Abstract
Purpose:
The last decade has witnessed a renewed interest in the study of corneas from cadavers to improve corneal surgeries. In keeping with this trend, the goal of this work is to study the random error of ultrasonic pachymetry of cadaver corneas, preserved either in moist chamber or in Optisol GS, by determining the coefficient of repeatability of the method.
Methods:
The central corneal thicknesses of 407 cadaver eyes were measured with the Sonomed PcScan 300AP ultrasound pachymeter, separated into two groups: 299 eyeballs preserved in a moist chamber for up to 5 days and 108 corneas preserved in Optisol GS for up to 15 days. The eyeballs were further divided into two subgroups: 151 right and 148 left eyes. Each cornea was measured twice with an interval of 10 seconds by the same operator, without knowledge of the results and without the possibility of correction. In the first group the corneas were measured "in situ" within the moist chamber; in the second, they were mounted on an artificial chamber under the controlled pressure of 25mmHg. The coefficient of repeatability (CR) was accessed using the method of Bland and Altman with the formula CR = 2.77 sw ; the standard error of the CR with the formula SE = 1.96 sw / (N-1)1/2 and its 95% CI with the formula CR ±1,98(SE).Using a one-way analysis of variance, with the cornea as a factor sw is the within-subject standard deviation.
Results:
The repeatability coefficients of both right and left eyeballs, preserved in moist chamber, were identical amounting 39μm. The same coefficient for the corneas preserved in Optisol GS was 44 μm. The differences of each pair of measurements did not show significant correlation with the magnitude of the center thickness in all groups..
Conclusions:
The random error of the ultrasonic pachymetry of the corneas preserved in the moist chamber should be between 35 to 45µm and the corresponding error for corneas preserved in Optisol GS between 40 and 50 µm, with a probability of 95%.There does not appear to be any relation between the error and the thickness of the central cornea.