Intraoperative PAR corneal topography system (CTS): comparison of its keratometric readings to manual keratometer, auto-keratometer, EyeSys Corneal Analysis system, and slit lamp PAR CTS in healthy eyes.

Purpose: To compare the keratometric readings obtained from Intraoperative PAR Corneal Topography System (IOPAR) to those produced by manual keratometer (Mnl-Km), autokeratometer (Auto-Km), EyeSys CAS (EyeSys Corneal Analysis System) and slit lamp PAR CTS in healthy eyes.

Methods: All instruments were calibrated prior to use and only data from the best image obtained was used for statistical analysis. Simulated keratometry readings obtained from the central 3-mm zone of the corneas by IOPAR, including flat (K1) and steep (K2) keratometry readings, average keratometric power (AK), astigmatism (As) (difference between steep and flat keratometry readings) and the axis of the steep meridian (Ax) were compared to those from four other units. The latter units were also compared among themselves. Statistical analysis was done for right and left eyes separately. For each variable, average differences between the measurements taken from pairs of instruments were estimated, with corresponding 95% confidence intervals. The degree of agreement between pairs of instruments on individual measurements was additionally assessed, via the use of "Bland-Altman"-type plots, and estimates of the proportion of cases achieving satisfactory agreement. Additionally, for every variable, the average of the measurements taken from the different instruments were compared.

Results: Forty-five (22 right and 23 left) normal corneas of 26 volunteers were examined. On the average, IOPAR tended to measure K1 higher than slit lamp PAR CTS system. Because K2 measurements taken by the IOPAR were higher than that of all other instruments, the As measurements, on the average, were also higher than that of others, with the exception of the PAR CTS. For the same reasons, the IOPAR produced average AK readings that were higher than those taken by EyeSys CAS and PAR CTS. When the individual measurements taken by the IOPAR were compared with each of the other units, according to the arbitrary designation of satisfactory agreement within +/-0.5 D (for K1, K2, AK, and As) and +/-20 degrees (for Ax), for almost all parameters, proportion of differences that were within the agreement range varied from 0.33 to 0.82, with wide confidence intervals (confidence interval lower limits ranging from 0.20 to 0.61 and upper limits ranging from 0.62 to 0.94).

Conclusions: IOPAR is a clinically useful topographic system, producing qualitative and quantitative data in the operating environment that, in normal corneas, on the average, matches those produced by the other units in the clinic. When individually analyzed, its keratometric measurements may show greater variations with respect to other units. Further studies with multiple examiners, in corneas with high or irregular astigmatism are required to establish its reproducibility and efficacy.