Interexaminer reproducibility for subjective refractions for an ametropic participant

To investigate interexaminer reproducibility of non-cycloplegic subjective refractions. Subjective refractions are frequently determined, and it is important to know whether differences in refractive state over time constitute meaningful, non-random change.

Fifty registered and experienced (≥5 years) optometrists from a single geographic region performed non-cycloplegic subjective refractions for a participant with moderate left eye(OS) to severe right eye (OD) ametropia. Subjective refractions were transformed to power matrices for analysis with stereopairs, distribution ellipsoids and polar profiles of variance of dioptric power. Absolute 95% limits of reproducibility ([Formula: see text](SD)) for excesses of subjective refractions for the right and left eyes separately from mean subjective refractions were determined.

Mean subjective refractions were -7.68-4.50×10 and -4.59-1.85×178 for the right and left eyes, respectively. The 95% absolute reproducibility limits for the stigmatic coefficients (spherical equivalents) were ≤1.71 D and ≤0.75 D for the right and left eyes, but corresponding limits for astigmatic coefficients were smaller (≤0.69 D).

Removal of possible outliers for OD and OS, respectively, reduces the absolute 95% reproducibility limits for the stigmatic and astigmatic coefficients to ≤0.97 D and ≤0.49 D, thus improving interexaminer reproducibility. However, these results suggest caution with analysis of refractive data where subjective rather than objective methods are applied for longitudinal and epidemiological studies.