Purpose: Most retrospective reviews convert Snellen visual acuity measurements obtained during routine clinic visits to logarithm of the minimum angle of resolution (logMAR) units so that statistical manipulations can be performed. However, visual acuity measurements expressed as logMAR units are not intuitively interpretable by clinicians. A more intuitive approach is presented here which uses the conversion of Snellen visual acuity fractions to Early Treatment Diabetic Retinopathy Study (ETDRS) letter scores for statistical manipulations.
Methods: Snellen visual acuity measurements were converted to approximate ETDRS (approxETDRS) letter scores for statistical manipulations and then converted back to Snellen equivalent fractions. The formula to convert Snellen visual acuity measurements to approxETDRS letter scores is 85 + 50 x log (Snellen fraction), which may be rounded to the nearest letter.
Results: A linear relationship exists between true ETDRS letter scores, approxETDRS letter scores, and logMAR units. The interconversion between Snellen visual acuity measurements, logMAR units, and approxETDRS letter scores was prepared in a tabular form for easy reference. The same outcomes (in Snellen fractions) were obtained with statistical manipulation of either approxETDRS letter scores or logMAR conversions.
Conclusion: Conversion of Snellen visual acuity fractions to approxETDRS letter scores for the purpose of performing statistical manipulations provides more readily interpretable outcomes compared with the current strategy of converting Snellen visual acuity fractions to logMAR units.