Fluctuation of the Differential Light Threshold at the Border of Absolute Scotomas Comparison between Glaucomatous Visual Field Defects and Blind Spots IV AN 0. HAEFLIGER, MD,t· 2 JOSEF FLAMMER, MD 2
Abstract: The outcome of repeated measurements of the differential light sen sitivity fluctuates slightly. This fluctuation can be markedly increased in glaucoma patients but has also been described at the border of the blind spot of healthy subjects. The question arises, therefore, whether in glaucoma this increase may be due to the fact that many test locations may touch the border of (detected or undetected) scotomas. This study compares the behavior of the threshold at the border of glaucomatous defects and blind spots. The threshold was measured with program F8 of the Octopus 201 (lnterzeag, Inc, Schlieren, Swit zerland) automated static perimeter. The results revealed that the borders of glaucomatous defects are less steep than those of blind spots, the fluctuation is largest just at the border of the scotomas in both groups, and that the fluc tuation is significantly larger at the edge of glaucomatous defects than at the edge of blind spots of normal subjects. Ophthalmology 1991; 98:1529-1532
In automated static perimetry, the differential light sensitivity threshold is usually measured by means of a bracketing procedure. 1 Although this measurement is taken under standardized conditions, repeated assess ments of this threshold, done at the same location of a visual field, show a scatter of threshold values. 2 This scat ter, measured during the same session, is called the "short-
Originally received: March 15, 1991. R.evision accepted: May 28, 1991. 1
2
Clinique d'Ophtalmologie, H6pital Cantonal Universitaire de Geneve, Switzerland. Universitats-Augenklinik, Basel, Switzerland.
The authors have no proprietary interest in the development or marketing of any equipment used in this study. Reprint requests to Ivan 0. Haefliger, MD, Clinque d'Ophthalmologie, H6pital Cantonal Universitaire de Geneve, 29 Rue Micheli-du-Crest, 1211 Geneve 9, Switzerland.
term fluctuation." 2 Whereas a small amount of short-term fluctuation is expected to be found in healthy subjects, a marked increase of this phenomenon has been described in a number of conditions and especially in glaucoma. 3.4 However, recently it has been shown that an increase of the short-term fluctuation could also be found in healthy subjects if measurements were taken at the border of the blind spot. It has been suggested that, in this situation, the increase of the short-term fluctuation was simply re lated to the fact that these measurements were taken at the border of a scotoma. 5 Therefore, one can wonder whether an increase of short-term fluctuation found in pathologic conditions is simply related to the fact that measurements are taken at the border of detected or un detected scotomas or if additional factors may also have an influence on this increase. In the current study, we have compared the short-term fluctuation measured at the border ofabsolute glaucoma tous scotomas with that measured at the border of blind spots.
1529
OPHTHALMOLOGY
•
OCTOBER 1991
SUBJECTS AND METHODS
•
VOLUME 98
1530
NUMBER 10
B1i nd spots : LMS as a funct 1on of test location
40
LMS (dB)
Two groups of subjects were selected. The first group was a control group and consisted of eight volunteers in whom the blind spots were considered to be physiologically absolute scotomas. The second group was made up of six primary open-angle glaucoma patients who had isolated absolute scotomas in their visual fields. All these scotomas had a clear-cut border and an abrupt loss of sensitivity when selected on Octopus program G 1, which has a grid density between 2° and 6°. 6 It must be noted that, among the large number ofglaucomatous visual tiel~ we screened, this kind of defect was rather uncommon. For each patient only one visual field was selected, and in each visual field only one scotoma was studied. Perimetry was done with an automated perimeter, Oc topus 201, using the standard Octopus parameters of background illumination (4 asb ), of stimulus presentation time (100 msec), and ofstimulus intensity (0 to 1000 asb). Tests were performed with a stimulus size of 0.43 degree of diameter. Assessment of the short-term fluctuation was done with the Octopus program F8. With this program, the differ ential light sensitivity threshold was measured at regular intervals (1 °) along short segments of a straight line. At each tested location, measurements were repeated eight times. In this way, a profile perimetry was performed that crossed perpendicularly the border of the scotoma. In control subjects, profiles had a horizontal axis that cut the blind spot into two halves. In glaucomatous sub jects, profiles had an axis that was radial and located in between the 45°-135° or 225°-315° meridian. For each scotoma or blind spot, only one profile was assessed, and measurements were taken outside (in the "normal" area), at the border of, and finally inside the scotoma. Therefore, each profile began 1o before the de crease of sensitivity of the scotoma and ended within 1o of the area with a complete loss of sensitivity. To avoid the influence of eccentricity on short-term fluctuation, 7 all measurements were taken within the cen tral 15 ° of the visual field. Furthermore, to rule out the question of a possible learning effect, all subjects had had previous experience with automated perimetry. Further more, only subjects who had a very high reliability rate were taken into account (rate of false responses in catch trials less than 5%). For each tested location, the mean and the standard deviation of the eight measures of differential light sen sitivity were calculated. Here, the local mean sensitivity refers to the mean of these eight measures, taken at one test location, and the local short-term fluctuation refers to the corresponding standard deviation. All values oflocal mean sensitivity and local short-term fluctuation were plotted as a function of their location along the profiles (Figs 1, 2). For each profile, the width of the border of the scotoma, namely the width of the transition zone be tween the "normal" area (outside the scotoma) and the
•
a a
Iii
30
I
a a a a a
20 10
a
.,
0
-2
outside the
0
a
a 4
2
6
B
test location (deg.)
bllnd spot
Glaucomatous scotomas: LMS as a function of test 1ocat ion
40
LMS (dB)
a
I!!
30
I li
20
II 0
10
. El
"
II a Cl
II
0
0 outside the scotoma
-2
0
+
2
0
a
0
4
6
B
test location (deg.)
Fig 1. For each location tested, the differential light threshold is assessed eight times. The average of these eight measures is called the local mean sensitivity (LMS). Each square corresponds to one value of local mean sensitivity. For primary open-angle glaucoma scotomas (bottom) and for healthy subjects' blind spots (top), values ofLMS are represented as a function of their location along short segments cutting perpendicularly the border of the scotoma. The point 0 on the X-axis is the last test location with a normal value of LMS before the loss of sensitivity of the scotoma. Highest values ofLMS represent measurements taken outside the scotoma.
area with the absolute defect (inside the scotoma), was measured. For statistical comparison between the two groups (Mann-Whitney UTest), the mean sensitivity (the average oflocal mean sensitivity values) as well as the short-term fluctuation (the average of local short-term fluctuation values) has been calculated for each patient.
RESULTS For both groups, individual values of local mean sen sitivity and local short-term fluctuation, measured along the border of scotomas, are presented graphically in Fig
HAEFUGER AND FLAMMER
•
Blind spots: LSF as a funct1on of test locat1on 15
LSF (dB) 10
5
0
lil
I a
.____
-2 outside the blind spot
Ill 1!1
2
0
4
6
8
to the center of the scotoma, in an area of decreased sen sitivity, local short-term fluctuation values are much lower. Although this relationship between local short-term fluctuation and test location is true for both groups, values oflocal short-term fluctuation found in glaucomatous eyes are much different from those ofcontrol eyes. At the bor der of glaucomatous scotomas, values of local short-term fluctuation can go up to lO decibels (dB) or even more, while at the border of blind spots, local short-term fluc tuation values do not exceed 5 dB. This difference was statistically significant (P < 0.01; Table 1).
test location (deg.)
DISCUSSION
Glaucomatous scotomas : LSF as a function of test location
IS
LSF (dB)
II
II
10
" II 5
" .. "" 1!1
-2 outside the
scotoma
DIFFERENTIAL LIGHT THRESHOLD
.____
a
"
.
"
a
.. .. " .. D
" El
"
0
2
4
6
8
test location (deg.)
Fig 2. For each location tested, the differential light threshold is assessed eight times. The standard deviation of the mean of these eight measures is called the local short-term fluctuation (LSF). Each square corresponds to one value ofLSF. For primary open-angle glaucoma scotomas (bottom) and for healthy subjects' blind spots (top), values of LSF are represented as a function of their location along the profile of the border of the scotomas. The point 0 on the X-axis is the last test location with a normal value of LMS before the loss of sensitivity of the scotoma. Negative values on the X-axis represent test locations outside of the scotoma.
ures l and 2. Furthermore, in Table l , the mean sensitivity and short-term fluctuation as well as the width of the bor der of the scotoma are given in a numerical form for each subject, with statistical comparison between the two groups. Although there is no significant difference in the mean sensitivity between the two groups (P > 0.05; Table 1), it appears that the border of glaucomatous scotomas is quite different from the border ofblind spots (Fig 1). The width of the border tends to be larger in the group of glaucoma tous eyes than in the other group (P < 0.01; Table 1). In Figure 2, it appears that values of local short-term fluctuation vary as a function of their location along the profile. Highest values oflocal short-term fluctuation can be observed at the border of the scotoma, whereas outside the scotoma, in an area of high sensitivity, as well as closer
It is known that, in glaucomatous visual fields and es pecially at the border of scotomas, an increase of short term fluctuation can be observed. 3•4 Nevertheless, in healthy subjects, an increase of short-term fluctuation has also been reported if measurements were taken at the bor der of the blind spot. 5 •8 Because no disease could explain these findings, it has been suggested that this phenomenon could be partly related to test locations where measure ments have been taken. For example, outside the blind spot, where the slope of the island of the vision is flat, it is very easy to assess the differential light sensitivity threshold with a bracketing procedure. Schematically, the stimulus can either be below the threshold (i.e., "seen") or above it and, consequently, undetectable. On the other hand, at the border of the blind spot, where the slope is very steep, it is much more difficult to assess whether the stimulus is above or under the threshold, as the stimulus has a tendency to straddle the threshold. Therefore, it has been suggested that the increase of short-term fluctuation observed at the border of blind spots could be mainly related to the fact that measurements are taken in the transition zone where the slope of the sensitivity threshold is very steep. 5 Results found in the group of normal eyes confirmed these previous findings. 5•8 The highest values oflocal short term fluctuation seen in Figure 2 (top) correspond to measurements taken just at the border of the blind spot, where the slope of sensitivity is the steepest (Fig 1, top). If one would assume that the increase of local short term fluctuation is only related to the slope of sensitivity where measurements are taken, one would expect to find higher values oflocal short-term fluctuation at the border ofthe blind spots that have a s·,eeper slope than the border ofglaucomatous scotomas (Fig I). However, the opposite is observed: higher values of local short-term fluctuation are found at the border of glaucomatous scotomas (Fig 2) where the loss of sensitivity is less steep (Fig 1), a dif ference which is statistically significant (P < 0.0 l; Table l ). Therefore, it is obvious that, in the group of glauco matous scotomas, additional factors other than the slope must have an influence on the increase of the local short term fluctuation.
1531
OPHTHALMOLOGY
•
OCTOBER 1991
•
VOLUME 98
•
NUMBER 10
Table 1. Mean Sensitivity, Short-term Fluctuation and Scotoma Border Width (Transition Zone Width) at the Border of Blind Spots and Open-angle Glaucoma Scotomas Blind Spots
Glaucomatous Scotomas
Case No.
MS (dB)
SF (dB)
Transition Zone Width (degree)
Case No.
MS (dB)
SF (dB)
Transition Zone Width (degree)
N1 N2 N3 N4 N5 N6 N7 N8
17.3 14.7 16.5 20.0 13.6 19.1 16.7 17.1
1.4 2.0 1.2 1.3 1.5 1.6 2.1 1.8
4 4 3 3 3 3 3 4
G1 G2 G3 G4 G5 G6
17.8 17.2 13.7 21.7 18.4 17.3
7.3 4.9 3.1 4.6 4.7 3.6
7 7 3 4 5 5
MS = average of local mean sensitivity values calculated for each patient; dB = decibel; SF = average of local short-term fluctuation values calculated for each patient. Statistical comparison between the two groups (Mann-Whitney U test): MS: P > 0.05; SF: P < 0.01; transition zone width: P < 0.01.
This study also shows that the border of glaucomatous scotoma is much wider than that of the blind spot (P < 0.01; Table 1), even though we selected the rare cases in which, on visual field printouts, the border ofthe defects appeared to be very similar to the border of blind spots. The reason why the border of these scotomas is not so distinct is not known. It could reflect a succession of rel ative or absolute microscotomas at the junction between normal and pathologic areas. It could also represent a temporal variation of sensitivity threshold at the border of these scotomas. It has to be remembered that this study was conducted on selected cases and on a small number of patients. Therefore, these results cannot be representative of all visual field defects found in primary open-angle glaucoma. The question whether such an increase of short-term fluc tuation can also be found at the border of other types of glaucoma defects or in other diseases remains to be clar ified.
ACKNOWLEDGMENT The authors thank Patricia Dumont for her statistical advice.
1532
REFERENCES 1. Bebie H, Fankhauser F, Spahr J. Static perimetry: strategies. Acta Ophthalmol 1976; 54:325-38. 2. Bebie H, Fankhauser F, Spahr J. Static perimetry: accuracy and fluc tuations. Acta Ophthalmol1976; 54:339-48. 3. Werner EB, Drance SM. Early visual field disturbances in glaucoma. Arch Ophthalmol1977; 95:1173-5. 4. Flammer J, Drance SM, Zulauf M. Differential light threshold. Short and long-term fluctuation in patients with glaucoma, normal controls, and patients with suspected glaucoma. Arch Ophthalmol1984; 102: 704-6. 5. Haefliger 10, Flammer J. Increase of the short-term fluctuation of the differential light threshold around a physiologic scotoma. Am J Ophthalmol1989; 107:417-20. 6. Flammer J, Jenni F, Bebie H, Keller B. The Octopus glaucoma G1 program. Glaucoma 1987; 9:67-72. 7. Rutishauser C, Flammer J, Haas A. The distribution of normal values in automated perimetry. Graefes Arch Clin Exp Ophthalmol 1989; 227:513-17. 8. Henson DB, Bryson H. Is the variability in glaucomatous field loss due to poor fixation control? In Mills RP, Heijl A, eds. Perimetry Update 1990/91: Proceedings of the IX International Perimetric Society Meet ing, Malmo, June 1990. Amsterdam: Kugler & Ghednin, 1991; 217 20.
Abstract: The outcome of repeated measurements of the differential light sen sitivity fluctuates slightly. This fluctuation can be markedly increased in glaucoma patients but has also been described at the border of the blind spot of healthy subjects. The question arises, therefore, whether in glaucoma this increase may be due to the fact that many test locations may touch the border of (detected or undetected) scotomas. This study compares the behavior of the threshold at the border of glaucomatous defects and blind spots. The threshold was measured with program F8 of the Octopus 201 (lnterzeag, Inc, Schlieren, Swit zerland) automated static perimeter. The results revealed that the borders of glaucomatous defects are less steep than those of blind spots, the fluctuation is largest just at the border of the scotomas in both groups, and that the fluc tuation is significantly larger at the edge of glaucomatous defects than at the edge of blind spots of normal subjects. Ophthalmology 1991; 98:1529-1532
In automated static perimetry, the differential light sensitivity threshold is usually measured by means of a bracketing procedure. 1 Although this measurement is taken under standardized conditions, repeated assess ments of this threshold, done at the same location of a visual field, show a scatter of threshold values. 2 This scat ter, measured during the same session, is called the "short-
Originally received: March 15, 1991. R.evision accepted: May 28, 1991. 1
2
Clinique d'Ophtalmologie, H6pital Cantonal Universitaire de Geneve, Switzerland. Universitats-Augenklinik, Basel, Switzerland.
The authors have no proprietary interest in the development or marketing of any equipment used in this study. Reprint requests to Ivan 0. Haefliger, MD, Clinque d'Ophthalmologie, H6pital Cantonal Universitaire de Geneve, 29 Rue Micheli-du-Crest, 1211 Geneve 9, Switzerland.
term fluctuation." 2 Whereas a small amount of short-term fluctuation is expected to be found in healthy subjects, a marked increase of this phenomenon has been described in a number of conditions and especially in glaucoma. 3.4 However, recently it has been shown that an increase of the short-term fluctuation could also be found in healthy subjects if measurements were taken at the border of the blind spot. It has been suggested that, in this situation, the increase of the short-term fluctuation was simply re lated to the fact that these measurements were taken at the border of a scotoma. 5 Therefore, one can wonder whether an increase of short-term fluctuation found in pathologic conditions is simply related to the fact that measurements are taken at the border of detected or un detected scotomas or if additional factors may also have an influence on this increase. In the current study, we have compared the short-term fluctuation measured at the border ofabsolute glaucoma tous scotomas with that measured at the border of blind spots.
1529
OPHTHALMOLOGY
•
OCTOBER 1991
SUBJECTS AND METHODS
•
VOLUME 98
1530
NUMBER 10
B1i nd spots : LMS as a funct 1on of test location
40
LMS (dB)
Two groups of subjects were selected. The first group was a control group and consisted of eight volunteers in whom the blind spots were considered to be physiologically absolute scotomas. The second group was made up of six primary open-angle glaucoma patients who had isolated absolute scotomas in their visual fields. All these scotomas had a clear-cut border and an abrupt loss of sensitivity when selected on Octopus program G 1, which has a grid density between 2° and 6°. 6 It must be noted that, among the large number ofglaucomatous visual tiel~ we screened, this kind of defect was rather uncommon. For each patient only one visual field was selected, and in each visual field only one scotoma was studied. Perimetry was done with an automated perimeter, Oc topus 201, using the standard Octopus parameters of background illumination (4 asb ), of stimulus presentation time (100 msec), and ofstimulus intensity (0 to 1000 asb). Tests were performed with a stimulus size of 0.43 degree of diameter. Assessment of the short-term fluctuation was done with the Octopus program F8. With this program, the differ ential light sensitivity threshold was measured at regular intervals (1 °) along short segments of a straight line. At each tested location, measurements were repeated eight times. In this way, a profile perimetry was performed that crossed perpendicularly the border of the scotoma. In control subjects, profiles had a horizontal axis that cut the blind spot into two halves. In glaucomatous sub jects, profiles had an axis that was radial and located in between the 45°-135° or 225°-315° meridian. For each scotoma or blind spot, only one profile was assessed, and measurements were taken outside (in the "normal" area), at the border of, and finally inside the scotoma. Therefore, each profile began 1o before the de crease of sensitivity of the scotoma and ended within 1o of the area with a complete loss of sensitivity. To avoid the influence of eccentricity on short-term fluctuation, 7 all measurements were taken within the cen tral 15 ° of the visual field. Furthermore, to rule out the question of a possible learning effect, all subjects had had previous experience with automated perimetry. Further more, only subjects who had a very high reliability rate were taken into account (rate of false responses in catch trials less than 5%). For each tested location, the mean and the standard deviation of the eight measures of differential light sen sitivity were calculated. Here, the local mean sensitivity refers to the mean of these eight measures, taken at one test location, and the local short-term fluctuation refers to the corresponding standard deviation. All values oflocal mean sensitivity and local short-term fluctuation were plotted as a function of their location along the profiles (Figs 1, 2). For each profile, the width of the border of the scotoma, namely the width of the transition zone be tween the "normal" area (outside the scotoma) and the
•
a a
Iii
30
I
a a a a a
20 10
a
.,
0
-2
outside the
0
a
a 4
2
6
B
test location (deg.)
bllnd spot
Glaucomatous scotomas: LMS as a function of test 1ocat ion
40
LMS (dB)
a
I!!
30
I li
20
II 0
10
. El
"
II a Cl
II
0
0 outside the scotoma
-2
0
+
2
0
a
0
4
6
B
test location (deg.)
Fig 1. For each location tested, the differential light threshold is assessed eight times. The average of these eight measures is called the local mean sensitivity (LMS). Each square corresponds to one value of local mean sensitivity. For primary open-angle glaucoma scotomas (bottom) and for healthy subjects' blind spots (top), values ofLMS are represented as a function of their location along short segments cutting perpendicularly the border of the scotoma. The point 0 on the X-axis is the last test location with a normal value of LMS before the loss of sensitivity of the scotoma. Highest values ofLMS represent measurements taken outside the scotoma.
area with the absolute defect (inside the scotoma), was measured. For statistical comparison between the two groups (Mann-Whitney UTest), the mean sensitivity (the average oflocal mean sensitivity values) as well as the short-term fluctuation (the average of local short-term fluctuation values) has been calculated for each patient.
RESULTS For both groups, individual values of local mean sen sitivity and local short-term fluctuation, measured along the border of scotomas, are presented graphically in Fig
HAEFUGER AND FLAMMER
•
Blind spots: LSF as a funct1on of test locat1on 15
LSF (dB) 10
5
0
lil
I a
.____
-2 outside the blind spot
Ill 1!1
2
0
4
6
8
to the center of the scotoma, in an area of decreased sen sitivity, local short-term fluctuation values are much lower. Although this relationship between local short-term fluctuation and test location is true for both groups, values oflocal short-term fluctuation found in glaucomatous eyes are much different from those ofcontrol eyes. At the bor der of glaucomatous scotomas, values of local short-term fluctuation can go up to lO decibels (dB) or even more, while at the border of blind spots, local short-term fluc tuation values do not exceed 5 dB. This difference was statistically significant (P < 0.01; Table 1).
test location (deg.)
DISCUSSION
Glaucomatous scotomas : LSF as a function of test location
IS
LSF (dB)
II
II
10
" II 5
" .. "" 1!1
-2 outside the
scotoma
DIFFERENTIAL LIGHT THRESHOLD
.____
a
"
.
"
a
.. .. " .. D
" El
"
0
2
4
6
8
test location (deg.)
Fig 2. For each location tested, the differential light threshold is assessed eight times. The standard deviation of the mean of these eight measures is called the local short-term fluctuation (LSF). Each square corresponds to one value ofLSF. For primary open-angle glaucoma scotomas (bottom) and for healthy subjects' blind spots (top), values of LSF are represented as a function of their location along the profile of the border of the scotomas. The point 0 on the X-axis is the last test location with a normal value of LMS before the loss of sensitivity of the scotoma. Negative values on the X-axis represent test locations outside of the scotoma.
ures l and 2. Furthermore, in Table l , the mean sensitivity and short-term fluctuation as well as the width of the bor der of the scotoma are given in a numerical form for each subject, with statistical comparison between the two groups. Although there is no significant difference in the mean sensitivity between the two groups (P > 0.05; Table 1), it appears that the border of glaucomatous scotomas is quite different from the border ofblind spots (Fig 1). The width of the border tends to be larger in the group of glaucoma tous eyes than in the other group (P < 0.01; Table 1). In Figure 2, it appears that values of local short-term fluctuation vary as a function of their location along the profile. Highest values oflocal short-term fluctuation can be observed at the border of the scotoma, whereas outside the scotoma, in an area of high sensitivity, as well as closer
It is known that, in glaucomatous visual fields and es pecially at the border of scotomas, an increase of short term fluctuation can be observed. 3•4 Nevertheless, in healthy subjects, an increase of short-term fluctuation has also been reported if measurements were taken at the bor der of the blind spot. 5 •8 Because no disease could explain these findings, it has been suggested that this phenomenon could be partly related to test locations where measure ments have been taken. For example, outside the blind spot, where the slope of the island of the vision is flat, it is very easy to assess the differential light sensitivity threshold with a bracketing procedure. Schematically, the stimulus can either be below the threshold (i.e., "seen") or above it and, consequently, undetectable. On the other hand, at the border of the blind spot, where the slope is very steep, it is much more difficult to assess whether the stimulus is above or under the threshold, as the stimulus has a tendency to straddle the threshold. Therefore, it has been suggested that the increase of short-term fluctuation observed at the border of blind spots could be mainly related to the fact that measurements are taken in the transition zone where the slope of the sensitivity threshold is very steep. 5 Results found in the group of normal eyes confirmed these previous findings. 5•8 The highest values oflocal short term fluctuation seen in Figure 2 (top) correspond to measurements taken just at the border of the blind spot, where the slope of sensitivity is the steepest (Fig 1, top). If one would assume that the increase of local short term fluctuation is only related to the slope of sensitivity where measurements are taken, one would expect to find higher values oflocal short-term fluctuation at the border ofthe blind spots that have a s·,eeper slope than the border ofglaucomatous scotomas (Fig I). However, the opposite is observed: higher values of local short-term fluctuation are found at the border of glaucomatous scotomas (Fig 2) where the loss of sensitivity is less steep (Fig 1), a dif ference which is statistically significant (P < 0.0 l; Table l ). Therefore, it is obvious that, in the group of glauco matous scotomas, additional factors other than the slope must have an influence on the increase of the local short term fluctuation.
1531
OPHTHALMOLOGY
•
OCTOBER 1991
•
VOLUME 98
•
NUMBER 10
Table 1. Mean Sensitivity, Short-term Fluctuation and Scotoma Border Width (Transition Zone Width) at the Border of Blind Spots and Open-angle Glaucoma Scotomas Blind Spots
Glaucomatous Scotomas
Case No.
MS (dB)
SF (dB)
Transition Zone Width (degree)
Case No.
MS (dB)
SF (dB)
Transition Zone Width (degree)
N1 N2 N3 N4 N5 N6 N7 N8
17.3 14.7 16.5 20.0 13.6 19.1 16.7 17.1
1.4 2.0 1.2 1.3 1.5 1.6 2.1 1.8
4 4 3 3 3 3 3 4
G1 G2 G3 G4 G5 G6
17.8 17.2 13.7 21.7 18.4 17.3
7.3 4.9 3.1 4.6 4.7 3.6
7 7 3 4 5 5
MS = average of local mean sensitivity values calculated for each patient; dB = decibel; SF = average of local short-term fluctuation values calculated for each patient. Statistical comparison between the two groups (Mann-Whitney U test): MS: P > 0.05; SF: P < 0.01; transition zone width: P < 0.01.
This study also shows that the border of glaucomatous scotoma is much wider than that of the blind spot (P < 0.01; Table 1), even though we selected the rare cases in which, on visual field printouts, the border ofthe defects appeared to be very similar to the border of blind spots. The reason why the border of these scotomas is not so distinct is not known. It could reflect a succession of rel ative or absolute microscotomas at the junction between normal and pathologic areas. It could also represent a temporal variation of sensitivity threshold at the border of these scotomas. It has to be remembered that this study was conducted on selected cases and on a small number of patients. Therefore, these results cannot be representative of all visual field defects found in primary open-angle glaucoma. The question whether such an increase of short-term fluc tuation can also be found at the border of other types of glaucoma defects or in other diseases remains to be clar ified.
ACKNOWLEDGMENT The authors thank Patricia Dumont for her statistical advice.
1532
REFERENCES 1. Bebie H, Fankhauser F, Spahr J. Static perimetry: strategies. Acta Ophthalmol 1976; 54:325-38. 2. Bebie H, Fankhauser F, Spahr J. Static perimetry: accuracy and fluc tuations. Acta Ophthalmol1976; 54:339-48. 3. Werner EB, Drance SM. Early visual field disturbances in glaucoma. Arch Ophthalmol1977; 95:1173-5. 4. Flammer J, Drance SM, Zulauf M. Differential light threshold. Short and long-term fluctuation in patients with glaucoma, normal controls, and patients with suspected glaucoma. Arch Ophthalmol1984; 102: 704-6. 5. Haefliger 10, Flammer J. Increase of the short-term fluctuation of the differential light threshold around a physiologic scotoma. Am J Ophthalmol1989; 107:417-20. 6. Flammer J, Jenni F, Bebie H, Keller B. The Octopus glaucoma G1 program. Glaucoma 1987; 9:67-72. 7. Rutishauser C, Flammer J, Haas A. The distribution of normal values in automated perimetry. Graefes Arch Clin Exp Ophthalmol 1989; 227:513-17. 8. Henson DB, Bryson H. Is the variability in glaucomatous field loss due to poor fixation control? In Mills RP, Heijl A, eds. Perimetry Update 1990/91: Proceedings of the IX International Perimetric Society Meet ing, Malmo, June 1990. Amsterdam: Kugler & Ghednin, 1991; 217 20.
