Improving on ETDRS acuities: Design and results for a computerised thresholding device

Abstract

Aims. All visual acuity data are subject to test-retest variability (TRV). This measurement error obscures true clinical change and reduces the statistical power of clinical trials using acuity as a primary outcome measure. This study was designed to assess whether a computerised system can reduce TRV by taking repeated acuity measurements and averaging them. A computerised system (PC-test) was developed for this purpose and compared in terms of TRV with the current Gold Standard ETDRS logNIAR chart. Methods. A total of 19 subjects with a mean acuity of +0.16 logMAR (range +0.49 to -0.10 logMAR) were recruited. The performance of two computerised tests (one averaging 10 repeats and one five) was compared with that of the ETDRS logMAR chart in terms of TRV and agreement of acuity data. Results. The 10 and five repeat computerised tests (PC-tests) produced a TRV of ±0.11 and ±0.10 logMAR, respectively, compared with ±0.18 logMAR for the ETDRS chart. No significant bias was observed between PC-test and ETDRS acuities. Conclusions. A computerised system that takes repeated acuity measurements and averages them is subject to less TRV than a single ETDRS acuity measurement. A reduced TRV of visual acuity data allows earlier detection of true clinical change in individual patients. It also allows smaller differences between groups to be detected in clinical trials for a given degree of statistical confidence and power.