Eight-second test reliably evaluates neurovisual dysfunction

This experimental study found that an eight-second screening test using rarebits may better detect neural substrate dysfunction than both visual acuity and automated perimetry.

The authors developed a quick, computer-based neural test that uses rarebits ‑ briefly flashed receptive field-size test targets ‑ arranged as segmented white digits against a black backdrop. They administered it to 47 patients with optic nerve or visual pathway lesions of low to moderate severity and 30 healthy subjects.

The test variable was the number of rarebits per segment; digit size was fixed. The test task was to call out verbally all digits seen during an eight-second presentation sequence. Test outcomes were contrasted with the results of an optotype acuity test and automated perimetry.

All subjects easily understood the test task. Healthy subjects read, on average, 5.6 of the 6 digits in the test sequence, while patients averaged 3.3 digits. The numbers of digits read correlated modestly with the acuity and perimetry results. Analysis of receiver operating characteristic curves indicated that the multiple rarebit test provided the best discrimination.

The authors note that when asked to try the Internet version of the test, several independent investigators raised concerns about excessive difficulty. As such, they reiterate the importance of carefully adhering to the specified test conditions (a stand-alone computer screen, 4-m test distance, and dark room conditions). They write that when viewed at a reading distance, the test is almost undoable because the rarebits will be spaced too far apart to generate an impression of digits, and also because targets appearing close to the screen borders will have visual field eccentricities too large to decipher.

They conclude that this multiple rarebit test seems to be highly capable of detecting neurovisual dysfunction. Its simplicity and short duration make it a useful tool in screening settings. It is available for free at www.oft.gu.se/webdiagnos/multibit/multibit.html.