Reproducibility of corneal astigmatism measurements with a hand held keratometer in preschool children.

AIMS To evaluate the overall accuracy and reproducibility of the Alcon portable autokeratometer (PAK) measurements in infants and young children. METHODS The accuracy of the Alcon PAK in measuring toric reference surfaces (1, 3, 5, and 7 D) under various suboptimal measurement conditions was assessed, and the reproducibility of PAK measurements of corneal astigmatism in newborn infants (n = 5), children (n = 19, age 3-5 years), and adults (n = 14) was evaluated. RESULTS Measurements of toric reference surfaces indicated (a) no significant effect of distance (17-30 mm) on accuracy of measurements, (b) no systematic relation between amount of toricity and accuracy of measurements, (c) no systematic relation between angle of measurement and accuracy, (d) no difference in accuracy of measurements when the PAK is hand held in comparison with when it is mounted, (e) no difference in accuracy of measurements when axis of toricity is oriented obliquely than when it is oriented horizontally, with respect to the PAK, and (f) a small positive bias (+0.16 D) in measurement of spherical equivalent. The PAK did not prove useful for screening newborns. However, measurements were successfully obtained from 18/19 children and 14/14 adults. There was no significant difference in median measurement deviation (deviation of a subject's five measurements from his/her mean) between children (0.21 D) and adults (0.13 D). CONCLUSIONS The PAK produces accurate measurements of surface curvature under a variety of suboptimal conditions. Variability of PAK measurements in preschool children is small enough to suggest that it would be useful for screening for corneal astigmatism in young children.