[Accuracy and reproducibility of Nidek's photokeratoanalyser]

We studied the accuracy and reproducibility of the keratoscope PKS 1,000 and keratoanalyser PKA 1,000 (Nidek laboratory). When calibrated steel balls are studied without any precaution, the system is unreliable. The photokeratoscope and the keratoanalyser provide good results only if they are used with care. To use the system, we propose the following schedule. Firstly, take a photograph of a steel ball, then set this picture under the video camera and vary the illumination until the accuracy remains unchanged during repeated calibration. With homogenous and constant illumination, the accuracy is about 0.5 diopter.