To assess the agreement of corneal power, corneal astigmatism
, axis location, and astigmatic vector component measurements using a Lenstar
LS900, a T-cone add-on, and an ARK 1S autorefractor, furthermore, to assess instrument agreement in a subset of astigmatic eyes.
Retrospective analysis of keratometric readings from 3 instruments for 66 eyes undergoing cataract surgery at the Maryland Veterans Affairs Medical Center from November 2014 to 2015. Agreement in corneal power, mean corneal power, axis location, and astigmatic vector components (J0, J45) between devices was evaluated. Comparability was assessed using intraclass correlation, analysis of variance (ANOVA) one-way analysis, Bland–Altman analysis, and power vector analysis. Intraclass correlation was then assessed for a subset of 42 eyes with corneal astigmatism
greater than or equal to 1 D.
The intraclass correlation between instruments was 0.941 for corneal power, 0.788 for corneal astigmatism
, 0.932 for J0, and 0.809 for J45. The ANOVA one-way repeated P
values comparing Kf
, corneal astigmatism
, axis location, J0, and J45 were all greater than 0.05. Power vector analysis showed nearly all values between instruments to be within ±0.5 D. Finally, the intraclass correlation between instruments in eyes with ≥1 D of corneal astigmatism
as measured by the T-cone was 0.946 for corneal power, 0.837 for corneal astigmatism
, 0.948 for J0, and 0.950 for J45.
The high level of agreement between all devices indicates the instruments may be used interchangeably. Our data suggest that this may also be true for the subset of eyes with greater than or equal to 1 D in astigmatism
, highlighting the toric intraocular lens population.