Related Experiment Video
Updated: Jun 8, 2026
Comparison of Agreement and Accuracy using Binocular Wavefront Optometer with Autorefractor and Phoropter
Published on: September 16, 2025
Visual acuity testing using autorefraction or pinhole occluder compared with a manual protocol refraction in
Jennifer K Sun1, Lloyd Paul Aiello, Jerry D Cavallerano
1Beetham Eye Institute & Eye Research Section, Joslin Diabetes Center, Boston, Massachusetts, USA.
Purpose:
To compare visual acuity (VA) scores obtained after autorefraction or using a pinhole occluder to scores obtained after refraction according to a standard clinical research protocol.
Design:
Prospective, comparative case series.
Participants:
One hundred ten study participants (209 eyes) with diabetes mellitus and a broad range of diabetic retinopathy severity and VA.
Methods:
We measured VA after autorefraction by a Topcon KR-8000 autorefractor as well as after a Diabetic Retinopathy Clinical Research Network (DRCR.net) protocol manual refraction. The order of testing was randomized and examiners were masked to the source of each refraction. A second VA measurement, utilizing an identical DRCR.net manual refraction, was made in a subset of eyes (n = 144; 69%) to establish test-retest variability for comparison purposes. All eyes underwent VA testing using a pinhole occluder.
Main Outcome Measures:
Best corrected VA as measured by the Electronic Early Treatment Diabetic Retinopathy Study Visual Acuity Test (E-ETDRS).
Results:
In all eyes, the median E-ETDRS VA letter score (EVA) obtained after manual refraction (MR-EVA) was 69 (Snellen equivalent 20/40), ranging from 4 to 93 (20/800 to 20/16). The median MR-EVA was slightly better than the median EVA obtained after autorefraction (AR-EVA), with a median difference (AR-EVA - MR-EVA) of -1 letter (25th, 75th percentiles, -4, 2 letters). The absolute difference between AR-EVA and MR-EVA was similar to the test-retest variability of MR-EVA alone. In contrast, MR-EVA was better than EVA obtained using a pinhole occluder (PH-EVA; median PH-EVA - MR-EVA, -4 letters [-9, 0]), and had significantly less test-retest variability (P<0.001). Generally, the spherical equivalent of autorefraction was slightly more hyperopic (or less myopic) than the spherical equivalent of manual refraction (median difference, +0.25 diopters [0, +0.63]).
Conclusions:
Given the substantial time and effort required for training and certification of study protocol refractionists, and the similarity between AR-EVA and MR-EVA, further evaluation of autorefraction, but not pinhole occluder testing, as an alternative to the current clinical research gold standard of ETDRS protocol manual refraction in study participants with diabetic retinopathy is warranted.
