Sharma Meghan, Savatovsky Eleonore, Huertas Laura, O'Brien Robert, Grajewski Alana, Bitrian Elena
Bascom Palmer Eye Institute at the University of Miami Miller School of Medicine, 900 NW 17 Street, Miami, Florida 33136.
Ophthalmol Sci. 2024 Apr 16;4(5):100534. doi: 10.1016/j.xops.2024.100534. eCollection 2024 Sep-Oct.
To test the use of a virtual reality visual field headset (VRVF) for implementation of the Esterman visual field (EVF) test as compared with standard automated perimetry (SAP) among people with glaucoma.
Experimental design.
Patients with mild to severe glaucoma ranging from 10 to 90 years who presented for follow-up at a glaucoma clinic in Miami, Florida were eligible.
Participants performed the EVF test on both SAP and VRVF. Five glaucoma-trained ophthalmologists were then asked to rate all anonymized SAP and RVF tests as a "pass" or "failure" based on Florida state law.
Point-by-point concordance between original VRVF EVF test results and SAP EVF test results was calculated using the Kappa statistic. Concordance between SAP and VRVF was secondarily assessed with a conditional logistic regression based on the pass-failure determinations by the glaucoma-trained ophthalmologists. Interrater agreement on test pass-failure determinations was also calculated. Finally, test results on SAP versus VRVF were compared based on Esterman efficiency score (EES), the number of correct points divided by the number of total points, and duration of testing.
Twenty-two subjects were included in the study with ages ranging from 14 to 78 years old. Concordance between VRVF and SAP test using point-by-point analysis was poor ( = 0.332, [95% confidence intervals {CI}: 0.157, 0.506]) and somewhat increased using pass-failure determinations from ophthalmologists ( = 0.657, [95% CI: 0.549, 0.751]). Ophthalmologists were more likely to agree amongst themselves on pass-failure determinations for VRVF tests ( = 0.890, [95% CI: 0.726, 0.964]) than for SAP ( = 0.590, [95% CI: 0.372, 0.818]); however, VRVF demonstrated significantly lower EES than SAP (median EES difference: 4.5 points, = 0.021).
This pilot study is the first to assess the implementation of the EVF test using a virtual reality headset. Based on the weak overall agreement between VRVF and SAP, the current VRVF EVF test is not an acceptable determinant of driver's licensing. However, ophthalmologists were more likely to agree amongst themselves on VRVF test reports than on SAP reports. With further testing and improvement, virtual reality may eventually become a portable and convenient method for administering the EVF test.
Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
与标准自动视野计（SAP）相比，测试虚拟现实视野头戴设备（VRVF）在青光眼患者中实施埃斯特曼视野（EVF）测试的情况。
实验设计。
年龄在10至90岁之间、在佛罗里达州迈阿密的青光眼诊所接受随访的轻至重度青光眼患者符合条件。
参与者在SAP和VRVF上均进行EVF测试。然后，邀请5名经过青光眼培训的眼科医生根据佛罗里达州法律，将所有匿名的SAP和VRVF测试评定为“通过”或“未通过”。
使用Kappa统计量计算原始VRVF-EVF测试结果与SAP-EVF测试结果之间的逐点一致性。基于经过青光眼培训的眼科医生的通过/未通过判定，通过条件逻辑回归对SAP和VRVF之间的一致性进行二次评估。还计算了评估者之间在测试通过/未通过判定上的一致性。最后，根据埃斯特曼效率得分（EES，正确点数除以总点数）、测试持续时间，比较SAP与VRVF的测试结果。
22名受试者纳入研究，年龄在14至78岁之间。使用逐点分析时，VRVF和SAP测试之间的一致性较差（κ = 0.332，[95%置信区间{CI}：0.157，0.506]），而根据眼科医生的通过/未通过判定，一致性有所提高（κ = 0.657，[95% CI：0.549，0.751]）。眼科医生在VRVF测试的通过/未通过判定上彼此之间的一致性（κ = 0.890，[95% CI：0.726，0.964]）高于SAP（κ = 0.590，[95% CI：0.372，0.818]）；然而，VRVF的EES显著低于SAP（中位数EES差异：4.5分，P = 0.021）。
这项初步研究是首次评估使用虚拟现实头戴设备实施EVF测试的情况。基于VRVF和SAP之间总体一致性较弱，当前的VRVF-EVF测试不是驾驶执照可接受的判定方法。然而，眼科医生在VRVF测试报告上彼此之间的一致性高于SAP报告。随着进一步测试和改进，虚拟现实最终可能成为实施EVF测试的一种便携且方便的方法。
本文末尾的脚注和披露中可能会有专有或商业披露信息。
