Accuracy of kinetic perimetry assessment with the Humphrey 850; an exploratory comparative study

To compare kinetic perimetry on the Humphrey 850 and Octopus 900 perimeters for assessment of visual fields, uniocular rotations and fields of binocular single vision.

Prospective cross section study comparing Humphrey 850 kinetic perimetry to kinetic perimetry using the Octopus 900. Results were compared for both perimeters for the measurement of visual field boundaries, uniocular rotations and fields of binocular single vision in subjects with normal visual function, with comparisons of mean vector extremity values and duration of testing. A visual field boundary overlay was used to assess detection potential of Humphrey 850 kinetic perimetry using I4e and I2e targets in results of known abnormal visual fields.

Fifteen subjects (30 eyes) with normal parameters of visual function underwent dual perimetry assessment. Mean visual field boundaries and ocular rotation extremity values were similar for Humphrey and Octopus kinetic perimetry along horizontal meridians. Measurements for Humphrey perimetry were significantly smaller for superior and inferior visual field and rotations with ceiling effects at approximately 40 and 50 degrees, respectively. Use of visual field boundary overlays for 140 patient results showed high detection of the known abnormal visual field results by the Humphrey 850 perimeter (91.4% with I4e target; 95% with I2e target) but with notable exceptions for peripheral superior visual field defects.

The Humphrey perimeter's aspheric bowl introduces a ceiling effect for measurements in the superior and inferior visual field at approximately 40 and 50 degrees respectively. This results in potential diagnostic accuracy issues when measuring uniocular rotations, fields of binocular single and visual field boundaries in conditions that specifically impair superior and/or inferior ocular motility (e.g., thyroid eye disease) or visual fields (e.g., chiasmal compression).