In the present study, the repeatability and reproducibility of the corneal power and astigmatism measurements using placido-based video keratography were evaluated and the agreement with other keratometers were compared.

This prospective study included 45 patients (45 eyes) scheduled to undergo cataract surgery between November 2015 and January 2016. Three sets of corneal power and astigmatism were measured using placido-based video keratometer (Keratograph® 5M), automatic keratometer (KR-8900®), manual keratometer (B×L manual keratometer®), Placido-scanning-slit keratometer (ORBscan II®), Scheimpflug keratometer (Pentacam®), and low coherence interferometry (Lenstar LS900®). Reliability of each device was analyzed using the coefficient of variation, standard deviation and intraclass correlation coefficient. Repeated measures analysis of variance was used to analyze the interdevice comparison of mean absolute difference. The agreement between the devices was evaluated with 95% limits of agreement (LoA) and Bland-Altman plots.

The mean corneal power, astigmatism and power vector analysis (J₀, J₄₅) were not significantly different among devices ( p > 0.05). In the Bland-Altman plot analysis, the 95% LoA of corneal power, J₀, and J₄₅ when comparing Keratograph® 5M with others ranged from -0.78 to 0.55 D, from -0.42 to 0.45 D, and from -0.39 to 0.47 D, respectively.

Keratograph® 5M showed good repeatability and reproducibility of corneal power and astigmatism measurements and was interchangeable with other keratometers.