Using computers to test visual acuity.

BACKGROUND This paper describes a computer-generated acuity test that can be produced on IBM compatible equipment in the Windows 3.1 environment. METHODS The acuity test was applied to 14 observers using two psychophysical methods (staircase and method of constant stimuli [MOCS]) to determine the validity and reliability of the outcomes. Comparisons were made between test methods and a hard copy LogMAR chart using statistics suited to intra-class comparisons. The effect that screen luminance has on outcomes and hardware warm-up requirements were also considered. RESULTS The computer staircase was found to provide a reliable estimate of acuity. The hard copy chart and computer staircase have an Intraclass Correlation Coefficient (ICC) of 0.98. The threshold derived from the different psychophysical methods is also highly correlated (ICC = 0.98). The monitor provides a stable luminance 30 minutes after a cold start, and a 1 logunit (10 times) change in screen luminance has little effect on outcomes. CONCLUSIONS A computer-driven acuity staircase provides a good estimate of threshold if a 30-minute warm-up period is allowed to stabilize monitor luminance. It also gives a measure of variability. A staircase is adequate for clinical applications as it provides a reasonable compromise between test time and accuracy. An extra 30 seconds is needed to obtain a staircase result (compared with hard copy), which is offset by the provision of this error term.