November 2, 2022

1:30 pm / 3:00 pm

TITLE: Humphrey Visual Field Measurements in Glaucoma Patients: Error Model and Implications for Clinical Trial Design: Wednesday, 2 November 2022


Laurel Stell (1)

Jeffrey Goldberg (2)

Gala Beykin (2)

  1. Biomedical Data Science
  2. Ophthalmology

DATE: Wednesday, 2 November 2022

TIME: 1:30–3:00 PM

LOCATION: Conference Room X303, Medical School Office Building, 1265 Welch Road, Stanford, CA



The Data Studio Workshop brings together a biomedical investigator with a group of experts for an in-depth session to solicit advice about statistical and study design issues that arise while planning or conducting a research project. This week, the investigator(s) will discuss the following project with the group.


The Humphrey Visual Field (HVF) exam is widely used to diagnose glaucoma and its progress.  It measures the sensitivity to light of varying brightness in an array of points on the retina.  Its test-retest variability increases considerably as sensitivity decreases.  Furthermore, the mapping of the retina locations to the optical nerve, where glaucoma damage occurs, results in unusual spatial relationships in the sensitivity measurements.  Clinical trials of potential glaucoma treatments seek to show improvement–or at least less decline–in sensitivity at a “sufficient” number of locations compared to untreated controls, but glaucoma generally progresses slowly.  All of these factors can result in prohibitively large sample sizes or long trial times.


We have performed exploratory analysis of HVF exams, seeking to model the test-retest variability and spatial relationships.  We hope to leverage such insights to improve clinical trial inclusion criteria and statistical tests for treatment effect.


We have HVF data from 140 glaucomatous eyes of 92 patients acquired by multiple sources.  Thirty of the eyes were in a test-retest study that performed weekly exams for three months (Artes et al, 2014).  Another 51 eyes of 30 patients were in a Phase 1b trial of eye drops (Goldberg et al, 2022).  The other exams were selected opportunistically.  Each eye has at least two exams within 105 days.


We will describe test-retest variability and spatial relationships in HVF sensitivity data based on our dataset.  We will also discuss potential measures of treatment effect.  We are seeking advice on statistical models for testing treatment effect.


(1) What are optimal trial inclusion criteria, taking into account the test-retest variability?

(2) How can we leverage spatial relationships and multiple exams to check the inclusion criteria?

(3) Which statistical models are appropriate for treatment effect?

(4) Any suggestions for clinical trial design and sample size calculations?