Purchase this article with an account.
A.M. Haun, P. Gunvant, M. Baskaran, L. Vijaya; Central corneal thickness measurement using a pachometer: Mean or lowest values? . Invest. Ophthalmol. Vis. Sci. 2004;45(13):137.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose:Central corneal thickness (CCT) is important in cornea and glaucoma clinics and is usually measured using an ultrasonic pachometer. CCT measurement technique with an ultrasonic pachometer is not yet standardized. Some studies have performed 3 to 5 measurements and have calculated the mean of the values and used that as estimation of CCT. If the pupil is used as an anatomical landmark to define central cornea, the lowest measured would be the most indicative of CCT as it would be obtained when the probe is placed most perpendicular to the cornea, whereas higher values will be obtained due to oblique placement of the probe (Copt et al., 1999, Gunvant et al., 2003). We sought to investigate how much error is induced if the mean of the readings is taken instead of the lowest value obtained as CCT. Additionally we also investigated minimum number of measurements required to obtain a reading equivalent to the lowest of 10 readings. Methods:An experienced practitioner measured the CCT with an ultrasonic pachometer using pupil as an anatomical landmark to identify the central cornea in 50 eyes of 25 subjects. Ten measurements were made in rapid succession for each eye. The means of the first 3, 5 and 10 measurements were calculated and compared with the lowest measurement recorded in an individual. The cumulative odds ratio was calculated to find the minimum number of measurements required to obtain a reading equivalent to the lowest of 10 readings. Results:The mean values obtained with 3, 5 and 10 measurements were 515.8, 516.0 and 516.3 microns respectively. The mean of lowest value recorded was 512.3 microns. The difference in mean values obtained by different measurement techniques was significant (repeated measures ANOVA, F =75.66, p<0.0001). Bonferroni contrast revealed that the taking of average values, i.e. mean of 3, 5 or 10, yielded significantly higher measurements than the lowest values obtained. Comparison of the means of 3, 5, and 10 CCT measurements was not significant (p>0.05). The error induced due to taking the mean CCT values instead of lowest values was up to 10.33 microns. There is only a 56% chance of obtaining the lowest of 10 values within the first 3 readings and a 72% chance in the first 5 readings. There was a 94% chance of finding the lowest CCT within the first 8 measurements. Conclusions:Using the mean of recorded values for CCT instead of the lowest value could lead to an error of up to 10 microns in an individual.
This PDF is available to Subscribers Only