Repeatability of corneal thickness measurements made by a scanning slit topography system

Two sets of corneal thickness (CT) measurements at five corneal locations were made with Orbscan on 33 normal young adults. No significant intravisit and intervisit differences at any of the corneal locations were found (Repeated measures ANOVA, F < 0.49, p > 0.49). The 95% limits of agreement (95% LA) for between-measurement differences for the central cornea was about +/-19 mum (or 3%), indicating good repeatability; for the inferior, temporal and nasal corneal locations, the 95% LA were close to +/-30 mum (approximate to4-5%); and for the superior corneal location, they were about +/-35 mum (or 5.5%). The estimated number of repeated measurements that should be taken for accurate CT measurements (at any of the five corneal locations and excluding outliers of >3 S.D.) of 2% (standard error) is less than or equal to2, and of 3% is 1. This study shows that only central CT measurements with the Orbscan are repeatable. The role of the Orbscan pachometry is therefore limited and is recommended for central CT measurements only.