Senthil Sirisha, Chary Raghava, Ali Mohammed Hasnat, Choudhari Nikhil, Badakere Swathi, Krishnamurthy Rashmi, Dikshit Siddharth, Garudadri Chandrasekhar
VST Center for Glaucoma Care, LV Prasad Eye Institute, Hyderabad, India.
Center for Biostatistics and Epidemiology, LV Prasad Eye Institute, Hyderabad, India.
Cornea. 2019 Sep;38(9):1117-1123. doi: 10.1097/ICO.0000000000002033.
To study the agreement between scleral intraocular pressure (IOP) measurements using the Schiotz, Icare, and Icare PRO tonometers versus Goldmann applanation tonometer (GAT) in eyes with nonscarred corneas.
This cross-sectional study included 83 eyes of 55 subjects. The order of IOP estimation was the corneal GAT, followed by ICare, ICare PRO, and Schiotz on the corneal and temporal sclera. The agreement between different tonometers and the 95% limits of agreement (LoA) were assessed using Bland-Altman plots. The repeated measures correlation coefficient was calculated between GAT IOP and scleral Schiotz IOP, and the 95% confidence intervals were calculated by the bootstrap method. The linear mixed effects model was used (adjusted for both eyes of the subjects) to generate an equation to predict GAT IOP from scleral Schiotz IOP. The prediction model was validated with new data from 60 eyes. Statistical analyses were performed using "R" software (version 3.3.2).
Comparing the scleral IOP measurements, the mean IOP difference (95% LoA) was the lowest with Schiotz, underestimating by -1.21 mm Hg (7.32, -9.74). Both ICare and ICare PRO significantly overestimated GAT IOP: ICare, 24.6 mm Hg (53.2, -3.97); and ICare PRO, 21.56 mm Hg (52.9, -9.79). The correlation coefficient between scleral Schiotz and GAT IOP was 0.92 (95% bootstrap confidence interval: 0.89, 98). The derived predictive equation was (Equation is included in full-text article.). The mean difference between the predicted GAT IOP and the actual GAT IOP was 0.96 mm Hg with narrow LoA (+1.79, -3.71), validating the prediction model.
Among the tonometers tested, the scleral IOP measurements with Schiotz had the best agreement with the GAT although LoA were wide. The predictive equation may have great potential to predict GAT IOP from scleral IOP readings in eyes with scarred/prosthetic corneas.
研究在无瘢痕角膜的眼中,使用西奥兹眼压计、Icare眼压计和Icare PRO眼压计测量巩膜眼压(IOP)与Goldmann压平眼压计(GAT)之间的一致性。
这项横断面研究纳入了55名受试者的83只眼。眼压估计的顺序为角膜GAT,然后是Icare、Icare PRO,以及角膜和颞侧巩膜的西奥兹眼压计测量。使用Bland-Altman图评估不同眼压计之间的一致性以及95%一致性界限(LoA)。计算GAT眼压与巩膜西奥兹眼压之间的重复测量相关系数,并通过自助法计算95%置信区间。使用线性混合效应模型(对受试者的双眼进行校正)生成一个从巩膜西奥兹眼压预测GAT眼压的方程。使用来自60只眼的新数据对预测模型进行验证。使用“R”软件(版本3.3.2)进行统计分析。
比较巩膜眼压测量值,西奥兹眼压计的平均眼压差异(95% LoA)最小,低估了-1.21 mmHg(7.32,-9.74)。Icare和Icare PRO均显著高估了GAT眼压:Icare为24.6 mmHg(53.2,-3.97);Icare PRO为21.56 mmHg(52.9,-9.79)。巩膜西奥兹眼压与GAT眼压之间的相关系数为0.92(95%自助置信区间:0.89,0.98)。推导的预测方程为(方程包含在全文中)。预测的GAT眼压与实际GAT眼压之间的平均差异为0.96 mmHg,LoA较窄(+1.79,-3.71),验证了预测模型。
在测试的眼压计中,尽管LoA较宽,但使用西奥兹眼压计测量巩膜眼压与GAT的一致性最佳。该预测方程在从有瘢痕/人工角膜的眼中的巩膜眼压读数预测GAT眼压方面可能具有很大潜力。