bims-librar Biomed News
on Biomedical librarianship
Issue of 2026–07–05
forty-two papers selected by
Thomas Krichel, Open Library Society
  1. Natural Language Processing (NLP) tools are superior to ICD-10 codes in capturing clinically relevant information.
  2. Is "Nurse" Necessary? Questioning Redundancies in Breastfeeding Research Searches.
  3. IDR searcher: a search engine solution for public image resources.
  4. Launching the Re-Engineered tala-med Medical Search Engine: CMS Integration and Index-Optimized ETL Pipeline.
  5. The role of artificial intelligence in academic citation: A study on lens, cataract, and anterior segment research.
  6. Large language model chatbots as sources of pediatric anesthesia health advice: An evaluation of reliability and readability.
  7. What Do Biomedical NER and Entity Linking Benchmarks Measure? A Corpus-Centric Diagnostic Framework.
  8. Health Information Sources Within Low Income Black Communities: A Families SHARE Study.
  9. Navigating digital health: Development and validation of the online health information seeking scale (OHISS).
  10. Skin Health Information Seeking on Short Video Platforms in Indonesia: Mixed Methods Approach.
  11. Improving core facility service discovery with an AI assistant grounded in institutional web content.
  12. A large language model pipeline for automated citation quality scoring across engineering journal quartiles with expert validation.
  13. Readability and Linguistic Characteristics of Alzheimer's Disease and Related Dementias Prevention, Symptom, and Treatment Information from Generative Artificial Intelligence Chatbots.
  14. Performance of large language models as an information resource on functional hypothalamic amenorrhea for patients and healthcare professionals.
  15. Readability, Quality, Understandability, and Actionability of ChatGPT Generated GI Patient Education Versus AGA Patient Center.
  16. Artificial Intelligence Answering Femoroacetabular Impingement Patient Questions: Helpful Tool or Harmful Risk? Evaluating NIPRGPT Answers to Frequently Asked Questions About Femoroacetabular Impingement.
  17. Readability and Quality of Chatbot Responses to Periodontal Patient Queries: A Cross-Sectional Evaluation of Three Publicly Accessible Large Language Models.
  18. Assessment of the accuracy and readability of artificial generative intelligence for patient questions on heavy menstrual bleeding.
  19. Gonarthrosis Advisor vs ChatGPT-5: quality and readability of artificial intelligence-generated patient education for knee osteoarthritis.
  20. Assessing the efficacy of LLMs in neuroscience: A performance analysis of ChatGPT and gemini on synaptic plasticity.
  21. Assessing The Ability of ChatGPT 5.2 to Answer Patient Questions Regarding Hip Avascular Necrosis.
  22. Can artificial intelligence applications be utilised in the management of dental trauma? A comparison of ChatGPT, Gemini, and DeepSeek.
  23. The Challenge of Understanding Gilmore's Groin: An Analysis of Online Patient Resources.
  24. Evaluating Online Information for Young Adult Cancer Symptoms: A Content Analysis of Top Google-Ranked Webpages.
  25. Quality of Allergic Rhinitis Content Across Social Media Platforms.
  26. Performance of large language models in mitral valve surgery patient education: a comparative analysis.
  27. Can social media videos meet patients' information needs for benign prostatic hyperplasia? A repeated cross-sectional study.
  28. Is YouTube a reliable source for toothbrushing information for children with special healthcare needs? A cross-sectional content analysis.
  29. Comment on "Content quality and reliability of youtube exercise videos for parkinson's disease".
  30. The role of digital resources in surgical education: An analysis of YouTube videos on dynamic stabilization.
  31. YouTube As an Information Source for Bioceramic-Based Root Canal Sealers in Endodontics: A Cross-Sectional Study.
  32. Comparative analysis of bias and quality in YouTube videos on bariatric surgery and weight-loss medications.
  33. ASO Visual Abstract: Quality, Reliability, and Viewer Engagement of YouTube Videos on Esophageal Cancer: A Cross-Sectional Study.
  34. YouTube and TikTok as patient resources: assessment of online video content on extracorporeal shock wave lithotripsy.
  35. Dissemination of misleading and biased information about benign prostate hyperplasia on YouTube in the Republic of Korea.
  36. The quality and reliability of short videos about liver transplantation and lung transplantation on BiliBili and TikTok: cross-sectional study.
  37. The quality and reliability of short videos about cervical spondylosis on TikTok and Redbook: a cross-sectional study.
  38. Information quality and reliability of pheochromocytoma-related videos on tiktok and Kwai: a cross-sectional study.
  39. Quality and reliability of rectal prolapse short-video content on tiktok and bilibili: A cross-sectional content analysis.
  40. Evaluation of the quality and reliability of infertility information on Chinese social media platforms Bilibili and TikTok : a cross-sectional study.
  41. Assessment of pelvic floor muscle training video information quality and associated factors on Douyin and Bilibili: a cross-sectional study.
  42. Credibility, educational quality, and specialty-specific depth of meniscal injury information on Douyin: a cross-sectional study.
  1. Pediatr Res. 2026 Jul 03.
    Natural Language Processing (NLP) tools are superior to ICD-10 codes in capturing clinically relevant information.
    Max J Coppes.
      
    DOI:  https://doi.org/10.1038/s41390-026-05277-w
  2. J Hum Lact. 2026 Jun 28. 8903344261460339
    Is "Nurse" Necessary? Questioning Redundancies in Breastfeeding Research Searches.
    Eman Sharara, Bahia Abdallah, Rachael Eastham, Mark Limmer.
       BACKGROUND: Rigorous systematic reviews in breastfeeding research depend on precise and transparent search strategies. However, the term nursing has dual meaning-referring both to breastfeeding and to the nursing profession-creating potential ambiguity in literature searches.
    RESEARCH AIM: To examine the methodological impact of including the terms nurse and nursing in breastfeeding-related search strategies, and to highlight implications for research quality and policy development.
    METHODS: During a global systematic review on partner support and breastfeeding self-efficacy, seven databases were searched from inception to January 2025. Screening revealed a substantial number of irrelevant records related to the nursing profession. A modified search excluding the truncation "nurs" from titles and abstracts was subsequently conducted in Medline Complete (EBSCOhost) and Embase to assess its impact on retrieval volume and relevance.
    RESULTS: Initial searches retrieved 21,214 records, with considerable screening burden attributable to irrelevant nursing-profession literature. When "nurs*" was excluded, records in Medline Complete and Embase substantially decreased by 65% and 63%, respectively, without observed loss of relevant breastfeeding studies. The findings demonstrate that ambiguous terminology can substantially inflate retrievals without enhancing comprehensiveness.
    CONCLUSION: Routine inclusion of nurse and nursing in breastfeeding searches may reduce efficiency and should be carefully considered in relation to the review context.Implications for Practice and Policy:Precision in search terminology is essential to maintain methodological rigor. Researchers, librarians, editors, and peer reviewers should critically evaluate term selection and promote validated, breastfeeding-specific search filters to strengthen evidence informing lactation policy and clinical practice.
    Keywords:  breastfeeding; lactation research; methodological rigor; nursing; search strategy; systematic review
    DOI:  https://doi.org/10.1177/08903344261460339
  3. Bioinformatics. 2026 Jul 01. pii: btag486. [Epub ahead of print]
    IDR searcher: a search engine solution for public image resources.
    Khaled Mohamed, William Moore, Dominik Lindner, Josh Moore, Jason R Swedlow, Petr Walczysko, Frances Wong, Jean-Marie Burel.
       MOTIVATION: In recent years, public image resources have emerged, but finding quality data efficiently remains a challenge, therefore limiting reuse.
    RESULTS: IDR searcher is an open-source search engine designed to facilitate the exploration of datasets hosted in public bioimaging resources. The application offers a fast, efficient, cost-effective solution for discovering datasets and has the potential to address current disparities in finding quality datasets for exploratory research and can be combined with metadata visualization tools to enhance usability for the scientific community.
    AVAILABILITY: IDR searcher is deployed using Ansible playbooks and released under the GPL v2 license. The source code associated with this manuscript is available at https://doi.org/10.5281/zenodo.20641515.
    SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
    DOI:  https://doi.org/10.1093/bioinformatics/btag486
  4. Stud Health Technol Inform. 2026 Jun 29. 338 264-268
    Launching the Re-Engineered tala-med Medical Search Engine: CMS Integration and Index-Optimized ETL Pipeline.
    Shreya Ponnamma, Simjon Radloff, Rudolf Korhummel, Raphael Schmitt.
      The web offers an abundance of health-related information, but quality and trustworthiness are often lacking. To address this, the tala-med search engine was introduced as a non-commercial platform for German evidence-based health information, drawing on hand-picked, quality-assessed sources. An early prototype was positively evaluated, but technical limitations hindered further development. The system was therefore re-engineered using open, modular technologies and a scalable Apache Nutch-based crawler. Key components include an index-optimized ETL pipeline with boilerplate removal, de-duplication, and metadata enrichment, as well as content integration via a WordPress-based CMS. The platform is deployed across two virtual servers to separate crawling from the user-facing service, ensuring stability and scalability. The resulting system demonstrates a modular and maintainable architecture that supports scalable ingestion and flexible content management. The public instance of the platform is accessible at https://suche.tala-med.info. Future work will address automated index evaluation and fine-grained metadata enrichment to support filter-based search and domain adaptation.
    Keywords:  CMS integration; Index-optimized ETL; Information retrieval; Medical search engine; System deployment; Web crawling
    DOI:  https://doi.org/10.3233/SHTI260843
  5. Indian J Ophthalmol. 2026 07 01. 74(7): 1073-1076
    The role of artificial intelligence in academic citation: A study on lens, cataract, and anterior segment research.
    Mustafa Civelekler, Mehmet Çıtırık.
       PURPOSE: To evaluate the accuracy and reliability of four artificial intelligence (AI) models-ChatGPT, Copilot, DeepSeek, and Gemini-in generating PubMed citations for literature related to lens disease, cataracts, iris disorders, and anterior chamber pathology.
    DESIGN: Comparative accuracy assessment study.
    METHODS: Forty standardized clinical paragraphs from The Review of Ophthalmology (4 th edition) were used as test inputs. Each AI model was prompted to generate AMA-11-style PubMed references. Citation accuracy was assessed using predefined criteria, including PubMed verifiability, DOI concordance, and bibliographic accuracy. Two expert reviewers independently classified the citations as fully cited, partially cited, or not cited, and assessed inter-rater reliability.
    RESULTS: The citation accuracy varied significantly among the models. DeepSeek demonstrated the highest accuracy (52.5%), followed by ChatGPT (32.5%) and Copilot (20.0%), whereas Gemini demonstrated the lowest accuracy (2.5%) ( P < 0.001). DOI mismatches were the most common errors across all models. Expert validation confirmed these findings, with DeepSeek producing the highest number of fully cited references. Inter-rater agreement was substantial (Cohen's κ = 0.65).
    CONCLUSION: Domain-specific AI models, particularly DeepSeek, outperform general-purpose models in generating PubMed citations from ophthalmic literature. However, all the evaluated models exhibited citation errors, underscoring the necessity of human verification. AI tools may enhance academic workflows as assistive systems but should not be used autonomously for reference generation in medical research.
    Keywords:  AI hallucinations; AI models; ChatGPT; Copilot; DeepSeek; Gemini; PubMed citations; anterior segment; artificial intelligence; cataract; citation generation; iris disorders; medical literature; ophthalmology; reference management
    DOI:  https://doi.org/10.4103/IJO.IJO_3155_25
  6. Digit Health. 2026 Jan-Dec;12:12 20552076261464749
    Large language model chatbots as sources of pediatric anesthesia health advice: An evaluation of reliability and readability.
    Xue Zhang, Yuchen Dai, Xin Zhao, Lin Wu, Boming Shao, Xisheng Shan, Fuhai Ji, Runzhi Deng, Baojian Zhao.
       Background: Large language models are increasingly used to obtain health information, but their quality in pediatric anesthesia remains insufficiently evaluated. This study aimed to assess the reliability and readability of four widely used AI chatbots in this context.
    Methods: This cross-sectional observational study developed 18 pediatric anesthesia-related questions using Medical Subject Headings terms, online search trend analysis, and commonly queried topics reflecting parental information needs. Each question was submitted under standardized conditions to four generative AI-driven chatbots: OpenAI's GPT-5.1 Thinking, Google's Gemini 3 Pro, Anthropic's Claude Opus 4.5 Extended Thinking, and DeepSeek-V3.2-Speciale. Models were accessed in their vendor-deployed configurations without task-specific fine-tuning. The generated responses were evaluated for information reliability using the Ensuring Quality Information for Patients (EQIP) instrument, DISCERN tool, Global Quality Score (GQS), and Journal of the American Medical Association (JAMA) benchmark criteria. Readability was assessed using seven validated indices including Flesch Reading Ease Score, Flesch-Kincaid Grade Level, Gunning Fog Index, Simple Measure of Gobbledygook, Coleman-Liau Index, Automated Readability Index, and Linsear Write Formula.
    Results: A total of 72 chatbot-generated responses were included for analysis. Significant between-model differences were observed in DISCERN, EQIP, and GQS, while JAMA benchmark scores were consistently low across all models. DeepSeek and Gemini showed higher median reliability scores across several instruments, although significant pairwise differences mainly involved ChatGPT. None of the evaluated models achieved the recommended sixth-grade readability level across any index. Correlations between reliability and readability were non-significant, suggesting that these represent independent dimensions of information quality.
    Conclusions: Current LLM-based chatbots provided pediatric anesthesia information with variable reliability and consistently suboptimal readability. Although certain models demonstrated relatively higher information quality, limited transparency and excessive reading complexity may restrict their suitability for public-facing educational use. These findings highlight the need for improved quality control, enhanced transparency, and readability-focused optimization in pediatric perioperative education.
    Keywords:  digital health information; generative artificial intelligence; large language models; pediatric anesthesia; readability
    DOI:  https://doi.org/10.1177/20552076261464749
  7. ArXiv. 2026 May 19. pii: arXiv:2605.20537v1. [Epub ahead of print]
    What Do Biomedical NER and Entity Linking Benchmarks Measure? A Corpus-Centric Diagnostic Framework.
    Robert Leaman, Rezarta Islamaj, Zhiyong Lu.
      Biomedical named entity recognition (NER) and entity linking (EL) strongly depend on annotated corpora, but the utility of these resources for benchmarking is often assumed rather than characterized. We present a corpus-centric framework for diagnosing benchmark-relevant properties directly from corpus annotations, concept links, train-test splits, document metadata, and terminology mappings. The framework organizes standardized statistics into five families: (1) scale, density and label distribution, (2) lexical and conceptual structure, (3) train-test overlap, (4) metadata composition, and (5) terminology coverage where applicable. Applying the framework to nine corpora spanning diseases, chemicals, and cell types, we find that corpus properties can differ substantially, even when they address the same apparent task. We find differences in the evaluation signal they provide, the generalization demands they impose, the degree of train-test reuse they permit, and the regions of biomedical literature and concept space they represent. These differences suggest that commonly reported corpus statistics can be insufficient to characterize what biomedical NER and EL benchmarks evaluate. We argue that corpus-centric diagnostics provide a practical framework for analyzing corpora beyond surface descriptors such as corpus size and entity type, for identifying potential transfer risks, and for interpreting the scope of benchmarking conclusions. We release the framework as open-source code with an interactive dashboard to support reproducing our analyses and characterizing additional corpora.
  8. J Prim Care Community Health. 2026 Jan-Dec;17:17 21501319261463083
    Health Information Sources Within Low Income Black Communities: A Families SHARE Study.
    Kristen Judy, Sophia Zuniga, Raegan Bishop, Isabel Cordova Amador, Laura Koehly.
      BackgroundAccess to credible health information is critical for health literacy and informed health behaviors. Although digital platforms are widely used, limited technology access and lower eHealth literacy may constrain online health information seeking in some communities. Black and low-income individuals, in particular, may rely less on internet-based sources and more on in-person channels, including social networks and community-based organizations.AimsThis study examined community-based sources of health information used by Black individuals living in a low-income, urban area.MethodsSemi-structured interviews (n = 27) completed 4-weeks post a community education program were transcribed verbatim. Of these, interview transcripts (n = 25) that identified sources of health information or advice were included in the analytic sample. Two coders categorized responses describing sources of health information and advice using inductive content analysis.ResultsHeath information sources that were identified included health care services, community organizations and networks, media, national organizations, commercial pharmacies, and workplaces. Health care services were most frequently cited, with over half of participants naming physicians, clinics, hospitals, insurance carriers, or other providers, followed by community organizations and networks.DiscussionBlack individuals in low-income, urban areas access health information through diverse in-person and media-based sources. Identifying trusted and accessible community channels can inform targeted outreach efforts to improve health literacy and reduce health disparities.
    Keywords:  black communities; community health; health information; health-seeking behaviors; low-income communities
    DOI:  https://doi.org/10.1177/21501319261463083
  9. Int J Clin Health Psychol. 2026 Apr-Jun;26(2):26(2): 100703
    Navigating digital health: Development and validation of the online health information seeking scale (OHISS).
    Gursheel Kaur, Dinesh Chhabra.
      The increasing reliance on online platforms for health information establishes the need to understand people's engagement and attitude towards Online Health Information Seeking (OHIS). The current study aims to develop and validate a comprehensive multidimensional Online Health Information Seeking Scale (OHISS). Items were generated using both inductive and deductive approaches, followed by multiple content and face validity screenings that incorporated qualitative and quantitative evaluations. Data were collected in two stages: Study 1 (N=364) for item development and identifying the preliminary factor structure using Exploratory Factor Analysis (EFA) and Study 2 (N=236) for validating the obtained factor structure using Confirmatory Factor Analysis (CFA). Following the descriptive analysis, EFA using Promax rotation yielded a 20-item scale comprising three dimensions: Perceived Benefits and Engagement (11 items), Perceived Risks (6 items), and Social Influences on OHIS (3 items), collectively accounting for 51.1% of the total variance. This three-factor structure was subsequently confirmed through CFA with DWLS estimator (χ²(167)=271.55, χ²/df=1.63, CFI=0.986, TLI=0.984, RMSEA=0.052, SRMR=0.071, and GFI=0.978). The scale has good internal consistency as demonstrated by Cronbach's alpha (α=.833), McDonald's omega (⍵=.824), Spearman brown split-half (rxx=.886), Guttman's lambda (λ=.885) and test-retest (ICC2,1=.728) coefficient. Construct validity was supported by significant correlations with the Attitude towards Online Health Information and Information-Seeking subscales, as well as Average Variance explained, Composite Reliability, and the Heterotrait-Monotrait Ratio. The scale demonstrated measurement invariance across male and female groups. The results conclude that the OHISS is a psychometrically standardized measure. The OHISS may be used by health practitioners to assess individuals' attitudes toward online health information, identify barriers in accessing digital health content, and support both health-related communication and research.
    Keywords:  Digital health; Online health information benefits; Online health information risks; Online health information seeking; Psychometric validation; Scale construction
    DOI:  https://doi.org/10.1016/j.ijchp.2026.100703
  10. JMIR Dermatol. 2026 Jul 03. 9 e93461
    Skin Health Information Seeking on Short Video Platforms in Indonesia: Mixed Methods Approach.
    Fathi Qushoyyi Ahimsa, Kelvin Saputra, Michelle Elizabeth Amanda Hutasoit, Putu Wuri Handayani, Hana Fitriani.
       Background: Many users have now switched to using short video platforms as the main channel in their search for skin health information. With high internet penetration and a large market for the skincare industry, short video platforms play an important role in the "beauty discovery" process and purchase decisions. However, the increasing consumption of skin health content is also accompanied by the risk of misinformation, uneven content quality, and the dominance of creators who are not health professionals. Therefore, it is important to determine what factors affect the use of short video platforms in the search for skin health information.
    Objective: By adopting the stimulus-organism-response framework, health belief model, and media richness theory, this study aims to analyze the factors that influence the use of short video platforms in the search for skin health information.
    Methods: This study used a mixed methods approach by distributing an online survey to 603 respondents and conducting interviews with 30 interviewees. Survey data were analyzed using the covariance-based structural equation modeling method, and qualitative data were analyzed using the thematic analysis method.
    Results: The results of this study found that perceived usefulness (P=.01), attitude (P=.02), perceived severity (P=.009), and perceived susceptibility (P=.02) directly affected the behavior of seeking skin health information on short video platforms. Health content expressiveness (P=.001) and personalized health insights (P<.001) directly affected perceived usefulness. These findings support the media richness theory, which shows that the expressiveness of health content and personalization can increase the perception of the usability of short video platforms. Perceived interactivity has an influence on attitude (P<.001), which then affects skin health information seeking on short video platforms (SHEs; P=.02). Upward skin comparison also had a direct influence on skin stigmatization (P<.001). Moreover, perceived severity (P=.009) and perceived susceptibility (P=.02) have an effect on SHEs. These findings confirm the health belief model's theory that perception of the severity of a skin problem and the perception of a person's likelihood of developing a skin problem can improve skin health seeking behavior. However, no effect of source credibility (P=.17) and skin stigmatization (P=.30) was found on SHEs. This is due to the user's willingness to exchange aspects of trust in information sources such as the functionality of the platform and familiarity with the platform or the existence of other internal cognitive or psychological aspects that can be investigated in the future.
    Conclusions: This study can provide guidance for the development of more effective health communication strategies in the digital era using short video platforms.
    Keywords:  Indonesia; health information; information seeking; short video platform; skin; social media
    DOI:  https://doi.org/10.2196/93461
  11. J Biomol Tech. 2026 ;37(2): 40-49
    Improving core facility service discovery with an AI assistant grounded in institutional web content.
    Eviatar Weizman, Dan Ben-Avraham, Robert Fluhr.
      Core facilities source advanced technologies and expertise but can remain under-utilized because researchers, students and early-career scientists, struggle to identify relevant units or cannot articulate appropriate technical inquiries. To enhance access to users, we implemented a domain grounded conversational application based on retrieval augmented generation (RAG). It combines advanced general AI-chat behavior with restricted alignment to core-facility services by uniting a Google Gemini File Search as a structured knowledge base and a Perplexity powered web agent for conceptual scientific queries. The system workflow constrains the agents to institutional domains and routes specific operational questions to a file search knowledge base. As a result, it relies on existing core facility websites and is updated in line with them. This article describes the design of the workflow, how the system is optimized including necessary guardrails to prevent general purpose chat. It proposes evaluation metrics such as veracity, cost per interaction, latency, and examples of usage. The chatbot can help researchers better define their experimental needs, discover relevant facilities they had not previously considered, and thereby increase the visibility and accessibility of institutional research infrastructures.
    Keywords:  AI; Chatbot; RAG; core facilities
    DOI:  https://doi.org/10.7171/001c.162898
  12. Sci Rep. 2026 Jul 03.
    A large language model pipeline for automated citation quality scoring across engineering journal quartiles with expert validation.
    Murat Isik, Sultan Guleryuzlu.
      This study proposes and evaluates a two-stage large language model (LLM)-based pipeline for automated citation quality scoring in academic manuscripts. The pipeline operates as follows: in Stage 1, citation sentences are extracted from full-text PDFs and matched to their referenced articles using the Gemini 2.5 Flash model; in Stage 2, each citation-reference pair is scored for semantic relevance on a continuous 0-10 scale by a second LLM inference call operating under a structured five-tier rubric and a skeptical reviewer prompt persona. The pipeline was applied to a corpus of 121 Web of Science (WOS)-indexed engineering articles drawn from journals spanning all four Journal Citation Reports quartile strata (Q1-Q4), yielding 5,615 scored citation-reference pairs. Descriptive analysis revealed an overall mean relevance score of 7.76 (SD = 2.36), with 74.7% of citations rated as Strong or Excellent. A Kruskal-Wallis test confirmed statistically significant score differences across quartile groups (H(3) = 157.10, p < 0.001), though the overall effect size was small (ε² = 0.028). Post-hoc Mann-Whitney U tests with Bonferroni correction identified Q2 articles as recording the highest mean scores (M = 8.04), significantly outperforming Q1 (M = 7.52), Q3 (M = 7.73), and Q4 (M = 7.74). The Q3 versus Q4 comparison was the sole non-significant pairing (p = 0.756), indicating these strata are statistically indistinguishable in citation quality. Spearman correlation yielded a weak negative rank correlation (ρ = -0.105, p < 0.001), with Q1 recording the highest proportion of Irrelevant citations (10.7%). These findings challenge the assumption that citation quality improves monotonically with journal prestige. The lower mean score of Q1 coexists with one of the highest proportions of highly relevant citations, indicating a bimodal rather than uniformly weaker profile, and a systematic annotation showed that context-dependent pointer citations are disproportionately concentrated in the Q1 Irrelevant set. We therefore attribute Q1's pattern to the broader interdisciplinary scope of top-tier articles together with a measurement effect, rather than to any single cause such as AI-assisted writing. The proposed pipeline offers a scalable, content-aware complement to existing academic integrity tools, with practical applications in editorial pre-screening and automated peer review support. An inter-rater reliability study on a stratified subsample of 150 citation-reference pairs showed strong ordinal agreement between the LLM and expert majority vote (Spearman ρ = 0.643, p < 0.001), with exact-category agreement of 48.0% rising to 77.3% under ± 1 adjacent-category tolerance, and highest agreement at the Irrelevant (80.0%) and Excellent (71.0%) poles.
    Keywords:  Citation analysis; Machine learning; Natural language processing; Scientific publishing; Text analysis
    DOI:  https://doi.org/10.1038/s41598-026-60947-3
  13. J Gerontol Nurs. 2026 Jul;52(7): 7-11
    Readability and Linguistic Characteristics of Alzheimer's Disease and Related Dementias Prevention, Symptom, and Treatment Information from Generative Artificial Intelligence Chatbots.
    John Robert Bautista, Olivia Goth, Allison B Anbari, Kimberly R Powell.
       PURPOSE: To examine the readability and linguistic characteristics of Alzheimer's disease and related dementias (ADRD) prevention, symptom, and treatment information from generative artificial intelligence (GenAI) chatbots.
    METHOD: We analyzed 66 outputs from free-to-use GenAI chatbots. We extracted readability (word count, Fleisch Reading Ease, and Fleisch-Kincaid Grade Level) and linguistic scores (analytical thinking, clout, authenticity, and emotional tone) using Microsoft Word and the Linguistic Inquiry and Word Count software. Data were analyzed using descriptive statistics, t tests, analysis of variance, and multivariate analysis of variance.
    RESULTS: ADRD information from GenAI chatbots, especially treatment information, had college-level readability. Linguistic analyses indicate a high analytical thinking score and low scores for clout, authenticity, and emotional tone.
    CONCLUSION: Our sample of ADRD GenAI information exceeded recommended reading levels for patient education materials. Although the outputs exhibited logical thinking, they also included uncertain, inauthentic, and negative tones. ADRD caregivers should be aware of these characteristics when using GenAI chatbots for ADRD information-seeking.
    DOI:  https://doi.org/10.3928/00989134-20260608-03
  14. Front Artif Intell. 2026 ;9 1788928
    Performance of large language models as an information resource on functional hypothalamic amenorrhea for patients and healthcare professionals.
    Nancy Safwan, Jana Karam, Sarah L Berga, Maria D Hurtado Andrade, Kristin Cole, Stacey J Winham, Stephanie S Faubion, Chrisandra L Shufelt.
       Introduction: To assess and compare the accuracy, readability, and overall performance of large language models (LLMs) in answering questions about functional hypothalamic amenorrhea (FHA) for patients and healthcare professionals.
    Methods: A total of 11 patient-level and 15 clinician-level FHA-related questions were entered separately into four LLMs: ChatGPT 3.5 (free version), ChatGPT 4.0 (updated, paid subscription), Gemini, and OpenEvidence. OpenEvidence was used only for clinician-based questions. Responses were evaluated by three expert reviewers blinded to the LLM used who rated them as accurate and complete, accurate but incomplete, or inaccurate. A fourth reviewer resolved discordant scores. Readability for patient-level questions was assessed using the Flesch Reading Ease Score (FRES) and word count. Lower FRES scores indicate more difficult reading. Accuracy and completeness were compared using odds ratios (95% CI) with ChatGPT 3.5 as the reference model, and differences in readability were analyzed using Friedman's test.
    Results: LLM performance varied across question types. For patient-level questions, ChatGPT 4.0 achieved the highest accuracy (9 of 11; 82%), followed by ChatGPT 3.5 and Gemini (each 8 of 11; 73%), with no statistically significant differences. Among clinician-level questions, OpenEvidence demonstrated perfect accuracy (15 of 15; 100%), compared with 93% for and 80% for ChatGPT 4.0 and Gemini. Completeness followed similar patterns, with OpenEvidence providing the most complete clinician responses (93%) and ChatGPT 4.0 the most complete patient-level responses (89%). Readability differed significantly among models (p = 0.012), with Gemini producing the most readable patient-level content (median FRES 43.5 [IQR 36.8-53.4]) compared with ChatGPT 3.5 (30.6 [16.8-48.4]) and ChatGPT 4.0 (28.8 [22.1-37.6]). Word counts did not differ significantly (p = 0.39).
    Discussion: LLMs demonstrated good overall performance in answering FHA-related questions but often provided incorrect or incomplete information. Fine tuning field-specific data, engineered prompts, and obtaining human-in-the-loop feedback may help improve the accuracy of these models.
    Keywords:  functional hypothalamic amenorrhea (FHA); healthcare professional (HCP); large language model (LLM); patient education; patient information
    DOI:  https://doi.org/10.3389/frai.2026.1788928
  15. Dig Dis Sci. 2026 Jul 02.
    Readability, Quality, Understandability, and Actionability of ChatGPT Generated GI Patient Education Versus AGA Patient Center.
    Shivam Chandra, Vineet Kumar, Abhin Sapkota, Robert Kwei-Nsoro, Anas Almoghrabi.
       BACKGROUND AND AIMS: Patients increasingly use the internet and artificial intelligence chatbots to obtain health information, yet the readability, quality, understandability, and actionability of AI-generated gastrointestinal patient education remain unclear. This study compared gastrointestinal patient education from a professional society website with content generated by ChatGPT using validated health literacy instruments.
    METHODS: In this cross-sectional comparative study, 50 gastrointestinal patient education topics from the American Gastroenterological Association patient information website were paired with ChatGPT-generated responses using standardized prompts. Readability was assessed using the Flesch-Kincaid Grade Level, quality of treatment information was evaluated using the DISCERN instrument, and understandability and actionability were assessed using the Patient Education Materials Assessment Tool; scoring was performed by two blinded reviewers. Paired t tests were used to compare mean scores between sources, and intraclass correlation coefficients (ICCs) were used to assess interrater reliability between reviewers.
    RESULTS: Fifty paired topics were analyzed. The mean Flesch-Kincaid Grade Level was higher for ChatGPT than GI website materials (10.33 ± 1.5 vs 8.72 ± 1.7; mean difference, 1.61; P < .001). Differences in DISCERN scores (63.5 ± 5.7 vs 64.3 ± 5.4; mean difference, - 0.8), PEMAT understandability (87.9% ± 6.9% vs 86.5% ± 7.8%; mean difference, 1.4%; P = .33), and PEMAT actionability (78.6% ± 9.8% vs 77.9% ± 10.2%; mean difference, 0.6%; P = .73) were not statistically significant. Inter-rater reliability was excellent across all measures, with intraclass correlation coefficients of 0.97 (95% CI, 0.95-0.99) for PEMAT understandability, 0.96 (95% CI, 0.94-0.98) for PEMAT actionability, and 0.99 (95% CI, 0.98-0.99) for DISCERN.
    CONCLUSION: ChatGPT-generated gastrointestinal patient education demonstrated similar quality, understandability, and actionability compared with professional society materials but was written at a significantly higher reading level. Improving readability may enhance accessibility and support the safe integration of AI-generated patient education.
    Keywords:  Artificial intelligence; Gastroenterology; Health literacy; Patient education; Readability
    DOI:  https://doi.org/10.1007/s10620-026-10087-5
  16. Mil Med. 2026 Jun 30. pii: usag296. [Epub ahead of print]
    Artificial Intelligence Answering Femoroacetabular Impingement Patient Questions: Helpful Tool or Harmful Risk? Evaluating NIPRGPT Answers to Frequently Asked Questions About Femoroacetabular Impingement.
    William J Ferris, Michael G Johnston, Harrison E Diaz, Mark S Katsma, Andrew P Hurvitz, Brian T Barlow, Lucas S McDonald.
       INTRODUCTION: As patients increasingly seek medical information online, artificial intelligence (AI) chatbots like NIPRGPT-the most widely available AI tool for Department of Defense (DOD) computer users-offer a novel resource for addressing queries about femoroacetabular impingement (FAI). To date, there have not been any studies evaluating NIPRGPT responses to orthopedic medical questions. The primary objective of this study was to evaluate the accuracy, comprehensiveness, and readability of NIPRGPT's responses to common FAI-related questions.
    MATERIALS AND METHODS: Twelve frequently asked questions (FAQs) regarding FAI were selected from a curated list and posed to NIPRGPT. The accuracy and adequacy of the responses were graded by a panel of board-certified surgeons as excellent (not requiring clarification), satisfactory (requiring minimal clarification), satisfactory (requiring moderate clarification), or unsatisfactory (requiring substantial clarification). Additionally, readability was assessed using the Flesch-Kincaid readability score.
    RESULTS: Of the 12 responses, four (33.3%) were excellent, requiring no clarification, seven (58.3%) were satisfactory, requiring minimal clarification, and one (8.3%) was satisfactory, requiring moderate clarification. No responses were deemed unsatisfactory. The average quality score was 3.38/4.0. However, the average Flesch-Kincaid readability score was a 19.6 Grade Level, indicating a reading level suited for postgraduate or specialized academic backgrounds. Interobserver agreement was low, with a Krippendorff's alpha of 0.046.
    CONCLUSIONS: NIPRGPT provides answers to FAQs about FAI that are generally accurate and reliable. However, the responses are generated at a complexity level far exceeding the recommended reading level for patient education. While a potentially useful adjunct in military healthcare settings where access may be limited, clinicians must be aware of the high literacy demand placed on patients using this tool.
    DOI:  https://doi.org/10.1093/milmed/usag296
  17. Int J Dent. 2026 ;2026 3188084
    Readability and Quality of Chatbot Responses to Periodontal Patient Queries: A Cross-Sectional Evaluation of Three Publicly Accessible Large Language Models.
    Nicola Alberto Valente, Lorenzo Floris, Chiara Cinquini, Zuhair S Natto, Lorenzo Mordini.
       Objective: The increasing use of large language models (LLMs) as sources of health information raises concerns regarding the quality and readability of patient-directed content. This study aimed to evaluate and compare the readability and quality of responses generated by three publicly accessible LLMs, ChatGPT-4o, Google Gemini 2.0 Flash, and DeepSeek-R1, to frequently asked patient questions related to periodontology.
    Materials and Methods: In this cross-sectional study, 48 real-world periodontal questions were retrieved from Reddit and Quora and entered verbatim into each chatbot (February 2025, default settings). Readability was assessed using Flesch Reading Ease (FRE) and Flesch-Kincaid Grade Level (FKGL). Quality and accuracy were evaluated using the Quality Analysis of Medical Artificial Intelligence (QAMAI) tool by three independent expert periodontists. Mean scores were compared using one-way ANOVA with Tukey post-hoc tests (α = 0.05).
    Results: FKGL differed significantly among models (p = 0.007), with ChatGPT-4o producing the most complex text (11.18 ± 1.55) compared to DeepSeek-R1 (9.84 ± 1.78) and Gemini (9.91 ± 1.89). FRE did not differ significantly (p = 0.144), although all scores (42.37-49.55) fell within the "difficult" readability range. Mean QAMAI scores were comparable across models (p = 0.078), ranging from 18.31 ± 1.90 (Gemini) to 19.10 ± 1.78 (DeepSeek-R1), placing the mean scores at the lower end of the predefined "good quality" range. Subgroup analysis revealed significant readability differences only within the "pockets" domain (p = 0.001), while quality remained broadly consistent across topics.
    Conclusions: All evaluated LLMs produced responses with mean QAMAI scores at the lower end of the predefined good-quality range; however, their readability exceeded recommended standards for public health materials. While LLMs may serve as supplementary educational tools, language simplification strategies and continued professional oversight remain essential to ensure safe and accessible patient information.
    DOI:  https://doi.org/10.1155/ijod/3188084
  18. J Obstet Gynaecol Can. 2026 Jun 29. pii: S1701-2163(26)00244-6. [Epub ahead of print] 103442
    Assessment of the accuracy and readability of artificial generative intelligence for patient questions on heavy menstrual bleeding.
    Olivia Carere, Lauren Clarfield, M Alix Murphy, Azra Shivji, Chelsie Warshafsky, Corey Sermer, Kinshuk Kumar, Ally Murji.
       OBJECTIVES: To evaluate the quality, patient-centeredness, clinician endorsement, and readability of generative artificial intelligence (AI) responses to patient questions about heavy menstrual bleeding (HMB) compared with high-quality patient-facing websites.
    METHODS: A cross-sectional study compared responses generated by ChatGPT and Google Gemini to excerpts from the five highest-quality HMB patient resources identified through a Google Trends-informed search and the QUality Evaluation Scoring Tool (QUEST). Five layperson-style questions representing HMB subtopics (definition, causes, investigations, management, and safety) were submitted to each model. Responses were de-identified and independently evaluated by five gynecologists, blinded to source, using 5-point Likert scales for accuracy/comprehensiveness (quality), empathetic and validating communication style (patient-centeredness), and expert comfort with patient use of the resource to guide understanding (clinician endorsement). Readability was assessed using the Flesch-Kincaid Grade Level (FKGL). Inter-rater reliability was measured using intraclass correlation coefficients (ICCs).
    RESULTS: Thirty-six responses were reviewed (16 AI-generated and 20 web-based). Compared with web-based excerpts, AI responses showed similar quality (3.94/5±0.70 vs. 3.53/5±1.21; p=0.21), lower patient-centeredness (2.81/5±0.82 vs. 3.52/5±1.13; p=0.04), and comparable clinician endorsement (3.56/5±0.92 vs. 3.44/5±1.35; p=0.52). Without impacting content quality, AI responses were written at a lower grade level (7.6±2.4) than web-based excerpts (11.1±5.2; p=0.042). Inter-rater reliability was high (ICC=0.82-0.84).
    CONCLUSIONS: Compared with high-quality web-based resources, AI-generated responses to HMB questions were more inclusive for varying health literacy levels, comparable in quality and clinician endorsement, but performed worse in patient-centeredness. In an evolving digital landscape for health information acquisition, AI-generated responses represent a valuable and safe supplementary or first-line resource for patients.
    Keywords:  Artificial Intelligence; Consumer Health Information; Health Literacy; Information Seeking Behavior; Menorrhagia; Patient Education as Topic
    DOI:  https://doi.org/10.1016/j.jogc.2026.103442
  19. Acta Orthop Traumatol Turc. 2026 Apr 08. 60(2):
    Gonarthrosis Advisor vs ChatGPT-5: quality and readability of artificial intelligence-generated patient education for knee osteoarthritis.
    Süleyman Kaan Öner, Nihat Demirhan Demirkiran, Enes Alptekin Canlı, Arda Bilir.
       OBJECTIVE: The primary objective of this study is to compare the quality and readability of patient education materials generated by a general-purpose large language model (ChatGPT-5) versus a guideline-based, fine-tuned model (the Gonarthrosis Advisor). The study aims to quantify the performance gains achieved through domain-specific, fine-tuning, and reinforcement learning using osteoarthritis clinical guidelines.  Methods: Thirty frequently asked patient questions regarding knee osteoarthritis were compiled from Google's "People Also Ask" feature and outpatient clinical observations in May 2025. Responses were generated in Turkish by both the Gonarthrosis Advisor and ChatGPT-5 to reflect real-world patient education materials. Content quality was assessed by 2 independent orthopedic surgeons who were blinded to model identity to minimize bias. Both reviewers were co-authors of the article yet did not participate in model construction or data analysis. The assessments utilized the DISCERN instrument, a validated 16-item measure for evaluating the reliability and quality of treatment-related information. Readability was analyzed using the Flesch-Kincaid Grade Level (FKGL), Flesch Reading Ease Score (FRES), and Turkish specific indices (Ateşman and Çakır-Demir). For comparability, English-based indices were applied to translated versions of the responses, whereas Turkish indices were applied to the original texts. All responses were anonymized and randomized prior to evaluation. Model identifiers were removed, and each response was presented in a standardized format to ensure blinding of reviewers. Inter-rater reliability was measured using Cronbach's α. Normality assumptions were tested, and Wilcoxon signed-rank tests were used for statistical comparisons. As no human subjects or personal data were involved, ethical approval was not required.  Results: Mean DISCERN scores corresponded to the "good" category (66.4) for the Gonarthrosis Advisor and the 'moderate' category (54.2) for ChatGPT-5, according to established cut-off thresholds. The Gonarthrosis Advisor achieved significantly higher DISCERN scores than ChatGPT-5 (66.4 ± 4.8 vs. 54.2 ± 5.6; P < .001) with high inter-rater reliability (Cronbach's α = 0.86). Readability metrics favored the Gonarthrosis Advisor across all indices: lower FKGL (7.8 ± 0.7 vs. 9.6 ± 0.9) and higher FRES (54.3 ± 3.4 vs. 46.7 ± 3.7), Ateşman (92.0 ± 4.2 vs. 84.3 ± 4.9), and Çakır-Demir (111.7 ± 5.1 vs. 106.9 ± 5.4) scores (all P <.001).  Conclusion: Fine-tuning large language models with guideline-based content and reinforcement learning improves the quality, neutrality, and accessibility of artificial intelligence-generated patient education materials, offering a scalable tool to enhance health literacy and support shared decision-making in knee osteoarthritis care.    Cite this article as: Öner SK, Demirkiran ND, Canlı EA, Bilir A. Gonarthrosis Advisor vs ChatGPT-5: quality and readability of AI-generated patient education for knee osteoarthritis. Acta Orthop Traumatol Turc., 2026, 60(2), 0618, doi: 10.5152/j.aott.2026.25618.
    DOI:  https://doi.org/10.5152/j.aott.2026.25618
  20. Work. 2026 Jul 01. 10519815261462135
    Assessing the efficacy of LLMs in neuroscience: A performance analysis of ChatGPT and gemini on synaptic plasticity.
    Melek Altunkaya, Ercan Babur, Emine Cihan, Cansu Sahbaz Pirincci.
      BackgroundSynaptic plasticity, which plays a critical role in fundamental neurological processes, is a complex subject to master. Therefore, large language models (LLMs) are increasingly being used to facilitate the learning of such complex topics. However, these models have limitations, including producing inaccurate information and failing to capture the nuances of scientific terminology.ObjectivesThis study aimed to evaluate the accuracy, quality and readability of LLM responses to questions on synaptic plasticity.MethodsThe widely used LLMs ChatGPT-4 and Gemini 2.5 were selected in the study. Ten questions were posed to each LLM, and the initial responses were recorded. Five neurophysiologists evaluated the responses qualitatively using a 4-point Likert scale. Readability level of the answers was analyzed using Flesch-Kincaid Grade Level test.ResultsIn the qualitative assessment, both models generally provided accurate and acceptable information. Within the limited scope of the questions analyzed, Gemini received higher median scores in certain instances; however, no statistically significant difference was observed between the two models across most of the question set. Linguistic analysis showed that Gemini's responses were longer and featured a higher Flesch-Kincaid Grade Level, suggesting a structure more aligned with academic or technical discourse.ConclusionFor the specific neuroscientific inquiries examined in this study, both LLMs demonstrated a high capacity for generating accurate content. While Gemini's responses exhibited a more technical linguistic profile, the findings are context-specific and further research is needed to determine if these trends persist across broader scientific domains and larger datasets.
    Keywords:  education; generative artificial intelligence; large language models; learning; physiology; synaptic plasticity
    DOI:  https://doi.org/10.1177/10519815261462135
  21. Acta Orthop Traumatol Turc. 2026 Mar 09. 60(2):
    Assessing The Ability of ChatGPT 5.2 to Answer Patient Questions Regarding Hip Avascular Necrosis.
    Ercan Şahin, Çağatay Baltacı, Hakan Kocaoğlu, Mahmut Kalem.
       OBJECTIVE: This study aimed to evaluate the quality and readability of Chat Generative Pretrained Transformer (ChatGPT) 5.2's responses to frequently asked patient questions about hip avascular necrosis (AVN), a challenging condition often requiring clear communication and shared decision-making.  Methods: Sixteen commonly asked patient questions regarding hip AVN were submitted to ChatGPT 5.2 without follow-up queries. Each response was independently evaluated by 2 orthopedic surgeons with over 20 years of experience in hip arthroplasty. The quality of responses was assessed using the grading system proposed by Mika et al. Readability was analyzed using the Flesch-Kincaid Reading Ease Score (FRES) and Flesch-Kincaid Reading Level (FKRL). Interrater reliability (IRR) was calculated using Cohen's kappa test.  Results: Reviewer 1 rated 11/16 responses as "excellent-no clarification required" and 5/16 as "satisfactory-minimal clarification needed." Reviewer 2 rated 10/16 responses as excellent and 6/16 as satisfactory. The mean FRES score was 27.1 (range: 12.4-45.6), indicating the content was "difficult to read." The FKRL scores corresponded to college or college graduate reading levels. The IRR between reviewers was moderate (κ = 0.59, 95% CI: 0.09-1.00).  Conclusion: ChatGPT 5.2 provided overall satisfactory to excellent responses regarding hip AVN. However, the high reading level required to understand these answers may limit their effectiveness in patient education unless simplified language is employed.    Cite this article as: Şahin E, Baltacı Ç, Kalem M, Kocaoğlu H. Assessing the ability of ChatGPT 5.2 to answer patient questions regarding hip avascular necrosis. Acta Orthop Traumatol Turc., 2026; 60(2), 0629, doi:10.5152/j.aott.2026.25629.
    DOI:  https://doi.org/10.5152/j.aott.2026.25629
  22. BMC Oral Health. 2026 Jun 30.
    Can artificial intelligence applications be utilised in the management of dental trauma? A comparison of ChatGPT, Gemini, and DeepSeek.
    Dolu Kaya.
       BACKGROUND: This study aimed to evaluate the accuracy and reliability of the responses provided by the artificial intelligence applications (chatbots) ChatGPT 4o Plus, Google Gemini Pro, and DeepSeek V3 to questions regarding traumatic dental injuries.
    METHODS: 14 open-ended questions were prepared to ask the chatbots regarding crown-root fractures, luxation injuries, and avulsion. Each question was asked to each application three times a day for three days. Responses were evaluated using the Global Quality Score (GQS) and modified DISCERN (mDISCERN) scales. Data were analysed using the SPSS programme. The Mann-Whitney U and Kruskal-Wallis tests were used in the evaluation. The significance level was set at p < 0.05.
    RESULTS: The average GQS scores for DeepSeek, ChatGPT, and Google Gemini were 4.26, 4.36, and 4.43, respectively. The mean mDISCERN scores were 3.86, 4.01, and 3.85, respectively. A significant difference was found between the GQS and mDISCERN scores for the chatbots (p < 0.05). There was no significant difference between the GQS and mDISCERN scores for crown-root fractures (p > 0.05). While there was no significant difference in GQS scores for luxation injuries and avulsion (p > 0.05), a significant difference was found in mDISCERN scores (p < 0.05). Gemini received the highest score for most questions regarding GQS, while ChatGPT received the highest score regarding mDISCERN.
    CONCLUSIONS: The findings of this study indicate that all three chatbots provide valuable support to dentists in the management of dental trauma. Although chatbots appear to facilitate emergency management in dental trauma, they are still under development, and additional information sources should also be used.
    Keywords:  Artificial intelligence; ChatGPT; Deepseek; Endodontics; Gemini; Large language models; Tooth injuries
    DOI:  https://doi.org/10.1186/s12903-026-08625-8
  23. Cureus. 2026 May;18(5): e109872
    The Challenge of Understanding Gilmore's Groin: An Analysis of Online Patient Resources.
    Victoria Byrne, Shane C Irwin, John P Gibbons.
      Objective This study aimed to assess the quality and readability of online information available to patients on Google regarding Gilmore's groin. Methods This descriptive cross-sectional study evaluated webpages identified through Google searches using the terms "sports hernia", "athletic pubalgia", "Gilmore's groin", "sportsman's hernia", and "hockey hernia". The first page of results for each search term was screened. Duplicate links, non-functioning pages, and irrelevant results were excluded. Unique webpages meeting the eligibility criteria were analysed. Readability was assessed using the Gunning Fog Index (GFI), Flesch-Kincaid Grade Level (FKGL), and Flesch Reading Ease (FRE) score. Each webpage was further evaluated for source type, intended audience, presence of relevant media, inclusion of key clinical information, and quality using the Journal of the American Medical Association (JAMA) benchmark criteria. Descriptive statistics were used to summarise the findings. Results A total of 26 unique webpages were included. Hospital or clinic websites accounted for 13 (50%) webpages, and 16 (62%) were primarily directed toward patients. Relevant images were present in 11 (42%) webpages and relevant videos in three (11.5%). Information on cause and symptoms was provided in 26 (100%) webpages, investigations in 22 (85%), treatment in 25 (96%), and prognosis in 15 (58%). With respect to JAMA benchmarks, authorship was reported in 16 (61.5%) webpages, attribution in 13 (50%), disclosure in 21 (81%), and currency in 17 (65%). Mean readability scores were 11.5 for GFI, 9.9 for FKGL, and 43.5 for FRE, indicating that the material was generally written above the recommended reading level for patient education resources. Conclusion Online patient information on Gilmore's groin is widely available but is typically written at a reading level that is too advanced for the general public. Improving readability while maintaining accuracy may enhance patient understanding, support shared decision-making, and improve access to health information.
    Keywords:  accessible healthcare; athletic pubalgia; gilmore’s groin; online medical information; sports hernia; sportsman’s hernia
    DOI:  https://doi.org/10.7759/cureus.109872
  24. J Adolesc Young Adult Oncol. 2026 Jul 02. 21565333261464961
    Evaluating Online Information for Young Adult Cancer Symptoms: A Content Analysis of Top Google-Ranked Webpages.
    Natasha C Allard, Heather Orom.
       BACKGROUND: Among young adults (20-39), cancer is the fifth leading cause of death. Delayed diagnoses in this population are frequent, contributing to reduced survival and higher morbidity. Delays may be driven by individuals attributing symptoms as nonserious and failing to seek timely medical care. Google search is commonly used for health information seeking, but we do not know the current online symptom content quality that young adults may encounter.
    PURPOSE: We aimed to answer 1. What is the content quality of top-ranked webpages for common young adult cancer symptom searches? 2. Does quality differ by website type (e.g., academic/health care vs. for profit)?
    METHODS: Using 18 young adult cancer symptoms as input into the SEMRush Keyword Magic Tool, we generated a list of the most common keyword searches and the top-ranked webpages (i.e., first three pages listed in Google output). We evaluated 162 pages on 9 quality metrics, including the JAMA benchmark criteria.
    RESULTS: Two-thirds of pages (66.7%, n = 108) were written at less than a 9th-grade reading level, and three-quarters (72.8%, n = 118) provided actionable content about when to seek medical care for symptoms. However, only 13.6% (n = 22) of pages included content framed for young adults. On average, pages met about half (2.33) of four JAMA criteria (authorship, disclosures, currency/up-to-date, and references).
    CONCLUSION: Academic/health and government organizations should devote resources to improving information about young adult cancer symptoms on their webpages and optimize these pages to appear higher in search result rankings.
    Keywords:  cancer; online information seeking; symptom appraisal; website optimization; young adult
    DOI:  https://doi.org/10.1177/21565333261464961
  25. Laryngoscope Investig Otolaryngol. 2026 Aug;11(4): e70490
    Quality of Allergic Rhinitis Content Across Social Media Platforms.
    Emily Kwon, Lucy Revercomb, Alexandra Filipkowski, Annie Xu, Brandon Gold, Rachel Kaye, Wayne D Hsueh.
       Objectives: To assess the quality and engagement of allergic rhinitis-related short-form content on social media.
    Methods: A search across TikTok, Instagram Reels, Facebook Reels, and YouTube Shorts was conducted for allergic rhinitis-related hashtags. Posts were categorized by content category, author, and popularity. Content was analyzed with the Patient Education Materials Assessment Tool for Audiovisual Material (PEMAT-AV) to assess understandability, Global Quality Scale (GQS) to measure quality, and Accuracy in Digital-health Information (ANDI) to measure accuracy. Videos were independently assessed by two reviewers per platform; a third reviewer resolved discrepancies.
    Results: Four hundred and sixty videos were analyzed. Most (69.1%) were educational and authored by medical professionals (38.5%) or lay individuals (48.9%). YouTube had the greatest proportion of medical professional content (56.7%). Instagram scored highest for PEMAT understandability (72.0%) versus YouTube (70.9%), TikTok (61.1%), and Facebook (45.9%) (p < 0.001). YouTube scored highest for PEMAT actionability (55.6%) versus Instagram (48.2%), TikTok (45.3%), and Facebook (35.4%) (p < 0.001). YouTube had the highest average GQS (3.12) compared with Instagram (2.70), TikTok (2.21), and Facebook (2.09) (p < 0.001). YouTube also had the highest average ANDI score (2.80) compared with Facebook (2.19), TikTok (1.91), and Instagram (1.68) (p ≤ 0.026). On multivariable analysis, medical professionals were associated with greater GQS, ANDI, PEMAT understandability, and PEMAT actionability scores (all p < 0.001).
    Conclusions: Among the platforms studied, Instagram and YouTube content just barely met the PEMAT understandability threshold (≥ 70%). However, none of the platforms on average contained adequately actionable or completely accurate content. YouTube Shorts host the greatest proportion of medical professional content and the highest-quality content relative to other platforms.
    Level of Evidence: 4.
    Keywords:  ANDI; GQS; PEMAT‐A/V; allergic rhinitis; hay fever; seasonal allergies; short‐form content; social media
    DOI:  https://doi.org/10.1002/lio2.70490
  26. BMC Med Inform Decis Mak. 2026 Jun 29.
    Performance of large language models in mitral valve surgery patient education: a comparative analysis.
    Banu Bahriye Akdag, Mehmet Senel Bademci, Ihsan Peker, Guven Okay Karaca, Cagrı Kandemir, Barcin Özcem, Huseyin Durmaz, Meryem Cakır, Irem Ozcetin, Hidayet Onur Selcuk.
       BACKGROUND: Large language models (LLMs), a form of artificial intelligence, are increasingly being utilized in healthcare to support patient education and information delivery. The aim of this study was to perform a comparative analysis of five different LLMs (i.e., ChatGPT-4o, Claude 3.7 Sonnet, Gemini 2.5 Pro Preview, DeepSeek-V3, and Microsoft Copilot) in terms of accuracy, completeness, and readability, based on their responses to frequently asked questions in preoperative patient education for mitral valve surgery (MVS).
    METHODS: A standardized questionnaire comprising seven frequently asked questions by patients prior to MVS was developed. Prompting procedures and model parameters were fully reported to support reproducibility. These questions were presented to each LLM in an identical manner. The responses were evaluated by two academic experts in cardiac surgery using structured assessment criteria across three main dimensions: accuracy, completeness, and readability. For the readability analysis, the Simplified Measure of Gobbledygook (SMOG) Index and the Flesch-Kincaid Grade Level (FKGL) scale were utilized.
    RESULTS: The ChatGPT-4o and Gemini 2.5 Pro Preview models received statistically significantly higher scores than Claude 3.7 Sonnet and Microsoft Copilot for both accuracy (median 5 for ChatGPT-4o and Gemini 2.5 Pro Preview vs. 4 for Claude 3.7 Sonnet and Microsoft Copilot, p < 0.001) and completeness (median 5 for Gemini 2.5 Pro Preview vs. 3 for Claude 3.7 Sonnet, p < 0.001). Claude 3.7 Sonnet achieved the highest readability scores, with significantly lower SMOG (10.90 for Claude 3.7 Sonnet vs. 12.24 for ChatGPT-4o, p = 0.006) and FKGL (8.0 for Claude 3.7 Sonnet vs. 9.04 for ChatGPT-4o, p = 0.004) scores, indicating simpler and more comprehensible sentence structures. Significant differences were observed among the evaluated models across all three assessment dimensions (p < 0.001 for all comparisons).
    CONCLUSIONS: The LLMs represent valuable supplementary tools in patient education processes. However, their implementation in clinical practice must be carefully evaluated, particularly with regard to accuracy and completeness. This study highlights the potential applicability of ChatGPT-4o and Claude 3.7 Sonnet models for preoperative patient education in MVS, while emphasizing that all LLMs should be used under the supervision and guidance of healthcare professionals. For LLMs to be reliably utilized in the medical field, improvement in medical accuracy and standardization are essential.
    Keywords:  Accuracy; Artificial intelligence; Large language models; Mitral valve surgery; Patient education; Readability
    DOI:  https://doi.org/10.1186/s12911-026-03662-3
  27. BMC Urol. 2026 Jun 29.
    Can social media videos meet patients' information needs for benign prostatic hyperplasia? A repeated cross-sectional study.
    Chengyang Tian, Weibo Wang, Junyue Tao, Guangyue Luo, Jun Zhou.
       BACKGROUND: Benign prostatic hyperplasia (BPH) is a common urological condition among older men and can cause lower urinary tract symptoms (LUTS). As patients increasingly seek disease-related information through social media, evaluations of the quality, reliability, and transparency of patient-facing BPH content remain limited.
    METHODS: Repeated cross-sectional searches were conducted on Bilibili and TikTok on December 1, 16, and 31, 2025. Basic video characteristics were extracted. Video reliability, quality, and transparency were assessed using the modified DISCERN (mDISCERN), Global Quality Score (GQS), and the Journal of the American Medical Association (JAMA) benchmark criteria, respectively. Exploratory supplementary assessments of content coverage and accessibility-related features were conducted using the content coverage checklist (CCC) and video accessibility checklist (VAC). Potentially misleading or harmful information signals were also descriptively assessed. Spearman correlation analysis and ordinal logistic regression were performed to examine associations between video characteristics and assessment scores.
    RESULTS: A total of 217 videos were included. Overall, the median (IQR) scores were 3.00 (2.00-3.00) for mDISCERN, 3.00 (2.00-3.00) for GQS, and 2.00 (2.00-2.00) for JAMA. In exploratory supplementary analyses, the median (IQR) scores were 3.00 (2.00-4.00) for CCC and 6.00 (5.00-6.00) for VAC. Potentially misleading or harmful information signals were observed in 29 videos (13.4%). Videos published by professional individuals and professional institutions had significantly higher mDISCERN, GQS, and JAMA scores than those published by non-professional individuals (all P < 0.001). Spearman correlation analysis and ordinal logistic regression indicated that engagement metrics were not independent predictors of assessment scores.
    CONCLUSIONS: Overall, BPH-related videos on social media showed moderate reliability, quality, and transparency, but consistently high-value content remained limited. A minority of videos also contained potentially misleading or harmful information signals. Videos from professional sources tended to provide higher-quality information, whereas user engagement was not a reliable indicator of content quality. These findings suggest that urologists and healthcare institutions may have an important role in providing more structured, understandable, and patient-oriented digital education for men seeking BPH information online.
    Keywords:  Benign prostatic hyperplasia; Digital health communication; Health information quality; Patient education; Social media videos
    DOI:  https://doi.org/10.1186/s12894-026-02234-x
  28. J Indian Soc Pedod Prev Dent. 2026 May 01. 44(3): 252-258
    Is YouTube a reliable source for toothbrushing information for children with special healthcare needs? A cross-sectional content analysis.
    Sairab Manzoor, Ramsiya Rahim, Mugilan Ravi, Tundup Dolker, Rhea Sudesh Gandham, Virat Galhotra.
       AIM: To evaluate the characteristics, engagement, content, and quality of YouTube videos addressing toothbrushing considerations for children with special healthcare needs (CSHCN).
    SETTINGS AND DESIGN: A cross-sectional content analysis of information provided on YouTube regarding toothbrushing information for special children.
    METHODOLOGY: Two independent reviewers screened 200 YouTube videos using two search terms based on Google Trends, followed by screening and scoring of the included videos using a 13-point customized scale and a 5-point Global Quality Scale.
    STATISTICAL ANALYSIS USED: Kruskal-Wallis and Chi-square tests were used.
    RESULTS: Thirty-one videos met the inclusion criteria. The majority (48%) of the videos were found to have moderate information content. Less than 10% of the videos received a good quality rating, highlighting a general lack of high-quality content. Nonprofessional sources performed better in terms of views, likes, interaction indices, and Video Power Index; however, these results were statistically significant only for likes ( P = 0.048).
    CONCLUSION: The findings of this study reveal that while the majority of uploaded YouTube videos contain useful information on oral health maintenance for CSHCN, the content is often incomplete and varies significantly in quality.
    CLINICAL SIGNIFICANCE: The lack of high-quality professional content for oral care for CSHCN on YouTube necessitates that clinicians actively vet digital resources for caregivers. Dental professionals must evolve into curators of digital health information to counter the popularity of nonprofessional, low-quality content.
    Keywords:  Children with special healthcare needs; YouTube; content analysis; oral hygiene; toothbrushing
    DOI:  https://doi.org/10.4103/jisppd.jisppd_156_26
  29. Acta Neurol Belg. 2026 Jul 01.
    Comment on "Content quality and reliability of youtube exercise videos for parkinson's disease".
    Pulen Das, Saon Sanyal Bhowmik, Tiyasa Hoom.
      
    Keywords:  Digital health education; Exercise videos; Health information quality; Parkinson’s disease; Physiotherapy; Reliability; YouTube
    DOI:  https://doi.org/10.1007/s13760-026-03122-9
  30. Technol Health Care. 2026 Jun 30. 9287329261464404
    The role of digital resources in surgical education: An analysis of YouTube videos on dynamic stabilization.
    Sezer Onur Gunara, Baris Aslanoglu, Mehmet Yigit Akgun, Ozkan Ates.
      BackgroundThe use of social media platforms like YouTube has surged among patients and their families seeking medical information. Despite the widespread use of video content for health education, no previous study has systematically evaluated the quality of YouTube videos on dynamic spinal stabilization.MethodsA YouTube search using the keyword "dynamic stabilization" was conducted in December 2024. Thirty eligible videos were assessed independently by two neurosurgery specialists using the DISCERN scale. Video characteristics and content quality were analyzed using descriptive statistics, correlation analysis, and regression models.ResultsThe overall quality of the videos was low, with a mean DISCERN score of 39.15. A negative correlation was found between the number of views and DISCERN score (r = -0.28), while positive correlations were observed between DISCERN scores and video duration (r = 0.19), and between DISCERN-1 and DISCERN-2 scores (r = 0.957). Videos including: a medically trained speaker,explanation of dynamic stabilization differences,surgical procedure details, andpre/post-operative information scored significantly higher, with each factor contributing to an average increase of 7.09 points (p < 0.05). ConclusionMost YouTube videos on dynamic stabilization offer low-quality and potentially misleading information. Medically trained contributors consistently produce more reliable and informative content, yet such videos often receive less viewer engagement. Increasing the visibility and clarity of evidence-based content is essential to improve patient education through social media.
    Keywords:  DISCERN; YouTube; dynamic stabilization; patient education; social media
    DOI:  https://doi.org/10.1177/09287329261464404
  31. Aust Endod J. 2026 Jul 02.
    YouTube As an Information Source for Bioceramic-Based Root Canal Sealers in Endodontics: A Cross-Sectional Study.
    Sevda Durust Baris, Seyma Atli, Rumeysa Betul Gumus.
      This study aimed to evaluate the content and quality of YouTube videos on bioceramic-based root canal sealers. A total of 43 videos were assessed using the Journal of the American Medical Association (JAMA), Global Quality Score (GQS) and Modified DISCERN (mDISCERN) measurement tools. Statistical analyses were conducted using the Kruskal-Wallis test, Spearman correlation analysis and multiple linear regression (p < 0.05). Videos uploaded by dentists had significantly higher GQS, mDISCERN, JAMA and Total Content Score (TCS) scores than those from other sources (p < 0.05). Videos with rich content showed higher GQS, mDISCERN and JAMA scores than those with poor content (p < 0.05). Although the number of comments demonstrated an individual association with GQS, the overall regression model was not statistically significant (p > 0.05). In conclusion, most videos on bioceramic-based root canal sealers were of poor to moderate quality and could have limitations in reliability.
    Keywords:  YouTube; bioceramic‐based root canal sealers; endodontics; internet; social media
    DOI:  https://doi.org/10.1111/aej.70105
  32. Surg Endosc. 2026 May;40(5): 4404-4409
    Comparative analysis of bias and quality in YouTube videos on bariatric surgery and weight-loss medications.
    Julia Botvinov, Mason Henrich, Qichen Deng, Aram Jawed, Aziz M Merchant.
       BACKGROUND: Weight-loss surgery (WLS) remains the most effective treatment for clinically severe obesity, while weight-loss medications (WLM), particularly GLP-1 agonists, have recently gained widespread public attention. With YouTube serving as a major source of health information, the quality and tone of videos discussing these interventions warrant evaluation. This study compares the source, content quality, and perceived tone of YouTube videos related to WLS and WLM.
    METHODS: Using an incognito browser, the first 200 "most relevant" YouTube videos for the search terms "weight loss surgery" and "weight loss medicine" were screened in October 2023. Eligible videos were categorized by source and content type. Video quality was assessed using the DISCERN instrument, and tone was classified as positive, neutral, or negative. Two independent reviewers scored all videos, with a third reviewer resolving disagreements. Descriptive statistics and nonparametric tests were used for comparisons, with statistical significance set at p < 0.05.
    RESULTS: A total of 129 WLS and 111 WLM videos were analyzed. WLM videos demonstrated significantly higher view ratios (p < 0.0001) and more comments (p = 0.0014). WLM videos were predominantly commercial (76%) and informational (97%). Mean DISCERN scores were higher for WLM videos (46.69 ± 8.75; fair) compared with WLS videos (40.32 ± 8.70; poor; p = 0.0366). In contrast, WLS videos more frequently originated from academic centers (50%) and included patient experience content (49%). Tone differed significantly between groups (p < 0.01): 60% of WLS videos were positive, whereas WLM videos exhibited fewer positive (44%) and more negative tones (14%).
    CONCLUSIONS: WLM videos were of higher informational quality but conveyed a less positive overall tone, whereas WLS videos presented more positively but were of lower quality. These findings highlight substantial variation in publicly accessible YouTube content on weight-loss interventions and underscore the need for improved, high-quality patient education materials.
    Keywords:  Bariatric surgery; GLP-1 agonists; Patient education; Video quality; Weight-loss medications; YouTube
    DOI:  https://doi.org/10.1007/s00464-026-12795-5
  33. Ann Surg Oncol. 2026 Jul 03.
    ASO Visual Abstract: Quality, Reliability, and Viewer Engagement of YouTube Videos on Esophageal Cancer: A Cross-Sectional Study.
    Sedat Carkit, Mustafa Karaagac.
      
    DOI:  https://doi.org/10.1245/s10434-026-20096-3
  34. Cent European J Urol. 2026 ;79(2): 178-184
    YouTube and TikTok as patient resources: assessment of online video content on extracorporeal shock wave lithotripsy.
    Ayberk Iplikci, Kenan Toprak, Ozgur Arikan, Asif Yildirim.
       Introduction: Our objective was to determine the content, reliability, and quality of YouTube and TikTok videos related to extracorporeal shock wave lithotripsy (SWL).
    Material and methods: The key word "shock wave lithotripsy" was searched on YouTube and TikTok in July 2025 and the first 100 videos on both platforms were watched. Video characteristics were documented, and each video was independently evaluated by two urologists using a custom-designed comprehensiveness scale, along with the modified DISCERN and Global Quality Scale (GQS).
    Results: Sixty-five YouTube and 23 TikTok videos met the inclusion criteria. While YouTube videos were longer in duration (p <0.001), TikTok videos achieved a significantly greater median view count (p = 0.048). YouTube videos demonstrated significantly higher scores on the comprehensiveness scale, modified DISCERN, and GQS compared to TikTok videos, respectively (p = 0.001, p <0.001, p = 0.002). Most narrators were urologists (54.5%), and their videos scored significantly higher GQS compared to others (p = 0.001). Comprehensiveness scale modified DISCERN and GQS scores were significantly positively correlated with duration of video (p <0.001). The scales demonstrated positive intercorrelations among themselves (p <0.001).
    Conclusions: YouTube videos about SWL were of higher quality, reliability, and comprehensiveness than TikTok videos, although TikTok content attracted more viewers. The results emphasize that popularity does not equate to accuracy and underscore the importance of increasing the availability of high-quality, expert-led educational content across social media platforms.
    Keywords:  TikTok; YouTube; extracorporeal shock wave lithotripsy; internet; social media; urolithiasis
    DOI:  https://doi.org/10.5173/ceju.2025.0306
  35. Prostate Int. 2026 Jun;14(2): 151-155
    Dissemination of misleading and biased information about benign prostate hyperplasia on YouTube in the Republic of Korea.
    Min S Song, Jeongwan Yoo, Hyun K Jang, Byeong J Jeon, Bum S Tae, Jae H Bae, Jae Y Park.
       Background: YouTube, a social media platform with over one billion users worldwide, has become a prominent source for health information among the public. Many individuals now turn to YouTube when seeking medical knowledge. This study aimed to evaluate the quality of health information presented in videos concerning benign prostatic hyperplasia (BPH) on YouTube.
    Methods: We conducted a quality assessment of BPH-related content on YouTube in the Republic of Korea, analyzing the 100 most-viewed videos. The validated DISCERN instrument, a standardized tool for evaluating consumer health information, was used. DISCERN consists of 16 questions, each scored on a five-point scale (1 = poor, 2 = generally poor, 3 = moderate, 4 = good, 5 = excellent).
    Results: The top 100 videos collectively accumulated 24.7 million views, with a median of 89,764 views per video and a median of 1,100 thumbs up per video. The mean DISCERN score for overall quality was 2.8, indicating a moderate level of information quality. Videos uploaded by urologists working in university hospitals and those posted by laypersons had a comparable number of views per video, while the latter received more thumbs up per video (2,752 vs 4,012). Fifty-two videos (52%) contained potentially misleading or commercially biased content, accumulating a total of 16.4 million views (3,15,456 views per video) and 5,182 thumbs up per video. A significant negative correlation was found between overall quality and views per month (r = -0.38, P < 0.01), whereas no significant correlation was observed between scientific quality and the thumbs up/views ratio (r = 0.06, P = 0.58).
    Conclusions: Low quality and layperson-generated YouTube videos on BPH gained more engagement than high-quality content. This highlights the paradoxical appeal of misleading and biased information and the need for stronger involvement of healthcare professionals to promote accurate and unbiased resources online.
    Keywords:  Consumer Health Information; Prostatic Hyperplasia
    DOI:  https://doi.org/10.1016/j.prnil.2025.12.008
  36. Sci Rep. 2026 Jul 01.
    The quality and reliability of short videos about liver transplantation and lung transplantation on BiliBili and TikTok: cross-sectional study.
    Fan Yang, Gang Ren, Yixing Li, Ruichao Wu, Xibing Zhang, Guangjian Zhang.
      Liver and lung transplantation are life-saving treatments for end-stage organ failure. While short-video platforms have become crucial health information sources, the educational quality and reliability of transplantation-related content remains uncertain. This cross-sectional study evaluated 416 liver (223) and lung (193) transplantation-related videos from TikTok and Bilibili in China (December 2025) using four validated instruments: Global Quality Score (GQS), modified DISCERN (mDISCERN), Medical Quality Video Evaluation Tool (MQ-VET), and Video Information and Quality Index (VIQI). Overall content quality was moderate to low. Videos by medical practitioners achieved statistically significantly higher quality scores than non-medical videos, though absolute differences were modest. TikTok videos achieved higher overall quality scores, while Bilibili videos showed better content reliability. Notably, user engagement metrics were inversely associated with medical professional quality scores in liver transplantation videos (a correlational observation that does not imply causation), with no such association in lung transplantation content. These findings, situated within the unique regulatory and content-governance environment of the Chinese digital ecosystem, highlight the need to optimize platform content governance and promote specialist-led educational content to improve the overall educational quality of transplantation-related short videos.
    Keywords:  Health education; Liver transplantation; Lung transplantation; Public health; Social media; Video quality
    DOI:  https://doi.org/10.1038/s41598-026-59850-8
  37. BMC Musculoskelet Disord. 2026 Jul 03.
    The quality and reliability of short videos about cervical spondylosis on TikTok and Redbook: a cross-sectional study.
    Jie-Min Lin, Kun-Hui Li, Cheng Lin.
       OBJECTIVE: To systematically evaluate the quality, reliability, understandability, and actionability of short videos related to cervical spondylosis on TikTok and Redbook, and to compare differences across platforms, uploader sources, and treatment modalities.
    METHODS: On February 1, 2026, the top 100 videos ranked by comprehensive relevance were retrieved from both platforms using the Chinese keyword "cervical spondylosis". Two researchers independently screened videos and assessed video quality and reliability using the Global Quality Scale (GQS) and the modified DISCERN tool (mDISCERN), while the Patient Education Materials Assessment Tool (PEMAT) was employed to evaluate understandability (PEMAT-U) and actionability (PEMAT-A). Basic video characteristics and engagement metrics were recorded, and stratified analyses were performed based on uploader professional background and video content.
    RESULTS: A total of 143 videos were included (80 from TikTok, 63 from Redbook). Overall, video quality was moderate, with median GQS and mDISCERN scores of 3.00, and median PEMAT-U and PEMAT-A scores of 75.00% and 67.00%, respectively. TikTok videos demonstrated significantly higher engagement metrics, GQS, and mDISCERN scores compared to Redbook videos (all P < 0.05). Orthopedic professionals were the primary contributors (59.4%), and videos uploaded by orthopedic and rehabilitation professionals achieved significantly higher GQS and PEMAT-U scores than those from other-field professionals (both P < 0.05). Rehabilitation intervention videos (accounting for 80.3%) scored significantly higher across all quality assessment domains compared to surgical intervention videos (all P < 0.05). Correlation analysis revealed only weak to moderate associations between engagement metrics and quality scores (ρ = 0.22-0.41), indicating that video popularity does not reflect content quality.
    CONCLUSIONS: The overall quality of cervical spondylosis-related short videos on TikTok and Redbook is moderate, with significant variations across platforms and uploader sources. Content from professional sources and rehabilitation interventions demonstrates superior quality, but video popularity does not equate to quality. The public should not rely solely on engagement metrics to judge video reliability. There is an urgent need to encourage more professional medical involvement in health science popularization and to strengthen platform content review mechanisms.
    Keywords:  Cervical spondylosis; Redbook; Social media; TikTok; Video quality
    DOI:  https://doi.org/10.1186/s12891-026-10105-7
  38. BMC Urol. 2026 Jul 03.
    Information quality and reliability of pheochromocytoma-related videos on tiktok and Kwai: a cross-sectional study.
    Jun Luo, Ziliang Chen, Runfu Cao, Kaihong Wang.
       BACKGROUND: Short-video platforms are major sources of health information, yet the quality of pheochromocytoma-related content is unclear. We assessed and compared the quality and reliability of related videos on TikTok and Kwai.
    METHODS: Using the Chinese keyword "," we searched TikTok and Kwai (24-25 Dec 2025) with a newly created account and retrieved the first 120 videos from each platform under the default "comprehensive" ranking. After exclusions, 178 videos were included (TikTok, n = 100; Kwai, n = 78). Video metrics, uploader characteristics, and content features were extracted. Quality and reliability were assessed using GQS (1-5) and mDISCERN (0-5) by two blinded reviewers, with disagreements resolved by a third reviewer. We compared video characteristics and GQS/mDISCERN scores between TikTok and Kwai, and performed subgroup analyses by uploader identity and physician specialty. Nonparametric tests and Spearman correlation analyses were conducted using R 4.1.3.
    RESULTS: Median video length was 73.0 s; median GQS and mDISCERN were both 2.0, indicating generally low quality and reliability. Most videos were posted by healthcare personnel (81.5%), predominantly urologists (50.6%). TikTok videos were longer and had higher likes, comments, and collections (all P < 0.001). mDISCERN was higher on TikTok (P = 0.047), while GQS did not differ (P = 0.580). Video length correlated with GQS (r = 0.46) and mDISCERN (r = 0.41), whereas engagement metrics were not reliable proxies for credibility.
    CONCLUSIONS: Pheochromocytoma-related short videos on TikTok and Kwai show limited quality and reliability despite professional authorship. Improvements should prioritize verifiable sources and decision-relevant risk framing, supported by platform-level quality labeling and distribution strategies.
    Keywords:  Health communication; Information quality; Pheochromocytoma; Reliability; Short videos
    DOI:  https://doi.org/10.1186/s12894-026-02242-x
  39. Digit Health. 2026 Jan-Dec;12:12 20552076261462738
    Quality and reliability of rectal prolapse short-video content on tiktok and bilibili: A cross-sectional content analysis.
    Xiaowen Wang, Zuowu Xi, Mingliao Niu, Zongyue Gao.
       Objective: To evaluate engagement metrics, content coverage, information quality, and reliability of rectal prolapse-related short videos on TikTok and Bilibili, and to assess the influence of uploader type.
    Methods: We conducted a cross-sectional content analysis. Using visitor mode and the default comprehensive ranking algorithm of each platform, two researchers searched for "rectal prolapse" from October 14 to 22, 2025. The first 150 videos per platform were screened, yielding 256 unique eligible videos. Extracted variables included platform, upload date, duration, and engagement metrics (likes, comments, favorites, shares). Video quality and reliability were independently assessed by two colorectal surgeons using the Global Quality Scale (GQS), modified DISCERN (mDISCERN), and the JAMA benchmark; discrepancies were adjudicated by a senior expert.
    Results: TikTok outperformed Bilibili across engagement metrics (all p<0.001); video duration did not differ significantly (p=0.068). Median GQS and JAMA scores were higher on TikTok (both p≤0.001); mDISCERN differences were smaller yet significant (p=0.008). Videos from specialists scored higher on GQS, mDISCERN, and JAMA (all p<0.001). Content emphasized treatment and symptoms, whereas prevention and differential diagnosis were under-covered. Engagement metrics were highly inter-correlated but showed very weak correlations with quality scores.
    Conclusions: Video volume is increasing, whereas overall quality remains moderate. TikTok showed higher engagement and quality than Bilibili. Specialists play a key role in improving content. Platforms should incorporate reliability-weighted ranking signals and promote an essential-elements checklist for patient education.
    Keywords:  bilibili; health information quality; rectal prolapse; short-video platforms; tiktok
    DOI:  https://doi.org/10.1177/20552076261462738
  40. BMC Public Health. 2026 Jul 02.
    Evaluation of the quality and reliability of infertility information on Chinese social media platforms Bilibili and TikTok : a cross-sectional study.
    Jianjie Zhang, Jixia He, Quanzhi Wei, Meng Rao.
       BACKGROUND: Infertility has become a major global health issue, affecting approximately one in six couples worldwide; this problem is particularly pronounced in China. With the advent of the digital age, an increasing number of people struggling with infertility are turning to social media platforms for health information. However, the quality and reliability of such video content remain unclear. Therefore, this study aims to assess the overall quality and reliability of infertility-related information on mainstream social media platforms in China.
    METHOD: This study used the Chinese term "Infertility" as a search term to retrieve the top 100 videos on Bilibili and TikTok. Video quality was assessed using the DISCERN tool and the Global Quality Scale (GQS), and statistical analysis was conducted to explore the relationships among various variables.
    RESULTS: This study ultimately included 163 videos in the analysis, comprising 72 videos from Bilibili and 91 from TikTok. The results showed that the average daily engagement metrics on Bilibili were significantly higher than those on TikTok (P < 0.05). The proportion of medical personnel on the TikTok platform was as high as 90.1%, significantly higher than the 27.8% on Bilibili (P < 0.001). Videos on the TikTok platform were more concentrated in the medium-quality category, while the Bilibili platform had a higher proportion of low-quality videos (GQS low quality: 41.7% vs. 21.1%, adjusted P = 0.024). Multivariate regression analysis showed that the DISCERN score was significantly positively correlated with the number of shares (β = 0.045, 95% CI: 0.002-0.088, P = 0.040). Both the number of likes and the number of comments were negatively correlated with video duration (β = -0.001, P = 0.040; β = -0.001, P = 0.022).
    CONCLUSION: The overall quality of videos on infertility-related topics on TikTok needs improvement. Bilibili has a significant gap in this content, suggesting that high-quality content creators-such as medical professionals-should not limit their science communication efforts to a single platform. Creators of science communication videos may consider adopting narrative techniques or focusing on specific topics for in-depth explanations to enhance the accessibility of their content.
    Keywords:  Health information; Infertility; Quality assessment; Reliability; Science communication; Short videos
    DOI:  https://doi.org/10.1186/s12889-026-28306-z
  41. Sci Rep. 2026 Jun 29.
    Assessment of pelvic floor muscle training video information quality and associated factors on Douyin and Bilibili: a cross-sectional study.
    Yu Cheng, Qianbing Li, Xiaoqing Gu, Wenni Xie, Xiangfei Wang.
      This study evaluated the information quality of pelvic floor muscle training (PFMT) videos on Douyin (the Chinese version of TikTok) and Bilibili and examined factors associated with higher-quality content. A cross-sectional analysis was conducted on 200 PFMT videos (100 per platform). Dissemination and engagement metrics were collected. Information quality was assessed using modified DISCERN (mDISCERN), the Global Quality Score (GQS), and the Patient Education Materials Assessment Tool for Audiovisual Materials (PEMAT-A/V). Group differences were tested using the Kruskal-Wallis and Mann-Whitney U tests; associations were assessed using Spearman's rank correlation; and factors associated with quality scores were examined using multiple linear regression. Median scores were 2.00 for mDISCERN, 3.00 for GQS, 66.67 for PEMAT-A/V understandability, and 100.00 for PEMAT-A/V actionability, indicating moderate overall quality. Multiple linear regression showed that publisher category was associated with GQS and mDISCERN scores. Compared with fitness or exercise influencers, science communicators had higher GQS scores (β = 0.298, P = 0.002), whereas individual users had lower GQS scores (β = - 0.361, P < 0.001). Healthcare professionals had higher mDISCERN scores (β = 1.169, P < 0.001), whereas individual users had lower mDISCERN scores (β =  - 0.491, P < 0 .001). The video topic was associated with mDISCERN, PEMAT-A/V understandability, and PEMAT-A/V actionability scores. Compared with scientific and educational information videos, experiential narratives and self-management videos (β = - 1.037, P < 0.001) and pelvic floor dysfunction and symptom management videos (β = - 0.745, P = 0.002) had lower mDISCERN scores. Pelvic floor training and exercise guidance videos had higher PEMAT-A/V understandability (β = 14.767, P = 0.014) and actionability scores (β = 22.658, P = 0.010). These findings suggest that professionally produced and evidence-based PFMT videos may help users access more reliable and actionable health information.
    Keywords:  Associated factors; Information quality; Pelvic floor dysfunction; Pelvic floor muscle training; Short videos
    DOI:  https://doi.org/10.1038/s41598-026-59992-9
  42. PeerJ. 2026 ;14 e21471
    Credibility, educational quality, and specialty-specific depth of meniscal injury information on Douyin: a cross-sectional study.
    Wen Zhang, Chang Hu, Yijie Xi, Lihong Xie.
       Background: We evaluate the credibility, overall educational quality, and specialty-specific depth of Douyin videos on meniscal injury, examined the relationship between information quality and dissemination metrics, and proposed improvement strategies.
    Methods: On July 31, 2025, two Chinese keywords were searched under three sorting modes; the top 50 results per query were collected (n = 300). Videos were included if they substantially addressed meniscal injury, symptoms, diagnosis, treatment, rehabilitation, or care-seeking. After de-duplication and exclusions, 143 videos remained. A three-tier framework was applied using the Journal of the American Medical Association (JAMA) benchmarks, the modified DISCERN tool, Global Quality Score (GQS), and the American Academy of Orthopaedic Surgeons (AAOS) Meniscus-specific Score (MSS). Two trained sports-medicine raters independently and blindly scored all videos with adjudication for disagreements. Kruskal-Wallis tests compared scores across video source and content categories; Spearman correlation assessed associations among tools.
    Results: Median (IQR) scores were JAMA 2 (2-2), DISCERN 2 (2-2), GQS 2 (1-3), and MSS 3 (1-8), indicating low transparency, evidence disclosure, and educational value, with variable meniscus-specific depth. Across source categories, JAMA, DISCERN, and MSS differed significantly (all p < 0.001), whereas GQS did not. A cross content categories, JAMA (p = 0.014) and MSS (p = 0.028) differed significantly, whereas DISCERN and GQS did not. Videos by professional physicians and rehabilitation therapists outperformed those by fitness enthusiasts and patients on MSS and some credibility metrics. By content, "treatment pathway and surgical options" achieved higher MSS, whereas "rehabilitation training and postoperative recovery" more often missed key points. Correlations were positive across tools; GQS correlated most strongly with MSS (r = 0.638, p < 0.01), while JAMA and DISCERN showed moderate correlations with MSS, suggesting complementarity between professional depth and credible disclosure. Among dissemination metrics, only comments differed significantly across video source and content categories (p = 0.041; p = 0.023); likes, saves, and shares did not consistently reflect quality, indicating a decoupling between engagement and quality.
    Conclusions: Using a three-tier framework, we found low professionalism, transparency, and educational value in Douyin videos on meniscal injury. Engagement appears decoupled from quality. We recommends a minimum disclosure checklist, platform-level structured publishing and credibility labels with weighted distribution, and standardized short videos by medical institutions integrated with offline education. The findings provide empirical evidence and practical implications for improving health information quality and informing governance strategies on Chinese short-video platforms.
    Keywords:  Chinese; Douyin; Health communication; Information quality; Meniscal injury
    DOI:  https://doi.org/10.7717/peerj.21471
This page is being maintained by Thomas Krichel. It was last updated on 2026‒07‒06 at 11:58.