Reliability and repeatability of the Pentacam on corneal curvatures

The purpose of this study was to assess the intra-observer reliability and inter-session repeatability of the Pentacam system (Oculus, Inc.) on the simulated keratometry readings (Sim K) and peripheral corneal curvatures (PCC) at both the anterior and posterior corneal surfaces.

One eye was randomly selected from each of the 39 normal subjects from the School of Optometry at The Hong Kong Polytechnic University. In the first session, three readings were taken for reliability analysis. Measurements were repeated in a second session scheduled one to two weeks later for repeatability analysis. Two approaches were used, comparing the first reading between the two sessions and comparing the average from three readings between the two sessions. Both the anterior and posterior Sim K and PCC were calculated from the tangential and axial maps. The inter-session repeatability was reported as the coefficient of repeatability (COR = +/- 1.96 x standard deviation of differences). Relative repeatability (RR) was a percentage of the ratio of COR to the mean.

All parameters showed good intra-observer reliability (Cronbach's alpha > or = 0.925; intra-class correlation coefficient, ICC > or = 0.804), except for the posterior power vector J45 (= 0.896; ICC = 0.742) and tangential PCC at the superior region (= 0.891; ICC = 0.732). The inter-session repeatability on anterior Sim K (COR < or = +/-0.21 D; RR < or = 0.48 per cent) and posterior Sim K (COR < or = +/-0.09 D; RR < or = 1.39 per cent) was good. Axial PCC showed better repeatability than the tangential PCC. The COR from using the first approach for all the posterior PCC was within 0.23 D, while the COR reduced to within 0.16 D from using the second approach.

The Pentacam showed good intra-observer reliability on Sim K and PCC measurements. Averaging three readings from one visit could improve repeatability between visits.