ABSTRACT
PURPOSE: The visual acuity of visually impaired patients has been reported to improve after a refraction, despite pinhole test results that show a decline or no change in acuity. Our aim was to investigate whether the pinhole-induced reduction in retinal illuminance accounted for these unreliable predictions of best-corrected acuity. METHODS: Participants were 64 adult patients referred for low-vision rehabilitation. Neutral density filters reproduced the pinhole-induced luminance loss, allowing pinhole test and postrefraction acuities to be measured at essentially equivalent levels of retinal illuminance. The following data were collected in random order from each subject's better eye: (1) habitual visual acuity, (2) habitual visual acuity with filter, (3) habitual visual acuity with pinhole, (4) best-corrected/postrefraction visual acuity, (5) postrefraction visual acuity with filter. RESULTS: On average, the pinhole test under-estimated postrefraction visual acuity by six letters (95% confidence limits = +/- 20). The pinhole test underestimated postrefraction visual acuity with the filter by two letters (95% confidence limits = +/- 16). Among subjects whose acuity improved with the pinhole test (N = 24), 83% experienced better postrefraction visual acuity. Among subjects whose acuity declined or remained unchanged with the pinhole test (N = 40), 50% achieved better postrefraction visual acuity. CONCLUSIONS: The pinhole-induced luminance loss contributed to inadequate predictions of postrefraction visual acuity. Pinhole test results were enormously variable, underestimating and overestimating postrefraction visual acuity. The pinhole test was less reliable when improvements in postrefraction visual acuity were small. Visually impaired patients deserve periodic refractions, and the pinhole test result should not be used as a dichotomizer for clinical decisions regarding the need for a refraction.
