Lower threshold estimates at the onset of automated perimetry causing artefacts in perimetrically naive subjects.

Reduction in the threshold value of the primary points resulting in depressed points on the pattern deviation plot of Humphrey visual fields was observed in a series of four perimetrically naive subjects. The defect disappeared on repeated testing. This primary point defect could be attributed to the patient's poor understanding of the test, response bias or poor instruction by the perimetrist.