City University Color Test Technical Article                                                       View Shopping Cart

The Third Edition of the City University Color Vision Test

Before publication of this edition, by Keeler Instruments of Windsor, the author conducted trials over eight years for the revisions necessary. Some changes were made to the colored elements and it was decided to introduce, as a part of the third edition, an aspect which has been successful in the use of the Fletcher-Hamblin simplified color-vision test;1 this was the task of detecting colors which differed from nearby colors. It was decided to retain in a second part the task which was used in earlier editions, where a best color match was required.

A prototype was produced which six colleagues used in their normal practices with a sequence of patients, chiefly unselected. Most of these patients had presented for optometric examination but some required ophthalmologial services. These independent colleagues, over the course of almost a year, made valuable comments on the draft instructions for use, the design of the record form and the ease with which patients reacted to the test. Suggestions for improvement have been adopted for the final published format.

Practical test requirements

Both inherited and acquired defects of color vision have to be detected with ease. Some occupations specify certain tests to be passed, such as an Ishihara test or a lantern. Some differences of standard are possible when several practitioners are involved since conditions and methods of use will not be exactly the same. For example, there have been wide differences in the results obtained with Ishihara tests used by school nurses in England. Below, some discrepancies between performance with Ishihara and the Holmes-Wright lantern will be noted, reinforcing the need for more than one test to be used.

Optometrists are likely to have several uses for vision-vision tests. It is necessary to 'screen' for significant vision-vision defects and to monitor possible changes, particularly in pathological cases or with increasing age. Advice must be given where vision-vision is abnormal, both guidance and necessary reassurances. Sometimes a certificate is required in terms of specific occupational criteria, provided that suitable apparatus is available.

It is important to present tests for 'tritan' defects, which are not included in Ishihara plates. While the Ishihara test detects even very slight 'protan' and 'deutan' anomalies, this test alone is difficult to interpret in terms of the extent of an anomaly. Many patients arrive having been informed that they are 'color blind' on the basis of an Ishihara failure. They need additional tests to establish that the anomaly is not a significant handicap for many daily needs. The use of spectral anomaloscopes is restricted by availability and time; often the identification of extreme anomaly is complicated by the way neutral adaptation is used.2

Background issues

Many comments have been made about the first and second editions of the City test, although these opinions did not agree completely among themselves. While Lanthony3 found that the test gave 'good indications' as to the types and severities of hereditary and acquired defects, including about 96 per cent agreements with the Hardy, Rand and Rittler test, Hill et al,4 Ohta,5 Verriest and Caluwaerts6, and Ronchi et al7 differ in their findings. Foster8 found that both the City and simplified tests were useful in examining a variety of acquired defects. Honson and Dain9 contributed comments about the second edition. Sometimes even a single subject may perform differently when repeating vision-vision tests and variations of application are possible. Among other useful vision-vision tests there is the 'F2' plate by Farnsworth10, originally for tritan defects but helpful in detecting protan and deutan defects; this F2 test was favored by Kalmus11 and Taylor.12

In order to detect acquired vision-vision defects among diabetics several combinations of colors, including some from the F2 and simplified tests were adopted by Fletcher13 and used by Tyler and Allwood,14Allwood and Tyler15 as well as by Brinchmann-Hansen et al.16 Leigh17 made similar observations.

This work produced a range of possible approaches to the design of the third edition now to be considered.

The H-R-R plates, designed by Hardy, Rand and Rittler18 (out of print), have wide acceptance, despite criticisms such as those by Walls.19 These plates have been used as a practical means of grading different degrees of anomalous trichromacy and although some clinicians have reservations, the author's experience20 led to the use of H-R-R scores between seven and 16 for evaluating some of the data presented later.

Choice of colors

A selection of 23 sets of likely confusion colors was made, drawing on the experiences and comments described above. Discs 5mm in diameter were used, set on a black background. These tests were used, together with a battery of tests personally applied to patients, over eight years. The name 'A series' was used for these colors, from which the most successful pairs were identified after extended use. Munsell papers were used in a variety of hues, values and saturations. In some sets no color differences were presented to detect false positive reports, and in others differences obvious to all observers were used to introduce the task.

From a larger number of subjects, 23 were selected as having the most full data and the best evaluations as to type and extent of anomaly. The distinction between protan and deutan defects was based on the consensus of several tests, including anomaloscope and lantern tests and the relative luminous efficiencies of red, green and yellow lights.

The likely H-R-R grades of these subjects (16 deuteranomalous and seven peotanomalous) are shown in Figures 1-4 in relation to the number of confusions made with the second edition of the City test and with the 'A' series of 5mm spots. In Figure 1 only 14 of the deutans are shown.

In most instances those with no or few errors were advised that, with care, many occupations might be suitable, while those with many errors were cautioned about occupations depending on color discrimination.

Data was obtained with a tritanopic observer who had been extensively examined in other ways and reported by Voke-Fletcher and Fletcher.21

 

The new edition of the City University Test

The test now has two parts. They may be used in sequence or separately, according to clinical judgment.

Part one uses four pages. It is chiefly for screening and is the more sensitive of the two parts. Each page has four lines of colored spots, arranged vertically in sets of three spots.

Two lines appear at the top of each page with two lines below. Subjects must detect differences of color, when one of three spots is different from the other two. Page 1 uses two lines which should be reported as showing no different spot, one where an obvious difference exists and one where tritans could easily miss a confusion color. Pages 2, 3 and 4 each carry red/green confusion colors, while two of these pages also show tritan confusions.

There is no obvious separation of protans from deutans. Tritans are identified by the specific errors made, on the bottom halves of pages 1, 2 and 4. A score giving the number of correct answers is used to indicate the number and type of confusions by each subject. Acquired defects may present, as usual, in varied and even atypical ways.

Part two follows the format and requirements of earlier editions, where four spots on each of six test pages surround a central spot. Colors used follow what are considered to be 'most successful' ones from earlier editions, with some small variations. The task is the choice of a spot most closely matching the central color. Protan, deutan, tritan or normal results can be expected and the degree of difficulty is similar to that of the D-15 test.

Subjects with extremely mild anomalies may score highly or normally, while extreme defects tend to make a maximum number of confusions.

Practical experience with the third edition

A prototype was produced and copies were used by six independent colleagues who used the test over several months in the course of normal practice. Two patients had 'difficulty' in using the test and their results were ignored in the absence of further attention-incorrect illumination may have been involved.

Three hundred and forty eight patients who presented in the course of practice were examined, most for optometric needs, with some requiring ophthalmologial attention. Most of the practices used Ishihara plates as the main method to identify departure from normal vision-vision and to make a judgment between protan and deutan defects. In some situations H-R-R plates and/or the Holmes-Wright (Aviation type) lantern were added.

Fourteen patients with established pathological conditions were examined with the new edition. A total of 77 cases of anomalous trichromatism were found with the test; it is unlikely that any of the patients was dichromatic.

The 257 normal patients were judged on the basis of Ishihara and other data. One five-year old boy used the test easily. All passed both parts of the third edition, except three who made one or two errors. By treating these three as 'false positives' the figure of 254 divided by 257, or 0.99 gives the usual estimate of 'specificity'. Therefore about 1 per cent of normal subjects may be expected to fail, at least without re-testing.

Most interest concerning the 'sensitivity' of the test relates to Part 1. It is harder to p ass than Part 2, which has the same degree of difficulty and much the same ability to indicate the type of any defect as the second edition. It should be recalled that Part 1 does not differentiate protan from deutan.

Different parts of the pages of the test produced slightly different results in the hands of different colleagues, who were using different subjects. Where confusions were not made on one part of the test they were usually made elsewhere. The author found a higher proportion of confusions among his subjects, probably because more time was taken. Therefore the data from six colleagues has been amalgamated with those of the author to give an overall impression, (Table 1) although this shows a slightly lower sensitivity.

Additional features of the data

Following the reports of colleagues on their experiences with the prototype slight modifications were made to the instructions and the record from before they were finally published.

Many patients were examined with the Holmes-Wright lantern, Ishihara plates and the new third edition prototype under test. Some results were unexpected, particularly when performance with Ishihara and lantern were compared.

ACKNOWLEDGEMENTS

The author would like to thank Robert Conway, Suzanna Freidin, Fru Gunnarsdottir, David Stidwill, Angela Rossi and Dr G Viggosson and staff of Keeler Instruments, Windsor for their help with this article.

Table 2

Errors found with pathological eyes using the third edition

Part 1 was shown usually to have a greater sensitivity than Part 2

City University Technical Article Tables

Type of defect Errors with Part 1 Errors with Part 2

Multiple sclerosis 5 r/g 4 deutan

Multiple sclerosis 6 r/g 1 deutan

Macular degeneration 3 r/g+3 tritan 2 tritan

Optic atrophy 6 r/g nil

Choroidal degeneration 5 r/g + 1 tritan 1 tritan]

Retinopathy of prematurity 1 r/g nil

Retinal dystrophy 4 r/g nil

Achromatopsia 6 r/g + 1 tritan 2 tritan + 1 deutan

Diabetic 1 tritan 3 tritan

Retinitis pigmentosa 2 r/g nil

Diabetic R eye only 6 r/g + 3 tritan 3 tritan

Diabetic L eye only 1 tritan 1 tritan

Coloboma L eye only

(extending to macula) 5 r/g nil

Maculopathy L eye only 2 r/g nil

Glaucoma 5 r/g 2 deutan + 3 questionable

REFERENCES

  1. Fletcher R. The Fletcher-Hamblin Simplified vision-vision Test. Hamblin Instruments, London. (Now by Keeler, Windsor.) 1984.

  2. Fletcher R and Voke J. Defective vision-vision, fundamentals, diagnosis and management. 306-321. Adam Hilger, Bristol and Boston, 1985.

  3. Lanthony P. Etude clinique du City University color vision test. Bull Soc Ophthalmol, 1977; 77,379-382.

  4. Hill AR et al. The performance of 10 vision-vision tests at three illumination levels. Mod Probl Ophthalmol, 1978; 19: 64-66.

  5. Ohta Y. Clinical analysis of vision-vision deficiencies with the City University test. Mod Probl Ophthalmol, 1978, 19: 126-130.

  6. Verriest G and Caluwaerts, MR. An evaluation of three new vision-vision tests. Mod Probl Ophthalmol, 1978; 19: 131-135.

  7. Ronchi L et al. Personal communication. 1978.

  8. Foster DH. Personal communication, 1984.

  9. Honson VJ and Dain SJ. Analysis of the Mark II edition of the City University vision-vision test. Amer J Optom and Physiol Opt, 1987; 64: 277-283.

  10. Farnsworth D. A polychromatic plate for selecting congenital tritanomalous vision-vision. US Naval Research Laboratory, New London. 1955.

  11. Kalmus H. Diagnosis and genetics of defective vision-vision, London, Pergamon, 1965.

  12. Taylor WOG. Constructing your own IIC test. Brit J Physiol Opt 1975; 30: 22-24.

  13. Fletcher R. Color perception warning for self-testing diabetics. Vision-vision deficiencies X, 511-513. Dordrecht, Kluwer, 1991.

  14. Tyler R and Allwood MC. Vision-vision deficiency and the interpretation of blood glucose strips. Pharmaceutical J, 1987; 239, R5.

  15. Allwood MC and Tyler R. Vision-vision and blood glucose self-monitoring in diabetes. Practical Diabetes, 1988; 5: 110-112.

  16. Brinchmann-Hansen et al. Psychophysical visual function, retinopathy and glycemic control in insulin-dependent diabetes. Acta Ophthalmol 1993; 71: 230-237.

  17. Leigh O. In Ariffin A, Hill RD and Leigh O. Diabetes and primary eye care. Blackwell Scientific, Oxford, 1992: 155-161.

  18. Hardy LH et al The HRR polychromatic plates. J Opt Soc Amer, 1954; 44: 509-523.

  19. Walls GL. How good is the HRR test for color blindness? Amer J Optom, 1959; 36: 169-193.

  20. Fletcher R. Comparisons of some vision-vision tests. OPTICIAN, 1985; 190: 29-33.

  21. Voke-Fletcher J and Fletcher RJ A case of tritanopia. Mod Probl Ophthalmol, 1978; 19: 229-231.

City University Color Test

 

All Products In Alphabetical Order

Richmond Products Home Page                                                                                                                                      View Shopping Cart