620 Rejwrts Investigative Ophthalmology

Transcription

1 Rejwrts Investigative Ophthalmology August D. E.: Retinal dystrophy in the rat a pigment epithelial disease, INVEST. OPHTHALMOL. :,. Color vision: blue deficiencies in? ANTHONY J. ADAMS,* RICHARD BAL- LIET, * AND MARK MCADAMS.* Recent publicized reports based on the use of the Farnsworth Panel D- test suggest that a large percentage of young have a deficiency of blue vision (tritan type). In our study, school (ages to ) were tested with both the Farnsworth Panel D- test, as well as the A.O. H-R-R plates. None of the failed either test for blue-yellow vision when traditional scoring instructions were observed. As in previous reports, we find that the make a number of minor errors which adults rarely make. These errors show marked age-related patterns, being more frequent in younger. However, further analysis of these errors revealed that the relative frequency with which particular error types were made on the D- test was significantly correlated with the existing perceived color difference data for the visually normal adult population. In addition, retesting significantly reduced all error types and reversing the test sequence demonstrated that most of the minor errors were made in the last half of the test regardless of the color vision task. The overall increase in the number of minor test errors seen with young seems unrelated to color defects. The modified scoring methods in conjunction with the characteristics of the Panel D- test design account for the high percentage of errors classified as errors of blue vision. Recently, there has been an increased interest in defining the color vision of young. This interest has been heightened by the trend toward the use of color coding in teaching mathematics and English in the early school years. Consequently, recent suggestions that a larger percentage of young have a deficiency of blue vision - have caused some concern among educators and health-care practitioners as well as with the public through news media. In a study by Adams and Harwood, none of the, grade school failed the "blue-yellow" screening plates in the A.O. H-R-R test or the Farnsworth F- plate for tritanomalous vision. In a recently released study of a representative sample (,) of the in the United States between the ages of and, only. per cent of the failed "blue-yellow" plates in the A.O. H-R-R test, and most of them also had accompanying mild deficiencies in the redgreen area. Table I. Sixteen ( males and female) of the (. per cent) failed the A.O. H-R-R screening test. Eleven of the screening test failures passed both the diagnostic plates and the Panel D- test. The who failed the screening and diagnostic plates also failed the Panel D- test. The A.O. H-R-R test classified them as follows: medium unclassified, strong protan, and strong deutans. The protan and deutan classifications were consistent with the~ D- classification of the same individuals. The D- failures amount to. per cent (/) of the males. These figures are consistent with the sex-linked recessive transmission of red-green color defects reported in a large number of studies of color blindness. On the other hand, minor red-green errors (between caps and ) clearly do not follow the same pattern ( males and females) «> Q -s: Farnswor Pass Fail A.O. H-R-R Pass Fail The Farnsworth Panel D- test is often used in evaluating color vision deficiencies; the test involves the serial placement of colored caps. Sassoon and Tolder and Sassoon report that between the ages of and years make a number of D- test errors which they interpret as indicating blue-vision deficiencies. Of the between and years, per cent were judged deficient in their blue vision. The authors showed a clear age-related reduction in errors. Over per cent of the three-year olds made errors which were considered typical of a mild blue-vision deficiency while only per cent of ten-year olds made such errors. Many practitioners find the D- test extremely difficult to administer to small and, even after careful instruction, a large percentage of three-year olds are unable to understand what is required to perform the test. The nature of the test and the reported decrease in prevalence of test errors with age suggest that nonvisual factors play an important role. It is our hypothesis that the reported - blue deficiencies are unrelated to the color vision of and are a consequence of the way in which the test was administered and scored. s. Color vision testing. The Farnsworth D- test and the A.O. H-R-R pseudoisochromatic plates Downloaded From: on //

2 Volume Number Reports NORMAL. la TRITAN (blue - yellow) I C Fig.. Figs., a and b show error configurations typical of normal color vision according to the D- test instructions. Fig., a represents an "errorless" performance. Fig., b shows minor errors typical of those found in young. Fig., c shows the tritan (blue-yellow) defect with cap displacements across the circle. The connecting lines indicate the cap arrangements made by a color vision defective; the orientation of the connecting lines suggests the diagnosis of the type defect. The D- is intended to test color saturation, the confusion of caps across the circle being designed to identify the various color vision anomalies. Hue differences are ordered around the color circle (i.e., No. through No. ) and minor errors of ordering, while possibly demonstrating poor color discrimination, have not been related to specific types of color vision defects. were administered to school between the ages of and. Both tests were given under a Macbeth easel lamp (artificial daylight C source, K.). Both tests were presented essentially as recommended by the manufacturers except that no time limits were imposed on the. Two hundred and thirty-five were given the Farnsworth D- test with the manufacturer's fixed reference cap ( A). We theorized that many lose interest toward the end of the Panel D-, and thus errors are more likely to occur on this part of the test. Such error localization can only be checked by counterbalancing. Therefore, were tested with the end color cap as the fixed starting cap, requiring the reversed serial arrangement of caps ( B). Scoring procedures. Both tests were scored ac- Downloaded From: on //

3 Reports Investigative Ophthalmology August Table II. Farnsworth D- test results for between ages and. The table indicates the total number of at each age who were categorized as BPD (see scoring procedures) when the test was administered in the conventional ( A forward) and a reversed sequence ( B). The results of Sassoon's study are included for comparison (numbers derived from reported percentages) A (Traditional) B (Reversed) ( A and B) Sassoon Age BPD BPD BPD Blue error cording to the manufacturer's recommendation (traditional scoring, Fig. ). In order to compare our results to previous studies using the same test, we adopted an additional modified scoring procedure proposed by Sassoon. s The tritan (blueyellow) errors were defined as follows : "the tritan axis was deemed to cover single-place reversals up to caps - and from caps - through -, and also three-place mismatches up to caps - (but not -) and from caps - (but not -) through -." All other mismatches up to threeplace were operationally defined as red-green errors. Further, we followed Sassoon's definition of blue perception decrement (BPD) as at least four single place reversals (on the tritan axis) and/ or at least one two-place error on the tritan axis. Color difference (&E) survey of D-. We hypothesized that our results could have been predicted from the relative difficulty experienced by individuals with normal color vision in discriminating the D- test caps, and that it was unnecessary to presume a selective lowering of blue vision discrimination in. To test this, we checked to see if color difference formulas, based on normal color vision, would predict the rank order for all errors made by the. From the many current color difference formulas we chose the CIE and the cube-root formulas because they are widely used and are intended to predict the medium-sized color differences of the Munsell chips used in the Farnsworth Panel D- test. Results. The A.O. H-R-R and Farnsworth D- tests traditional scoring. Of a total of who were examined with the A.O. H-R-R test, were also examined with the D- test ( were unable to respond to the D- test). Table I shows that the prevalence of red-green color vision defects is consistent with the known incidence of hereditary color vision defects. No blue-yellow failures were found in either test. Modified D- scoring. Table II illustrates that the results using each of our methods (A and B) were very similar to those of Sassoon; in both studies the error frequency clearly decreases with age (Fig. ). The rank order correlations with his results were: A, r =., p <.; B, r =., p <.; s A and B combined, r =., p <.. s A and B differ only in the sequence in which colored caps are placed in order. When all single place ( per cent of all errors) and two-place errors ( per cent of all errors) are counted, an interesting result emerges in the comparison of s A and B. In A, the conventional method, most of the errors ( per cent) were made between caps numbers and and all such errors, by definition, fall on the tritan axis of Sassoon. s When the test order is reversed, in B, only per cent of all the errors are made., between cap Nos. and. Again, the last caps to be placed in order account for most of the errors ( per cent). Apparently nonvisual factors, such as lack of concentration and fatigue, are playing a major role in the production and localization of test errors. Retesting. All who were classified as BPD using B were retested. This amounted to per cent ( of ) of the. After retesting, none of the could be classified as BPD. This lack of reliability in classification after retesting indicates that the child's initial performance is at best poorly related Downloaded From: on //

4 Volume Number Reports Table III. Rank order of D- test errors according to increasing difficulty predicted by CIE color differences ( AE) between caps for color normal adults; the smaller AE between caps, the greater the predicted difficulty in color discrimination for those caps. Included are the total number of test errors made by aged between and years for all testing (s A and B) as well as for A and B separately. The who failed the D- and the A.O. H-R-R tests for conventional scoring procedures were excluded from this table Rank according to magnitude of CIE color difference (smaller to larger) Rank Cap error R R - All testing (s A & B) No. of errors A B Table III. Cont'd Rank according to magnitude of CIE color difference (smaller to larger) Rank Cap error R R All testing (s A & B) No. of errors A B to his color perception and that the modified scoring procedure is not a valid estimate of color discrimination. Color difference (&E) analysis of D- test results. Table III shows the total number of D- test errors produced by the as well as a ranking of error types according to the degree of difficulty predicted by CIE color differences between caps for color normal adults. Almost all of the errors ( per cent) were in the first most difficult discriminations for normal subjects. A Spearman rank correlation of r s =., between the frequency of error types for both methods combined and the color difference between caps, is significant at the p <. level (t =., N = ). The same significance is attained when cube-root rather than CIE calculations are used (r s =.; p <.). Thus the relative frequency of D- error types produced by the in this study were predicted by the relative difficulty of the test cap combinations for the color normal adult. Discussion. In this study of - to -year olds neither the A.O. H-R-R nor the Farnsworth D- test produced blue-yellow vision failures for any of the when the tests were scored by the traditional methods. The result is consistent with the extremely low prevalence of blueyellow vision deficiencies reported previously" for. Our adoption of the modified scoring techniques of Sassoon for the D- test revealed results similar to those reported by Sassoon and Tolder and Sassoon. However, the errors produced by the in our study can be ac- Downloaded From: on //

5 Reports Investigative Ophthalmology August AGE [YEARS] Fig.. Percentage of classified BPD on the Farnworth D- for school ages to. The crosses ( + ) indicate the results following the conventional method of presentation ( A) on. The triangles (A) indicate the combined results of A and a reverse-sequence method ( B) on. The filled circles ( ) show the percentage of classified as having "blue errors" in Sassoon's study (N = ). counted for by factors other than blue vision deficiencies. The D- test involves the serial placement of caps. If a young child has only partial understanding of the concept of serialization, he will make more errors than an adult. Our results and those of Sassoon indicate that the total number of errors is clearly a function of age (Fig. ). This is consistent with Piaget's suggestion that during the ages to, progress through developmental stages which allow them to perform serialization tasks with lesser difficulty. Nonvisual factors were demonstrated when the test sequence was reversed. Regardless of test sequence, forward or reversed, approximately per cent of all the errors were made in the last half of the test. We observed that tended to lose interest and concentration toward the end of the test and as a result made careless cap placements. The preponderance of errors in the last half of the test seems to reflect this behavior. There are almost three times as many D- test one- and two-place blue error possibilities () than for the red-green () providing a considerable blue loading on the test; we found that per cent of all the test errors were single or two-place reversals. Of far greater significance is the blue loading that results from the test design itself. For the normal observer, each cap is not separated by equally discriminable steps from its neighbor. Indeed, we have shown that our results can be predicted by the color difference formulas derived from the normal color vision functions of adults. In short, the relative frequency of different error types is highly correlated with the relative difficulty that normal subjects have in the hue discrimination of adjacent caps. The hypothesis that have a specific blue vision deficiency, when compared to adults, cannot be supported by our results. However, adults typically make fewer total errors on the D- test than we found for the in this study. This could be due to a general reduction across the spectrum in color aptitude or color acuity in. Although this is a possible explanation for the child's performance, it is unlikely since labeled BPD who were retested invariably lost this label after retesting, suggesting that test aptitude played an important role. There is no evidence in our study to support the hypothesis that have a selective deficiency of blue vision. Previous reports - of blue vision deficiencies in have been based on test methods and scores similar to ours and are subject to the same nonvisual factors. We have shown that these nonvisual factors have lead to the unwarranted conclusion that have blue vision deficiencies. From the ^School of Optometry, University of California, Berkeley, Calif., and the Downloaded From: on //

6 Volume Number Reports "^Smith-Kettlewell Institute of Visual Science and the Department of Visual Sciences, University of Pacific, Webster St., San Francisco, Calif.. Submitted for publication Jan.,. Key words: color vision, 's vision, color vision tests, blue-yellow deficiencies, tritan defects, color difference formulae, Farnsworth Panel D- test, A. O. H-R-R test. REFERENCES. Sassoon, H. F., and Tolder, M.: Blue vision and learning difficulties in, Fed. Proc. :,. Abstr.. Sassoon, H. F.: Blue vision in, Clin. Pediatr. :,.. Editorial: Chasing the blues away, Nation's Schools :,.. Editorial: "Blue-blindness" is more common than many think, JAMA :,.. Adams, A. J., and Harwood, L.: Color vision screening: a comparison of the AO-HRR and the Farnsworth F- pseudoisochromatic plate tests, in preparation.. National Center for Health Statistics, Color vision deficiencies in United States. Vital and Health Statistics, series, No., Health, Education, and Welfare Publication No. (HSM)-, Public Health Service, Washington, D. C, U. S. Government Printing Office.. Farnsworth, D.: The Farnsworth dichotomous test for color vision Panel D-, New York,, The Psychological Corporation, pp. -.. Sassoon, H. F.: s of administering and interpreting four tests for screening students at elementary and secondary schools. Private distribution, personal communication,.. Freile, L. F. C: A survey of some current color difference formulae, Colorimetrics, Proceedings of Helmholtz Memorial Symposium on Colorimetrics.,.. Flavell, J. H.: The Developmental Psychology of Jean Piaget. New York,, D. Van Norstrand Co., Inc., p.. Calculation of the optical power of intraocular lenses. S. N. FYODOROV,* M. A. GALIN,* AND A. LINKSZ.** A rather simple clinical approach has been used to derive formulas necessary to calculate the power of pupillary intracameral prosthetics and these have been applied in eyes. In eyes, the postoperative measurements were within one diopter of preoperative calculations. The replacement of the human, cataractous lens at the time of cataract extraction with a safe and dioptrically correct intraocular prosthetic device Table I. Age distribution of cataract patients. Note high incidence of younger patients due to traumatic cataracts Age (years) < > No. of patients has long been a desirable event. However, a variety of immediate and long-term complications associated with certain types of intracameral lenses led to the use of implants only in those institutions where the chemistry, physics, and mechanics of these lenses were well understood. - Extensive and successful clinical use of intraocular prostheses in such centers continued on the continent and stimulated both basic and clinical investigations in this country. " The data that has resulted from these studies clearly confirm the published results of European investigators and indicate that intraocular implants are excellent adjuncts in the rehabilitation of certain cataract patients. From onward, under the auspices of the Soviet-American Health Exchange, extensive clinical testing has been carried out to analyze and refine the technique of measurement and the mathematics employed in calculating the necessary power of an intraocular prosthetic lens of Soviet manufacture to permit lens extraction and simultaneous insertion of such a pupillary implant. These studies have involved solely the Fyodorov lens, and the present report is an analysis of clinical results in eyes in which such calculations have been carried out. Mathematical considerations. The calculation is based on measurements of the axial length of the eye, the refracting power of the cornea, and an estimate of anterior chamber depth with the implant in place. The following formula expresses the relationship between these constituents: Dp = n-adc (a-k)(l-kdc) () in which "a" represents the axial length (in meters); "k" anterior chamber depth with the pupillary implant in place (in meters); "D c " the refracting power of cornea (in diopters); "D P " the refracting power of the intraocular lens (in diopters and assuming a thin lens); and "n" the refractive index of aqueous and vitreous (.). This basic formula may appear formidable, but Downloaded From: on //