Department of Ophthalmology, University Medical Center of Saarland UKS, Homburg/Saar, Germany.
Acta Ophthalmol. 2013 Jun;91(4):e258-62. doi: 10.1111/aos.12056. Epub 2013 Apr 5.
To evaluate the diagnostic capacity of the Ocular Response Analyser's keratoconus match index (KMI) and keratoconus match probability (KMP) classification in a sample of keratoconus (KC) patients.
Keratoconus match index and KMP from 114 KC eyes, randomly selected from 114 patients with bilateral keratoconus (KCG), were compared with the corresponding ones from 109 normal eyes (CG). Keratoconus match index's predictive accuracy was assessed by receiver operating curves (ROC). Keratoconus match probability level of agreement was evaluated at the different KC stages of the Amsler-Krumeich classification. Correlations were estimated with topographic keratoconus classification (TKC), keratoconus index (KI), index of surface variance (ISV), vertical asymmetry (IVA), height asymmetry (IHA), height decentration (IHD), minimal radius (Rmin), central corneal thickness (CCT), thinnest corneal thickness (TCT) mean keratometry (Km) and intraocular pressure (IOPg).
Mean KMI in KCG and CG was 0.20 ± 0.38 and 0.98 ± 0.25, respectively (p < 0.01). Significant KMI differences (p < 0.01) were detected in different KC groups [range: 0.62 ± 0.38 (KC 1), -0.62 ± 0.04 (KC 4)]. Significant correlation was detected between KC staging and KMI (r = -0.56, p < 0.0001). Keratoconus match probability identified 22.03% of the CG eyes as suspect. Moreover, KMP identified 7.01% and 23.68% of the KCG eyes as normal and suspect, respectively. Receiver operating curves analysis for KMI parameter indicated a predictive accuracy of 97.7% (cut-off point: 0.512, sensitivity: 91.18%, specificity: 94.34%).
Keratoconus match index seems to be a reliable index in keratoconus diagnosis and staging. Keratoconus match probability identifies a significant percentage of topographically defined KC and CG eyes as suspect. Diagnostic capacity of these novel indexes needs to be further explored.
评估眼前节分析系统（ORA）的圆锥角膜匹配指数（KMI）和圆锥角膜匹配概率（KMP）分类在圆锥角膜（KC）患者样本中的诊断能力。
随机选取 114 例双侧圆锥角膜（KCG）患者的 114 只患眼和 109 例正常眼（CG）的 ORA 圆锥角膜匹配指数和 KMP 进行比较。采用受试者工作特征曲线（ROC）评估圆锥角膜匹配指数的预测准确性。评估 Amsler-Krumeich 分类不同 KC 阶段的圆锥角膜匹配概率一致性水平。通过角膜地形图分类（TKC）、圆锥角膜指数（KI）、表面变异指数（ISV）、垂直不对称性（IVA）、高度不对称性（IHA）、高度偏心（IHD）、最小半径（Rmin）、中央角膜厚度（CCT）、最薄角膜厚度（TCT）、平均角膜曲率（Km）和眼内压（IOPg）对圆锥角膜匹配概率进行相关性估计。
KCG 和 CG 的平均 KMI 分别为 0.20 ± 0.38 和 0.98 ± 0.25（p < 0.01）。在不同的 KC 组中发现了显著的 KMI 差异（p < 0.01）[范围：0.62 ± 0.38（KC 1），-0.62 ± 0.04（KC 4）]。KC 分期与 KMI 之间存在显著相关性（r = -0.56，p < 0.0001）。圆锥角膜匹配概率将 22.03%的 CG 眼识别为可疑。此外，KMP 将 KCG 眼中的 7.01%和 23.68%分别识别为正常和可疑。KMI 参数的受试者工作特征曲线分析表明其预测准确率为 97.7%（截断点：0.512，敏感性：91.18%，特异性：94.34%）。
圆锥角膜匹配指数似乎是一种可靠的圆锥角膜诊断和分期指标。圆锥角膜匹配概率可将一定比例的经角膜地形图定义的 KC 和 CG 眼识别为可疑。这些新指标的诊断能力需要进一步探讨。
