Reproducibility and accuracy of corneal curvature measurements in patients with and without dry eye: a device-based study.

Purpose: To evaluate the reproducibility and accuracy of corneal curvature measurements using five different devices in patients with and without dry eye, and to identify the most reliable device for clinical use in dry eye patients.

Methods: This study included 68 eyes from dry eye patients (dry eye group) and 48 eyes from non- dry eye patients (non- dry eye group). Corneal curvature was measured with five devices: Corneal topography, ARK-1a, IOL Master 700, OPD-Scan III, Pentacam. A total of 580 examination reports were collected. Reproducibility was assessed using intraclass correlation coefficients (ICCs), and accuracy was evaluated using Bland-Altman analyses. Differences between devices were examined through paired-sample -tests or Wilcoxon signed-rank tests.

Results: In both groups, the ICCs for ARK-1a, IOL Master 700, OPD-Scan III, and Pentacam were all > 0.9, indicating high reproducibility. Compared to the dry eye group, the non- dry eye group generally had narrower confidence intervals. Bland-Altman analyses showed that the 95% limits of agreement (LOA) were wider in the dry eye group, indicating greater variability. However, the LOA of IOL Master 700 and OPD-Scan III remained below 1.0 in the dry eye group, suggesting higher reliability and potential clinical advantages.

Conclusion: Dry eye significantly affects the reliability of corneal curvature measurements, especially with optical reflection-based devices. Corneal topography and Pentacam are more sensitive to tear film abnormalities, while the IOL Master 700 and OPD-Scan III show more consistent results, making them preferable for clinical practice, such as clinical applications including refractive surgery planning, contact lens fitting, and preoperative cataract assessment.