Eye Disorders, Glaucoma, Visual Fields, Visual Field Tests
Conditions
Keywords
HVF, humphrey visual field test, Glaucoma, LLM, AI, GPT, ChatGPT, Education
Brief summary
The purpose of this study is to evaluate whether a large language model (LLM)-based audiovisual educational tool improves the test time and reliability of standard automated perimetry (SAP) using the SITA Standard 24-2 protocol in English-speaking glaucoma patients. Glaucoma is a disease that can lead to blindness if not properly monitored and treated. One of the most important tests for glaucoma is the visual field (VF) test, which checks how well a person can see in different directions. However, this test is difficult for many patients to perform correctly, especially if they don't fully understand how it works. Unreliable test results can lead to repeated visits, wasted time, and incorrect treatment decisions. This study is testing whether a computer-based educational tool, powered by artificial intelligence (AI), can help patients better understand the VF test before taking it. The study team want to see if this helps make the test results more reliable. The goal is to improve the quality of care while reducing the burden on patients and clinic staff. The LLMs will be used as an educational tool only, not for the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease.
Interventions
OTHERLLM-based Education
Participants will receive audiovisual education powered by a large language model (LLM) before their visual field test. The LLM will be presented using a 10 inch tablet or laptop device by a trained research team member. The interaction is intended to be self-guided, with no interference from the staff unless the LLM displays incorrect or hallucinated content. In such cases, the research staff will immediately correct any misinformation and record the occurrence, including details and frequency of the hallucination, for quality monitoring. The LLM module will deliver instructions, simulate the visual field test experience, and include a brief knowledge check. This LLM-based education is for research purposes only. Afterward, participants will proceed to their scheduled visual field test, which will include standard support from the clinic perimetrist.
Sponsors
Stanford University
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
TREATMENT
Masking
SINGLE (Caregiver)
Masking description
The research personnel consenting and enrolling potential study participants and the visual field testing technicians will be blinded to the next study group the potential study participant will be put in.
Intervention model description
* Intervention Group: Participants will receive audiovisual education powered by a large language model (LLM) before their visual field test. The LLM will be presented using a 10 inch tablet or laptop device by a trained research team member. The interaction is intended to be self-guided, with no interference from the staff unless the LLM displays incorrect or hallucinated content. In such cases, the research staff will immediately correct any misinformation and record the occurrence, including details and frequency of the hallucination, for quality monitoring. The LLM module will deliver instructions, simulate the visual field test experience, and include a brief knowledge check. This LLM-based education is for research purposes only. Afterward, participants will proceed to their scheduled visual field test, which will include standard support from the clinic perimetrist. * Control Group: Participants will receive standard-of-care perimetrist instructions only
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Adults (≥18 years) with a diagnosis or suspected diagnosis of glaucoma * English-speaking * Scheduled for Humphrey visual field testing 24-2 SITA standard * Having at least 2 prior visual field tests , with the most recent prior test performed using the 24-2 SITA Standard strategy within 2 years before enrollment. * Best corrected visual acuity in both eyes ≥ 20/40
Exclusion criteria
* Unable to comply with study and questionnaires * Participant has, in the opinion of the investigator, any physical or mental condition that would affect the participation in the study or may interfere with the study procedures, evaluations and outcome assessments.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Humphrey Visual Field Test Taking Duration | Same Day of Enrollment up to 2 hours | Test duration of Humphrey visual field test of the first tested eye will be compared to the same first tested eye of the study participant's previous test as a continuous variable. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Changes in False Positives and False Negatives (Reliability Indices) | Same Day of Enrollment up to 2 hours | Change in false positive percentage and false negative percentage for the first tested eye between the most recent prior unreliable Humphrey visual field and the study visit field, comparing intervention and control groups. Each index will be analyzed as a continuous variable, with higher values indicating worse reliability. |
| Proportion of Visual Field Test Standard Reliability | Same Day of Enrollment up to 2 hours | Proportion of participants whose first tested eye produces a reliable Humphrey visual field at the study visit, comparing AI based education plus standard perimetrist instruction versus standard perimetrist instruction alone. Reliability will be defined a priori using standard criteria, for example fixation losses less than 20 percent, false positives less than 15 percent, and false negatives less than 15 percent on the same day visual field report. The first tested eye at the study visit will be used for the primary analysis. |
| Patient satisfaction scale score as measured by Client Satisfaction Questionnaire-8 (CSQ-8) | Same Day of Enrollment up to 2 hours | 4-point ordinal Likert-type response options. The total CSQ 8 score will be used as a continuous outcome, with higher scores indicating greater satisfaction. |
| Scores on the brief post-education knowledge assessment | Same day of enrollment up to 2 hours | Score on a brief post education knowledge assessment about Humphrey visual field testing, administered immediately after the study visit test. The score will be calculated as the number or percentage of correct responses on a clinical visual field questionnaire on a nominal scale with response options such as yes, no, and not sure. Higher scores indicate better knowledge. |
| Second-tested eyes | Same Day of enrollment up to 2 hours | Test results of the second eye tested in the same session during the Humphrey visual field test (to explore effects of fatigue, learning, or test order). For participants who undergo testing in both eyes during the same session, the proportion of reliable visual fields for the second tested eye at the study visit, using the same predefined reliability criteria as the primary outcome. Secondary analyses will also evaluate false positive and false negative percentages for the second tested eye and their change from the most recent prior unreliable Humphrey visual field test. |
Countries
United States
Outcome results
None listed