[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"health-study-detail:100642942":3},{"organization":4,"armGroups":7,"interventions":10,"overallOfficials":12,"centralContacts":17,"locations":25,"responsibleParty":40,"collaborators":10,"id":42,"slug":43,"hasResults":44,"nctId":45,"briefTitle":46,"officialTitle":47,"acronym":48,"eligibilityCriteria":49,"healthyVolunteers":44,"sex":50,"minAge":51,"maxAge":10,"enrollmentInfo":52,"targetDuration":10,"studyType":55,"phases":10,"briefSummary":56,"conditions":57,"keywords":61,"overallStatus":27,"whyStopped":10,"lastUpdateSubmitDate":63,"lastUpdatePostDateStruct":64,"startDateStruct":67,"completionDateStruct":69,"leadSponsor":71,"locationsCount":72},{"fullName":5,"class":6},"Marmara University Pendik Training and Research Hospital","OTHER",[8],{"label":9,"type":10,"description":11,"interventionNames":10},"Emergency Department Patient Cohort",null,"Consecutive adult patients presenting to the emergency department with a fully documented electronic anamnesis note and a definitive primary ICD-10 diagnosis",[13],{"name":14,"affiliation":15,"role":16},"Emir Ünal","Marmara University","PRINCIPAL_INVESTIGATOR",[18,23],{"name":19,"role":20,"phone":21,"phoneExt":10,"email":22},"Emir Ünal, Assistant Professor","CONTACT","+905327766010","emirunal@gmail.com",{"name":24,"role":20,"phone":10,"phoneExt":10,"email":22},"Emir Unal, Assistant Professor",[26],{"facility":5,"status":27,"city":28,"state":28,"zip":29,"country":30,"countryCode":10,"cosmosGeoPoint":31,"geoPoint":36,"contacts":37},"RECRUITING","Istanbul","34899","Turkey (Türkiye)",{"type":32,"coordinates":33},"Point",[34,35],28.94966,41.01384,{"lat":35,"lon":34},[38],{"name":39,"role":20,"phone":10,"phoneExt":10,"email":22},"Emir ünal",{"type":16,"investigatorFullName":14,"investigatorTitle":41,"investigatorAffiliation":5,"oldNameTitle":10,"oldOrganization":10},"MD, Assistant Professor","100642942","diagnostic-accuracy-of-gpt-4o-and-claude-46-sonnet-in-turkish-ed-anamnesis-notes-100642942",false,"NCT07632859","Diagnostic Accuracy of GPT-4o and Claude 4.6 Sonnet in Turkish ED Anamnesis Notes","Diagnostic Accuracy of Large Language Models From Emergency Department Anamnesis Notes: A Comparison of GPT-4o and Claude 4.6 Sonnet With Emergency Medicine Specialists","LLM-ED-DX-TR","INCLUSION CRITERIA:\n\n* Adult patients (aged 18 years and older) presenting to the emergency department.\n* Complete electronic health record available in the hospital information system (HBYS) containing a detailed anamnesis note with chief complaint, symptom duration, associated symptoms, and relevant medical history.\n* A definitive primary diagnosis recorded by the treating emergency physician using ICD-10 codes at the time of patient file closure.\n\nEXCLUSION CRITERIA:\n\n* Emergency department anamnesis notes containing fewer than 50 words or completely lacking substantive clinical content\\[cite: 1\\].\n* Pediatric cases (age under 18 years)\\[cite: 1\\].\n* Patients critically ill and triaged to high-acuity resuscitation areas (Emergency Severity Index \\[ESI\\] level 1)\\[cite: 1\\].\n* Clinical notes containing residual identifying information that cannot be fully de-identified, preventing compliance with data privacy regulations\\[cite: 1\\].\n* Non-independent clinical notes consisting solely of a brief cross-reference to a prior hospital visit without a new history entry\\[cite: 1\\].","ALL","18 Years",{"count":53,"type":54},600,"ESTIMATED","OBSERVATIONAL","This retrospective diagnostic accuracy study evaluates the ability of two large language models (LLMs) - GPT-4o (gpt-4o-2024-11-20; OpenAI) and Claude 4.6 Sonnet (claude-sonnet-4-6; Anthropic) - to generate correct diagnoses from anonymized Turkish-language emergency department (ED) anamnesis notes, and compares their performance with the diagnosis entered by the treating emergency physician. A consensus gold standard is established by three independent board-certified emergency medicine specialists who blindly review each note and vote on the primary diagnosis using ICD-10 three-character codes; the majority vote (at least 2 of 3 specialists agreeing) constitutes the reference standard. Both LLMs are evaluated using a standardized zero-shot direct prompting strategy (temperature=0, stateless API sessions). The primary outcome is diagnostic accuracy (proportion of ICD-10 chapter-level matches) and Cohen's kappa for each LLM against the gold standard. Secondary outcomes include top-3 accuracy, treating physician accuracy, inter-model agreement, and subgroup analyses by ESI triage level and ICD-10 chapter. Inter-rater reliability among the three specialists is quantified using Fleiss' kappa. Analyses are performed in Jamovi. This study represents the first evaluation of LLM diagnostic accuracy using Turkish-language clinical notes and the first to benchmark LLM performance against an independent three-specialist majority-vote gold standard rather than against the treating physician's own diagnosis.",[58,59,60],"Emergency Medicine","Diagnostic Errors","Artificial Intelligence (AI) in Diagnosis",[62],"Large Language Model; GPT-4o; Claude 4.6 Sonnet; ICD-10; Clinical Coding; Turkish; Emergency Department; Diagnostic Accuracy; STARD; STARD-AI","2026-06-22",{"date":65,"type":66},"2026-06-25","ACTUAL",{"date":68,"type":54},"2026-06",{"date":70,"type":54},"2026-10",{"name":5,"class":6},1]