Purpose
To examine whether glaucomatous central visual field abnormalities can be more effectively detected using a qualitative, expert evaluation of the 10-2 test compared with the topographically corresponding central 12 locations of the 24-2 test (C24-2).
Design
Cross-sectional study.
Methods
Eyes with a glaucomatous optic nerve appearance or ocular hypertension ( n = 523) and healthy eyes ( n = 107) were included as cases and control subjects, respectively. The 10-2 and C24-2 visual field results of all eyes were graded by 4 glaucoma specialists for the probability that central visual field abnormalities were present.
Results
The sensitivity of the 10-2 and C24-2 tests for detecting the cases at 95% specificity were not significantly different (e.g., 32.2% and 31.4%, respectively, for grader 1, P = .87; all graders P ≥ .25). At 95% specificity, the pattern standard deviation values from these tests had a similar sensitivity to the qualitative evaluation for the C24-2 test for all graders ( P ≥ .083), but it had a significantly higher sensitivity than the qualitative evaluation for the 10-2 test for 3 graders ( P ≤ .016).
Conclusions
The similarity in performance of the 10-2 and C24-2 test suggests that the increased sampling density of the former does not significantly improve the detection of central visual field abnormalities, even when based on expert assessment. These findings should not be taken to mean that the 10-2 test is not useful, but it underscores the need for its utility to be clearly established before incorporating it as routine glaucoma standard of care.
VVisual field testing remains one of the most important tools for the diagnosis and monitoring of glaucoma. Its results are crucial for estimating the current level and future risk of functional impairment for individuals affected by this disease. , Although visual field damage in glaucoma predominantly affects the peripheral field in most patients, glaucomatous damage to the central field has been found to occur more often than conventionally expected, and even in the early stages of this disease. In addition, the central visual field is an important predictor of vision-related quality of life. Therefore, being able to detect glaucomatous central visual field damage effectively is important in the clinical management of patients with glaucoma.
Numerous studies have suggested that targeted testing of the central visual field region (e.g., the 10-2 strategy on the Humphrey Field Analyzer [HFA]; Carl Zeiss Meditec, Dublin, California, USA) or adding additional test locations within this region to conventional stimulus patterns (such as done with the new 24-2C testing strategy on the HFA) , could improve the detection of central visual field abnormalities. However, none of these studies have compared the newly proposed approaches with conventional strategies at topographically equivalent locations and at matched specificities.
Instead, we recently observed that pattern standard deviation (PSD) values derived from 10-2 tests and the topographically equivalent central 12 locations of 24-2 tests (C24-2) performed similarly at detecting central visual field abnormalities in a large cohort of eyes with either a glaucomatous optic nerve appearance or ocular hypertension. Other recent studies have also confirmed our findings of a similar performance of the 10-2 and C24-2 PSD in smaller cohorts. , However, we previously cautioned about the interpretation of these findings, given that PSD values are not typically used in isolation to detect visual field abnormalities in clinical practice. Indeed, a recent study showed how the reliance on summary measures may result in underdetection of glaucomatous damage on optical coherence tomography imaging when compared with an expert qualitative evaluation of the imaging results. This difference is likely related to the fact that summary measures may be unable to identify distinguishing patterns known to be associated with glaucomatous damage.
We hypothesized that a qualitative evaluation of the visual field results could provide a more effective and clinically relevant means for understanding the potential advantages of 10-2 visual field testing for detecting central visual field abnormalities. As such, this study sought to compare the performance of the 10-2 and 24-2 visual field tests for detecting central visual field abnormalities based on an expert qualitative evaluation of its results.
Methods
This study included participants that were examined in 2 prospective observational studies of optic nerve structure and visual function in glaucoma—the Diagnostic Innovations in Glaucoma Study and the African Descent and Glaucoma Evaluation Study 17 —and these studies were conducted in adherence with the Declaration of Helsinki and to the Health Insurance Portability and Accountability Act. Institutional review board approvals were obtained from all sites included in these studies and written informed consent was obtained from all participants.
The full description of all the testing procedures for these studies have been provided in a previous publication. Briefly, all participants underwent a medical history review, best-corrected visual acuity measurements, slit-lamp biomicroscopy, dilated fundus examination and optic disc stereophotography, and intraocular pressure and pachymetry measurements at annual intervals. Visual field testing for the participants included in this study was performed using the Swedish Interactive Thresholding Algorithm Standard 24-2 and 10-2 strategy on the HFA II-i (Carl Zeiss Meditec) at approximately semiannual intervals.
This study included eyes with a glaucomatous optic nerve appearance (based on masked grading of stereophotographs by ≥2 graders, using a methodology described further previously ) or ocular hypertension (defined as having an intraocular pressure ≥22 mm Hg) as cases and healthy eyes as control eyes in a cross-sectional study. The control eyes included eyes from healthy participants who were otherwise unremarkable on a comprehensive ophthalmologic examination and with an intraocular pressure <22 mm Hg. All participants were recruited from optometry or ophthalmology clinics or from the general population through advertisements or community presentations. All eyes in the case and control groups were required to also have a best-corrected visual acuity of 20/40 or better, open angles on gonioscopy, to be ≥18 years of age. Participants with any other ocular or systemic disease that could affect the optic nerve head or the visual field were excluded from this study. Note that visual field results were not used for classifying eyes as either cases or controls, to avoid potential biases in evaluating the performance of the 10-2 and C24-2 results (which could occur if information from the latter could be used to define the diagnostic categories).
Visual Field Testing
Eyes included in this study were all required to have reliable 10-2 and 24-2 visual field tests that were performed on the same day, and unreliable tests were defined as those with >33% fixation loss, >33% false negative errors (except when the mean deviation [MD] was worse than −12 dB), or >15% false positive errors. Tests with artifacts, such as inappropriate fixation, fatigue, inattention or learning effects, eyelid or rim artifacts, or evidence that the results were affected by a condition other than glaucoma (e.g., homonymous hemianopia) were excluded. The assessment of these artifacts were performed by experienced graders at the University of California, San Diego Visual Field Assessment Center.
Qualitative Evaluation of Visual Field Test Results
To compare the performance of the 10-2 and 24-2 visual fields at topographically matched regions, only the results of the central 12 locations of the 24-2 visual field was made visible for the qualitative evaluation. An example of the 10-2 and C24-2 reports generated for the qualitative evaluation is shown in Figure 1 . All the visual field reports were then deidentified and presented in a randomized order in a custom-written program, and each report was graded for the probability of the presence of visual field damage in the central visual field region consistent with glaucomatous damage. This grading was performed along a continuous scale (0%-100%), with higher values indicating a higher graded probability of central visual field abnormalities. Four experienced glaucoma specialists (A.O., G.S.A., A.S.C., and D.S.W.) graded all the visual field reports.
Statistical Analysis
The performance of the qualitative evaluation of the 10-2 and C24-2 visual field test results for detecting the cases was compared using a Wald test of the difference in their sensitivity at 95% specificity (i.e., based on a cutoff of the graded probabilities that led to 5% of the control eyes being flagged as abnormal), using a bootstrap resampling procedure (n = 1000 resamples). Of interest, the performance of the qualitative grading was also compared with the performance of the PSD measure (calculated using methods as described in further detail previously ), similarly using a Wald test for the 10-2 and C24-2 visual field tests separately. The agreement between the cases detected by these two methods at 95% specificity was also examined using the Cohen kappa.
Results
Participant Characteristics
A total of 523 eyes with a glaucomatous optic nerve appearance or ocular hypertension from 300 participants were included as cases, and these participants were on average 70 ± 11 years of age (range 27-99 years of age). Another 107 eyes from 67 healthy participants were included as control eyes, and these participants were on average 64 ± 11 years of age (range 34-95 years of age). The characteristics of these eyes based on the 24-2 and 10-2 visual fields are summarized in Table 1 . Among the eyes included as cases, 272 (52%) eyes had a Glaucoma Hemifield Test result that was “outside normal limits” and 309 (59%) eyes had a PSD value at a probability of <5% based on the whole 24-2 visual field test. In addition, 408 (78%) eyes had an 24-2 visual field MD ≥ −6 dB, 55 (11%) eyes had an MD < −6 dB and ≥ −12 dB, and 60 eyes had an MD < −12 dB.
| Control Eyes (n = 107) | Cases (n = 523 Eyes) | |||
|---|---|---|---|---|
| 24-2 VF | 10-2 VF | 24-2 VF | 10-2 VF | |
| MD | −0.26 (−1.44 to 0.72) | 0.17 (−1.17 to 0.83) | −1.76 (−4.98 to 0.09) | −1.25 (−4.42 to 0.25) |
| PSD | 1.78 (1.52–2.32) | 0.83 (0.69–1.10) | 2.33 (1.67–6.20) | 1.11 (0.79–3.64) |
| Graded probability of central VF abnormalities (%) a | ||||
| Grader 1 | 2 (0–2) | 2 (0–3) | 2 (0–65) | 4 (2–80) |
| Grader 2 | 5 (5–5) | 5 (5–5) | 5 (5–50) | 5 (5–50) |
| Grader 3 | 10 (0–25) | 0 (0–10) | 25 (10–100) | 10 (0–50) |
| Grader 4 | 0 (0–10) | 5 (0–20) | 10 (0–60) | 10 (5–70) |
a On a scale of 0%-100%. Note that only the central 12 locations for the 24-2 VF were shown during the qualitative grading process.
Qualitative Grading Results
The median qualitatively graded probabilities of central visual field abnormalities were similar between the case eyes with a glaucomatous optic nerve appearance or ocular hypertension and the control eyes for the C24-2 (e.g., 2% and 2%, respectively, for grader 1) and 10-2 visual fields (e.g., 4% and 2%, respectively, for grader 1), reflecting how the case group had a large proportion of eyes with central visual field graded as being relatively normal. Instead, the 75th percentile of the graded probabilities showed a greater expected difference between the case and control groups for the C24-2 (e.g., 65% vs 2%, respectively, for grader 1) and 10-2 visual fields (e.g., 80% and 3%, respectively, for grader 1). Of interest, the cutoffs of the graded probabilities used to determine sensitivity at 95% specificity for the 10-2 and C24-2 visual field tests (or the 95th percentile of the graded probabilities in the control group) were 20% and 20%, respectively, for grader 1, 25% and 15%, respectively, for grader 2, 75% and 50%, respectively, for grader 3, and 60% and 30%, respectively, for grader 4.
Diagnostic Performance of the 10-2 and C24-2 Visual Field Qualitative Grading
Based on the qualitative grading of the visual field test results, there was no significant difference in the sensitivity of detecting the cases at 95% specificity between the 10-2 and C24-2 visual fields ( P ≥ .25 for all graders). The absence of a significant difference remained when considering only cases with a 24-2 visual field MD of ≥ −6 dB ( P ≥ .27 for all graders) or ≥ −12 dB ( P ≥ .24 for all graders); these findings are summarized in Table 2 . There were also no significant differences when examining only cases with an MD < −6 dB and ≥ −12 dB (all P ≥ .19), or cases with an MD < −12 dB (all P ≥ .43).
