[Variability of findings in light sense and ring perimetry]

The variability of the interpretation of visual fields by physicians has so far been thoroughly examined only with reference to light sense perimetry. Therefore, we studied the variability in assigning visual fields to topical diagnoses of the visual pathway with the ring perimeter. For this purpose we examined the proportion of correct assignments of physicians and the reproducibility of their assignments. Two experienced ophthalmologists had to assign 76 selected pairs of right and left visual fields to one of the following diagnosis without any information about further clinical findings: (1) normal (n = 5), (2) media opacity (n = 5), (3) retinal lesion (n = 5), (4) optic disc lesion (n = 6), (5) optic nerve lesion (n = 6), (6) lesion of the chiasma (n = 5) and (7) retrochiasmal lesion (n = 6). The assignment was repeated after intervals of 6 weeks and 10 months. Our statistical analysis of the number of correct assignments revealed no significant differences between the perimeters or between the physicians. The amount of correct assignments with both perimeters decreased from 75% and 66% during the first run through 60% and 70% during the second run to 58% and 60% during the third. Neither perimeter showed any major differences in the reproducibility of the assignments (Fig. 4). During the three runs only 47% (physician 1) and 58% (physician 2) of the cases were assigned three times to the same diagnoses with the octopus. The corresponding figures for the ring perimeter were 58% and 60%. This is regarded as an indicator of the unreliability in the interpretation of visual fields.