Grating visual acuity with Teller cards compared with Snellen visual acuity in literate patients
B. J. Kushner, N. J. Lucchese and G. V. Morton
University of Wisconsin Hospital and Clinic, Madison, USA.
OBJECTIVE: To determine the usefulness of Teller Acuity Cards for detecting
three levels of vision deficit, the cutoff for amblyopia (20/40 or poorer),
vision impairment (20/70), or legal blindness (20/200). DESIGN: We compared
grating visual acuity with the Teller cards with Snellen visual acuity (our
gold standard) in 69 literate patients with amblyopia or other cause of
vision loss in a prospective masked study. RESULTS: Teller card visual
acuity and distance Snellen visual acuity correlated significantly (r =
.508, P < .001); however, Teller card visual acuity explained only 26%
of the variation in distance Snellen visual acuity. Teller card visual
acuity had a low sensitivity for detecting vision deficit of 20/40 or
poorer (58%), vision deficit of 20/70 or poorer (39%), or legal blindness
(24%), but somewhat more accurately reflected near Snellen visual acuity
than distance visual Snellen acuity. Teller cards had a higher positive
predictive value--80% for 20/70 visual acuity and 43% for legal blindness,
as determined by near Snellen visual acuity. Specificity of Teller cards
was 88% for detecting visual acuity loss of 20/70 and 98% for legal
blindness. Negative predictive value of Teller cards for detecting visual
acuity loss of 20/70 was 50% and for legal blindness was 71%. CONCLUSIONS:
Teller Acuity Cards may underestimate the presence of amblyopia of all
types, legal blindness, and a specified level of vision impairment (20/70).
Even in the presence of normal visual acuity measurements with Teller
cards, significant visual loss as assessed by standard Snellen optotypes
may be anticipated in many patients.
