Precision of the Nidek Tonoref II autokeratometer: how many repeated measurements are required?

To determine the minimal number of repeated measurements required for precise Nidek Tonoref II autokeratometry.

This prospective, non-intervention study was performed at the Department of Ophthalmology, Randers, Denmark. We used the Nidek Tonoref II autokeratometer to perform 10 successive measurements on 100 right eyes of cooperative individuals. Each keratometry was converted to the spherical equivalent power (SE), while the net astigmatism was converted to polar values along zero (KP(0)) and 45 degrees (KP(45)). All units were in dioptres (D). The precision was calculated as the mean absolute difference between paired measurements, using one or the average of two, three, four or five autokeratometries. Statistical assessment was performed with Dunn's test for repeated measurements with a Bonferroni correction.

The precision of SE, KP(0) and KP(45) increased statistically significantly from one to three measurements, with no significant improvement for autokeratometries based on four or five measurements. There was no significant precision difference between one and two measurements.

A single keratometry is inadequate, but the vector average of three measurements is sufficient for precise autokeratometry with the Tonoref II device. The consistent use of three keratometries with this device may increase the precision of spherical and toric IOL calculation.