Limits of Agreement Between the Optical Pachymeter and a Noncontact Specular Microscope

Abstract

To determine the limits of agreement between central corneal thickness (CCT) measurements made with the slit lamp-attached optical pachymeter and the SP2000P noncontact specular microscope. Triplicate readings for CCT were obtained for each of 130 (right) eyes of 130 patients, using the slit lamp-attached optical pachymeter and then the SP2000P noncontact specular microscope. The average CCT measured by each method was compared. Subsequently, the mean difference between both sets of measurements was assessed, and the 95% confidence interval (limits of agreement) between both techniques was determined. The mean +/- SD CCT measured by the optical pachymeter was 543 +/- 34 microm and 532 +/- 34 microm for the specular microscope. We found a statistically significant (P < 0.001) mean bias of 10 mum between CCT values measured with both types of equipment, with the optical pachymeter returning the higher values. The coefficient of variation was 6.3% for the optical pachymeter and 6.4% for the specular microscope. The right eye CCT measurements made by the optical pachymeter are, on average, 10 mum thicker than those made with the SP2000P specular microscope, which suggests that both pieces of equipment cannot be used interchangeably to monitor CCT changes in patients. Excluding left eye measurements, the reliability of the optical pachymeter is identical to that of the noncontact specular microscope.