Advertisement

Comparison of 10-2 and 24-2C Test Grids for Identifying Central Visual Field Defects in Glaucoma and Suspect Patients

  • Jack Phu
    Correspondence
    Correspondence: Jack Phu, MPH, PhD, Centre for Eye Health, Gate 14 Barker St., Rupert Myers Building South Wing, University of New South Wales Sydney 2052, New South Wales, Australia.
    Affiliations
    Centre for Eye Health, University of New South Wales, Kensington, New South Wales

    School of Optometry and Vision Science, University of New South Wales, Kensington, New South Wales
    Search for articles by this author
  • Michael Kalloniatis
    Affiliations
    Centre for Eye Health, University of New South Wales, Kensington, New South Wales

    School of Optometry and Vision Science, University of New South Wales, Kensington, New South Wales
    Search for articles by this author
Open AccessPublished:March 12, 2021DOI:https://doi.org/10.1016/j.ophtha.2021.03.014

      Purpose

      To compare the ability of 24-2C and 10-2 test grids in measuring visual field global indices, identifying central visual field defects, and facilitating macular structure-function analysis with OCT scans in glaucoma and glaucoma suspect patients.

      Design

      Prospective, cross-sectional study.

      Participants

      One eye from 131 glaucoma and 57 glaucoma suspect patients recruited from a referral-only, university-based glaucoma clinic.

      Methods

      Each subject underwent perimetric testing using 24-2C SITA-Faster and 10-2 SITA-Fast in random order, and Cirrus OCT macular imaging (Ganglion Cell Analysis) for structure-function correlations.

      Main Outcome Measures

      Visual field global indices (mean deviation, pattern standard deviation, binarized “cluster” pass/fail, and central mean sensitivity), number and proportion of visual field defects, and structure-function concordance with the Cirrus OCT deviation map following visual field location displacement for correspondence with underlying retinal ganglion cell position.

      Results

      Global indices (mean deviation, pattern standard deviation, and central mean sensitivity) were similar between both grids. The 10-2 detected more defects compared with the 24-2C (P < 0.0001 for all patients, P = 0.006 for glaucoma patients). This was preserved when analyzing the proportion of defects in the central visual field for all patients (P = 0.02) but was not significantly different for glaucoma patients (P = 0.051). The 10-2 identified more central “clusters” of 2+ contiguous points of deficit (P < 0.0001). Structure-function comparisons performed at locations where visual field and OCT test locations were colocalized revealed greater concordance of structural and functional deficits using the 10-2 (P < 0.0001). The 10-2 took a median of 201 seconds, and the 24-2C took a median of 154 seconds, corresponding to the different thresholding algorithms.

      Conclusions

      The 24-2C and 10-2 test grids return similar global indices of visual field performance and proportionally similar amounts of central visual field loss. The additional points in the 10-2 grid return more “clusters” of defects and a greater rate of structure-function concordance compared with the 24-2C test grid. Thus, the 24-2C can identify the presence of a clustered central visual field defect using similar probability criteria, whereas the 10-2 may be more useful in comprehensively characterizing the defect and predicting central visual function.

      Keywords

      Abbreviations and Acronyms:

      dB (decibels)
      Visual field testing is a critical component of the glaucoma assessment.
      • Jampel H.D.
      • Singh K.
      • Lin S.C.
      • et al.
      Assessment of visual function in glaucoma: a report by the American Academy of Ophthalmology.
      ,
      • Phu J.
      • Khuu S.K.
      • Yapp M.
      • et al.
      The value of visual field testing in the era of advanced imaging: clinical and psychophysical perspectives.
      The current clinical standard for static automated perimetry is using the 24-2 grid, which has fixed test locations that are spaced 6 degrees apart. The arrangement of these test locations is thought to adequately capture areas in visual space typically affected in early glaucoma.
      • Heijl A.
      Perimetric point density and detection of glaucomatous visual field loss.
      ,
      • Zeyen T.G.
      • Zulauf M.
      • Caprioli J.
      Priority of test locations for automated perimetry in glaucoma.
      Several studies have suggested the role of 10-2 in early, and not just late, glaucoma, citing the poor ability of 24-2 testing to identify or adequately describe central visual field loss.
      • Grillo L.M.
      • Wang D.L.
      • Ramachandran R.
      • et al.
      The 24-2 Visual field test misses central macular damage confirmed by the 10-2 visual field test and optical coherence tomography.
      ,
      • De Moraes C.G.
      • Hood D.C.
      • Thenappan A.
      • et al.
      24-2 Visual fields miss central defects shown on 10-2 tests in glaucoma suspects, ocular hypertensives, and early glaucoma.
      Identification of central vision loss is important for disease staging and management, especially because of its impact on patient quality of life and activities of daily living.
      • Yamazaki Y.
      • Sugisaki K.
      • Araie M.
      • et al.
      Relationship between vision-related quality of life and central 10 degrees of the binocular integrated visual field in advanced glaucoma.
      ,
      • Blumberg D.M.
      • De Moraes C.G.
      • Prager A.J.
      • et al.
      Association between undetected 10-2 visual field damage and vision-related quality of life in patients with glaucoma.
      Accurately describing central visual function in glaucoma is important for understanding an individual’s ability to undertake essential tasks such as reading, near work, and recognizing distant objects for navigation and driving.
      • Murata H.
      • Hirasawa H.
      • Aoyama Y.
      • et al.
      Identifying areas of the visual field important for quality of life in patients with glaucoma.
      • Cheng H.C.
      • Guo C.Y.
      • Chen M.J.
      • et al.
      Patient-reported vision-related quality of life differences between superior and inferior hemifield visual field defects in primary open-angle glaucoma.
      • Chun Y.S.
      • Sung K.R.
      • Park C.K.
      • et al.
      Vision-related quality of life according to location of visual field loss in patients with glaucoma.
      Preservation of central vision is an important component of glaucoma management.
      Despite the advantages afforded by the higher resolution 10-2 grid, recent studies have offered contrarian views on its additional usefulness compared with the 24-2.
      • Wu Z.
      • Medeiros F.A.
      • Weinreb R.N.
      • Zangwill L.M.
      Performance of the 10-2 and 24-2 Visual field tests for detecting central visual field abnormalities in glaucoma.
      ,
      • West M.E.
      • Sharpe G.P.
      • Hutchison D.M.
      • et al.
      Value of 10-2 visual field testing in glaucoma patients with early 24-2 visual field loss.
      A question for clinicians is which test to deploy in routine glaucoma assessments, because both central and peripheral tests may provide useful information for the individual patient, depending on their phenotype. Simultaneously performing testing using both grids may be valuable in comprehensively characterizing the visual field with the complementary regions of testing.
      • Jung K.I.
      • Ryu H.K.
      • Hong K.H.
      • et al.
      Simultaneously performed combined 24-2 and 10-2 visual field tests in glaucoma.
      • Shin H.Y.
      • Park H.L.
      • Park C.K.
      Comparison of visual field tests in glaucoma patients with a central visual field defect.
      • Wu Z.
      • Medeiros F.A.
      • Weinreb R.N.
      • et al.
      Comparing 10-2 and 24-2 Visual fields for detecting progressive central visual loss in glaucoma eyes with early central abnormalities.
      However, this is typically impractical in a busy clinical practice, which is often unable to even do more than 1 test per year.
      • Fung S.S.
      • Lemer C.
      • Russell R.A.
      • et al.
      Are practical recommendations practiced? A national multi-centre cross-sectional study on frequency of visual field testing in glaucoma.
      More recently, a new test grid has been proposed, the 24-2C, which incorporates a selection of 10 asymmetrically distributed test points derived from the 10-2 into the 24-2 grid, and so clinicians are presented with the options of deploying the 24-2C as a catchall method for examining the central and peripheral field, or a more comprehensive assessment using the 10-2.
      In this study, we performed a head-to-head comparison of the 24-2C and 10-2 test grids in glaucoma and suspect patients for identifying central visual field defects. We first aimed to compare the global indices, test duration, pointwise sensitivity, and pointwise probability scores across the patterns in identifying and characterizing glaucomatous visual field loss. Our second aim was to assess the concordance of the visual field results against macular OCT scans obtained using an instrument provided by the same manufacturer.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      Given that half of the 10 central points within the 24-2C grid are outside of a 7 × 8.4 degree macular scan area, our prediction was that the higher test density of the 10-2 would facilitate a greater proportion of test locations demonstrating improved structure-function concordance for eventual correlation with functional central vision.

      Methods

       Study Design and Subjects

      This study was a prospective, cross-sectional study that took place at the Centre for Eye Health, University of New South Wales. Ethics approval for the study was provided by the Human Research Ethics Committee of the University of New South Wales. The study adhered to the tenets of the Declaration of Helsinki. Subjects provided written informed consent before inclusion in the study.
      We recruited subjects with glaucoma or high-risk glaucoma suspects, as per the criteria used in our recent studies.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      ,
      • Phu J.
      • Khuu S.K.
      • Agar A.
      • Kalloniatis M.
      Clinical evaluation of Swedish interactive thresholding algorithm-faster compared with Swedish interactive thresholding algorithm-standard in normal subjects, glaucoma suspects, and patients with glaucoma.
      Glaucoma was diagnosed as per current clinical guidelines,
      • Prum Jr., B.E.
      • Rosenberg L.F.
      • Gedde S.J.
      • et al.
      Primary Open-Angle Glaucoma Preferred Practice Pattern® Guidelines.
      including clear characteristic structural anomalies at the optic nerve head (including but not limited to increased cup-to-disc ratio, cup-to-disc asymmetry, and neuroretinal rim thinning or notching), or retinal nerve fiber layer defects, with or without corresponding visual field loss (defined using the 24-2 SITA-Faster result: a pattern standard deviation result at P < 0.05, Glaucoma Hemifield Test outside normal limits or a “cluster” fail described next). Glaucoma suspect patients were those in whom 1 or more of these signs were present, but in whom the signs were insufficient for a diagnosis of glaucoma. The diagnosis was made by at least 1 examining clinician with remote review by at least 1 other clinician. For inclusion in the study, the diagnosis was also agreed upon by a third clinician.
      Other inclusion criteria included age ≥18 years; having provided consent to use their clinical data for research and teaching; no other ocular, systemic, or neurologic comorbidities that would confound the visual field test result; no history of ocular surgery aside from uncomplicated selective laser trabeculoplasty, laser peripheral iridotomy or cataract surgery and intraocular lens implantation; and spherical equivalent refractive error between +8.00 diopters and −8.00 diopters.

       Visual Field Testing and Data Extraction

      The recruited subjects underwent testing using the 10-2 (SITA-Fast) and the 24-2C (SITA-Faster) test grids on the Humphrey Field Analyzer (Carl Zeiss Meditec). The order of testing was randomized between the 2 grids. If a subject had glaucoma in 1 eye only, that eye was chosen for the study. Otherwise, in cases where glaucoma was present in both eyes, or if both eyes were suspicious for glaucoma, a random eye was chosen for the study.
      The reliability criteria for inclusion in the present study were false-positive rate <15%, no seeding point errors (as we have previously defined),
      • Phu J.
      • Kalloniatis M.
      A Strategy for Seeding Point Error Assessment for Retesting (SPEAR) in perimetry applied to normal subjects, glaucoma suspects, and patients with glaucoma.
      <20% of instances where the gaze tracker deviation exceeded 6 degrees during the test or no gross or frequent eye movements as monitored by the perimetrist using the on-screen eye monitor, and the absence of other technician-related errors (including incorrect trial lens use, lens rim artefacts, poor patient position, and others). These criteria were similar to our recent visual field studies.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      ,
      • Phu J.
      • Khuu S.K.
      • Agar A.
      • Kalloniatis M.
      Clinical evaluation of Swedish interactive thresholding algorithm-faster compared with Swedish interactive thresholding algorithm-standard in normal subjects, glaucoma suspects, and patients with glaucoma.
      ,
      • Phu J.
      • Kalloniatis M.
      A Strategy for Seeding Point Error Assessment for Retesting (SPEAR) in perimetry applied to normal subjects, glaucoma suspects, and patients with glaucoma.
      Due to the use of SITA-Faster, elevated false-negative rates and fixation losses were not used as measures of low test reliability.
      • Heijl A.
      • Patella V.M.
      • Chong L.X.
      • et al.
      A new SITA perimetric threshold testing algorithm: construction and a multicenter clinical study.
      Subjects who did not have a reliable OCT (Cirrus OCT, Carl Zeiss Meditec) macular scan for structure-function analysis (including, but not limited to, the following reasons: signal strength <6, motion artefacts, segmentation errors, and image degradation
      • Alshareef R.A.
      • Dumpala S.
      • Rapole S.
      • et al.
      Prevalence and distribution of segmentation errors in macular ganglion cell analysis of healthy eyes using Cirrus HD-OCT.
      ) were excluded from further analysis.
      Mean deviation was not used to stage the level of glaucoma because there remains debate regarding the relative contribution of central visual field test locations in the staging process, and that in some grading systems, the presence of any central visual field defect may be regarded as advanced glaucoma, in the absence of significant mean deviation loss.
      • De Moraes C.G.
      • Sun A.
      • Jarukasetphon R.
      • et al.
      Association of macular visual field measurements with glaucoma staging systems.

       Aim 1: Comparison of Conventional Visual Field Indices

      We compared only those results that were mutually available across the 24-2C and 10-2: mean deviation, pattern standard deviation, the Glaucoma Hemifield Test, and a “cluster” criterion.
      Because mean deviation and pattern standard deviation are derived using the entirety of the 24-2C test grid, we also examined the central mean sensitivity.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      ,
      • Hood D.C.
      • Kardon R.H.
      A framework for comparing structural and functional measures of glaucomatous damage.
      Central mean sensitivity was linearized by taking the sensitivity values at each location within the central 20 degrees (in diameter) in decibels (dB), then converting it into (L) luminance in cd.m-2, where L = 3183/(10ˆ(dB/10)).
      • Khuu S.K.
      • Kalloniatis M.
      Standard automated perimetry: determining spatial summation and its effect on contrast sensitivity across the visual field.
      These luminance values were then averaged (Laverage) before converting the final score back into dB units (mean sensitivity = 10×log10(3183/Laverage). The linearization process was to ensure that artefacts arising from averaging logarithmic units, which may underestimate the depth of deficit in visual field results affected by disease by assuming equidistance between averaged values,
      • Hood D.C.
      • Kardon R.H.
      A framework for comparing structural and functional measures of glaucomatous damage.
      may be minimized and better represented. The same calculation was performed for all points within the 10-2 to enable like-for-like comparison.
      The “cluster” criterion (at least 3 points at the P < 0.05 level, where at least 1 point is at the P < 0.01 level noted on the pattern deviation map) is used in clinical studies and in practice to identify areas of visual field defect. In addition, we also used a statistically comparable criterion that has been recently reported by others:
      • Alluwimi M.S.
      • Swanson W.H.
      • Malinovsky V.E.
      • King B.J.
      A basis for customising perimetric locations within the macula in glaucoma.
      a pair of contiguous points both at the P < 0.01 level at minimum (which has a compounded probability of P < 0.0001), which is probabilistically comparable, given the probability scores on the Humphrey Field Analyzer, to the “3 or more points” criterion (which has a compounded probability of P < 0.000025). This was implemented because of the sparse of sampling by the 24-2C. Because the additional 10 points of the 24-2C grid were asymmetrically distributed, the contiguity included points that were directly adjacent to each other only, that is, the nearest distance in degrees. The “cluster” criterion was binarized: pass (no cluster reaching statistical significance) or fail (the opposite). Although a single visual field location with sufficiently reduced sensitivity may be deemed clinically suspicious, the lowest probability score on the Humphrey Field Analyzer is the P < 0.001 level. This is probabilistically less stringent compared with the pair of points criterion described earlier. Therefore, without neighboring points of reduction, we did not specifically examine a single point of reduction as a criterion for a significant visual field defect. As we focused on the central visual field, we only report on the “cluster” criterion outcomes occurring within the central 20 degrees (in diameter) of the visual field.
      The test duration was directly extracted from the instrument printout. Although the 10 additional points of the 24-2C are assessed at the end of the test, the test times for these points are not reported separately. Thus, the entirety of the 24-2C test and the 10-2 was directly compared.

       Aim 2: Structure-Function Analysis

      We compared the ability of the 24-2C and 10-2 test locations to detect colocalized structural and functional deficits within the central visual field as per our previous methods.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      As per our previous work,
      • Tong J.
      • Phu J.
      • Khuu S.K.
      • et al.
      Development of a spatial model of age-related change in the macular ganglion cell layer to predict function from structural changes.
      ,
      • Yoshioka N.
      • Zangerl B.
      • Phu J.
      • et al.
      Consistency of structure-function correlation between spatially scaled visual field stimuli and in vivo OCT ganglion cell counts.
      we corrected the visual field test locations to account for relative displacement of Henle’s fibers.
      • Drasdo N.
      • Millican C.L.
      • Katholi C.R.
      • Curcio C.A.
      The length of Henle fibers in the human retina and a model of ganglion receptive field density in the visual field.
      The corrected positions were superimposed upon macular OCT scans of the ganglion cell–inner plexiform layer thickness values, the Ganglion Cell Analysis printout of the Cirrus OCT (Carl Zeiss Meditec), as per our previous methods.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      The anatomic space described by the Ganglion Cell Analysis printout is 7 by 8.4 degrees. Accordingly, not all visual field test locations were within this area; therefore, we performed structure-function analysis using 2 different conditions as detailed below. Deviation maps for both structure and function were used to account for variations in individual subject age.
      We identified points that were shown to be statistically reduced (in sensitivity or in retinal thickness) at the P < 0.05 level or P < 0.01 level. The possible outcomes for each overlapping location were “neither” (no structural or functional loss), “both” (structural and functional loss present), “structure only,” or “function only.”
      As mentioned above, there are instances in which the visual field test locations are not colocalized with anatomic analyses performed on the Cirrus OCT.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      We examined the structure-function relationship at locations where visual field test locations were superimposed on the structural scan (the “structure grid” condition) and separately at locations across all visual field test locations within the central 20 degrees (the “function grid” condition). These conditions are shown in Figure 1. Note that although the “function grid” includes points that are, by definition, unable to have the structure-function relationship analyzed in the present study, we reported results at these locations to provide additional guidance on potential limitations arising from specific structural scan protocols and visual field test grids (see “Discussion”).
      Figure thumbnail gr1
      Figure 1A, Humphrey Field Analyzer test locations. The red circles delineate the locations tested in the 24-2C located within the central 20 degrees of visual field. These are also assessed in the 10-2. The rest of the 10-2 locations are shown in black circles. B, The visual field test locations when shifted according to ganglion cell displacement. C, The visual field test locations when flipped to match anatomic space. The light blue shaded region bounded by the solid dark blue lines indicate the area of anatomic space statistically assessed using the Cirrus OCT Ganglion Cell Analysis protocol.

       Statistical Analysis

      We used descriptive statistics to analyze the demographic characteristics of the cohorts. The distributions of quantitative data were assessed for normality (D’Agostino and Pearson test), with parametric and nonparametric statistics then applied as appropriate. The chi-square test was used to assess proportional data, except for the structure-function colocalization that was assessed using the McNemar test upon the quantitative values. Analyses were conducted using GraphPad Prism version 8 (GraphPad) and SPSS Statistics version 26 (IBM Corporation).

      Results

       Demographic and Clinical Characteristics of Cohorts

      Table 1 shows the characteristics of the 188 subjects with reliable results on both test grids and with appropriate OCT results. There was no significant difference in age or the distribution of ethnicities between the glaucoma and suspect cohorts. There was a greater proportion of male subjects in the glaucoma group (P = 0.0108) and a greater proportion of left eyes in the suspect group (P = 0.0458). There was also a greater tendency for the subjects in the glaucoma group to be more myopic compared with the suspect group (P = 0.0014).
      Table 1Demographics and Clinical Characteristics of Patients Included in the Study
      Glaucoma (n = 131)Suspect (n = 57)P Value
      Age, yrs (median, IQR)65.0 (59.5–71.0)64.0 (57.0–70.0)0.3070
      Gender, male:female (n)80:5123:340.0108
      Ethnicity (n)0.3009
       White6435
       East Asian5517
       South Asian85
       Hispanic30
       Other10
      Eye, right:left (n)62:6936:210.0458
      Spherical equivalent refractive error, D (median, IQR)−0.63 (−3.56 to +0.50)+0.25 (−1.00 to +1.00)0.0014
      D = diopters; IQR = interquartile range.
      Table 2 shows the distribution of “clusters” of visual field defects found within the central 20 degrees of visual field diameter identified using 24-2C only, 10-2 only, both, or neither. In total, 53 of 188 (4 on 24-2C only, 49 on both; totaling 28.2%) subjects had “clusters” of visual field defects identified using the 24-2C (at least 2 contiguous points), excluding points in the more peripheral regions of the grid. When including the other 52 points of the 24-2C test grid, an additional 86 subjects (totaling 139/188, 73.9%) had “clusters” of visual field defects (Fig S1, available at www.aaojournal.org). In total, 108 of 188 (59 on 10-2 only, 49 on both; totaling 57.4%) subjects had a visual field defect failing the “cluster” criterion (at least 2 contiguous points) on the 10-2 grid, and, by definition, all of these were central visual field defects. The McNemar test showed a significant difference between grids in detecting clusters of central visual field defects, with 59 (31.4%) identified only by the 10-2 and 4 (2.1%) by only the 24-2C (and the rest mutually detected, 49 [26.1%], or mutually absent, 76 [40.4%]) (P < 0.0001). Table 2 shows the results for different minimum cutoffs (3+, 4+, and others) of a number of contiguous and significantly defective points and for the glaucoma patient group (n = 131), demonstrating the tendency for the 10-2 to be able to identify more cases of contiguous clusters of defects (P < 0.0001 for all conditions). The McNemar test showed that the 10-2 identified significantly more defects compared with the 24-2C for all contiguity conditions (P < 0.0001; this was preserved in a within-gender analysis, Table S1, available at www.aaojournal.org). When divided by gender (between-gender analysis), there was a tendency for a greater proportion of male patients with contiguous deficits compared with female patients in the whole cohort for the 10-2, but this was not apparent for the other conditions (Table S2, available at www.aaojournal.org, also discussed next).
      Table 2Number and Proportion of Subjects Demonstrating at Least Two Contiguous Points of Visual Field Loss within the Central 20 Degrees Found Using the 24-2C and 10-2 Grids Only, Both Grids, or Neither Grid
      All Subjects (n = 188)Glaucoma Subjects Only (n = 131)
      24-2C Only10-2 OnlyBothNeither24-2C Only10-2 OnlyBothNeither
      2+ contiguous points
      For the 2+ contiguous points criterion, both points needed to be identified as reduced at the P < 0.01 level or worse. For 3+ contiguous points or more, the criterion was all points at the P < 0.05 level, with at least 1 at the P < 0.01 or worse.
      4 (2.1%)59 (31.4%)49 (26.1%)76 (40.4%)1 (0.5%)43 (22.9%)50 (26.6%)37 (19.7%)
      3+ contiguous points3 (1.6%)69 (36.7%)37 (19.7%)79 (42.0%)1 (0.5%)52 (27.7%)39 (20.7%)39 (20.7%)
      4+ contiguous points1 (0.5%)68 (36.2%)24 (12.8%)95 (50.5%)0 (0%)59 (31.4%)23 (12.2%)49 (26.1%)
      5+ contiguous points1 (0.5%)68 (36.2%)16 (8.5%)103 (54.8%)1 (0.5%)63 (33.5%)15 (8.0%)52 (27.7%)
      6+ contiguous points1 (0.5%)65 (34.6%)10 (5.3%)112 (59.6%)1 (0.5%)60 (31.9%)10 (5.3%)60 (31.9%)
      For the 2+ contiguous points criterion, both points needed to be identified as reduced at the P < 0.01 level or worse. For 3+ contiguous points or more, the criterion was all points at the P < 0.05 level, with at least 1 at the P < 0.01 or worse.

       Comparison of Conventional Visual Field Indices

      Mean deviation and pattern standard deviation results for the 24-2C and 10-2 grids for each subject are shown in Figure 2A and B . Linear regression analysis showed a significant relationship between grids (P < 0.0001). There was also a significant relationship between grids when focusing on the central mean sensitivity (P < 0.0001), with a higher coefficient of determination (R2 = 0.78, Fig 2C).
      Figure thumbnail gr2
      Figure 2Linear regression analyses and distributions of global indices mean deviation (A), pattern standard deviation (B), and mean central sensitivity (C). Regression analysis results are shown in the insets. The frequency distributions for the 24-2C results are shown in green, and the 10-2 results are shown in orange; Gaussian curves have been fitted to the distributions. The bottom row shows the Bland–Altman plots comparing 10-2 and 24-2C global index results. The black dotted line in the plots show y = 0, the red solid line indicates the mean bias, the red dashed lines indicate the 95% limits of agreement, and the black solid lines show the linear regression of the Bland–Altman plots. The inset results within the Bland–Altman plots show the results of the linear regression analysis.
      Bland–Altman analyses showed small biases for both mean deviation (mean bias 0.70, 95% limits of agreement −3.8 to 5.2 dB) and pattern standard deviation (mean bias −0.92, 95% limits of agreement −5.0 to +3.2 dB) (Fig 2). The bias was smaller for central mean sensitivity (mean bias 0.02, 95% limits of agreement −6.6 to 6.6 dB). Regression analysis showed that the magnitude of difference tended to increase with a greater defect in the case of mean deviation (P = 0.004), but there was no significant trend found for pattern standard deviation and mean sensitivity.
      The slopes of the regression analyses for each global index were not significantly different when separated by gender (Figs S2 and S3, available at www.aaojournal.org). Linear regression analysis also showed no significant relationship between the spherical equivalent refractive error and the global indices (Fig S4, available at www.aaojournal.org), except for the number of defects within a cluster for the 10-2 (P = 0.01). Although this relationship reached statistical significance, the coefficient of determination suggested that refractive error only minimally affected the number of defects within the present cohort, accounting for only 3% of the variance (R2 = 0.03) (Table S3, available at www.aaojournal.org).
      Test duration was longer for the 10-2 (median, 201.0; interquartile range, 188.0–223.8; full range, 129.0–433.0 seconds) compared with the 24-2C (median, 154.0; interquartile range, 138.0–183.0; full range, 113.0–335.0 seconds). This was primarily due to the different test algorithms (SITA-Fast vs. SITA-Faster), but also likely a contribution arising from the larger number of test locations for the 10-2 test grid (64 in the 24-2C vs. 68 in the 10-2).
      There were more defective points within the largest “cluster” of visual field defects meeting the minimum statistical requirement for the “cluster” criterion found using the 10-2 compared with the 24-2C across all subjects (P < 0.0001) and when considering only glaucoma subjects (P = 0.006) (Fig 3, top row). The difference was preserved when considering the proportion of locations showing a defect across the whole cohort (P = 0.02), but the difference in the glaucoma cohort no longer reached statistical significance (P = 0.051) (Fig 3, middle row). Within individual subjects, the 10-2 test grid identified a greater proportion of defective test locations across the whole cohort (60.1%) and for the glaucoma subjects (58.9%) (Fig 3, bottom row).
      Figure thumbnail gr3
      Figure 3The proportion of test locations found to be defective at the P < 0.05 across each test pattern (the full 24-2C grid, black; only the peripheral, outside the central 20 degrees, points of the 24-2C, blue; the 14 central points tested in the 24-2C, red; and the entirety of the 10-2, green) for the totality of subjects (A) and for glaucoma subjects only (B). Each datum point represents the result from an individual subject, and the box and whiskers represent the median, interquartile range, and full range. The asterisks indicate statistically significant differences between groups (∗∗∗P < 0.001; ∗∗∗∗P < 0.0001). The bottom row shows an intra-subject comparison of the proportion of central defects found using the 10-2 (y-axis) and the 24-2C (x-axis) for the totality of subjects (C) and for glaucoma subjects only (D). Points falling above the line of parity in red favor the 10-2 in detecting more defects, and points below favor the 24-2C.
      To further examine whether the difference in defect detection rate was a property of the test grid characteristics or if the number of test locations contributed to this finding, we performed a binary logistic regression, with the outcome variable a binarized defect present or absent, and the independent variables the nominal grid type and the scalar number of central visual field test locations (14 for the 24-2C and 24-2 and 68 for the 10-2). The model identified both the test grid as a significant factor (P < 0.0001) and the number of test points (P = 0.04), with a calculated odds ratio of 1.004 favoring defect detection by the 68 points of the 10-2.

       Structure-Function Analysis

      The distributions of structure-function concordance outcomes were significantly different in a head-to-head comparison of the 24-2C and 10-2 test grids for both the “structure grid” (locations where visual field test locations were superimposed on the structural scan) and the “function grid” (all visual field test locations) conditions (Fig 4). The results demonstrated more instances of concordance (“both” and “neither”) using the 10-2, as well as a greater proportion of functional defects when examining all visual field test points (P < 0.0001 for all conditions, except for the structural grid at P = 0.001). Where only locations with defects were analyzed, the 10-2 identified more instances of functional loss in the absence of apparent structural loss using the “function grid” and greater structure-function concordance when using the “structure grid” compared with the 24-2C (P < 0.0001 for all conditions). Results for this analysis when performed at the P < 0.01 level showed similar results (Fig S5, available at www.aaojournal.org).
      Figure thumbnail gr4
      Figure 4Proportion of instances of each possible structure-function comparison outcome (function defective only, red checks; structure defective only, blue diagonals; both defective, green solid; neither defect, black solid) across all tested locations (A for the entire cohort, C for the glaucoma subjects only) and at locations with structural or functional defects (B for the entire cohort, D for the glaucoma subjects only) at the P < 0.05 level. The columns indicate the type of grid analyzed, with the “F” grids representing all locations tested using the functional visual field grid, and the “S” grids representing only those locations tested by the structural scan of the OCT.

      Discussion

      We performed a head-to-head comparison of the 24-2C and 10-2 test grids for detecting central visual field defects in glaucoma suspect and glaucoma subjects, with overall similar results between the test grids for the identification of visual field anomalies. As expected, there was an obvious difference between grids in detecting the absolute number of defects, reflecting the greater test density and resolution of the 10-2 grid.

       Differences in Visual Field Indices

      The discordance between grids in the global indices was more pronounced with greater magnitude of visual field defects. This was not unexpected, because the variability of global indices tends to increase with greater severity of loss.
      • Rabiolo A.
      • Morales E.
      • Kim J.H.
      • et al.
      Predictors of long-term visual field fluctuation in glaucoma patients.
      Another reason for the discordance is that mean deviation and pattern standard deviation results are to be weighted by the variability at each test location,
      • Asman P.
      • Heijl A.
      Weighting according to location in computer-assisted glaucoma visual field analysis.
      ,
      • Heijl A.
      • Lindgren G.
      • Asman P.
      A package for the statistical analysis of visual fields.
      which increases in the periphery.
      • Phu J.
      • Khuu S.K.
      • Nivison-Smith L.
      • et al.
      Pattern recognition analysis reveals unique contrast sensitivity isocontours using static perimetry thresholds across the visual field.
      ,
      • Heijl A.
      • Lindgren G.
      • Olsson J.
      Normal variability of static perimetric threshold values across the central visual field.
      As such, grids assessing peripheral and central locations are not directly comparable. Specifically, differences in the patterns and locations of glaucomatous visual field defects would produce different scores in those with primarily peripheral or central loss. In the present cohort, more subjects had peripheral visual field loss found on the 24-2C compared with the number with central field loss upon examination of the additional 10 test locations, compared with the 10-2 test grid. The comparative frequency of central and peripheral defects varies significantly across individuals, and the prevalence of central or peripheral defects occurring first in early glaucoma has been debated in the literature.
      • Odden J.L.
      • Mihailovic A.
      • Boland M.V.
      • et al.
      Evaluation of central and peripheral visual field concordance in glaucoma.
      ,
      • Traynis I.
      • De Moraes C.G.
      • Raza A.S.
      • et al.
      Prevalence and nature of early glaucomatous defects in the central 10 degrees of the visual field.
      However, it is acknowledged that the likelihood of defect detection is in large part dependent on the study population.
      • Ekici E.
      • Moghimi S.
      • Hou H.
      • et al.
      Central visual field defects in patients with distinct glaucomatous optic disc phenotypes.
      ,
      • Kim J.M.
      • Kyung H.
      • Shim S.H.
      • et al.
      Location of initial visual field defects in glaucoma and their modes of deterioration.
      In a head-to-head comparison of the central visual field, we found similar results between 24-2C and 10-2 test grids in terms of their central mean sensitivity. Thus, on average, both test grids would return similar levels of visual function. This result is similar to the recent work of Wu et al.
      • Wu Z.
      • Medeiros F.A.
      • Weinreb R.N.
      • Zangwill L.M.
      Performance of the 10-2 and 24-2 Visual field tests for detecting central visual field abnormalities in glaucoma.

       Test Density and Sampling for Identifying Contiguous Points of Visual Field Deficit

      Identification of central visual field defects is possible using sparse test grids (24-2 and 24-2C), even if a traditional “cluster” criterion cannot be met, which could then signal the clinician to perform additional focused central testing.
      • Wu Z.
      • Medeiros F.A.
      • Weinreb R.N.
      • Zangwill L.M.
      Performance of the 10-2 and 24-2 Visual field tests for detecting central visual field abnormalities in glaucoma.
      ,
      • West M.E.
      • Sharpe G.P.
      • Hutchison D.M.
      • et al.
      Value of 10-2 visual field testing in glaucoma patients with early 24-2 visual field loss.
      ,
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      ,
      • Sullivan-Mee M.
      • Karin Tran M.T.
      • Pensyl D.
      • et al.
      Prevalence, features, and severity of glaucomatous visual field loss measured with the 10-2 achromatic threshold visual field test.
      ,
      • Park H.Y.
      • Hwang B.E.
      • Shin H.Y.
      • Park C.K.
      Clinical clues to predict the presence of parafoveal scotoma on Humphrey 10-2 visual field using a Humphrey 24-2 visual field.
      The 10 extra locations tested by the 24-2C grid are derived from the same Cartesian coordinates of the 10-2 grid. Unsurprisingly, the overall proportions of central test locations showing visual field defects were similar between the 10-2 and 24-2C.
      The 10-2 identifying more cases of contiguous clusters of deficit compared with the 24-2C was also not unexpected, because the relatively sparser sampling of the 24-2C central locations means that “gaps” between test points present a challenge for identifying contiguity, a notion highlighted by others who have sought to better describe glaucomatous defects.
      • Chong L.X.
      • McKendrick A.M.
      • Ganeshrao S.B.
      • Turpin A.
      Customized, automated stimulus location choice for assessment of visual field defects.
      ,
      • Ballae Ganeshrao S.
      • Turpin A.
      • Denniss J.
      • McKendrick A.M.
      Enhancing structure-function correlations in glaucoma with customized spatial mapping.
      This was also supported by the similarities in central mean sensitivity noted earlier. If there were fundamental differences in the scotoma characteristics described by the test grids, we might also expect greater differences in the global results. In individual cases where the central visual field defect deepens quicker than peripheral field loss, higher resolution perimetry would facilitate measurement of disease progression.
      • Schiefer U.
      • Papageorgiou E.
      • Sample P.A.
      • et al.
      Spatial pattern of glaucomatous visual field loss obtained with regionally condensed stimulus arrangements.
      ,
      • Nevalainen J.
      • Paetzold J.
      • Papageorgiou E.
      • et al.
      Specification of progression in glaucomatous visual field loss, applying locally condensed stimulus arrangements.
      The improvement to structure-function concordance found using the 10-2 was overall small, even when excluding those locations that were deemed to be both structurally and functionally normal. This finding was again consistent with the idea of “gaps” within consistent scotomata being largely filled by the additional points within the 10-2. Therefore, although there were 4 times more points tested within the central 20 degrees on the 10-2 compared with the 24-2C, it only detected 22% more instances of structure-function concordance and 31% more instances where functional deficits were identified without statistically significant structural loss.
      The logistic regression model also suggested the effect of the number of test locations, with an expectedly higher probability of identifying a significant defect compared with fewer test points. Given the limitations of only assessing 2 available test grids, we were unable to determine an optimal number of locations for testing the central visual field, and this is an area that requires further detailed study as suggested by recent groups.
      • Kucur S.S.
      • Hackel S.
      • Stapelfeldt J.
      • et al.
      Comparative study between the SORS and dynamic strategy visual field testing methods on glaucomatous and healthy subjects.
      However, this result further supports a “critical number” for defect identification and concordance with structure, which also has practical implications for test duration.
      The improvement in structure-function concordance would expectedly facilitate enhanced understanding of the characteristics of central vision loss. Several groups have identified specific regions of the visual field that may be better correlated with specific visual functions.
      • Murata H.
      • Hirasawa H.
      • Aoyama Y.
      • et al.
      Identifying areas of the visual field important for quality of life in patients with glaucoma.
      • Cheng H.C.
      • Guo C.Y.
      • Chen M.J.
      • et al.
      Patient-reported vision-related quality of life differences between superior and inferior hemifield visual field defects in primary open-angle glaucoma.
      • Chun Y.S.
      • Sung K.R.
      • Park C.K.
      • et al.
      Vision-related quality of life according to location of visual field loss in patients with glaucoma.
      To individualize management plans, which may also involve task-specific rehabilitative strategies,
      • Mathews P.M.
      • Rubin G.S.
      • McCloskey M.
      • et al.
      Severity of vision loss interacts with word-specific features to impact out-loud reading in glaucoma.
      ,
      • Harvey H.
      • Anderson S.J.
      • Walker R.
      Increased word spacing improves performance for reading scrolling text with central vision loss.
      would require an intimate understanding of central vision, better characterized using the 10-2 than the 24-2C.

       Practical Considerations for Central Visual Field Testing

      Similarities in global indices and proportions of central visual field defects and the modest increase in structure-function concordance suggest that there may be a critical point for test density after which additional testing may be superfluous. Although the contribution of additional sensitivity information may be valuable for mapping out central scotomata, the trade-off is the amount of time required to perform more comprehensive visual field testing on a separate test grid compared with just adding several points to the 24-2.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      Commercial OCT devices that are often used for structure-function comparisons may have differences in the scan protocol and normative databases. Mapping out suitable test locations in visual space has been suggested to be guided by structure-function matching between the visual field test grid and scan area.
      • Ganeshrao S.B.
      • McKendrick A.M.
      • Denniss J.
      • Turpin A.
      A perimetric test procedure that uses structural information.
      ,
      • Montesano G.
      • Rossetti L.M.
      • Allegrini D.
      • et al.
      Improving visual field examination of the macula using structural information.
      However, there were still many instances in which the functional grid identified contiguous patterns of visual field defect that were outside this specific test grid; thus, clinicians cannot solely rely on those colocalized positions for identifying glaucomatous vision loss.
      Because the 24-2C can in most cases identify the presence of central visual field defects, but not necessarily “clusters,” when might the 10-2 be clinically useful for the individual patient as part of a personalized management plan?
      • Tomairek R.H.
      • Aboud S.A.
      • Hassan M.
      • Mohamed A.H.
      Studying the role of 10-2 visual field test in different stages of glaucoma.
      ,
      • Phu J.
      • Agar A.
      • Wang H.
      • et al.
      Management of open-angle glaucoma by primary eye-care practitioners: toward a personalised medicine approach.
      Three representative examples of subjects used in the present study are shown in Figure 5, describing the diverse possible outcomes of glaucomatous visual field defects, despite instances of mutually identified or suspected defects. Therefore, identification of central visual field defects on 24-2 or 24-2C could provide an indication for conducting the 10-2 to confirm or fully characterize the defect and its depth or gradient of change.
      Figure thumbnail gr5
      Figure 5Three representative examples of subjects examined in the present study. Each column includes (from the top to bottom rows): the 10-2 deviation map result as per the instrument printout, the 10-2 visual field test locations when shifted according to ganglion cell displacement and flipped for anatomic comparisons with the Ganglion Cell Analysis macular scan, the 24-2C deviation map result as per the instrument printout, and the 24-2C visual field test locations when shifted according to ganglion cell displacement and flipped for anatomic comparisons with the Ganglion Cell Analysis macular scan. Mean deviation and pattern standard deviation for the 10-2 and 24-2C results are shown along the bottom row. MD = mean deviation; PSD = pattern standard deviation.

       Study Limitations

      Our consecutive sampling strategy was used to reduce the probability of spectrum biases by preselecting subjects with different stages of glaucoma or specific glaucoma-related visual field defects. Although this resulted in a diverse sample of subjects, in particular, the representation of subjects with very severe mean deviation scores was low, and there was a high frequency of test locations at which neither structural nor functional losses were apparent. This strategy also resulted in a distribution of central and peripheral visual field defects that was predominantly within an ethnically diverse, urban population consisting of mostly early or moderate open-angle glaucoma patients, and not necessarily comparable to others. Likewise, we incidentally found a greater number of defective locations within the 10-2 grid compared with the 24-2 in male compared with female patients across the entire cohort. The role of gender in glaucoma severity in open-angle glaucoma remains contentious in the literature;
      • Mathan J.J.
      • Patel D.V.
      • McGhee C.N.J.
      • Patel H.Y.
      Analysis of glaucoma subtypes and corresponding demographics in a New Zealand population.
      ,
      • Riva I.
      • Legramandi L.
      • Katsanos A.
      • et al.
      Influence of sociodemographic factors on disease characteristics and vision-related quality of life in primary open-angle glaucoma patients: the Italian Primary Open Angle Glaucoma Study (IPOAGS).
      thus, this particular result may be another characteristic of the present cohort. Likewise, we had a restricted range of refractive errors in the present study. High levels of refractive error are known to affect the optic nerve head integrity and structure-function relationship and would require further comprehensive study.
      • Denniss J.
      • Turpin A.
      • McKendrick A.M.
      Individualized structure-function mapping for glaucoma: practical constraints on map resolution for clinical and research applications.
      • Hirasawa K.
      • Matsuura M.
      • Fujino Y.
      • et al.
      Comparing structure-function relationships based on Drasdo's and Sjostrand's retinal ganglion cell displacement models.
      • Tan N.Y.Q.
      • Sng C.C.A.
      • Ang M.
      Myopic optic disc changes and its role in glaucoma.
      We limited our structure-function analysis to one OCT device produced by the same manufacturer as the perimeter. Other imaging modalities may provide different scan patterns and scan areas that may produce different levels of concordance, especially given areas where function, but not structure, was statistically analyzed.
      Specifically related to the results of the present study, the “function grid” condition highlighted many locations within the 10-2 and 24-2C that were outside the Cirrus OCT scan area, as we have previously similarly reported.
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      Inclusion of the “function grid” condition highlights the need to consider alternative structural measurement methodologies beyond that which is currently constrained by commercial hardware and software, because analyzing visual field test results based on their relevance in structural superimposition may spuriously exclude useful functional information. Characterizing and identifying such scotomata at the border of the central 20 degrees of vision could be missed by small retinal scan regions.
      Previous structure-function studies in glaucoma focus on the macular region and central visual field mapped using the 10-2 due to the advantageous offering of high test density relative to the 24-2.
      • Tong J.
      • Phu J.
      • Khuu S.K.
      • et al.
      Development of a spatial model of age-related change in the macular ganglion cell layer to predict function from structural changes.
      ,
      • Yoshioka N.
      • Zangerl B.
      • Phu J.
      • et al.
      Consistency of structure-function correlation between spatially scaled visual field stimuli and in vivo OCT ganglion cell counts.
      Investigators have focused efforts on optimizing the structure-function relationship with individualized mapping strategies and correction factors using these grids.
      • Hirasawa K.
      • Matsuura M.
      • Fujino Y.
      • et al.
      Comparing structure-function relationships based on Drasdo's and Sjostrand's retinal ganglion cell displacement models.
      ,
      • Turpin A.
      • Chen S.
      • Sepulveda J.A.
      • McKendrick A.M.
      Customizing structure-function displacements in the macula for individual differences.
      Development of structure-function maps using the 24-2C test locations, or other alternatively dense or bespoke test grid variants,
      • Phu J.
      • Kalloniatis M.
      Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
      ,
      • Hood D.C.
      • Nguyen M.
      • Ehrlich A.C.
      • et al.
      A test of a model of glaucomatous damage of the macula with high-density perimetry: implications for the locations of visual field test points.
      may benefit from additional focused study to further optimize this relationship.
      Finally, although the greater test density and number of test locations in the 10-2 would be expected to provide more information on the change in scotoma shape especially in more advanced disease,
      • Rao H.L.
      • Begum V.U.
      • Khadka D.
      • et al.
      Comparing glaucoma progression on 24-2 and 10-2 visual field examinations.
      ,
      • de Moraes C.G.
      • Song C.
      • Liebmann J.M.
      • et al.
      Defining 10-2 visual field progression criteria: exploratory and confirmatory factor analysis using pointwise linear regression.
      this requires a different study design.
      In conclusion, the 24-2C and 10-2 test grids return similar global indices of visual field performance and proportionally similar amounts of central visual field loss. The advantages afforded by the 10-2 test grid over the 24-2C appear to be related to the test density, wherein the additional points contribute to filling in gaps between test locations to return more “clusters” of defects. The increased test resolution provides a more comprehensive description of central vision loss, relevant for the clinician to develop an appropriate management plan, and the patient’s quality of life and activities of daily living. Also, the greater rate of structure-function concordance achieved using the 10-2 provides an avenue for better prediction of central visual function.

      Supplementary Data

      References

        • Jampel H.D.
        • Singh K.
        • Lin S.C.
        • et al.
        Assessment of visual function in glaucoma: a report by the American Academy of Ophthalmology.
        Ophthalmology. 2011; 118: 986-1002
        • Phu J.
        • Khuu S.K.
        • Yapp M.
        • et al.
        The value of visual field testing in the era of advanced imaging: clinical and psychophysical perspectives.
        Clin Exp Optom. 2017; 100: 313-332
        • Heijl A.
        Perimetric point density and detection of glaucomatous visual field loss.
        Acta Ophthalmol (Copenh). 1993; 71: 445-450
        • Zeyen T.G.
        • Zulauf M.
        • Caprioli J.
        Priority of test locations for automated perimetry in glaucoma.
        Ophthalmology. 1993; 100: 518-523
        • Grillo L.M.
        • Wang D.L.
        • Ramachandran R.
        • et al.
        The 24-2 Visual field test misses central macular damage confirmed by the 10-2 visual field test and optical coherence tomography.
        Transl Vis Sci Technol. 2016; 5: 15
        • De Moraes C.G.
        • Hood D.C.
        • Thenappan A.
        • et al.
        24-2 Visual fields miss central defects shown on 10-2 tests in glaucoma suspects, ocular hypertensives, and early glaucoma.
        Ophthalmology. 2017; 124: 1449-1456
        • Yamazaki Y.
        • Sugisaki K.
        • Araie M.
        • et al.
        Relationship between vision-related quality of life and central 10 degrees of the binocular integrated visual field in advanced glaucoma.
        Sci Rep. 2019; 9: 14990
        • Blumberg D.M.
        • De Moraes C.G.
        • Prager A.J.
        • et al.
        Association between undetected 10-2 visual field damage and vision-related quality of life in patients with glaucoma.
        JAMA Ophthalmol. 2017; 135: 742-747
        • Murata H.
        • Hirasawa H.
        • Aoyama Y.
        • et al.
        Identifying areas of the visual field important for quality of life in patients with glaucoma.
        PLoS One. 2013; 8e58695
        • Cheng H.C.
        • Guo C.Y.
        • Chen M.J.
        • et al.
        Patient-reported vision-related quality of life differences between superior and inferior hemifield visual field defects in primary open-angle glaucoma.
        JAMA Ophthalmol. 2015; 133: 269-275
        • Chun Y.S.
        • Sung K.R.
        • Park C.K.
        • et al.
        Vision-related quality of life according to location of visual field loss in patients with glaucoma.
        Acta Ophthalmol. 2019; 97: e772-e779
        • Wu Z.
        • Medeiros F.A.
        • Weinreb R.N.
        • Zangwill L.M.
        Performance of the 10-2 and 24-2 Visual field tests for detecting central visual field abnormalities in glaucoma.
        Am J Ophthalmol. 2018; 196: 10-17
        • West M.E.
        • Sharpe G.P.
        • Hutchison D.M.
        • et al.
        Value of 10-2 visual field testing in glaucoma patients with early 24-2 visual field loss.
        Ophthalmology. 2021; 128: 545-553
        • Jung K.I.
        • Ryu H.K.
        • Hong K.H.
        • et al.
        Simultaneously performed combined 24-2 and 10-2 visual field tests in glaucoma.
        Sci Rep. 2021; 11: 1227
        • Shin H.Y.
        • Park H.L.
        • Park C.K.
        Comparison of visual field tests in glaucoma patients with a central visual field defect.
        Can J Ophthalmol. 2019; 54: 489-494
        • Wu Z.
        • Medeiros F.A.
        • Weinreb R.N.
        • et al.
        Comparing 10-2 and 24-2 Visual fields for detecting progressive central visual loss in glaucoma eyes with early central abnormalities.
        Ophthalmol Glaucoma. 2019; 2: 95-102
        • Fung S.S.
        • Lemer C.
        • Russell R.A.
        • et al.
        Are practical recommendations practiced? A national multi-centre cross-sectional study on frequency of visual field testing in glaucoma.
        Br J Ophthalmol. 2013; 97: 843-847
        • Phu J.
        • Kalloniatis M.
        Ability of 24-2C and 24-2 grids to identify central visual field defects and structure-function concordance in glaucoma and suspects.
        Am J Ophthalmol. 2020; 219: 317-331
        • Phu J.
        • Khuu S.K.
        • Agar A.
        • Kalloniatis M.
        Clinical evaluation of Swedish interactive thresholding algorithm-faster compared with Swedish interactive thresholding algorithm-standard in normal subjects, glaucoma suspects, and patients with glaucoma.
        Am J Ophthalmol. 2019; 208: 251-264
        • Prum Jr., B.E.
        • Rosenberg L.F.
        • Gedde S.J.
        • et al.
        Primary Open-Angle Glaucoma Preferred Practice Pattern® Guidelines.
        Ophthalmology. 2016; 123: P41-P111
        • Phu J.
        • Kalloniatis M.
        A Strategy for Seeding Point Error Assessment for Retesting (SPEAR) in perimetry applied to normal subjects, glaucoma suspects, and patients with glaucoma.
        Am J Ophthalmol. 2020; 221: 115-130
        • Heijl A.
        • Patella V.M.
        • Chong L.X.
        • et al.
        A new SITA perimetric threshold testing algorithm: construction and a multicenter clinical study.
        Am J Ophthalmol. 2019; 198: 154-165
        • Alshareef R.A.
        • Dumpala S.
        • Rapole S.
        • et al.
        Prevalence and distribution of segmentation errors in macular ganglion cell analysis of healthy eyes using Cirrus HD-OCT.
        PLoS One. 2016; 11e0155319
        • De Moraes C.G.
        • Sun A.
        • Jarukasetphon R.
        • et al.
        Association of macular visual field measurements with glaucoma staging systems.
        JAMA Ophthalmol. 2019; 137: 139-145
        • Hood D.C.
        • Kardon R.H.
        A framework for comparing structural and functional measures of glaucomatous damage.
        Prog Retin Eye Res. 2007; 26: 688-710
        • Khuu S.K.
        • Kalloniatis M.
        Standard automated perimetry: determining spatial summation and its effect on contrast sensitivity across the visual field.
        Invest Ophthalmol Vis Sci. 2015; 56: 3565-3576
        • Alluwimi M.S.
        • Swanson W.H.
        • Malinovsky V.E.
        • King B.J.
        A basis for customising perimetric locations within the macula in glaucoma.
        Ophthalmic Physiol Opt. 2018; 38: 164-173
        • Tong J.
        • Phu J.
        • Khuu S.K.
        • et al.
        Development of a spatial model of age-related change in the macular ganglion cell layer to predict function from structural changes.
        Am J Ophthalmol. 2019; 208: 166-177
        • Yoshioka N.
        • Zangerl B.
        • Phu J.
        • et al.
        Consistency of structure-function correlation between spatially scaled visual field stimuli and in vivo OCT ganglion cell counts.
        Invest Ophthalmol Vis Sci. 2018; 59: 1693-1703
        • Drasdo N.
        • Millican C.L.
        • Katholi C.R.
        • Curcio C.A.
        The length of Henle fibers in the human retina and a model of ganglion receptive field density in the visual field.
        Vision Res. 2007; 47: 2901-2911
        • Rabiolo A.
        • Morales E.
        • Kim J.H.
        • et al.
        Predictors of long-term visual field fluctuation in glaucoma patients.
        Ophthalmology. 2020; 127: 739-747
        • Asman P.
        • Heijl A.
        Weighting according to location in computer-assisted glaucoma visual field analysis.
        Acta Ophthalmol (Copenh). 1992; 70: 671-678
        • Heijl A.
        • Lindgren G.
        • Asman P.
        A package for the statistical analysis of visual fields.
        Doc Ophthalmol Proc Ser. 1987; 49: 153-168
        • Phu J.
        • Khuu S.K.
        • Nivison-Smith L.
        • et al.
        Pattern recognition analysis reveals unique contrast sensitivity isocontours using static perimetry thresholds across the visual field.
        Invest Ophthalmol Vis Sci. 2017; 58: 4863-4876
        • Heijl A.
        • Lindgren G.
        • Olsson J.
        Normal variability of static perimetric threshold values across the central visual field.
        Arch Ophthalmol. 1987; 105: 1544-1549
        • Odden J.L.
        • Mihailovic A.
        • Boland M.V.
        • et al.
        Evaluation of central and peripheral visual field concordance in glaucoma.
        Invest Ophthalmol Vis Sci. 2016; 57: 2797-2804
        • Traynis I.
        • De Moraes C.G.
        • Raza A.S.
        • et al.
        Prevalence and nature of early glaucomatous defects in the central 10 degrees of the visual field.
        JAMA Ophthalmol. 2014; 132: 291-297
        • Ekici E.
        • Moghimi S.
        • Hou H.
        • et al.
        Central visual field defects in patients with distinct glaucomatous optic disc phenotypes.
        Am J Ophthalmol. 2020; 223: 229-240
        • Kim J.M.
        • Kyung H.
        • Shim S.H.
        • et al.
        Location of initial visual field defects in glaucoma and their modes of deterioration.
        Invest Ophthalmol Vis Sci. 2015; 56: 7956-7962
        • Sullivan-Mee M.
        • Karin Tran M.T.
        • Pensyl D.
        • et al.
        Prevalence, features, and severity of glaucomatous visual field loss measured with the 10-2 achromatic threshold visual field test.
        Am J Ophthalmol. 2016; 168: 40-51
        • Park H.Y.
        • Hwang B.E.
        • Shin H.Y.
        • Park C.K.
        Clinical clues to predict the presence of parafoveal scotoma on Humphrey 10-2 visual field using a Humphrey 24-2 visual field.
        Am J Ophthalmol. 2016; 161: 150-159
        • Chong L.X.
        • McKendrick A.M.
        • Ganeshrao S.B.
        • Turpin A.
        Customized, automated stimulus location choice for assessment of visual field defects.
        Invest Ophthalmol Vis Sci. 2014; 55: 3265-3274
        • Ballae Ganeshrao S.
        • Turpin A.
        • Denniss J.
        • McKendrick A.M.
        Enhancing structure-function correlations in glaucoma with customized spatial mapping.
        Ophthalmology. 2015; 122: 1695-1705
        • Schiefer U.
        • Papageorgiou E.
        • Sample P.A.
        • et al.
        Spatial pattern of glaucomatous visual field loss obtained with regionally condensed stimulus arrangements.
        Invest Ophthalmol Vis Sci. 2010; 51: 5685-5689
        • Nevalainen J.
        • Paetzold J.
        • Papageorgiou E.
        • et al.
        Specification of progression in glaucomatous visual field loss, applying locally condensed stimulus arrangements.
        Graefes Arch Clin Exp Ophthalmol. 2009; 247: 1659-1669
        • Kucur S.S.
        • Hackel S.
        • Stapelfeldt J.
        • et al.
        Comparative study between the SORS and dynamic strategy visual field testing methods on glaucomatous and healthy subjects.
        Transl Vis Sci Technol. 2020; 9: 3
        • Mathews P.M.
        • Rubin G.S.
        • McCloskey M.
        • et al.
        Severity of vision loss interacts with word-specific features to impact out-loud reading in glaucoma.
        Invest Ophthalmol Vis Sci. 2015; 56: 1537-1545
        • Harvey H.
        • Anderson S.J.
        • Walker R.
        Increased word spacing improves performance for reading scrolling text with central vision loss.
        Optom Vis Sci. 2019; 96: 609-616
        • Ganeshrao S.B.
        • McKendrick A.M.
        • Denniss J.
        • Turpin A.
        A perimetric test procedure that uses structural information.
        Optom Vis Sci. 2015; 92: 70-82
        • Montesano G.
        • Rossetti L.M.
        • Allegrini D.
        • et al.
        Improving visual field examination of the macula using structural information.
        Transl Vis Sci Technol. 2018; 7: 36
        • Tomairek R.H.
        • Aboud S.A.
        • Hassan M.
        • Mohamed A.H.
        Studying the role of 10-2 visual field test in different stages of glaucoma.
        Eur J Ophthalmol. 2020; 30: 706-713
        • Phu J.
        • Agar A.
        • Wang H.
        • et al.
        Management of open-angle glaucoma by primary eye-care practitioners: toward a personalised medicine approach.
        Clin Exp Optom. 2021; 104: 367-384
        • Mathan J.J.
        • Patel D.V.
        • McGhee C.N.J.
        • Patel H.Y.
        Analysis of glaucoma subtypes and corresponding demographics in a New Zealand population.
        Biomed Hub. 2016; 1: 1-8
        • Riva I.
        • Legramandi L.
        • Katsanos A.
        • et al.
        Influence of sociodemographic factors on disease characteristics and vision-related quality of life in primary open-angle glaucoma patients: the Italian Primary Open Angle Glaucoma Study (IPOAGS).
        J Glaucoma. 2018; 27: 776-784
        • Denniss J.
        • Turpin A.
        • McKendrick A.M.
        Individualized structure-function mapping for glaucoma: practical constraints on map resolution for clinical and research applications.
        Invest Ophthalmol Vis Sci. 2014; 55: 1985-1993
        • Hirasawa K.
        • Matsuura M.
        • Fujino Y.
        • et al.
        Comparing structure-function relationships based on Drasdo's and Sjostrand's retinal ganglion cell displacement models.
        Invest Ophthalmol Vis Sci. 2020; 61: 10
        • Tan N.Y.Q.
        • Sng C.C.A.
        • Ang M.
        Myopic optic disc changes and its role in glaucoma.
        Curr Opin Ophthalmol. 2019; 30: 89-96
        • Turpin A.
        • Chen S.
        • Sepulveda J.A.
        • McKendrick A.M.
        Customizing structure-function displacements in the macula for individual differences.
        Invest Ophthalmol Vis Sci. 2015; 56: 5984-5989
        • Hood D.C.
        • Nguyen M.
        • Ehrlich A.C.
        • et al.
        A test of a model of glaucomatous damage of the macula with high-density perimetry: implications for the locations of visual field test points.
        Transl Vis Sci Technol. 2014; 3: 5
        • Rao H.L.
        • Begum V.U.
        • Khadka D.
        • et al.
        Comparing glaucoma progression on 24-2 and 10-2 visual field examinations.
        PLoS One. 2015; 10e0127233
        • de Moraes C.G.
        • Song C.
        • Liebmann J.M.
        • et al.
        Defining 10-2 visual field progression criteria: exploratory and confirmatory factor analysis using pointwise linear regression.
        Ophthalmology. 2014; 121: 741-749