Skip to main content

  • VA Variability: Snellen Versus ETDRS Outcomes

    By Jean Shaw
    Selected By: Andrew P. Schachat, MD
    Add to My Bookmarks

    Journal Highlights

    Ophthalmology Retina, March 2021

    Download PDF

    The Snellen visual acuity (VA) chart is routinely used in clinical practice, while that developed in the Early Treatment Diabetes Retinopathy Study (ETDRS) is the gold standard for ophthalmic clinical trials. In a retrospective chart review, Yu et al. set out to compare the variability between the two measure­ments of best-corrected visual acuity (BCVA). They found that the ETDRS visual acuity scores were significantly better—and that the difference was more pronounced among eyes with worse visual acuity. The findings sug­gest that caution must be taken when comparing clinical trial VA results with routine clinical outcomes.

    For this study, the authors assessed data from 12 prospective clinical trials conducted at a large urban retina practice in the United States. Eyes were included if a Snellen VA measurement was performed at the visit preceding the initial trial screening and VA was better than counting fingers. All Snellen and ETDRS VA measurements were converted to logMAR units for direct comparison, and the variability be­tween measurements was calculated.

    All told, 413 patients (773 eyes) met the inclusion criteria for this study. Mean patient age was 62.8 years (range, 25-93 years), and there was a mean of 27.2 days between measurements. Out­come measures included absolute VA and VA variability among disease states.

    Mean Snellen VA was 0.40 logMAR (20/50 Snellen equivalent), and mean ETDRS measurement was 0.27 logMAR (20/40 Snellen equivalent). Overall, 76.6% of eyes correctly identified more letters with the ETDRS chart. When VA was assessed by subgroups, eyes with worse vision had a greater difference between Snellen and ETDRS scores: Eyes 20/25 or better were a mean +1.9 letters better on ETDRS testing, and eyes 20/160 or worse were a mean +12.6 letters better on ETDRS testing (p < .05 for both).

    The authors also conducted a subgroup analysis by disease state. Of the five conditions evaluated—diabetic retinopathy (DR) with diabetic macular edema (DME), DR without DME, wet age-related macular de­generation (AMD), dry AMD, and retinal vein occlusion—only those eyes that had DR without DME did not show a significant difference between VA scores. Thus, the authors said, the specific retinal disease state and extent of edema may play a role in the variability between the measurements. (Also see related commentary by Andrew P. Schachat, MD, and Marco A. Zarbin, MD, in the same issue.)

    The original article can be found here.