State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Guangdong Provincial Clinical Research Center for Ocular Diseases, Sun Yat-sen University, Guangzhou, China.
Ophthalmic Res. 2023;66(1):1191-1197. doi: 10.1159/000530992. Epub 2023 Jul 18.
The aim of the study was to compare and explore the agreement between the nomogram tool and ultrasound biomicroscopy (UBM) images method to calculate the ultrasound cycloplasty (UCP) probe model in Chinese glaucoma patients.
Retrospective analysis of Chinese glaucoma patients who visited Zhongshan Ophthalmic Center in Guangzhou from January to December 2019 and were eligible for UCP surgery. Visual acuity, intraocular pressure (IOP), ocular axial length (AL), and horizontal corneal diameter (white to white [WTW]) were measured. UBM images with clear ciliary body imaging and AL and WTW data were sent to trained personnel for probe model measurements. The data calculated by both methods were analyzed using unweighted and weighted κ statistics. The level of agreement refers to Landis and Koch's guideline for the strength of agreement indicated with weighted κ values.
1,061 eyes of 642 patients were involved, with a mean age of 61.66 ± 11.66 years. Their best-corrected visual acuity converted to logarithm of minimal-angle-of-resolution (logMAR) scores of -0.18-3.00 with a mean value of 0.69 ± 0.77. IOP was 22.0-60.0 mm Hg with a mean of 27.97 ± 5.66 mm Hg. The mean AL and WTW were 22.88 ± 1.33 (19.15-32.14) mm and 11.52 ± 0.49 (10.00-12.90) mm, respectively. The agreement between the two methods was fair (weighted κ = 0.299), matching in 62.86% of eyes (weighted κ = 0.299, κ = 0.264). The agreement in primary open angle glaucoma, acute primary angle-closure glaucoma, chronic primary angle-closure glaucoma, and secondary glaucoma patients was 60.85% (weighted κ = 0.336, κ = 0.301), 65.06% (weighted κ = 0.146, κ = 0.127), 62.26% (weighted κ = 0.204, κ = 0.184), and 57.97% (weighted κ = 0.332, κ = 0.280) of eyes, respectively.
The agreement between UBM images and the nomogram tool to calculate the UCP probe model of Chinese patients is at a fair level. The nomogram tool prefers to use larger probes. Improvements to the nomogram tool, such as including data from more ethnic groups and being able to calculate separately for different types of glaucoma, are needed to improve accuracy. The inclusion of parameters or images from more directions of the eye may help measure probe models more accurately for both the nomogram tool and the UBM image measurement.
本研究旨在比较和探讨在计算中国青光眼患者超声睫状体成形术（UCP）探头模型时，列线图工具和超声生物显微镜（UBM）图像方法之间的一致性。
回顾性分析 2019 年 1 月至 12 月在广州中山大学眼科中心就诊并符合 UCP 手术条件的中国青光眼患者。测量视力、眼压（IOP）、眼轴长度（AL）和水平角膜直径（白到白[WTW]）。将具有清晰睫状体成像和 AL 及 WTW 数据的 UBM 图像发送给经过培训的人员进行探头模型测量。使用未加权和加权 κ 统计分析两种方法计算的数据。一致性水平指的是 Landis 和 Koch 指南中表示一致性强度的加权 κ 值。
共纳入 642 例患者的 1061 只眼，平均年龄为 61.66±11.66 岁。他们的最佳矫正视力转换为最小分辨角对数（logMAR）分数为-0.18-3.00，平均值为 0.69±0.77。IOP 为 22.0-60.0mmHg，平均值为 27.97±5.66mmHg。平均 AL 和 WTW 分别为 22.88±1.33（19.15-32.14）mm 和 11.52±0.49（10.00-12.90）mm。两种方法之间的一致性为中等（加权 κ=0.299），62.86%的眼睛匹配（加权 κ=0.299，κ=0.264）。在原发性开角型青光眼、急性原发性闭角型青光眼、慢性原发性闭角型青光眼和继发性青光眼患者中，两种方法的一致性分别为 60.85%（加权 κ=0.336，κ=0.301）、65.06%（加权 κ=0.146，κ=0.127）、62.26%（加权 κ=0.204，κ=0.184）和 57.97%（加权 κ=0.332，κ=0.280）的眼睛。
UBM 图像与列线图工具计算中国患者 UCP 探头模型的一致性处于中等水平。列线图工具倾向于使用更大的探头。需要改进列线图工具，例如纳入更多种族的数据，并能够分别为不同类型的青光眼计算，以提高准确性。纳入来自眼睛更多方向的参数或图像可能有助于更准确地测量列线图工具和 UBM 图像测量的探头模型。
