Repeatability and comparability of keratometry measurements obtained with swept-source optical coherence and combined dual Scheimpflug-Placido disk-based tomography.

Purpose: To assess the repeatability of keratometric measurements obtained using a reflectance and swept-source optical coherence tomography (SS-OCT) based biometer and to compare these values with those obtained with a dual Scheimpflug-Placido (S-P) disk-based tomographer/topographer.

Setting: Department of Ophthalmology, Penn State College of Medicine, Hershey, Pennsylvania, USA.

Design: Prospective consecutive case series.

Methods: Patients undergoing cataract evaluation were prospectively enrolled. Surgery-naive eyes without corneal pathology were measured 3 times with the reflectance/SS-OCT biometer and once with the dual S-P disk tomographer/topographer. Intradevice repeatability for anterior (K), posterior (PK), and total corneal keratometry (TK) readings were evaluated using within-eye SDs, test-retest repeatability (TRT), coefficients of variance, and intraclass correlation coefficients (ICCs). Bland-Altman analysis was used to assess agreement between devices.

Results: Eighty-two eyes from 48 patients met inclusion criteria. The TRT for the anterior, posterior, and total corneal powers were 0.33, 0.08, and 0.36 diopter (D), respectively. TRT for the anterior, posterior, and total keratometric astigmatism were 0.58, 0.14, and 0.64 D, respectively. Anterior keratometric indices had higher ICCs than posterior indices. Bland-Altman analysis revealed that the SS-OCT consistently reported flatter posterior corneal curvatures than the S-P.

Conclusions: The SS-OCT posterior keratometric measurements contain more noise (lower ICCs), but still have lower within-eye SDs and TRTs than that of anterior measurements. This may be due to the small index of refraction change at the cornea-aqueous interface. Although keratometric indices from the 2 devices are highly correlated, measurements taken using SS-OCT and S-P devices were not interchangeable.