A C TA 0 P H T H A L M 0 L O G I C A
68 (1990) 145-150
A new glare test based on low contrast letters Evaluation in cataract patients Anna-Lena HArd, Maths Abrahamsson and Johan Sjostrand Department of Ophthalmology (Head:J. Sjostrand),University of Goteborg,Sweden
Abstract. A new simple glare test was designed and evaluated regarding clinical usefulness and reproducibility. The ability to recognize letters of equal size and varying contrast was determined with the absence and presence of glare sources above and below the letters. Ten patients with cataract, visual acuity of at least 0.3 and glare problems, and three age matched controls were tested, as well as one patient with glare complaints and exophoria, one with lens subluxation and one with cataract and no glare problem. The test was found to be cheap and simple to produce and useful for clinical testing. Normal eyes had no detectable reduction of letter contrast sensibility with glare. All cataractous eyes had a letter contrast sensivitity without glare that was well below that of the controls and under glare conditions they all had a drop in visual function that was unrelated to their visual acuity. The reproducibility was of a magnitude similar to that of other low contrast letter tests. Our conclusion is that this test will be a valuable tool in the evaluation of cataract patients providing information not only about glare-induced visual loss but also about contrast sensitivity, separating eyes with increased intraocular light scattering from normal eyes. Key words: glare -cataract - contrast sensitivity - low contrast letters.
The need for glare testing as a complement to visual acuity testing in the evaluation of cataract patients has lately been pointed out by several authors (Neumann et al. 1988;Jaf€e 1986; Koch 1988).The decision to operate on patients with dense cataracts and very low visual acuity is not difficult. There is, however, a group of patients who, in spite of relatively good visual acuity in the examining room, have severe visual complaints. Opacities of the optic 10 Acta Ophthal. 68.2
media scatter the incoming light and a so-called ‘veilingluminance’ is projected on the retina (Miller & Benedek 1973).Thereby the contrast of the retinal image is decreased. The patients complain about severe visual impairment on sunny days or when meeting car headlights at night. There are now a number of glare tests on the market. None has attained general acceptance and no standards have been set concerning glare source or type of visual function tested. Broad fields of light as well as single or multiple light sources of varying extension and position relative to the visual target are used to produce glare. Visual functions tested are either contrast sensitivity or visual acuity. Most of the commercially available tests are based on complicated electronic devices and are thus expensive. Our intention was to construct a cheap, simple, clinically useful test to measure glare disability in the evaluation of cataract patients. Letter identification is a task familiar to most patients and a correctly identified letter is unlikely to be a mere guess. Because of the contrast lowering effect of the scattered light, we decided to test the ability to recognize letters of low contrast. The aim of this study was to design a glare test based on low contrast letters and to evaluate its clinical usefulness and reprolducibility. A preliminary description of this glare test has been previously reported (Sjostrand et al. 1987). We also wanted to study the relation between visual acuity and glare-induced visual loss of contrast sensitivity. 145
Material and Methods Patients Patients in the study had been referred to the Eye Clinic, Sahlgren’s Hospital for evaluation and possible cataract surgery. Visual acuity (VA) of at least 0.3 in each eye and no sign of eye disease other than cataract were used as inclusion criteria for the study, together with positive answer to the question ‘Do you experience reduced vision under bright light conditions?’Ten patients (age 44 to 75) fulfilling the above criteria were tested. Two eyes were excluded. One because it had a clear lens and one because VA had deteriorated to 0.2 since referral. We also tested one cataract patient who did not experience reduced vision from glare and one patient with exophoria and severe glare complaints, as well as one patient with Marfan’s syndrome and lens subluxation. Three controls (age 63 to 69) with clear media and without evidence of eye disease were also tested. Refraction was determined for each subject and all wore spectacle lenses and had a natural pupil during the testing. All subjects were tested and examined by the same examiner (A-LH).
Apparatus A low contrast letter chart (Fig. 1) was constructed using an Apple Macintosh personal computer and a Linotronic 2000TMprinting device. Letters of equal size were made up of small dots and contrast was varied by changes in dot size and interdot distance. All dots were too small to be resolved by the eye at a distance of 4 m. This is a simple and cheap method to produce letters of selected sizes and contrast levels (0.44 0.3, 0.21, 0.18, 0.14, 0.11, 0.09, 0.064, 0.055 and 0.03). The two lowest contrasts were put on a card that could be inserted between two glare sources when needed. Eight letters (E, F, K, N, U, V, X & Y) of equal legibility (Hedin & Olsson 1984) were used. The optotype size was equivalent to an average stroke width of 3.3 min of arc at a viewing distance of 4 m which corresponds to a visual acuity of 0.3. There were eight letters of each contrast positioned in two rows of four letters with an interletter distance of one letter width. Fluorescent tubes (Philips PL 9W/ 82) covered with diffusing glass were used as horizontal glare sources above and below the letters. Letters and glare sources were put in a light box (LIC, Sweden).The luminance of the chart was 450
146
Fig 1. Low contrast letter chart.
cd/m2 and the luminance of the glare sources was 9000 cdlm‘. The visual acuity (VA) chart was mounted in a similar LIC-box. It exhibited the same letters (E, F, K, N, U, V, X & Y) as the low contrast letter chart. The acuity-chart had logarithmic steps, and ten letters in each row for acuities between 0.3 and 2.0. Procedure Visual acuity (VA) and letter contrast sensitivity without glare (LCS) and with glare (LCS,;) were tested in a similar way, at 4 m distance. The right eye was tested first then the left and finally both together.
VA measurements were started with determination of the smallest letter size which could be identified without errors. If the subject was unable to read all letters of any size, the level where the maximal number of letters could be read was identified instead. From this level the subject was asked to read subsequent letters of smaller size down to the row in which no letters could be read. All correctly identified letters were noted and the number of letters read on each row was plotted as a frequency of seeing curve. LCS was tested in a similar way. Instead of smallest letter size, the lowest contrast level in which all letters could be correctly identified was determined, and the subject was asked to read subsequent letters of lower contrast down to the row where no letters at all could be read. For LCS,,determination one glare source was lit above and one below the low contrast letters being read. Subjects were instructed to look between the glaresources and avoid direct fixation. There was no time limit. Subjects were allowed to look at the letters as long as they wanted. A slit-lamp examination followed the testing. Intraocular pressure was measured by applanation tonometry and the pupils were dilated with Tropicamide (Mydriacylm0.5%)or Cyclopentolate (Cyclogyl' l0/o). The central fundus was inspected through a 60 diopters lens. Cataracts were graded at the slit-lamp regarding nuclear features according to 'The Oxford Clinical Cataract Classification and Grading System' (Brown et al. 1987).The presence of posterior subcapsular cataract (PSC) was noted. Retroillumination photography with the 'Oxford Retro-lllumination Cataract Camera' (Holofax Limited, England) was used to assess the amount of cortical opacities in all subjects. Cataracts were grossly classified as nuclear (N), posterior subcapsular (PSC) and cortical (except PSC) (C).Seven of the patients and two of the controls were retested within a month after the initial testing. Evaluation LCS and glare-score When the lowest contrast level where the subject could identify all letters was determined, an assumption was made that all letters of higher contrast could be seen as well. Results were plotted as a frequency of seeing curve (Fig. 2a) where ni is the number of correctly read letters of the contrast ci. The area SLrsunder ,,:I:
the curve was calculated according to the following formula.
+
S,,, = n,x log (l/c1)+n, [ l ~ g ( l / ~ , ) - l ~ g ( l / ~ J ] n3 [log(l/c,)-log(l/cp)].... LCS is the value of log (l/c) that multiplied by 8 (the maximum value of n) is equal to S,*(\. Hence LCS
SICS = __
8 LCS, is the LCS determined with glare sources lit. Glare score is defined as LCS-LCSC7. Thus higher glare scores indicate more severe visual loss. Evaluation-VA The results of the VA-testingwere treated in a way similar to that of LCS results (Fig. 2b). The area S, under the frequency of seeing curve was calculated as follows.
S, = n , (log VA, (log VA,+l)]. ...
+ 1) + n2 [(log VA,+l)
-
The (log VA+l) value that multiplied by 10 (the maximum value of ni)equals S, yields the patients
sv-
estimated VA log VA
+ 1 =S,
VA = 10
1"
10
10
Results It was possible to obtain an LCS and an LCS,, value for all eyes in the study except two. These two eyes both had posterior subcapsular cataract and a low LCS. With glare these eyes saw none of the letters at the 0.44 (highest)contrast level. VA of these two eyes was 0.4 and 0.5, respectively. All control eyes saw most of the letters of the lowest contrast (0.03) both with and without glare (glarescore=O). All cataractous eyes had a LCS that was lower than that of the controls and it decreased substantially when the glare source was turned on. This drop in visual performance was unrelated to VA (Fig. 3). Range of glare scores in patients with glare complaints was for the worst eye 0.13-0.47, for the best eye 0.04-0.29 and binocularly 0.02-0.44. The cataract patient without glare problems had a glarescore of 0.15 for the right and 0.10 for the left eye. He had a combination of nuclear and cortical cataract, none of which was more predominant. 147
L
log (1 /Ci)
0.36
0.52
0.74
0.68
0.85
0.96
1.05
1 + log v A. 0.48
0.7
VA
=
10
+
+
+
VA
=
+
+
+
+
1.11
0.9
Fig. 2. contrast, q = number of correctly read letters. (0.68-0.52) X 8 (0.74-0.68) X 7 (0.85-0.74) X 6
Example of LCS calculation. a) ci = SLcs = 0.36 X 8 (0.52-0.36) X 8 LCS = 6.63/8 = 0.83. b) Example of VA calculation. ni = number of correctly read letters. S, = 0.48 X 10 (0.6-0.48) X 10 (0.7-0.6) X 6 (0.81-0.7) X 6 (0.9-0.81) 7.82-10
+
1.0
0.81
0.6
X
4
+ (0.96-0.85) X 1 = 6.63
+ (1.0-0.9) X 2 = 7.82.
0.61
Discussion
The exophoria patient with severe glare complaints had normal LCS and a glare score of 0 We have presented a new method to measure glare disability in cataract patients. The test is simple monocularly as well as binocularly. The patient with superiorly subluxated lens and cheap to produce and does not include comcould not be tested at 4 m distance because her vi- plicated electronic devices. The test task of identsion was too poor. While seated two meters from ifying letters was found to be easy to understand the chart she could read all letters of the highest and the test procedure rapid to perform in most contrast without glare. When both glare sources as cases. In agreement,Pelli and coworkers (Pelliet al. 1988)have concluded that for clinical contrast senwell as the one above the letters were on she could read none of the letters. When only the lower glare sitivity testing letters are more suitable than gratsource was on she could read the letters of 44% ings which we previously used for glare testing (Paulson & Sjostrand 1980; Abrahamsson & Sjoscontrast 'as easy as without glare'. Testhetest variations for VA, LCS and LCS
68 (1990) 145-150
A new glare test based on low contrast letters Evaluation in cataract patients Anna-Lena HArd, Maths Abrahamsson and Johan Sjostrand Department of Ophthalmology (Head:J. Sjostrand),University of Goteborg,Sweden
Abstract. A new simple glare test was designed and evaluated regarding clinical usefulness and reproducibility. The ability to recognize letters of equal size and varying contrast was determined with the absence and presence of glare sources above and below the letters. Ten patients with cataract, visual acuity of at least 0.3 and glare problems, and three age matched controls were tested, as well as one patient with glare complaints and exophoria, one with lens subluxation and one with cataract and no glare problem. The test was found to be cheap and simple to produce and useful for clinical testing. Normal eyes had no detectable reduction of letter contrast sensibility with glare. All cataractous eyes had a letter contrast sensivitity without glare that was well below that of the controls and under glare conditions they all had a drop in visual function that was unrelated to their visual acuity. The reproducibility was of a magnitude similar to that of other low contrast letter tests. Our conclusion is that this test will be a valuable tool in the evaluation of cataract patients providing information not only about glare-induced visual loss but also about contrast sensitivity, separating eyes with increased intraocular light scattering from normal eyes. Key words: glare -cataract - contrast sensitivity - low contrast letters.
The need for glare testing as a complement to visual acuity testing in the evaluation of cataract patients has lately been pointed out by several authors (Neumann et al. 1988;Jaf€e 1986; Koch 1988).The decision to operate on patients with dense cataracts and very low visual acuity is not difficult. There is, however, a group of patients who, in spite of relatively good visual acuity in the examining room, have severe visual complaints. Opacities of the optic 10 Acta Ophthal. 68.2
media scatter the incoming light and a so-called ‘veilingluminance’ is projected on the retina (Miller & Benedek 1973).Thereby the contrast of the retinal image is decreased. The patients complain about severe visual impairment on sunny days or when meeting car headlights at night. There are now a number of glare tests on the market. None has attained general acceptance and no standards have been set concerning glare source or type of visual function tested. Broad fields of light as well as single or multiple light sources of varying extension and position relative to the visual target are used to produce glare. Visual functions tested are either contrast sensitivity or visual acuity. Most of the commercially available tests are based on complicated electronic devices and are thus expensive. Our intention was to construct a cheap, simple, clinically useful test to measure glare disability in the evaluation of cataract patients. Letter identification is a task familiar to most patients and a correctly identified letter is unlikely to be a mere guess. Because of the contrast lowering effect of the scattered light, we decided to test the ability to recognize letters of low contrast. The aim of this study was to design a glare test based on low contrast letters and to evaluate its clinical usefulness and reprolducibility. A preliminary description of this glare test has been previously reported (Sjostrand et al. 1987). We also wanted to study the relation between visual acuity and glare-induced visual loss of contrast sensitivity. 145
Material and Methods Patients Patients in the study had been referred to the Eye Clinic, Sahlgren’s Hospital for evaluation and possible cataract surgery. Visual acuity (VA) of at least 0.3 in each eye and no sign of eye disease other than cataract were used as inclusion criteria for the study, together with positive answer to the question ‘Do you experience reduced vision under bright light conditions?’Ten patients (age 44 to 75) fulfilling the above criteria were tested. Two eyes were excluded. One because it had a clear lens and one because VA had deteriorated to 0.2 since referral. We also tested one cataract patient who did not experience reduced vision from glare and one patient with exophoria and severe glare complaints, as well as one patient with Marfan’s syndrome and lens subluxation. Three controls (age 63 to 69) with clear media and without evidence of eye disease were also tested. Refraction was determined for each subject and all wore spectacle lenses and had a natural pupil during the testing. All subjects were tested and examined by the same examiner (A-LH).
Apparatus A low contrast letter chart (Fig. 1) was constructed using an Apple Macintosh personal computer and a Linotronic 2000TMprinting device. Letters of equal size were made up of small dots and contrast was varied by changes in dot size and interdot distance. All dots were too small to be resolved by the eye at a distance of 4 m. This is a simple and cheap method to produce letters of selected sizes and contrast levels (0.44 0.3, 0.21, 0.18, 0.14, 0.11, 0.09, 0.064, 0.055 and 0.03). The two lowest contrasts were put on a card that could be inserted between two glare sources when needed. Eight letters (E, F, K, N, U, V, X & Y) of equal legibility (Hedin & Olsson 1984) were used. The optotype size was equivalent to an average stroke width of 3.3 min of arc at a viewing distance of 4 m which corresponds to a visual acuity of 0.3. There were eight letters of each contrast positioned in two rows of four letters with an interletter distance of one letter width. Fluorescent tubes (Philips PL 9W/ 82) covered with diffusing glass were used as horizontal glare sources above and below the letters. Letters and glare sources were put in a light box (LIC, Sweden).The luminance of the chart was 450
146
Fig 1. Low contrast letter chart.
cd/m2 and the luminance of the glare sources was 9000 cdlm‘. The visual acuity (VA) chart was mounted in a similar LIC-box. It exhibited the same letters (E, F, K, N, U, V, X & Y) as the low contrast letter chart. The acuity-chart had logarithmic steps, and ten letters in each row for acuities between 0.3 and 2.0. Procedure Visual acuity (VA) and letter contrast sensitivity without glare (LCS) and with glare (LCS,;) were tested in a similar way, at 4 m distance. The right eye was tested first then the left and finally both together.
VA measurements were started with determination of the smallest letter size which could be identified without errors. If the subject was unable to read all letters of any size, the level where the maximal number of letters could be read was identified instead. From this level the subject was asked to read subsequent letters of smaller size down to the row in which no letters could be read. All correctly identified letters were noted and the number of letters read on each row was plotted as a frequency of seeing curve. LCS was tested in a similar way. Instead of smallest letter size, the lowest contrast level in which all letters could be correctly identified was determined, and the subject was asked to read subsequent letters of lower contrast down to the row where no letters at all could be read. For LCS,,determination one glare source was lit above and one below the low contrast letters being read. Subjects were instructed to look between the glaresources and avoid direct fixation. There was no time limit. Subjects were allowed to look at the letters as long as they wanted. A slit-lamp examination followed the testing. Intraocular pressure was measured by applanation tonometry and the pupils were dilated with Tropicamide (Mydriacylm0.5%)or Cyclopentolate (Cyclogyl' l0/o). The central fundus was inspected through a 60 diopters lens. Cataracts were graded at the slit-lamp regarding nuclear features according to 'The Oxford Clinical Cataract Classification and Grading System' (Brown et al. 1987).The presence of posterior subcapsular cataract (PSC) was noted. Retroillumination photography with the 'Oxford Retro-lllumination Cataract Camera' (Holofax Limited, England) was used to assess the amount of cortical opacities in all subjects. Cataracts were grossly classified as nuclear (N), posterior subcapsular (PSC) and cortical (except PSC) (C).Seven of the patients and two of the controls were retested within a month after the initial testing. Evaluation LCS and glare-score When the lowest contrast level where the subject could identify all letters was determined, an assumption was made that all letters of higher contrast could be seen as well. Results were plotted as a frequency of seeing curve (Fig. 2a) where ni is the number of correctly read letters of the contrast ci. The area SLrsunder ,,:I:
the curve was calculated according to the following formula.
+
S,,, = n,x log (l/c1)+n, [ l ~ g ( l / ~ , ) - l ~ g ( l / ~ J ] n3 [log(l/c,)-log(l/cp)].... LCS is the value of log (l/c) that multiplied by 8 (the maximum value of n) is equal to S,*(\. Hence LCS
SICS = __
8 LCS, is the LCS determined with glare sources lit. Glare score is defined as LCS-LCSC7. Thus higher glare scores indicate more severe visual loss. Evaluation-VA The results of the VA-testingwere treated in a way similar to that of LCS results (Fig. 2b). The area S, under the frequency of seeing curve was calculated as follows.
S, = n , (log VA, (log VA,+l)]. ...
+ 1) + n2 [(log VA,+l)
-
The (log VA+l) value that multiplied by 10 (the maximum value of ni)equals S, yields the patients
sv-
estimated VA log VA
+ 1 =S,
VA = 10
1"
10
10
Results It was possible to obtain an LCS and an LCS,, value for all eyes in the study except two. These two eyes both had posterior subcapsular cataract and a low LCS. With glare these eyes saw none of the letters at the 0.44 (highest)contrast level. VA of these two eyes was 0.4 and 0.5, respectively. All control eyes saw most of the letters of the lowest contrast (0.03) both with and without glare (glarescore=O). All cataractous eyes had a LCS that was lower than that of the controls and it decreased substantially when the glare source was turned on. This drop in visual performance was unrelated to VA (Fig. 3). Range of glare scores in patients with glare complaints was for the worst eye 0.13-0.47, for the best eye 0.04-0.29 and binocularly 0.02-0.44. The cataract patient without glare problems had a glarescore of 0.15 for the right and 0.10 for the left eye. He had a combination of nuclear and cortical cataract, none of which was more predominant. 147
L
log (1 /Ci)
0.36
0.52
0.74
0.68
0.85
0.96
1.05
1 + log v A. 0.48
0.7
VA
=
10
+
+
+
VA
=
+
+
+
+
1.11
0.9
Fig. 2. contrast, q = number of correctly read letters. (0.68-0.52) X 8 (0.74-0.68) X 7 (0.85-0.74) X 6
Example of LCS calculation. a) ci = SLcs = 0.36 X 8 (0.52-0.36) X 8 LCS = 6.63/8 = 0.83. b) Example of VA calculation. ni = number of correctly read letters. S, = 0.48 X 10 (0.6-0.48) X 10 (0.7-0.6) X 6 (0.81-0.7) X 6 (0.9-0.81) 7.82-10
+
1.0
0.81
0.6
X
4
+ (0.96-0.85) X 1 = 6.63
+ (1.0-0.9) X 2 = 7.82.
0.61
Discussion
The exophoria patient with severe glare complaints had normal LCS and a glare score of 0 We have presented a new method to measure glare disability in cataract patients. The test is simple monocularly as well as binocularly. The patient with superiorly subluxated lens and cheap to produce and does not include comcould not be tested at 4 m distance because her vi- plicated electronic devices. The test task of identsion was too poor. While seated two meters from ifying letters was found to be easy to understand the chart she could read all letters of the highest and the test procedure rapid to perform in most contrast without glare. When both glare sources as cases. In agreement,Pelli and coworkers (Pelliet al. 1988)have concluded that for clinical contrast senwell as the one above the letters were on she could read none of the letters. When only the lower glare sitivity testing letters are more suitable than gratsource was on she could read the letters of 44% ings which we previously used for glare testing (Paulson & Sjostrand 1980; Abrahamsson & Sjoscontrast 'as easy as without glare'. Testhetest variations for VA, LCS and LCS
![[Glare test and bright light vision in cataract patients].](/assets/img/hold.png)
