PURPOSE To evaluate the repeatability of a new biometer using Scheimpflug technology combined with partial coherence interferometry (PCI) (Pentacam AXL) and its agreement with a device based on optical low-coherence… Click to show full abstract
PURPOSE To evaluate the repeatability of a new biometer using Scheimpflug technology combined with partial coherence interferometry (PCI) (Pentacam AXL) and its agreement with a device based on optical low-coherence reflectometry (OLCR), the Allegro Biograph. SETTING Oftalvist Centro Integral Ocular Jerez, Jerez de la Frontera, Spain. DESIGN Evaluation of a diagnostic test. METHODS The mean keratometry (K), central corneal thickness (CCT), anterior chamber depth (ACD), and axial length (AL) were measured with the 2 devices 3 times by the same examiner in 2 groups (patients with cataract and patients without cataract). The repeatability was determined using the within-subject standard deviation, test-retest repeatability, coefficient of variation, and intraclass correlation coefficient. The correlation was evaluated with the Pearson coefficient and interchangeability with the Bland-Altman plot. RESULTS Eighty eyes (40 eyes in each group) of 80 patients were analyzed. Significant differences were found between the Scheimpflug-PCI device and the OLCR device for mean K in the normal group (P < .001) and for CCT in the normal group (P < .05) and the cataract group (P < .001). There were no differences between devices in ACD and AL in either group. The repeatability between devices was similar. Although a significant correlation between devices was found for all measurements (all P < .001), wide limits of agreement were found in both groups for all biometric parameters. CONCLUSIONS The Scheimpflug-PCI and OLCR devices showed excellent intravisit repeatability and high correlation for mean K, CCT, ACD, and AL in healthy and cataractous eyes. No differences were found in AL; however, the 2 devices might not be interchangeable.
               
Click one of the above tabs to view related content.