Year : 2003 | Volume
: 51 | Issue : 4 | Page : 341--345
Comparative evaluation of teller and cardiff acuity tests in normals and unilateral amblyopes in under-two-year-olds
P Sharma, D Bairagi, Murli M Sachdeva, K Kaur, S Khokhar, R Saxena
Dr. Rajendra Prasad Centre for Ophthalmic Sciences, AIIMS, Ansari Nagar, New Delhi-110 029, India
Dr. Rajendra Prasad Centre for Ophthalmic Sciences, AIIMS, Ansari Nagar, New Delhi-110 029
Purpose: To compare and evaluate Teller Acuity Cards (TAC) and Cardiff Acuity Cards (CAC) to assess vision in children below the age of two.
Methods: The study evaluated TAC and CAC to assess visual acuity in 90 normal children divided into three age groups, 0 - 6 months (group I), 6 - 12 months (group II) and 12 - 24 months (group III). 30 cases of unilateral amblyopiogenic conditions, 10 cases each of unilateral refractive error, unilateral esotropia, and unilateral cataract, were also examined. Trained optometrists carried out binocular testing followed by monocular testing, and recorded the test time in each case.
Results: The mean visual acuity (in Snellen units) and standard deviation (in octaves) in the three age groups of normal children, I, II, III respectively were 6/44 ± 0.54, 6/21 ± 0.37 and 6/21 ± 0.41 (binocularly by TAC) and 6/46 ± 0.80, 6/21 ± 0.59 and 6/14.5 ± 0.84 (binocularly by CAC). Although the time taken for testing with CAC was less, its coefficient of variance was greater for all age groups as compared to TAC. Diminution of visual acuity could be assessed correctly by both the tests in cases of strabismus and cataract, but not in some cases of refractive error.
Conclusion: CAC is a useful and child-friendly test. It can be used clinically but may miss some cases of visually significant refractive errors. TAC is a more dependable test to assess amblyopiogenic conditions despite the use of gratings.
|How to cite this article:|
Sharma P, Bairagi D, Sachdeva MM, Kaur K, Khokhar S, Saxena R. Comparative evaluation of teller and cardiff acuity tests in normals and unilateral amblyopes in under-two-year-olds.Indian J Ophthalmol 2003;51:341-345
|How to cite this URL:|
Sharma P, Bairagi D, Sachdeva MM, Kaur K, Khokhar S, Saxena R. Comparative evaluation of teller and cardiff acuity tests in normals and unilateral amblyopes in under-two-year-olds. Indian J Ophthalmol [serial online] 2003 [cited 2023 Feb 5 ];51:341-345
Available from: https://www.ijo.in/text.asp?2003/51/4/341/14650
How much an infant or a toddler can see is a question that has intrigued parents as well as ophthalmologists for decades. Preferential looking techniques such as Teller Acuity Cards (TAC) have been found both convenient and reliable. ,,, However, these grating acuity tests have significant limitations. Subsequently, another card-based acuity test procedure (Cardiff Acuity Cards, CAC) has been recently introduced to measure visual acuity in toddlers. This is based on the "vanishing" optotype principle and has the advantage of bridging the gap between a resolution and a recognition task. ,,, Thus there is a need to compare Teller and Cardiff Acuity Cards. This study is designed to compare and evaluate the two tests in normal children and children with amblyopiogenic conditions below the age of two years.
Materials and Methods
Ninety normal children and 30 children with amblyopiogenic conditions were recruited for the study. The normal children were divided into three groups according to age: group I: up to 6 months, group II: 6-12 months and group III: 12-24 months. All these children had been born full term, i.e., their birth date coincided with their due date or was ± 2 weeks from the due date in which case the age was corrected. The birth weight ranged from 2.5 to 4.5 kg. Children with known neurological or developmental anomalies, birth trauma, family history of visual deficiencies or ocular disorders were excluded from the study. Children with significant refractive errors [(hyperopia of greater than 5 Dsphere (DS), myopia of greater than 3 DS. and astigmatism greater than 3 D cylinder (C)] were not included in the normal category.
Thirty children with amblyopiogenic conditions (causing diminution of vision and likely to cause amblyopia) included 10 cases each of unilateral esotropia, unilateral significant refractive error [(hyperopia above 5 DS, myopia above 3 DS and astigmatism above 3 DC and unilateral cataract (>3 mm in extent and dense enough to obstruct the red glow on distant direct ophthalmoscopy)].
All the subjects underwent a complete ophthalmic examination. A cover test for binocular alignment, a test for ocular motility, pupillary reactions, an examination of the anterior segment by flashlight, and an evaluation of media clarity and posterior segment by direct ophthalmoscopy were performed. Cycloplegic refraction was performed in all the cases to evaluate refractive errors. Each subject's vision was tested with Teller Acuity Cards and Cardiff Acuity Cards, first binocularly, and then monocularly for each test procedure. The order of the two tests was randomised. Trained optometrists, unaware of the clinical condition of the child carried out the testing. Only one test was performed at a time with at least a day's interval between the two tests. An adequate time interval was provided between binocular and monocular testing in each test to overcome fatigue and boredom.
The methods of testing were as per the manuals provided by the manufacturers (Vistech Consultant Inc.,US for TAC and Keeler, UK for CAC). Some salient features are mentioned here.
Teller Acuity Cards
The TACs were held in front of the gray (space average luminance matched with the cards) screen at the specified distance i.e., 38 cm for 0-6 months and 55 cm for older children. A 0.64 cy/cm card was used for group I, 1.3 cy/cm card for group II, and 2.4 cy/cm card for group III. The grating was increased or decreased stepwise depending on the child's response. The tester's task was to judge the child's visual acuity based on his behavioural reactions, i.e., eye movements, head movements, or pointing gestures. The most important guide was the initial eye movement. The cards were presented thrice. The left-right position was reversed by the presenter who remained unaware of the correct location of the grating. The subject's responses - consistent/inconsistent look or "refused to look" - were noted. The tester confirmed the location of the grating on the card only after making a definitive judgement of the child's response behavior. The highest spatial grating that generated a consistent look was judged as the visual acuity of the subject.
The child was subjected to binocular testing for each eye while the other eye remained patched. This was followed by a monocular testing. The total time taken for each test was recorded.
Cardiff Acuity Cards
CACs were used to test visual acuity in the child (held in the parent's lap) at a distance of 50 cm (0-12 months) and 100 cm ( 12-24 months). The three cards for each of the ten acuity levels were shown starting with 6/120 (the lowest acuity level) for 0-6 months and 6/60 for 6 months and older. The assessment pattern was similar to the one used for Teller Acuity Cards. If the child made one wrong assessment, the previous (lesser acuity level) was checked. The highest level at which at least two out of three cards scored correctly, was taken as the visual acuity of the subject. Monocular testing followed binocular testing and the test time of each was recorded.
Acuity obtained with Teller Acuity Cards (grating acuity) is expressed in cycles/cm or cycles/degree. Acuity obtained with Cardiff Acuity Cards is expressed in Snellen fractions. While clinicians are more familiar with Snellen notations, visual acuity also needs to be expressed in logarithmic values. (logarithm of minimum angle of resolution (log MAR)) for data analysis. This is preferred over the use of Snellen notations for arriving at the mean and standard deviation of visual acuity. Hence, in this study, the acuity score for TAC first converted to its Snellen equivalent subsequently, for both the tests the logarithm of base 10, relative to 20/20 or 6/6 was calculated. The mean, median, and standard deviations of the logs were calculated thereafter, and finally the anti-log of the end result gave the Snellen equivalents of the mean or median. The standard deviation in log units (base 10) was divided by 0.30 to yield the standard deviation in octaves (log 2 = log 10/0.301). The tables thus show the mean in Snellen fractions (derived as above) and the standard deviations in octaves (0.5 octave implies one line difference on ETDRS chart e.g.6/9, 6/7.5, 6/6 or 20/30, 20/25, 20/20). Statistical analysis was carried out using students' t test.
Norms of visual acuity
To establish the standard norms for each age group, the mean acuity was plotted (in logarithmic scale), with the tolerance limits calculated as ± 2 S.D. from the mean.
Evaluation of tests
To evaluate the tests, the coefficient of variance was calculated for both the tests, in the three age groups:- Coefficient of variance = standard deviation X 100/mean.
The mean ages in the three groups; I (0 to 6months), II ( 6-12 months) and III ( 12-24 months) of normal children, were 3.6 months, 8.8 months, and 18.3 months respectively. The ages of the children with amblyopiogenic conditions ranged from 3-17 months (mean 11.3), 6-23 months (mean 12.6) and 7-23 months (mean 12.2) respectively in the three subgroups. There were 53 male and 37 female babies in normal group and 17 male and 13 female babies in the amblyopiogenic case groups.
In the bar diagram in [Figure 1] the binocular visual acuity has been plotted on logarithmic scale in the three groups of normal children. In [Figure 2], the bar diagram shows the monocular visual acuities from the two tests in the corresponding groups of normal children.
[Table 1][Table 2] give the values of mean (Snellen fractions) ± S.D. (in octaves), median and range of the 90 normal children by Teller and Cardiff tests respectively. It may be noted that with Cardiff test monocular testing was not possible and though binocular visual acuity testing was possible, even for infants less than 6 months, the variation was high (S.D. = 0.80). This was also true for group III, thus making the Cardiff test comparable to Teller only in the over 6 - 12 months (group II) children.
The coefficient of variance was 18.6%, 20% and 21.8% in groups I, II and III respectively by Teller tests, and it was 27.3% (only binocularly), 30% and 65% for the Cardiff test. This clearly indicates that Teller Acuity Cards have less variability in a group and hence are more reliable.
The corresponding binocular Snellen visual acuity was 6/44, 6/21and 6/21 on Teller and 6/46, 6/21 and 6/14.5 on Cardiff tests in the three groups respectively, showing progressive improvement. The differences between the two tests in groups I and II were not significant, but in group III the Cardiff test appeared to significantly over-estimate the vision (P=0.0022). The Teller tests are considered reliable (because of their coefficient of variance).
It may be noted that monocular visual acuity could not be obtained by Cardiff in group I. In group II the difference between the two tests was not significant but in group III it was significant (P=0.0007), as is also reflected in the coefficient of variance.
The mean time taken for testing was 13.27 minutes, 12.1 minutes, and 11.8 minutes in the groups I, II and III respectively by the Teller test. Test durations with Cardiff tests were 4.3 minutes, (only binocular testing done as monocular not possible), 9 minutes and 9.4 minutes respectively. The difference was statistically significant in the three groups (P<0.001).
Visual acuity estimation in cases with abnormal vision was studied in three groups of 10 cases each. All these cases had a definite pathology present in one eye that would result in subnormal vision. The three groups were of unilateral strabismus (esotropia), high unilateral refractive error (seven cases with compound hypermetropia of greater than +5 D; three cases with compound myopia of greater than 4 D myopia) and unilateral cataract. The age group of these patients ranged from 3 to 23 months with a mean of 11.3, 12.6, and 12.2 months respectively in the three groups.
The visual acuity obtained by TAC in the abnormal eye was plotted in the bar diagram [Figure 3]. A bar corresponding to the mean visual acuity in the 12 months age group is also plotted with a two standard deviation variation to indicate the limits of tolerance. It was noted that all the abnormal cases were distinct from the norm in all the three amblyopiogenic conditions (P=0.001, P=0.019, and P=0.0001 respectively for refractive error, strabismus, and unilateral cataract groups). The lowest visual acuity was obtained in the unilateral cataract group.
Similarly, [Figure 4] plotted the visual acuity of the abnormal eye in the three groups using the Cardiff test with the reference of the visual acuity of an age matched normal group. The cases with strabismus and unilateral cataract could be clearly distinguished (P = 0.03 and P = 0.00001). However, the cases of significant unilateral refractive error had that less than normal vision though they were still not statistically significant (P =0.07, NS).
Assessment of visual acuity in infants and preschool children has always been difficult. The Teller Acuity Cards (TAC), based on the preferential looking (behavioral) principle has simplified the task to some extent. However, they utilise gratings, which at best tell us the resolution acuity and perhaps do not truly represent recognition acuity. Due to this drawback it has been suggested that it may not be a good test to screen for amblyopia. Another recent test, Cardiff Acuity Cards (CAC) test utilises pictorial optotypes, which are of the same size but made on the principle of the optotypes "vanishing" beyond the resolution distance. This was also based on the two-alternative-forced-choice-preferential-looking (2-AFCPL) method just like the TAC. It has been suggested that they, being pictorial, may bridge the gap of resolution and recognition acuity and generate more interest among children, especially those more than a year old. This could be an advantage of Cardiff test over Teller Acuity Cards. It has been said that vanishing optotypes provide a more precise endpoint compared to gratings in 1-5 months old infants.
Our evaluation indicates that Cardiff test is a valid and useful test for assessing visual acuity above 6 months of age, though others have found the test more suitable for children above 12 months of age. However, it has a tendency to overestimate vision compared to Teller's test. Moreover, the variations and standard deviations are large, indicating less precision. This could be because children in this age group are not very interested in pictures. The lack of precision of the Cardiff test could also be due to lack of proper standardisation, such as lack of a screen, lack of a peep hole and use of cards with hands which are visible along with the head popping up to see the fixation pattern. Although these features undoubtedly make the testing environment more child - friendly, they also adversely affect the previous levels. Even the luminance criteria with CAC is not as stringent as with the Teller test which uses a luminance matched screen.
On the positive side, the test time with CAC was significantly lower. This was also found to be more enjoyable and easy to administer. A realistic evaluation could suggest the use of Cardiff cards in clinical setting as a starter or for screening, but for more dependable visual assessment one would have to switch to the Teller Acuity Cards. Geer and Westall compared the visual acuity in 4 to 15-year-old children by Bailey-Lovie, Teller, and Cardiff Cards and found that Bailey-Lovie was the most sensitive test in detecting amblyopia. But this was applicable only in older children and not in infants and toddlers as done in this study. In infants and toddlers it would not be feasible to perform the Bailey Lovie test. Further Geer and Westall cautioned possible overestimation of vision with both Teller and Cardiff tests as well as the likelihood of failing to detect mild to moderate degree of amblyopia.
In our study, apart from comparing the use of the two tests in normal subjects, we evaluated their relative efficacy in the "abnormal" group of cases with amblyopiogenic conditions such as unilateral cataract, unilateral strabismus, and unilateral or asymmetric uncorrected refractive error. A good test should be able to detect the abnormal from the normal. Fixing the norm with age-matched normals with tolerance limits of two standard deviations, the lower limit of normal visual acuity for the group was fixed. The Teller Acuity Cards could distinguish cases with amblyopia with any of the above-mentioned three conditions from the normal. This indicates that grating tests are not so unreliable for amblyopia screening. However, as rightly observed by Kushner et al, since the Teller cards correlate well with near visual acuity (but not with distance visual acuity), the drawback of its testing distance has to be accepted. It may also miss low-grade amblyopias. But the Cardiff Acuity cards could reliably distinguish only the unilateral cataract (with marked diminution of vision) and unilateral strabismus but not all cases of uncorrected refractive error, which have lesser diminution of vision. This casts doubts on its ability to be used as a screening test for visually significant refractive errors. This is evidently due to a larger variation in the assessment by Cardiff test. Unfortunately, screening of these cases by the visual acuity test is more crucial as unilateral cataract or strabismus can be detected even otherwise by a corneal reflex test, a simple cover test, or a retinal reflex test.
Several studies have shown the usefulness of Teller Acuity Cards for evaluating visual acuity in 0-14 months age group, and some studies of assessment in infantile esotropia.
Our study clearly shows that the Teller Acuity Cards are more reliable, precise and dependable for assessment of the visual acuity of infants and under-two year olds and are useful even for detecting amblyopiogenic conditions like cataract, strabismus, and to a lesser extent, refractive errors. Cardiff Acuity Cards are enjoyable and child-friendly and may be used in the clinical setting though they are not as precise as the Teller Acuity Cards, and may not be suitable to screen visually significant refractive errors.
|1||Preston KL, McDonald MA, Sebris SL, Dobson V, Teller DY. Validation of the acuity card procedure for assessment of infants with ocular disorders. Ophthalmology 1987;94:644-53.|
|2||Dobson V, Salem D, Mayer DL,Moss C, Sebris SL. Visual acuity screening of children 6 months to 3 years of age. Invest Ophthalmol Vis Sci 1985;26:1057-63.|
|3||Getz LM, Dobson V, Luna B, Mash C. Interobserver reliability of the Teller Acuity Card procedure in pediatric patients. Invest Ophthalmol Vis Sci 1996;37:180-87.|
|4||Salomao SR, Ventura DF. Large sample population age norms for visual acuities obtained with Vistech Teller Acuity Cards. Invest Ophthalmol Vis Sci 1995;36:657-70.|
|5||Kushner BJ, Lucchese NJ, Morton GV. Grating visual acuity with Teller Cards compared with Snellen visual acuity in literate patients. Arch Ophthalmol 1995;113:485-93.|
|6||Adoh TO, Woodhouse JM. The Cardiff Acuity Test used for measuring visual acuity development in toddlers. Vision Res 1994;14:1063-66.|
|7||Howland B., Ginsburg AP and Campbell FW. High-pass spatial frequency letters as clinical optotypes. Vision Res 1978:1063-66.|
|8||Frisen L. New type of acuity test letters. Arch.Ophthalmol .1986; 104: 1194-1198.|
|9||Harris SJ, Hansen RM and Fulton AB. Assessment of acuity in human infants using face and grafting stimuli. Invest Ophthalmol Vis Sci 1984;25:782-86.|
|10||Moseley MJ, Jones HS. Visual acuity: Calculating appropriate averages. Acta Ophthalmologica 1993;71:296-300.|
|11||Geer I, Westall CA. Comparison of two preferential looking tests in the detection of amblyopia. Am Orthopt J 1995;45:79-89.|