A Combined Biomechanical and Tomographic Model for Identifying Cases of Subclinical Keratoconus

ATALAY E., ÖZALP O., EROL M. A., BİLGİN M., YILDIRIM N.

CORNEA, cilt.39, sa.4, ss.461-467, 2020 (SCI-Expanded) identifier identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 39 Sayı: 4
  • Basım Tarihi: 2020
  • Doi Numarası: 10.1097/ico.0000000000002205
  • Dergi Adı: CORNEA
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, CAB Abstracts, EMBASE, MEDLINE, Veterinary Science Database
  • Sayfa Sayıları: ss.461-467
  • Anahtar Kelimeler: keratoconus, subclinical keratoconus, corneal biomechanics, tomography, pentacam, ocular response analyzer, FORME-FRUSTE KERATOCONUS, IATROGENIC KERATECTASIA, STATISTICAL CORRECTION, INTRAOCULAR-PRESSURE, FELLOW EYES, CORNEAL, INDEXES, PARAMETERS
  • Eskişehir Osmangazi Üniversitesi Adresli: Evet

Özet

Purpose: To develop a combined biomechanical and tomographic model for identifying eyes with subclinical keratoconus (SKC) that are categorized as normal or borderline in the Pentacam Belin/Ambrosio Enhanced Ectasia Display. Methods: This case-control study comprised 62 eyes with SKC and randomly selected eyes of 186 age-matched healthy controls. SKC was defined as the presence of the following: 1) normal topography, topometric indices, and slit lamp; 2) normal or borderline Belin/Ambrosio Enhanced Ectasia Display D index, back and front elevation difference; and 3) keratoconus in the fellow eye. Stepwise logistic regression analysis was performed to identify the best variable combination for detecting SKC cases from Ocular Response Analyzer and Pentacam parameters. Receiver operating characteristic curve analysis was used to determine the predictive accuracy [area under the curve (AUC)] of the model. Based on the predictors in the final logistic regression model, a linear equation was derived using the discriminant function analysis. Results: The final model (AUC: 0.948, sensitivity: 87.1%, and specificity: 91.4%) chose corneal hysteresis (CH) and D index from a total of 63 candidate variables. The final model had a higher AUC compared with D (0.933, P = 0.053) and CH (0.80, P < 0.001) alone. According to the discriminant function analysis, a higher CH was required with increasing D index to classify an eye as normal. Conclusions: The proposed combined model provided varying cutoffs for CH and D as a function of the other. The probability plot as a function of CH and D index may be used for identifying eyes with SKC.