings suggest that while ChatGPT may serve as a supportive informational tool, clinician guidance remains essential to ensure appropriate patient understanding and prevent potential misinformation.

Keywords:  Artificial Intelligence; bone; fractures; olecranon process; quality control; readability; reproducibility of results

DOI:  https://doi.org/10.1177/10225536261469293
Health Sci Rep. 2026 Jul;9(7): e72706

Assessing Readability and DISCERN Quality of Osteoporosis Education Materials Generated by ChatGPT and Deepseek for Diverse Health Literacy Levels: A Cross-Sectional Study.

Junfang Miao, Guanghu Sun, Yichao Wang, Fangying Li, Fangli Li.

   Background and Aims: Large language models (LLMs) are increasingly used for patient education, but their output quality across health literacy levels is unknown. This cross-sectional study compared ChatGPT-4o, ChatGPT-5, DeepSeek-V3.1, and DeepSeek-R2 in generating osteoporosis education materials tailored to low, moderate, and high health literacy.
Methods: Six clinical domains were posed to each LLM, with prompts adapted for three literacy tiers per model (18 outputs per model). Outputs were aggregated into 12 composite texts (4 models × 3 tiers). Three blinded clinicians assessed information quality using DISCERN (0-80) and readability using Flesch-Kincaid (FKGL, lower = easier).
Results: Mean DISCERN scores ranged 36-52/80 ("fair"); no output reached "excellent" (> 70/80). DeepSeek-V3.1 provided the highest treatment detail for high literacy and best low-literacy readability (FKGL 3.99). ChatGPT-5 performed most consistently across tiers. Median readability ranged from grade 4.8 to 10.2. All outputs lacked citations, publication dates, quantitative risk data, and uncertainty statements.
Conclusion: LLMs can rapidly generate readable osteoporosis education, but current outputs require supplementation with references, risk statistics, and updated transparency before clinical use.

Keywords:  ChatGPT; DISCERN; DeepSeek; artificial intelligence; health literacy; osteoporosis; readability

DOI:  https://doi.org/10.1002/hsr2.72706
Front Public Health. 2026 ;14 1859078

Information quality, readability, and empathy of AI-generated public mental health information: a comparative evaluation of eight large language models.

Qilong Wang, Siheng Ma, Jian Sun, Xin Qi, Jianpeng Zhao, Dongmei Ma, Xiaofei Liu, Qiang An, Dongrong Zhao, Songkai He.

   Background: Mental disorders are a growing global health burden, yet healthcare resources remain scarce. Large language models (LLMs) may support public mental health information seeking, but their information quality, readability, and empathy in psychiatric contexts require validation.
Method: We developed a test bank of 48 public mental health questions from literature, Google Trends (2004-2025), and clinical consultations. Eight LLM chatbots were compared for information quality, source transparency, readability, and empathy using established rating instruments, readability indices, and psychiatrist-rated and user-perspective empathy assessments. Statistical analysis used Kruskal-Wallis and Dunn's post-hoc tests, with Spearman correlations.
Results: Gemini 3.0 Pro and GPT-5.2 Think showed relatively higher information quality and source transparency scores, whereas spontaneous source transparency was poor across models, with median JAMA scores of 0. No model met the sixth-grade readability standard; Claude Sonnet 4.5 generated relatively more readable responses, whereas Claude Sonnet 4.5 Think produced responses with the highest reading difficulty. In the psychiatrist-rated empathy assessment, Gemini 3.0 Pro and DeepSeek-V3 showed the highest high-empathy response rates, at 54.2% and 43.8%, respectively. User-perspective empathy ratings were generally lower, with DeepSeek-V3 and Gemini 3.0 Pro showing the highest user-perspective high-empathy response rates, at 31.2% and 27.1%, respectively. Information quality and source transparency metrics showed only weak correlations with readability metrics.
Conclusion: LLMs face important challenges in psychiatric information delivery. Gemini 3.0 Pro and GPT-5.2 Think showed higher information quality and source transparency scores and better information structuring, whereas Claude Sonnet 4.5 generated relatively more readable responses. However, source opacity and high reading difficulty limit direct patient-facing use. Empathy varied across models and differed between psychiatrist-rated and user-perspective assessments, suggesting that empathic communication requires separate optimization and validation with intended users. Future work should balance information quality, source transparency, readability, traceability, safety, and emotionally appropriate responses.

Keywords:  empathy; information quality; large language models; patient education; psychiatry; readability

DOI:  https://doi.org/10.3389/fpubh.2026.1859078
Ann Afr Med. 2026 Jul 07.

A Comparative Cross-sectional Study Assessing the Readability of ChatGPT and UpToDate Content on Multiple Endocrine Neoplasia Syndromes.

Yamini Saraswathi Gurram, Atharv B Deo, Aryan Jha, Shalini Jeyaraman, Jimena Lozano-Núñez, Prapti S Sonkusare.

   BACKGROUND: With growing reliance on digital health tools, the readability of online medical content is increasingly vital in patient-centered care. Artificial intelligence platforms like ChatGPT are widely used to access medical information, yet their suitability for conveying complex conditions such as multiple endocrine neoplasia (MEN) syndromes remains underexplored.
OBJECTIVES: This study aimed to compare the readability of MEN-related educational materials generated by ChatGPT with those from the evidence-based platform UpToDate (UTD), evaluating their appropriateness for patient education.
MATERIALS AND METHODS: Six related subjects that addressed MEN types 1 and 2 were chosen. The WebFX readability tool was used to examine texts produced by ChatGPT and the related UTD articles. Word count, sentence count, proportion of difficult words, Flesch Reading Ease (FRE), Flesch-Kincaid Grade Level (FKGL), and Simple Measure of Gobbledygook Index were among the metrics. R (v4.3.2) and SPSS (v25) were used for statistical analysis, and the Wilcoxon signed-rank test was used.
RESULTS: ChatGPT responses were significantly shorter (median word count: 594.0 vs. 1955.5) and contained fewer sentences (66.0 vs. 166.0), though sentence complexity differences were not significant. Both sources scored poorly in readability: ChatGPT had an FRE of 21.3 and FKGL of 15.3, while UTD scored 24.9 and 16.7, respectively. Although ChatGPT used fewer difficult words, it had a higher proportion of complex terms (35.6% vs. 30.1%; P = 0.0306).
CONCLUSIONS: Although both sources went over the suggested readability limits for patient materials, ChatGPT produced more condensed text. To improve understanding and promote health fairness, medical terminology must be made simpler.

Keywords:  Artificial intelligence in health care; ChatGPT; Flesch Reading Ease; Flesch–Kincaid Grade Level; Intelligence artificielle en santé; Simple Measure of Gobbledygook Index; UpToDate; facilité de lecture de Flesch; health literacy; indice SMOG; lisibilité; littératie en santé; multiple endocrine neoplasia; niveau scolaire de Flesch–Kincaid; néoplasies endocriniennes multiples; patient education; readability; éducation du patient

DOI:  https://doi.org/10.4103/aam.aam_367_26
Medicine (Baltimore). 2026 Jul 03. 105(27): e49658

Artificial intelligence-based patient information in rotator cuff injuries: A cross-sectional comparative study of ChatGPT and DeepSeek models.

Bahri Bozgeyik, Erman Öğümsöğütlü, Gazi Huri.

  This study aimed to comparatively evaluate the performance of the chat generative pretrained transformer (ChatGPT) and DeepSeek artificial intelligence (AI) models in patient information about rotator cuff injuries. This cross-sectional comparative study was conducted in May 28, 2025 using ChatGPT-4o (OpenAI) and DeepSeek V3 (DeepSeek Inc.) models. Sixteen frequently asked questions related to rotator cuff injuries were posed to both the AI models. The responses were then independently assessed by 2 experienced orthopedic surgeons using the Journal of the American Medical Association (JAMA), response rating system, DISCERN, and 4-point Likert scales. In addition, the readability of the responses was analyzed using the Flesch-Kincaid Readability Score (FRES) and Flesch-Kincaid Grade Level. The primary outcome was overall information quality, secondary outcomes included JAMA benchmark adherence and readability metrics. None of the models met JAMA criteria. In terms of response rating system, there was no statistically significant difference between the 2 models (P >.05). DeepSeek demonstrated higher DISCERN scores compared to ChatGPT (50.12 vs 47.03), with a mean difference of 3.09 (95% CI: 1.58 to 4.61; P = .001). While there was no significant difference in accuracy, clarity, and consistency criteria between the 2 models in the 4-point Likert evaluation (P >.05), DeepSeek scored significantly higher than ChatGPT in the completeness criterion, with a mean difference of 0.75 (95% CI: 0.46 to 1.04; P = .001). In terms of readability, both models showed similar performance (FRES, P >.05; Flesch-Kincaid Grade Level, P >.05). Both AI models deliver satisfactory and clinically relevant information for rotator cuff injury patient education. Although DeepSeek was superior to ChatGPT in terms of completeness of patient information regarding rotator cuff injuries, the results were similar for the other criteria. The responses from both the AI tools were considered promising. However, they require improvements in terms of adherence to scientific standards, transparency, citations, and readability.

Keywords:  ChatGPT; DeepSeek; artificial intelligence; rotator cuff injury

DOI:  https://doi.org/10.1097/MD.0000000000049658
Odontology. 2026 Jul 07.

Clinical safety, content coverage, and patient-centered language of AI responses to periodontal complaint-based queries: a comparative study of four large language models.

Çağrı Esen, Mehmet Gül, Fatih Karayürek.

  This study aimed to evaluate the clinical safety, informational completeness, and patient-centered language of responses generated by four large language model (LLM)-based systems to standardized periodontal complaint-based queries. Seven standardized symptom-based periodontal queries were developed based on commonly reported patient complaints and submitted in Turkish to Copilot, Gemini, Claude, and ChatGPT. Responses were evaluated using a structured rule-based framework consisting of a Content Coverage Score (CCS), Risk of Harm Score (RHS), and Patient Language Score (PLS). Assessments were performed by two blinded human reviewers and one blinded AI-based evaluator. Human consensus, AI consensus, and combined evaluator scores were calculated. Agreement between evaluators was assessed using intraclass correlation coefficients (ICC), while human-AI differences and inter-model comparisons were analyzed using non-parametric statistical tests. Excellent agreement was observed between human reviewers, repeated AI evaluations, and human-AI consensus scores (ICC range: 0.911-0.925; p < 0.001). No significant difference was found between overall human and AI consensus scores (p = 0.985). Across the four AI systems, no statistically significant differences were observed in CCS or PLS scores in the human, AI, or combined evaluator analyses (all p > 0.05). PLS scores were generally high across models, indicating good linguistic accessibility for patients. No clearly harmful guidance was identified by the human reviewers. Overall, LLM-based systems generated clinically safe, reasonably comprehensive, and generally patient-accessible responses to common periodontal complaint-based queries. Although these systems may serve as supplementary sources of periodontal health information, they cannot replace individualized clinical evaluation and professional dental consultation.

Keywords:  Artificial intelligence; Health communication; Health literacy; Natural language processing; Periodontal diseases

DOI:  https://doi.org/10.1007/s10266-026-01498-x
Digit Health. 2026 Jan-Dec;12:12 20552076261459596

Reliability and readability of adenoid hypertrophy information generated by five publicly accessible LLM chatbots: A default-setting snapshot study.

Xiaoming Qian, Zhishui Wu, Jing Li, Qiuyu Su, Qian Qin, Beibei Zhang.

   Objective: This study evaluated the performance of five major large language model (LLM) chatbots in generating patient-oriented information on adenoid hypertrophy, focusing on content reliability and readability.
Methods: Sixty-three frequently asked questions (FAQs) on adenoid hypertrophy were collected, covering seven domains including etiology, symptoms, and treatment. From October 1, 2025, to January 10, 2026, questions were submitted in English to five LLMs via their official web interfaces. Reliability was assessed using DISCERN, EQIP, JAMA benchmarks, and the Global Quality Scale (GQS). Readability was measured by six standard indices (ARI, CLI, FKGL, GFI, SMOG, FRES). Three otorhinolaryngology clinicians blindly scored all responses.
Results: Significant differences in reliability were found among models (P <0.001). Perplexity scored highest on DISCERN (41.98±1.87) and EQIP (58.40±3.67), followed by Copilot; ChatGPT and DeepSeek scored lowest. Only Copilot and Perplexity scored 1 point on JAMA benchmarks. No model met the recommended sixth-grade reading level. Gemini had the best readability (FRES: 61.95±9.64), while Copilot scored poorest (FRES: 24.27±10.77). All models failed to meet the recommended sixth-grade readability thresholds. (P <0.001).
Conclusion: Current LLMs show a notable imbalance between reliability and readability in generating adenoid hypertrophy information, with none excelling in both. In this default-setting, product-level snapshot, Perplexity showed higher information-quality scores, whereas Gemini generated comparatively easier-to-read responses. These findings should not be interpreted as a controlled benchmark of underlying base models. Limitations include potential prompt sensitivity, single-response sampling, and the snapshot nature of the assessment given rapid model updates. Future improvements should focus on source transparency, text simplification, and condition-specific evaluation to enhance AI-assisted health communication for pediatric care.

Keywords:  adenoid hypertrophy; large language models; patient education; quality of health information; readability; reliability

DOI:  https://doi.org/10.1177/20552076261459596
JMIR Cancer. 2026 Jul 09. 12 e86073

Evaluation of a Retrieval-Augmented Generation-Based Large Language Model for Evidence-Based Herb and Supplement Information in Cancer Care.

Yen-Nien Hou, Jyothirmai Gubili, Pulkit Jain, Chun Sing Lam, Avijit Chatterjee, Jun J Mao.

   Unlabelled: Our study describes the development and evaluation of a retrieval-augmented generation-based large language model to improve the quality of responses to provider questions about herbs and dietary supplements.

Keywords:  artificial intelligence; cancer; dietary supplements; health information; herbs

DOI:  https://doi.org/10.2196/86073
PLOS Digit Health. 2026 Jul;5(7): e0001493

Evaluating the quality, reliability and readability of digital and artificial intelligence resources for adults with cancer who have significant caregiving responsibilities for children.

Rohan Anand, Cherith J Semple, Lisa Strutt, Sally Paul, Jeffrey R Hanna.

Families often report searching the internet for guidance on how best to support children when a significant adult has cancer. This study aimed to identify and evaluate the quality, reliability, readability and content of websites, videos, and artificial intelligence (AI) resources available to adults with cancer who have caregiving responsibilities for children. Online platforms were searched using 10 phrases across Google web, YouTube, TikTok and four AI platforms. The mDISCERN instrument assessed reliability and quality, GQS assessed overall quality, and the NHS Medical Document Readability Tool assessed readability. Quantitative differences between sources were determined using pairwise analysis. Google web had significantly higher quality and reliability compared with AI and TikTok sources, with mean mDISCERN and GQS scores of 3.74 and 3.72, respectively. AI-generated resources showed lower mean mDISCERN and GQS scores of 2.77 (P < .05) and 2.32 (P < .05), respectively. TikTok videos had lower scores of 2.73 (P < .05) for mDISCERN and 2.49 (P < .05) for GQS. Estimated reading time was significantly longer (P < .05) for Google web (11:45mins) compared to AI (02:09mins). However, reading age did not differ (P = .31) at 15.09 years and 15.17 years respectively. There was a lack of accessible and inclusive resources for non-nuclear families, adults with neurodivergent children, culturally and ethnically diverse populations and families at end of life. Although Google web resources demonstrated higher overall quality and reliability, written resources across platforms often exceeded recommended reading levels, which may represent a significant health equity concern for individuals with lower health literacy and families experiencing deprivation. AI presents an opportunity whereby a single high-quality and evidence-based resource can be rapidly adapted into multiple formats, reading levels, languages and be culturally relevant. Future resources may benefit from co-production using a trusted, regulated, and centralised information hub, with supportive collaboration between health and social care professionals and technology providers.

DOI: https://doi.org/10.1371/journal.pdig.0001493
J Stomatol Oral Maxillofac Surg. 2026 Jul 06. pii: S2468-7855(26)00189-8. [Epub ahead of print] 102893

Comparing the Performance of Conversational Artificial Intelligence Models in Answering Patient Questions About Dental Implants.

Esra Mavi, Tuğçe Çevik Sönmez.

   OBJECTIVE: In this study, the aim was to compare three widely used AI-based chatbots, ChatGPT-4.5, DeepSeek, and Claude Sonnet 3.7. This study investigated how well they answered frequently asked questions by patients about dental implants. The answers were checked for reliability and usefulness for patients and were compared with the opinions of oral and maxillofacial surgeons.
MATERIALS AND METHODS: 27 patient-oriented questions about dental implants were selected. These questions were about who is suitable for implants, possible risks, symptoms of peri-implant diseases, what to do after tooth loss, preventive measures, and prognosis. Two independent and blinded oral and maxillofacial surgeons evaluated the answers with a modified three-point Likert scale.
RESULTS: The three chatbots gave answers that were similar in reliability and usefulness. There was no statistically significant difference in reliability or usefulness (p > 0.05). A significant difference was only found in the "Avoiding Medical Jargon" category (p < 0.001). Overall, Claude 3.7 Sonnet gave longer and more detailed answers. ChatGPT-4.5 gave shorter answers and used fewer medical terms, but it was not the easiest-to-read model by readability indices. DeepSeek showed higher Flesch Reading Ease scores than ChatGPT-4.5. Agreement between the two raters was high (Cronbach's α = 0.89-0.93).
CONCLUSIONS: The results of the study show that the three models are similar in reliability and usefulness when answering common dental implant-related questions. These findings suggest that such chatbots can be used as supportive tools to inform patients about medical procedures.

Keywords:  Artificial Intelligence; ChatGPT; Claude Sonnet; DeepSeek; Dental Implant

DOI:  https://doi.org/10.1016/j.jormas.2026.102893
Front Med (Lausanne). 2026 ;13 1830356

Trust and verification in AI-enabled physician chatbots for chronic disease management: evidence from digital health behavior.

Haitham Alzghaibi.

   Background: Advances in digital health technologies have transformed how individuals with chronic diseases seek and use health information. Patients increasingly rely on online sources, including search engines, social media, and messaging applications, to understand symptoms and manage chronic conditions. However, these digital environments can also expose users to misinformation or conflicting advice. Artificial intelligence (AI) enabled tools and mobile health (mHealth) services have emerged to assist patients in identifying symptoms, verifying health information, and supporting timely health decisions. Despite these developments, limited conceptual work has examined how individuals with chronic diseases integrate such tools into their health information-seeking and decision-making processes.
Objective: This study aimed to develop and empirically illustrate a model explaining how individuals with chronic diseases seek and verify digital health information using AI-enabled tools and how these processes influence trust and health-related decision-making.
Methods: A cross-sectional survey was conducted among adults aged ≥ 18 years diagnosed with diabetes or hypertension. Participants were recruited through chronic disease support groups on Facebook and WhatsApp. The survey assessed digital health information-seeking behavior, verification practices, trust in AI-enabled physician chatbots and national mHealth services, and their role in health-related decision-making. Descriptive statistics and visualization analyses were conducted using R.
Results: Health information seeking occurred across multiple digital platforms, with considerable overlap between messaging applications, social media, and web-based sources. Participants reported using AI physician chatbots and national mHealth services mainly to verify health information encountered online. Trust in AI diagnostic support tools was moderate, indicating cautious but active engagement. Most participants used these tools to support clinical consultations rather than replace professional medical advice.
Conclusion: Verification behavior and trust play key roles in how individuals with chronic diseases engage with digital health information. AI-enabled mHealth tools may function as complementary decision-support resources that help patients verify information and interpret symptoms while supporting informed health decisions alongside traditional healthcare services.

Keywords:  AI physician chatbots; chronic disease management; digital health verification; health information seeking; mHealth platforms

DOI:  https://doi.org/10.3389/fmed.2026.1830356
Cureus. 2026 Jun;18(6): e110481

Artificial Intelligence-Generated Versus Professional Society Patient Education Materials in Gastroenterology, Surgery, Ophthalmology, and Anesthesiology: A Comparative Analysis of Readability and Health Literacy Metrics.

Shivam Chandra, Abhin Sapkota, Vineet Kumar, Steven R Bonomo, Scott Schimpke.

   AIM: Professional society patient education materials frequently exceed recommended literacy levels, limiting equitable health information access. This study aimed to compare the readability, information quality, understandability, and actionability of artificial intelligence (AI)-generated patient education materials versus professional society materials across gastroenterology, surgery, ophthalmology, and anesthesiology.
SUBJECTS AND METHODS: We conducted a cross-sectional comparative analysis of 100 paired topics (25 per specialty), comparing professional society materials with the responses generated by ChatGPT (OpenAI, San Francisco, California, United States) under standardized conditions. Readability was assessed using the Flesch-Kincaid grade level, information quality with DISCERN, and understandability and actionability with the Patient Education Materials Assessment Tool (PEMAT). Paired two-sided t-tests assessed within-specialty differences.
RESULTS: In surgery, AI-generated materials had lower reading levels and higher quality, understandability, and actionability (all p<0.001). In anesthesiology, AI materials were more readable (p<0.001) with no differences in other measures. In ophthalmology, AI improved readability (p<0.001), while professional society materials had higher quality and understandability (p<0.01) with no difference in actionability. In gastroenterology, AI materials had higher reading levels (p<0.001) with no differences in quality or usability.
CONCLUSION: The performance of AI-generated patient education materials varied by specialty and appeared to depend on the structure and complexity of clinical content. AI improved readability in several domains, but these gains were not uniform across specialties, particularly in areas requiring more complex or longitudinal explanations.

Keywords:  artificial intelligence; health literacy; medical specialties; patient education; readability

DOI:  https://doi.org/10.7759/cureus.110481
J Surg Res. 2026 Jul 07. pii: S0022-4804(26)00384-7. [Epub ahead of print]325 728-736

Evaluation of Online Educational Content on Indocyanine Green Fluorescence for Flap Perfusion: Assessment of Quality, Reliability, and Engagement Metrics.

Rene Aleman, Rodrigo Arean-Sanz, Edward Dougherty, Martin Newman, Andres Mascaro-Pankova, Rutger Schols, Fernando Dip, Alberto Rancati.

   INTRODUCTION: Indocyanine green (ICG) fluorescence angiography is widely implemented for flap perfusion assessment. YouTube is commonly used for procedural learning despite variable educational rigor. This study evaluated the quality and reliability of YouTube videos on ICG-assisted flap perfusion and examined whether popularity metrics reflect educational value.
METHODS: Videos were systematically identified and assessed using the Global Quality Score, modified DISCERN, and a procedure-specific Medical Content Index. Popularity metrics included views per day, like ratio and video power index.
RESULTS: Twenty-nine videos were included. Perceived quality was generally high, but reliability and content completeness varied. Popularity metrics were strongly intercorrelated but did not correlate with overall educational quality. Higher like ratio was associated with greater reliability and completeness, whereas higher views per day predicted lower reliability.
CONCLUSIONS: Viewer engagement metrics do not reliably indicate educational quality of YouTube content on ICG-assisted flap perfusion. Peer-informed standards are needed for online surgical education.

Keywords:  Fluorescence imaging; Indocyanine green; Plastic surgery; Surgical education; Surgical flaps; Video-based learning

DOI:  https://doi.org/10.1016/j.jss.2026.06.018
J Community Health. 2026 Jul 08.

Readability of Online Chronic Traumatic Encephalopathy (CTE) Information: Health Literacy Implications for Parents Navigating Youth Tackle Football Decisions.

Danna Ethan, Corey H Basch, Grace C Hillyer.

  Tackle football is the most participated youth sport in the U.S. with leagues beginning as early as age 5. Exposure to cumulative repetitive head impacts (RHI) over years of play is increasingly viewed as a major contributor to chronic traumatic encephalopathy (CTE), a progressive neurodegenerative disease documented in contact sport athletes. Amid growing awareness of CTE, parents may turn to online information to guide decisions about youth tackle football participation. This cross‑sectional study examined the readability of online CTE information. Using the search term, 'CTE,' 68 URLs providing non‑technical information were identified after applying exclusion criteria. Online software was used to generate metrics from six widely-used readability formulas. Grade-level readability scores were categorized as ≤ Grade 8, 9-12, and ≥ 13 and summarized using descriptive statistics; distributions were compared by URL designation using chi-square tests (P < 0.05). Web page publication/revision date and presence of references were recorded. Median readability scores ranged from high school to early college with few pages meeting the recommended ≤ Grade 8 reading level for the general population. Levels were similarly high across non-commercial (.org,.gov,.edu) and commercial (.com) domains. Nearly 40% lacked clear publication or revision dates; fewer than half (47.1%) included references. Commonly accessed online CTE resources exceed recommended reading levels. This digital barrier impairs parents' functional health literacy and capacity for informed decision-making. As research on CTE and tackle football participation evolves, there is a need for plain‑language, clearly-sourced, updated online resources tailored to this decisional context.

Keywords:  CTE; Digital literacy; Health literacy; Readability; Youth tackle football

DOI:  https://doi.org/10.1007/s10900-026-01594-7
Aesthet Surg J Open Forum. 2026 ;8 ojag106

The Online Landscape of Brazilian Butt Lift Information: Readability, Quality, and Traffic Analysis.

Elisa Bertulla, Edoardo Raposio.

Background: The Brazilian Butt Lift (BBL) has become increasingly popular yet remains one of the highest-risk aesthetic procedures. As patients frequently use online sources to understand medical procedures, clear and reliable web-based information is essential.
Objectives: This study evaluates the readability, understandability, quality, and visibility of online BBL resources.
Methods: A Google search for "Brazilian Butt Lift Procedure" was conducted using Startpage to reduce bias. The first 10 eligible English-language websites were analyzed. Extracted text was evaluated using 6 readability indices. Understandability was assessed with the Patient Educational Material Assessment Tool (PEMAT), overall quality with the JAMA (Journal of the American Medical Association) Benchmark Criteria. Website visibility was measured using SpyFu estimates of monthly organic traffic. Spearman's correlation was used to explore the relationship between readability and traffic.
Results: Readability for all websites exceeded the recommended sixth-grade level (mean readability score 21.68), indicating complex content. Sixty percentage of the websites scored 83% in PEMAT due to long sentences and limited explanation of technical terms; the remaining 40% websites met all understandability criteria. Journal of the American Medical Association assessment revealed deficiencies in transparency, with few sites providing references, author credentials, or publication dates. The correlation between readability and organic traffic was weakly positive (ρ = .176), indicating that frequently visited websites were not necessarily easier to read.
Conclusions: In this review, online BBL information was generally difficult to understand and often lacks essential quality indicators. The weak relationship between readability and traffic suggests that popular websites may not provide accessible content. Improving linguistic clarity, structure, and transparency is needed to ensure patients receive reliable, comprehensible guidance on this high-risk procedure.
Level of Evidence 5 Therapeutic: For image description, please refer to the figure legend and surrounding text.

DOI: https://doi.org/10.1093/asjof/ojag106
Med Ref Serv Q. 2026 Jul 10. 1-18

Online patient education: results of a content analysis of disempowering language in patient education materials about type 1 diabetes.

Gwen Geiger Wolfe, Sally Bowler-Hill, Lisa M Acuff, Orrin B Myers.

  Global health organizations urge the use of empowering language in patient education for type 1 diabetes (T1D). Twenty-nine online patient education materials (PEMs) were assessed in the first phase of a content analysis examining how disempowering terminology is used in both frequency and context. All of the PEMs used disempowering terms, were primarily from the U.S. (70%), and authored by health care providers (34%). Imperative terms carried the highest word count (54%) and mean occurrence rates. Results suggest online PEMs may benefit from regular updates to improve health literacy, proving to be a growing area of focus for health librarians.

Keywords:  Diabetes; empowerment; health communication; language; patient education; stigma

DOI:  https://doi.org/10.1080/02763869.2026.2697173
Z Evid Fortbild Qual Gesundhwes. 2026 Jul 08. pii: S1865-9217(26)00120-0. [Epub ahead of print]

Brief health information leaflets in general practice: A cross-sectional survey on usage, common topics and sources in Austria.

Laura Plotho, Nicole Posch, Christina Radl-Karimi, Ulrike Spary-Kainz, Klaus Jeitler, Andrea Siebenhofer.

   INTRODUCTION: Health information materials are widely available, yet their reliability is often uncertain. General practitioners remain a trusted source of information for patients, and written materials such as leaflets are common in general practice. The topics that are potentially relevant for general practitioners have not yet been sufficiently researched. This study examined the use of brief health information leaflets in Austrian general practice, focusing on their topics, sources, and preferred formats.
METHODS: An online survey was distributed to Austrian general practitioners between November 2023 and January 2024. Data were analysed descriptively, and chi-square tests were applied to assess associations.
RESULTS: Of 4,053 invited general practitioners 891 responded (22.0%) and 501 completed the survey (12.4%). Among respondents, 41.3% reported using health information leaflets. Of the 272 respondents who provided more detailed information, 35.7% used only self-produced materials. External materials most often came from providers of pharmaceuticals and other medical products. In general, 54.1% prefer paper-only formats, while 29.2% prefer a mix of paper and digital formats. Health information on endocrine and metabolic diseases, nutrition, and infectious diseases proved to be particularly relevant - both in terms of current use and demand.
DISCUSSION: Health information leaflets are widely used, with self-produced materials playing a greater role than previously reported. At the same time, there is a clear need for additional materials. It is interesting to note that the demand for health information focuses on the same topics as the materials that are already most frequently used.
CONCLUSION: The results indicate a high demand for health information materials. The reasons for this demand remain unclear and warrant further research. Understanding these factors and the topics sought by GPs can help improve the development of short, evidence-based materials that are more frequently used in primary care.

Keywords:  Evidence-based health information; Evidenzbasierte Gesundheitsinformationen; General practice; Hausarztpraxis; Patient information materials; Patienteninformationsmaterialien; Primary care; Primärversorgung

DOI:  https://doi.org/10.1016/j.zefq.2026.05.010
J Surg Res. 2026 Jul 04. pii: S0022-4804(26)00374-4. [Epub ahead of print]325 720-727

Emergency Surgical Cricothyrotomy Videos Online: Reliability and Training Quality Evaluation.

Fatma Çakmak, Turgut Dolanbay, Bilal Kartal, Mustafa İbaş.

   INTRODUCTION: Emergency cricothyrotomy is a life-saving, high-stakes surgical airway procedure that, although rarely performed, must be executed rapidly and accurately when indicated. Online video-based educational resources are increasingly used to support training in such critical interventions; however, the instructional adequacy and procedural accuracy of these materials remain uncertain. We aimed to evaluate the educational value, reliability, and procedural competency of online videos related to cricothyrotomy and to investigate whether video-related characteristics are associated with content quality.
MATERIALS AND METHODS: Videos meeting predefined inclusion criteria were evaluated using the Global Quality Scale, DISCERN, The Journal of the American Medical Association benchmark criteria, and the cricothyrotomy-specific Cricothyrotomy-Objective Structured Assessment of Technical Skills (C-OSATS) scoring system. Two emergency medicine specialists independently assessed all videos. Inter-rater agreement was examined using the intraclass correlation coefficient. Associations between quality scores and video popularity indicators were analyzed using correlation analysis. In addition, a multivariable logistic regression analysis was performed to identify factors independently associated with high-quality videos.
RESULTS: A total of 83 videos were included in the analysis. A substantial proportion of the videos were insufficient in terms of educational quality and reliability. Video duration showed a significant positive association with quality scores. C-OSATS demonstrated strong correlations with DISCERN and Global Quality Scale and a moderate correlation with The Journal of the American Medical Association criteria. Inter-rater reliability was high across all instruments. In multivariable analysis, video duration (odds ratio [OR]: 2.867, 95% confidence interval [CI]: 1.650-4.979, P < 0.001), number of likes (OR: 1.012, 95% CI: 1.001-1.022, P = 0.026), and uploader type (OR: 0.092, 95% CI: 0.015-0.558, P = 0.009) were independently associated with high-quality videos, whereas views, subscriber count, and total number of videos on a channel were not.
CONCLUSIONS: Many online videos intended to teach emergency cricothyrotomy do not provide adequate educational content to effectively support competency acquisition for this critical procedure. Procedure-specific assessment approaches such as C-OSATS may complement general quality tools and provide a more clinically meaningful evaluation of instructional materials. Content-related characteristics and source credibility appear to be more important than popularity metrics in determining educational quality.

Keywords:  C-OSATS; Cricothyrotomy; Emergency airway management; YouTube

DOI:  https://doi.org/10.1016/j.jss.2026.05.073
BMC Oral Health. 2026 Jul 10.

Navigating digital resources in dental education: a cross-sectional comparison of the educational quality of rubber dam videos on Youtube and Instagram.

Eray Ceylanoglu, Dilek Hancerliogullari, Bengisu Yuksel.

   BACKGROUND: Applying a rubber dam (RD) is fundamental in both endodontic and restorative procedures to ensure a sterile working area, achieve optimal moisture control, and minimize microbial contamination. While platforms like YouTube and Instagram have become go-to resources for dental education, the actual quality of their content needs closer inspection. Therefore, this study sets out to assess how accurate and reliable the information on RD application techniques is across YouTube videos and Instagram Reels.
METHODS: Systematic searches for rubber dam application content were conducted on both platforms on January 18, 2026. After filtering, 39 relevant videos from each source were included. Three independent endodontists evaluated these videos using the Global Quality Scale (GQS), JAMA benchmarking criteria, and the modified Patient Educational Materials Evaluation Tool (mPEMAT). For statistical evaluation, Fleiss Kappa was used to check inter-rater reliability, Mann-Whitney U test was used to compare platforms, and Spearman correlation analysis (p < 0.05) was used.
RESULTS: A strong consensus was observed among the observers; inter-rater reliability across all measures fell within the "good" to "excellent" categories (p < 0.001). YouTube significantly outperformed Instagram Reels, scoring higher on the GQS, JAMA, and mPEMAT scales (p < 0.001 for all). Content uploaded by dentists received the highest scores, regardless of the platform. Longer videos tended to have higher GQS and mPEMAT scores on both platforms.
CONCLUSION: In this limited cross-sectional study, YouTube videos exhibited higher instructional quality and reliability scores compared to Instagram Reels. Rather than establishing an absolute superiority of one platform over the other, it appears that YouTube currently provides a more comprehensive learning resource.

Keywords:  Education; Instagram; Rubber dam application; Social media; YouTube

DOI:  https://doi.org/10.1186/s12903-026-09209-2
BMC Oral Health. 2026 Jul 09.

Educational quality and viewer engagement of Youtube videos on ceramic implants: a cross-sectional study.

Mehmet Zahit Baş, Halil İbrahim Durmuş, Tevfik Kızılseki, Nuran Dağ.

   BACKGROUND: Ceramic implants are increasingly used as metal-free alternatives to titanium implants because of their favorable esthetic properties and biocompatibility. As patients frequently seek dental information online, YouTube has become an important source of patient-oriented health information. This study aimed to evaluate the educational quality and ceramic implant-specific content coverage of YouTube videos on ceramic implants and to examine whether viewer engagement metrics were associated with video quality scores.
METHODS: A cross-sectional analysis was conducted on 63 English-language YouTube videos identified through six predefined keyword searches. Videos were screened independently by two calibrated evaluators according to predefined eligibility criteria. During scoring, evaluators were blinded to viewer engagement metrics. Video quality was assessed using the Video Information and Quality Index, Global Quality Score, and an exploratory, non-validated dentistry-specific Specific Content Checklist developed to evaluate ceramic implant-specific content coverage. Viewer engagement metrics, including views, likes, comments, and subscriber counts, were recorded. Inter-rater reliability was assessed using intraclass correlation coefficients, and relationships between variables were analyzed using Pearson correlation tests.
RESULTS: The mean Video Information and Quality Index, Global Quality Score, and Specific Content Checklist scores were 3.27 ± 0.56, 3.34 ± 0.57, and 3.29 ± 1.68, respectively. Inter-rater reliability showed good agreement between evaluators, with average-measure intraclass correlation coefficients (two-way mixed-effects, consistency) ranging from 0.779 to 0.870. No statistically significant correlations were found between video quality scores and viewer engagement metrics; exact p values for these correlations ranged from .129 to .921. The wider dispersion of Specific Content Checklist scores indicated variability in ceramic implant-specific content coverage across videos.
CONCLUSIONS: Viewer engagement metrics should not be interpreted as reliable indicators of educational quality or ceramic implant-specific content coverage in YouTube videos on ceramic implants. Because the Specific Content Checklist was exploratory and non-validated, findings should be interpreted as indicators of content coverage rather than definitive measures of factual accuracy or clinical adequacy. Clinician-guided digital health literacy remains essential for helping patients interpret online ceramic implant information.

Keywords:  Ceramic implants; Digital health; Implant dentistry; Patient education; YouTube; Zirconia implants

DOI:  https://doi.org/10.1186/s12903-026-09202-9
J Clin Pract Res. 2026 Jun;48(3): 261-270

Is YouTube an Accurate and/or Comprehensive Source of Information on Monosymptomatic Nocturnal Enuresis?

Halil Tosun, Yasar Issi, Ahsen Karagozlu Akgul, Emrah Kizilay, Numan Baydilli, Mesrur Selcuk Silay.

   Objective: Monosymptomatic nocturnal enuresis (MEN) is a common pediatric condition, and misleading or inaccurate information regarding its diagnosis and management is frequently encountered on social media platforms. Therefore, ensuring reliable and high-quality online health information is crucial. This study aimed to assess the accuracy, usefulness, and comprehensiveness of YouTube videos addressing MEN.
Materials and Methods: Five predefined MEN-related keywords were searched on YouTube. Eligible videos were analyzed using the Global Quality Scale (GQS), DISCERN, and Journal of the American Medical Association (JAMA) scoring systems in terms of diagnosis, management, and treatment. Videos were categorized as useful or non-useful according to content quality and were also grouped based on comprehensiveness scores. Comparative statistical analyses were performed between groups.
Results: A total of 153 videos were included. Significant differences were observed between useful and non-useful videos in DISCERN, GQS, total JAMA scores, comprehensiveness scores, and upload source (p<0.001). Videos with higher comprehensiveness scores also differed significantly in DISCERN, GQS, total JAMA scores, upload source, and video duration (p<0.001). No significant differences were found in the video power index, presenter type, or engagement parameters, such as likes and dislikes (p>0.05). Although most videos (79%) were useful, fewer than one-third (28%) were comprehensive.
Conclusion: YouTube provides partially useful information on MEN; however, its overall comprehensiveness remains insufficient. Collaboration with academic institutions and professional organizations may improve the quality and reliability of MEN-related information on social media platforms.

Keywords:  Health information quality; YouTube; monosymptomatic nocturnal enuresis

DOI:  https://doi.org/10.14744/cpr.2026.70082
Oral Health Prev Dent. 2026 Jul 07. 24 527-539

Content Gaps and Informational Quality of TikTok and Bilibili Videos on Pit and Fissure Sealing in China: A Cross-Sectional Study.

Feng Cai, Wending Gao, Linhu Wang.

   PURPOSE: To evaluate the disease-specific content coverage, overall quality, reliability, and transparency of Chinese-language short videos on pit and fissure sealing posted on TikTok and Bilibili.
METHODS AND MATERIALS: In this cross-sectional study, videos were retrieved from TikTok and Bilibili on December 20, 2025, using the Chinese search term '' (pit and fissure sealing). The first 150 results from each platform were screened, and 204 eligible videos were included. Overall quality, reliability, and transparency were assessed using the Global Quality Score (GQS), modified DISCERN (mDISCERN), and Journal of the American Medical Association (JAMA) benchmark criteria, respectively. A predefined disease-specific coding framework was used to evaluate content coverage. Group comparisons, Spearman's rank correlation, and multivariable ordered logistic regression were performed.
RESULTS: Benefits and indications were the most frequently covered domains, appearing in 87.75% and 78.92% of videos, respectively, whereas contraindications and cost were addressed in only 21.57% and 28.92%. The median GQS, mDISCERN, and JAMA scores were all 3.00. No significant differences in validated assessment scores were observed between TikTok and Bilibili. Videos uploaded by healthcare professionals, especially specialised healthcare professionals, achieved higher scores than those uploaded by individual users. Video duration showed weak positive correlations with validated assessment scores, whereas engagement indicators were not meaningfully associated with informational quality.
CONCLUSION: Short videos on pit and fissure sealing provide moderately useful information, but content coverage remains incomplete and uneven. Greater participation by healthcare professionals and better access to balanced, trustworthy preventive information are needed on short-video platforms.

Keywords:  Bilibili; TikTok; oral health communication; pit and fissure sealing; short-video platforms

DOI:  https://doi.org/10.3290/j.ohpd.c_2759
Front Public Health. 2026 ;14 1797405

Quality and reliability of cardiac rehabilitation-related short Chinese videos on Douyin and Bilibili: a cross-sectional content analysis.

Jie Yuan, Zhuo Chen, Yannan Ma, Song Lin, Xiaodong Feng, Yuzhen Sun.

   Background: Social media platforms have emerged as prominent channels for disseminating cardiovascular health information. However, the accuracy, completeness, and clinical reliability of cardiac rehabilitation (CR)-related content vary widely. Therefore, this study This study aims to identify upload sources, contents, and feature information of these videos on Bilibili and Douyin, and further evaluate the video quality.
Methods: A cross-sectional study was conducted on Bilibili and Douyin using the keywords "cardiac rehabilitation ()" and "postoperative coronary heart disease ()." A total of 200 videos were included. Data on video characteristics were collected, including title, uploader identity, upload time, video duration, content type, engagement metrics (likes, comments, and shares), presentation format, and video quality scores. Video quality and reliability were assessed using the modified DISCERN (mDISCERN), the Journal of the American Medical Association (JAMA) benchmark criteria, and the Global Quality Scale (GQS).
Results: Douyin videos were shorter in duration and more recently uploaded, and received significantly more likes and comments than Bilibili videos (P < 0.001). No significant differences were observed between platforms in terms of saves or shares. Engagement metrics differed significantly across uploader categories for likes and comments (P < 0.001). Videos uploaded by hospital departments achieved the highest mDISCERN, JAMA, and GQS scores (P < 0.01). Bilibili videos showed slightly higher JAMA scores than Douyin videos (P = 0.04). Spearman correlation analysis indicated that content quantity was positively associated with both GQS and JAMA scores. Overall, associations between video quality scores and user engagement metrics were weak and inconsistent. GQS showed a weak positive correlation with shares, whereas mDISCERN was negatively correlated with comment counts. No significant association was observed between JAMA scores and engagement indicators.
Conclusions: Our study shows that the quality of short videos on health information related to CR is poor on Bilibili and Douyin. However, videos uploaded by institutional and healthcare professional accounts demonstrate better performance in terms of information reliability and content quality. Therefore, CR information obtained from short-video platforms should be interpreted with caution, and viewers are advised to critically evaluate content credibility before using such information to guide health-related decisions.

Keywords:  cardiac rehabilitation; exercise rehabilitation; public education; public health; public media

DOI:  https://doi.org/10.3389/fpubh.2026.1797405
Medicine (Baltimore). 2026 Jul 03. 105(27): e49577

Quality, content, and reliability of lumbar spondylolisthesis-related videos on short-video platforms: A cross-sectional study of TikTok and Bilibili.

Keli Guo, Yanzheng Gao.

  Short-video platforms have become prominent sources of health-related information for musculoskeletal conditions, yet the quality and reliability of content concerning lumbar spondylolisthesis remain insufficiently evaluated. This study aimed to systematically assess the characteristics, content composition, educational quality, and reliability of lumbar spondylolisthesis-related videos on TikTok and Bilibili. A cross-sectional analysis was conducted on videos retrieved from TikTok and Bilibili on December 25, 2025. Eligible videos were evaluated for general characteristics, uploader type, presentation format, thematic content, and audience engagement. Educational quality and information reliability were assessed using the Global Quality Score, modified DISCERN (mDISCERN), and Journal of the American Medical Association (JAMA) benchmark criteria. Correlations between video characteristics, engagement metrics, and quality scores were analyzed. A total of 199 videos were included (TikTok: n = 105; Bilibili: n = 94). Overall, videos demonstrated moderate educational quality and limited reliability. Treatment-related content predominated, whereas epidemiology, prevention, and rehabilitation were infrequently addressed; notably, etiology ranked as the third most common topic on both platforms. Videos from orthopedic practitioners had higher mDISCERN and JAMA scores but low overall quality; Global Quality Score differed by uploader type and was not highest for orthopedic practitioners. Scientific popularization videos exhibited superior educational quality, while engagement metrics did not reliably reflect informational value. Video duration - particularly on Bilibili - was positively correlated with quality and reliability scores, whereas engagement indicators showed weak or negligible associations. Lumbar spondylolisthesis-related short videos on major platforms provide moderate educational value but show substantial deficiencies in reliability and content balance. Greater professional involvement and improved platform-level quality oversight are warranted to enhance the dissemination of accurate and high-quality health information.

Keywords:  TikTok and Bilibili; health information quality; lumbar spondylolisthesis; medical education; short-video platforms

DOI:  https://doi.org/10.1097/MD.0000000000049577
Front Digit Health. 2026 ;8 1874643

The quality evaluation of information in hospice and palliative care short videos on Douyin (TikTok's Chinese version) and Bilibili: a cross-sectional study.

CongBing Yan, Xiaoxia Zhu, Man Xu, Biyun Xia, Rendong Liang.

   Background: Due to the fast growth of short video services, more and more individuals use platforms like Douyin (the Chinese version of TikTok) and Bilibili to find out some health-related information concerning hospice and palliative care (HPC). Nevertheless, the credibility and reliability of HPC-related materials in these platforms are still ambiguous. Previous studies on health information in short videos have mainly focused on Western platforms such as YouTube and TikTok (international version), leaving a gap in systematic evaluation of HPC content on Chinese platforms.
Objective: The purpose of this paper is to assess the quality and reliability of the Chinese-language short videos related to HPC on Douyin and Bilibili and to analyze the associations between the characteristics of the content and user involvement.
Methods: The present cross-sectional research is a study of the top 100 most popular HPC-related videos on each platform according to the "comprehensive ranking" (a composite algorithm score integrating view counts, likes, comments, shares, and recency, as displayed by default on both platforms). Video quality and reliability were assessed with three validated scoring tools by three independent raters-Global Quality Score (GQS), modified DISCERN (mDISCERN), and JAMA benchmark criteria. Inter-rater reliability was assessed using intraclass correlation coefficients (ICC). The statistical analyses were done using the Mann-Whitney U tests, Kruskal-Wallis tests for multi-group comparisons, and Spearman correlation analyses with 95% confidence intervals (CI).
Results: The analysis covered 200 videos. In Douyin videos, the GQS scores were statistically significantly higher than in Bilibili (the median score was 4.75 compared with 4.40, p < 0.001, r = 0.41, 95% CI: 0.28-0.54); no statistically significant differences in mDISCERN and JAMA scores were identified in the two platforms. Health care professionals had videos with better quality ratings using all three assessment methods. Engagement measures of users (likes, comments, shares) were significantly greater on Douyin than on Bilibili (all p < 0.001). Likes and GQS scores had weak positive correlations (r = 0.376, p < 0.001, 95% CI: 0.24-0.51), indicating that popularity metrics alone are not reliable surrogates of information quality.
Conclusions: In general, the HPC-related short videos had a fair quality on both platforms with major areas to improve the level of reliability of information and the use of evidence in support. The healthcare professional-uploaded content proved to be high-quality. These results underscore the necessity of improving the regulation of content and promoting authoritative health information on the short video platforms.

Keywords:  Bilibili; Douyin (TikTok China); health communication; hospice and palliative care; information quality; short video platforms

DOI:  https://doi.org/10.3389/fdgth.2026.1874643
Medicine (Baltimore). 2026 Jul 03. 105(27): e49431

Quality and reliability of anxiety disorder-related short videos on TikTok and Bilibili: A cross-sectional study.

Biao Jiang, Qian-Ying Liao, Xue-Ping Yang, Yu-Hao Chu.

  In recent years, short-form videos have demonstrated substantial potential in disseminating health-related information, with TikTok and Bilibili emerging as major platforms for public access to such content. However, to the best of our knowledge, no studies have systematically evaluated the quality and reliability of anxiety disorder-related videos on short-video platforms. Therefore, this study aimed to assess the quality and reliability of anxiety-related short videos on TikTok and Bilibili. A total of 250 anxiety disorder-related videos were analyzed from TikTok and Bilibili. Data on video characteristics, uploader identity, and engagement metrics were collected. The Global Quality Score and modified DISCERN were used to assess quality and reliability. The Mann-Whitney U test and Kruskal-Wallis H test were used for group comparisons, and Spearman rank correlation was applied to assess the relationships between engagement and video quality. The study revealed significant differences between the 2 platforms in video length, quality, reliability, and engagement metrics. Specifically, TikTok videos were shorter (median: 96 seconds, interquartile range [IQR]: 56.00-171.50) and received higher engagement (median likes: 5916, median shares: 2016) compared with Bilibili videos (median: 564 seconds, IQR: 200.00-1004.50, median likes: 611, median shares: 135). Videos uploaded by specialists exhibited higher quality and reliability, with a median Global Quality Score of 3.00 (IQR: 2.00-3.00) and a median modified DISCERN score of 2.00 (IQR: 2.00-3.00), significantly outperforming those uploaded by nonspecialists and individual users (P < .05). Content analysis indicated that most videos focused on symptoms (82.4%) and treatment (53.6%), while crucial topics such as diagnosis (9.2%), prevention (9.6%), and epidemiology (4.8%) were less frequently discussed. Anxiety disorder-related videos on TikTok and Bilibili have moderate quality and reliability. Specialist-uploaded videos are superior in quality. Key topics such as diagnosis and prevention should be better covered, and content moderation and professional involvement should be strengthened.

Keywords:  Bilibili; TikTok; anxiety disorder; health communication; information quality; social media

DOI:  https://doi.org/10.1097/MD.0000000000049431
Int J Rheum Dis. 2026 Jul;29(7): e70725

The Quality and Reliability of Short Videos About Psoriatic Arthritis on Bilibili and TikTok: Cross-Sectional Study.

Zenan Tang, Yan Zhao, Xiaoyang Liu.

DOI: https://doi.org/10.1111/1756-185x.70725