Diagnostic accuracy of confrontation visual field tests.

OBJECTIVE To determine the diagnostic accuracy of confrontation visual field testing and to compare the accuracy of confrontation tests both individually and in combination. METHODS Patients were prospectively recruited from ophthalmology clinics over a 6-month period. All patients underwent SITA-standard 24-2 Humphrey visual field analysis. Two examiners, masked to the automated perimetry results and the results of the other examiner, assessed patients using 7 common confrontation visual field tests. The order of testing was randomized to reduce any learning effect. For each individual test and combination of tests, the sensitivity, specificity, positive predictive value, and negative predictive value were calculated. RESULTS A total of 301 eyes from 163 patients were included in the study. The average mean deviation was -5.91 +/- 7.72 (SD) dB. Most confrontation tests were insensitive to the identification of field loss. The sensitivity and specificity varied depending on the type, density, and cause of the visual field defect. Kinetic testing with a red target provided the highest sensitivity (74.4%) and specificity (93.0%) of any individual test and when combined with static finger wiggle testing achieved a sensitivity of 78.3% while retaining a specificity of 90.1%. CONCLUSIONS Confrontation visual field tests are insensitive at detecting visual field loss when performed individually and are therefore a poor screening test. Combining confrontation tests is a simple and practical method of improving the sensitivity of confrontation testing.