Reproducibility of keratometric measurements decreases with time after blinking

The reproducibility of keratometry measurements was investigated in relation to the elapsed time after blinking.

Thirty ophthalmologically healthy subjects were examined, using a standard corneal topographic instrument (17 women, 13 men, age 25.7+/-5.6 years). Photographs were taken in series of four: at 5, 15, 30, and 60 seconds after a complete blink. The series was repeated three times in each individual. The main outcome measures were the mean values, the standard deviation, and the reproducibility error of the simulated keratometric values (K1, K2), the surface regularity index (SRI), and the surface asymmetry index (SAI).

The mean of K 1 and K 2 values did not change (p=0.684 and p=0.982); however, the measurement error increased significantly for K1 (p=0.007) and for K2 (p=0.038). The mean values of SRI changed significantly (p=0.034) during the 1-minute pause in blinking together with a non-significant change in the standard deviation (p=0.106), without elevation of the measurement error (p=0.619). The elevation of SAI with time was not significant (p=0.093).

The break-up phenomenon of the tear film at the ocular surface induces significant deterioration of the reliability of keratometric measurements on prolonged gaze without blinking. The created error can exceed 0.6 D, which is unacceptably high in practice. Therefore care should be taken to avoid such circumstances during the examination.