Purpose
To assess the repeatability and comparability of corneal power and central corneal thickness (CCT) measurements obtained using Orbscan II (Bausch & Lomb), Pentacam (Oculus), and Galilei (Ziemer) tomographers.
Design
Prospective, comparative study.
Methods
setting: Departments of Ophthalmology, University of Auckland and Auckland District Health Board, Auckland, New Zealand. study population: Thirty eyes of 30 healthy participants. observations . CCT and corneal power measured using Orbscan II, Pentacam, and Galilei tomography. main outcome measures: Degree of agreement in and repeatability of CCT and corneal power measures.
Results
Orbscan II measured significantly lower CCT compared with Pentacam (20 μm; P < .0005) and Galilei (18 μm; P < .0005). The Orbscan II had wide limits of agreement when compared with both the Galilei (−11 to 47 μm) and Pentacam (−88 to 47 μm). For each device, the intraclass correlation coefficient for CCT was higher than 0.9. The coefficient of variation ranged from 0.33% to 0.93%. There was no significant difference in mean steep keratometry or mean flat keratometry between instrument pairs. However, there was poor agreement in flat keratometry and steep keratometry obtained by Orbscan II compared with those obtained by the Galilei and Pentacam.
Conclusions
The keratometry and pachymetry measurements obtained by Orbscan II, Pentcam, and Galilei tomographers were sufficiently disparate that the 3 devices could not be considered equivalent. All 3 devices demonstrated a high level of repeatability, although the Galilei exhibited the best repeatability.
The cornea is the major refractive element of the eye, and accurate measurements of corneal shape, refractive power, and corneal thickness are crucial to the design and ultimate success of vision-corrective procedures such as refractive and cataract surgery. Measurement precision also is of vital importance in the diagnosis and monitoring of a variety of corneal diseases. Central corneal thickness (CCT) forms part of the assessment of corneal endothelial cell health and therefore is useful in the evaluation and management of endothelial diseases such as endothelial dystrophies, bullous keratopathies, and corneal transplant failure. Moreover, the clinical applications of corneal tomography continue to expand with the development of novel treatments such as collagen corneal cross-linking for keratoconus.
It is important to note that in clinical practice, it is not feasible to obtain direct measurements of characteristics such as corneal thickness, shape, or refractive power. Therefore, indirect measurements are used as a surrogate. Therefore, over the past 30 years, a variety of sophisticated instruments have been developed that facilitate attempts to measure and quantify the living human cornea.
Early computerized topographers used only reflections from a Placido disc element. However, this technique largely is limited to an evaluation of the anterior surface of the cornea, and therefore precludes a comprehensive tomographic analysis of the cornea.
In 1995, the introduction of the Orbscan (Bausch & Lomb, Rochester, New York, USA) heralded a significant advance in the field of corneal topography. This device was founded on the principle of assessing the characteristics of a slit-scanning beam projected across the cornea. This development permitted an assessment of both the anterior and posterior surfaces, thereby allowing a 3-dimensional reconstruction of the cornea. Subsequently the Orbscan II device (Bausch & Lomb) incorporated both the Placido disc and slit-scanning technology—combining the advantages of both approaches.
Further contributions to the domain of corneal tomography were made with the exploitation of the Scheimpflug principle introduced in the form of the Pentacam (Oculus, Wetzlar, Germany). Scheimpflug imaging provides a wide depth of focus, allowing a planar subject that is not parallel to the image plane to be completely in focus. The Pentacam system uses a single rotating Scheimpflug camera and monochromatic slit-light source in combination with a static camera. The Scheimpflug camera scans across the cornea, obtaining multiple slit images that correspond to specific angles along the optical axis.
The Galilei Dual Scheimpflug Analyzer (Ziemer, Port, Switzerland) is a new hybrid device that combines dual rotating Scheimpflug cameras and a Placido disc to assess the anterior segment. The theoretical advantages of the Galilei over the pre-existing Pentacam are 2-fold: the simultaneously recording dual cameras are purported to produce more reliable pachymetry and posterior curvature data, and the addition of the Placido disc data may improve the topographic analysis of the anterior surface.
Previous studies suggest that the Galilei exhibits high reproducibility in measurements of corneal power and CCT. Although other studies have assessed the agreement of the Galilei measurements with the Pentacam separately and the agreement of the Galilei with the longer-established Orbscan, no studies to date have involved a direct comparison of all three devices. The aim of this study was to assess the repeatability and comparability of CCT measurements and corneal power values obtained from the Orbscan II, Pentacam, and Galilei Dual Scheimpflug Analyser.
Methods
The study protocol was approved prospectively by the Northern X Regional Ethics Committee, Auckland, New Zealand. All participants gave informed consent to participate in research. A prospective, comparative study was performed of healthy participants who were health-care colleagues invited to participate after explanation of the study and study purposes by one of the investigators (A.Z.C.). All examinations were performed at the Department of Ophthalmology, Greenlane Clinical Centre, Auckland District Health Board, Auckland, New Zealand. Exclusion criteria were: a history of ocular surgery or trauma, contact lens wear within the preceding 2 weeks, or ocular or systemic disease that may affect the cornea. A single experienced examiner (A.Z.C.) performed 3 sets of measurements per device for each eye in a random order. The 3 devices were the Orbscan II, Pentacam, and Galilei Dual Scheimpflug Analyzer. An acoustic factor of 0.92 was used for all Orbscan II data. Subjects were instructed to blink immediately before each measurement. All of the measurements were conducted in a darkened room. For each patient, all measurements were performed within a 30-minute period. The left eye of each participant was selected. In the event that any device was unable to obtain a scan of acceptable quality after 5 attempts, the right eye of the patient was used instead.
Both the Pentacam and Galilei provide a quality score of individual measurements. In contrast, the Orbscan II does not provide a quality score, but instead automatically discards measurements deemed to be of unacceptable quality. The Galilei breaks down the percentage quality of the image into 4 components: motion compensation, Placido, Scheimpflug, and motion distance. These components then are summarized as an overall quality score along with a reference minimum required percentage score. The Pentacam also provides an overall quality score and recommends that poor-quality measurements be repeated. All 3 tomographers were calibrated during the course of the study.
Statistical analysis was performed using Microsoft Office Excel 2007 (Microsoft, Redmond, Washington, USA) and SPSS software version 19.0 for Windows (SPSS Inc, Chicago, Illinois, USA). Where data were demonstrated to have a normal distribution, as shown by the 1 sample Kolmogorov-Smirnov test, parametric tests were used. Statistical P values of .05 or less were considered significant. The mean values for CCT, steep keratometry (Ks), and flat keratometry (Kf) were compared using a 1-way analysis of variance (ANOVA) for repeated measures, with a post hoc analysis for comparison among the 3 tomographers.
Results
Thirty eyes of 30 participants were enrolled for this study (7 males and 23 females). The mean age was 38.0 ± 11.5 years (range, 20 to 61 years). Twenty-nine left eyes and 1 right eye were assessed (it was not possible to obtain an acceptable Galilei quality score for measurements of the left eye in 1 patient; therefore, the right eye was used instead).
Agreement of Central Corneal Thickness Measurements
A comparison of CCT values recorded by the 3 devices is summarized in Table 1 . A repeated-measures ANOVA with a Greenhouse-Geisser correction determined that mean CCT values differed statistically significantly between instruments ( F (1.486, 43.107) = 45.726; P < .0005). Post hoc tests using the Bonferroni correction revealed that Orbscan II measured significantly lower CCT (524 ± 36 μm) compared with both the Pentacam ( P < .0005) and Galilei ( P < .0005). There was no significant difference in CCT measured by Pentacam (544 ± 30 μm) and Galilei (542 ± 26 μm; P = .629). Bland-Altman plots for agreement of CCT values ( Figure 1 ) showed that the Galilei and Pentacam systems demonstrated better agreement with each other than with the Orbscan II.
| Instrument | CCT (μm) | Ks (D) | Kf (D) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | Mean Δ from Galilei | Mean Δ from Pentacam | Mean ± SD | Mean Δ Galilei | Mean Δ from Pentacam | Mean ± SD | Mean Δ from Galilei | Mean Δ from Pentacam | |
| Galilei | 542 ± 26 | 0 | 44.5 ± 1.6 | 0 | 43.3 ± 1.6 | 0 | |||
| Orbscan | 524 ± 36 | 18 ± 15 | 20 ± 14 | 44.3 ± 1.9 | 0.2 ± 0.5 | 0.2 ± 0.5 | 43.3 ± 1.8 | 0.1 ± 0.4 | 0.1 ± 0.5 |
| Pentacam | 544 ± 26 | −2 ± 8 | 0 | 44.5 ± 1.8 | 0.0 ± 0.3 | 0 | 43.4 ± 1.6 | −0.1 ± 0.2 | 0 |
Repeatability of Central Corneal Thickness
Repeatability parameters for CCT are summarized in Table 2 . The Galilei produced the lowest measurement error and coefficient of variation (COV) and the highest intraclass correlation coefficient (ICC), indicating a higher degree of repeatability compared with the Pentacam and Orbscan II. The Pentacam and Orbscan II also demonstrated good repeatability with COV measurements of 0.7 and 0.93, respectively. The ICC was higher than 0.9 for all devices.
| Instrument | CCT | Ks | Kf | ||||||
|---|---|---|---|---|---|---|---|---|---|
| COV (%) | ICC | ME (μm) | COV (%) | ICC | ME (D) | COV (%) | ICC | ME (D) | |
| Galilei | 0.33 | 0.992 | 2.35 | 0.29 | 0.991 | 0.17 | 0.25 | 0.994 | 0.13 |
| Orbscan | 0.93 | 0.978 | 5.26 | 0.48 | 0.984 | 0.24 | 0.5 | 0.981 | 0.25 |
| Pentacam | 0.70 | 0.979 | 4.36 | 0.24 | 0.994 | 0.14 | 0.25 | 0.992 | 0.15 |
Agreement of Steep Axis Power Measurements
A repeated-measures ANOVA with a Greenhouse-Geisser correction determined that mean Ks values differed statistically significantly between instruments ( F (1.581, 45.862) = 3.781; P = .040). Post hoc tests using the Bonferroni correction revealed no significant difference in Ks measured by Pentacam and Galilei ( P = 1.0) or by Orbscan II and Galilei ( P = .223). The difference in Ks measured by Orbscan II and the Pentacam approached significance ( P = .078). Using Bland-Altman plots for agreement ( Figure 2 ), the Galilei and the Pentacam showed the greatest agreement, and the Pentacam and Orbscan II showed the least agreement.
Repeatability of Steep Axis Power Measurements
Repeatability parameters for K values are summarized in Table 2 . The Galilei exhibited the lowest measurement error and produced a similarly low COV compared with the Pentacam. The Orbscan II produced a higher COV of 0.48%. The ICCs were higher than 0.98 for all devices. The COVs and measurement error (ME) measurements were less than 0.48% and 0.24 diopters, respectively, for all 3 devices.
Agreement of Flat Axis Power Measurements
A comparison of Kf values recorded by the 3 devices is summarized in Table 1 . A repeated-measures ANOVA with a Greenhouse-Geisser correction determined that mean Kf measurements did not differ statistically significantly between instruments ( F (1.289, 37.381) = 1.373; P = .257). Using Bland-Altman plots for agreement ( Figure 3 ), the Galilei and the Pentacam showed the greatest agreement and the Pentacam and Orbscan II showed the least agreement.
Repeatability of Flat Axis Power Measurements
Repeatability parameters for Kf are summarized in Table 2 . The ICC was higher than 0.9 for all devices, indicating a high degree of repeatability. The Galilei produced the lowest measurement error, and both Galilei and Pentacam obtained COVs of 0.25%.
Results
Thirty eyes of 30 participants were enrolled for this study (7 males and 23 females). The mean age was 38.0 ± 11.5 years (range, 20 to 61 years). Twenty-nine left eyes and 1 right eye were assessed (it was not possible to obtain an acceptable Galilei quality score for measurements of the left eye in 1 patient; therefore, the right eye was used instead).
Agreement of Central Corneal Thickness Measurements
A comparison of CCT values recorded by the 3 devices is summarized in Table 1 . A repeated-measures ANOVA with a Greenhouse-Geisser correction determined that mean CCT values differed statistically significantly between instruments ( F (1.486, 43.107) = 45.726; P < .0005). Post hoc tests using the Bonferroni correction revealed that Orbscan II measured significantly lower CCT (524 ± 36 μm) compared with both the Pentacam ( P < .0005) and Galilei ( P < .0005). There was no significant difference in CCT measured by Pentacam (544 ± 30 μm) and Galilei (542 ± 26 μm; P = .629). Bland-Altman plots for agreement of CCT values ( Figure 1 ) showed that the Galilei and Pentacam systems demonstrated better agreement with each other than with the Orbscan II.
