Evaluation of AI Large Models for Diagnosis and Treatment in Real-World Cases: Multicenter Retrospective Study

NCT07378358 | Recruiting

• 1 other identifier: MRCTA,ECFAH of FMU[2025]902
• Study Type: observational
• Target: 800
• Locations: 1 country
• Sites: 1

Brief Summary

This multicenter retrospective study aims to evaluate the diagnostic and therapeutic performance of three large language models-ChatGPT, Gemini and Deepseek-using 800 archived inpatient medical records from urology departments across four tertiary hospitals. The study will focus on the accuracy and applicability of these models in disease recognition, preliminary diagnosis and treatment recommendation generation, in order to explore their potential value and limitations in supporting clinical decision-making in real-world settings.

Conditions

Urologic Diseases

Keywords

Large Language Models; Urologic Diseases; Clinical Decision Support; Retrospective Study

Outcome Measures

Primary Outcomes (6)

• Diagnostic Accuracy: Assessed by Top-1 accuracy

  Top-1: Proportion of cases where the model's first diagnosis matches the true primary diagnosis.

  Through study completion, an average of 3 months

• Diagnostic Accuracy: Assessed by Top-3 accuracy

  Top-3: Proportion of cases where the true diagnosis appears in the model's top 3.

  Through study completion, an average of 3 months

• Diagnostic Completeness

  Proportion of the model's diagnoses that overlap with all diagnoses (primary and secondary) in the case.

  Through study completion, an average of 3 months

• Differential Diagnosis Quality

  Evaluated by experts using a Likert 5-point scale, considering factors like common disease coverage, logical clarity, and specificity

  Through study completion, an average of 3 months

• Treatment Plan Quality

  Assesses whether the model's treatment suggestions align with clinical guidelines, scored by experts on completeness, appropriateness, and safety.

  Through study completion, an average of 3 months

• Analysis Time

  5.Time taken by the AI model to provide diagnoses and treatment suggestions (in seconds), reflecting real-time capability.

  Through study completion, an average of 3 months

Interventions

Large Language Model Assessment (ChatGPT, Gemini, DeepSeek) OTHER

De-identified inpatient medical records were retrospectively collected from the urology departments of four tertiary hospitals (200 cases per site, 800 in total). Each case included standardized clinical information such as demographics, chief complaint, history of present illness, past medical history, physical examination, laboratory and imaging findings, discharge diagnosis and treatment plan. To simulate the role of an AI system in a "first-visit physician" scenario, all diagnostic conclusions, differential diagnoses and treatment plans were removed before being input into the models. Three large language models (ChatGPT, Gemini and DeepSeek) were prompted with a standardized instruction: "Based on the above clinical information, provide your preliminary diagnosis, differential diagnoses and treatment recommendations." Each model generated outputs including (i) primary and secondary diagnoses, (ii) differential diagnosis lists with reasoning and (iii) preliminary treatment suggesti

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

The study population was drawn from the following institutions: The First Affiliated Hospital of Fujian Medical University, The Second Affiliated Hospital of Fujian Medical University,Shishi City Hospital and Shaowu City Hospital

You may qualify if:

• The case data is sourced from the four hospitals involved in the study, with complete and authentic diagnosis and treatment records.
• Patients must be 18 years or older, with no gender restrictions.
• Complete medical records, including the following core information: patient' s basic information, present illness history, past medical history, physical examination, and auxiliary examinations (including laboratory and imaging tests).
• A clear discharge diagnosis and treatment plan (including therapeutic measures and follow-up arrangements).
• Medical records have been archived, with objective and accurate information that has not been altered.
• The patient or their legal representative has provided informed consent, agreeing to the use of their anonymized medical data for research analysis.

You may not qualify if:

• Medical records with significant missing information, such as key clinical details (present illness history, diagnostic or treatment records, etc.).
• Cases where the diagnosis or treatment plan is unclear, or where treatment has not been fully completed for an initial diagnosis.
• Cases where the primary diagnosis is not urological.
• Cases with major errors or inconsistencies in the records that could affect further assessment.
• Medical records in special formats or images that are not readable (e.g., handwritten notes, non-standard documentation).
• Patients who have not signed the informed consent form or who refuse to allow their medical data to be used for research.

Sponsors & Collaborators

• First Affiliated Hospital of Fujian Medical University: lead

Study Sites (1)

The First Affiliated Hospital of Fujian Medical University

Fuzhou, China

RECRUITING

MeSH Terms

Conditions

Urologic Diseases

Condition Hierarchy (Ancestors)

Female Urogenital Diseases; Female Urogenital Diseases and Pregnancy Complications; Urogenital Diseases; Male Urogenital Diseases

Central Study Contacts

Ning Xu

CONTACT

+86-13235907575; drxun@fjmu.edu.cn

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: RETROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: December 9, 2025
• First Posted: January 30, 2026
• Study Start: January 1, 2026
• Primary Completion: April 1, 2026
• Study Completion: June 1, 2026
• Last Updated: January 30, 2026

Record last verified: 2026-01

Data Sharing

• IPD Sharing: Will not share

Locations

CN (1)