Assessing the Reliability of Humphrey Visual Field Testing in an Urban Population

Abstract

Purpose: The purpose of this study was to assess the reliability of Humphrey visual field testing based on the three reliability indices defined by Humphrey Instruments, Inc. In addition, the study aims to assess the correlation between visual field reliability with severity of glaucoma, visual acuity, intraocular pressure, pupil diameter, and number of medications.

Methods: This was a retrospective comparative study. Medical records were reviewed to determine patient reliability in Humphrey automated visual field testing. Three hundred seventy-eight randomly selected electronic patient charts were evaluated, including patients with and without glaucoma, providing 1,091 visual field tests. Reliability criteria were established by Humphrey Instruments, Inc. as less than 20% fixation losses or less than 33% false negative errors or false positive errors.

Results: Overall, patients performed reliably in 52% of visual field tests. The most common cause of poor reliability was fixation loss, with 43% of patient tests deemed unreliable due to a fixation loss rate greater than 20%. False positive responses (1.7%) and false negative responses (4.2%) were much less common causes of poor reliability. There was a statistically significant difference between visual field reliability and severity of glaucoma (p<0.0001). Fifty-four percent of patients with mild glaucoma (MD >-7db) and 50% of patients with moderate glaucoma (-15<MD <-7db) were significantly more reliable compared to 29% with severe glaucoma (MD<-15), (p<0.0001). There was a significant difference between reliability and visual acuity and number of glaucoma medications (p<0.05), however, there was not a statistical difference between reliability and IOP (p=0.06).

Conclusions: A significantly higher number of Humphrey visual field tests were reliable in the mild to moderate glaucoma groups compared with the severe group within this population of patients. The majority of unreliable fields were due to fixation losses. Increased severity of glaucoma correlated with reduced reliability. Decreased visual acuity, higher numbers of glaucoma medications, and decreased pupil size were also associated with reduced reliability; however, IOP was not significantly associated with glaucoma severity.

Keywords: 758 visual fields • 642 perimetry • 465 clinical (human) or epidemiologic studies: systems/equipment/techniques