Embed Size (px)
Citation preview
Color Dissociation Artifacts in Double Maddox Rod Cyclodeviation Testing
Kurt Simons, PhD, Kyle Arnoldi, CO, COMT, Mary H. Brown, CO
Background: The double Maddox rod test, based on a red Maddox rod in front of one eye and a clear Maddox rod in front of the other, is used to measure cyclodeviation, typically in patients with superior oblique muscle pareses. Discrepant results between the double Maddox rod test and other torsion measures, and reports of "paradoxic" cyclodeviation in the normal eye of some patients with superior oblique paresis, suggest the two-color format of the double Maddox rod test may produce artifactual torsion measures.
Methods: Forty patients with superior oblique paresis were tested twice using the double Maddox rod test, reversing the red and white Maddox rods between eyes for the second test, and 18 were tested further with same-color red or clear Maddox rods in front of both eyes.
Results: With the standard double Maddox rod test, 33 (83%) of 40 patients localized their cyclodeviation to the eye viewing through the red Maddox rod, irrespective of laterality of the paresis or fixation preference. In all 33 patients, laterality of the perceived torsion changed between eyes when testing was repeated with red and white Maddox rods interchanged between eyes. With same-color Maddox rods before both eyes, 17 (94%) of 18 patients localized extorsion to the paretic eye. There was a 7.6:1 ratio of luminance transmission and a 1.6:1 ratio of grating spatial frequency bandpass in the plano meridian between the clear and red Maddox rods, which appear to be responsible for the double Maddox rod test artifact.
Conclusion: The traditional double Maddox rod test may produce artifactual cyclodeviation measurements. An alternative version of the test, based on same-color Maddox rods in front of both eyes, is proposed. The relatively high spatial frequency bandpass characteristics of the plano meridian of the Maddox rod (as high as 20/25 Snellen equivalent resolution through the clear Maddox rod) also suggests double Maddox rod testing should be conducted in a dark room to avoid biases from visual environment cues. Ophthalmology 1994;101:1897-1901
Superior oblique muscle paresis is a common form of paretic strabismus. 1
- 3 The standard test for laterality and size of the
Originally received: September 24, 1993. Revision accepted: May 2, 1994.
From the Wilmer Ophthalmological Institute, Johns Hopkins Hospital, Baltimore.
Ms. Arnoldi currently is affiliated with the Children's Eye Center, St. Louis Children's Hospital, St. Louis.
Supported in part by National Institutes of Health grants EY07577 and EY07990, Bethesda, Maryland.
resulting cyclodeviation is the double Maddox rod test, in which a red Maddox rod is placed in front of one eye and a "white" (clear) Maddox rod in front of the other eye in a trial frame. 1
,4 The Maddox rods are oriented vertically to produce white and red horizontal lines when the patient views a point light source. One or both lines may appear to be tilted off the horizontal if torsion is present. The orientations of the Maddox rods then are adjusted until the two lines are seen as parallel by the patient, and the amount of torsion in degrees is read from the trial frame, using axis marks scribed onto the rods for this purpose.
Reprint requests to Kurt Simons, PhD, Wilmer Ophthalmological Institute, BI-35, Johns Hopkins Hospital, Baltimore, MD 21287-9009.
The difficulty with the double Maddox rod test is that its results have been shown to conflict with those
1897
Ophthalmology Volume WI, Number 12, December 1994
of other measures of cyclodeviation, both subjective (e.g., double Bagolini lenses5,6) and objective (fundus photographs5,7). For instance, only 15.7% ofa large (n = 231) group of patients with superior oblique paresis reported awareness of image tilting under normal viewing conditions, yet 78% of those who could be tested (n =
154) perceived cyclodeviation on the double Maddox rod test.2 Another study found discrepancies between fundus photographs and double Maddox rod test results in five of seven patients with partially or fully masked bilateral superior oblique paresis in whom both measures were made. The double Maddox rod test indicated unilateral extorsion, whereas objective testing indicated bilateral extorsion, and vice versa (see ref. 7 for Figs 1 and 2). Perhaps most noteworthy, however, are reports of "paradoxic" extorsion seen in the nonparetic eye of some patients with unilateral superior oblique paresis in double Maddox rod testing.2,8,9
The double Maddox rod test presents a different color image to each eye to provide a color cue that allows the patient to specify which eye's line appears torted and hence, in principle, which eye is torted, However, the redgreen two-color format has been found to introduce artifacts in stereopsis and fusion and suppression testing. 10- 12 Does the double Maddox rod test two-color (i.e., red and white) format introduce an artifact in this test's dissociated condition as well?
If the two-color format does introduce such an artifact, the result of that artifact might be expected to be affected by which eye views through the clear Maddox rod and which through the darker red Maddox rod. Surprisingly, there appears to be no testing standard in this respect. Some reference works suggest the convention of placing the red Maddox rod in front of the right eye,I.13 another in front of the eye suspected of being paretic,4 and yet others offer no recommendation. 14- 16 In addition, the laterality of color of the two Maddox rods is, with few exceptions,6 typically not specified in reports of studies using the double Maddox rod test.2,3,7-9,17-19
The current study investigated the artifact potential of the standard double Maddox rod test by comparing cyclodeviation measurement with the clear and red Maddox rods interchanged between eyes, and with same-color Maddox rods positioned in front of both eyes.
Methods
Forty consecutive patients with isolated unilateral superior oblique muscle paresis and subjective extorsion as identified by the "three-step" test,20 version testing, and the double Maddox rod test were included in the study. Patients with a previous history of eye muscle surgery were excluded from the study. Special care was taken to exclude bilateral superior oblique muscle paresis using specific criteria3,7; that is, patients with any of the following findings were eliminated from the study: (1) reversal of the hypertropia, however mild, in any field of gaze, including head tilt; (2) a positive Bielschowsky head tilt in which the difference in hyperdeviations between right and left
1898
tilt was less than 5 prism diopters; (3) subjective cyclodeviation over 10°; (4) a "V" pattern of20 prism diopters or more; (5) a chin-down compensatory head posture; (6) presence of bilateral objective fundus torsion; or (7) appearance of an overaction of the contralateral inferior oblique muscle within 6 months of surgery for unilateral superior oblique muscle paresis. If there was no history of trauma, neoplasm, vascular event, or other probable cause, patients were presumed to have a congenital superior oblique muscle paresis when there was photographic evidence of longstanding head tilt, increased vertical fusion amplitUdes, or evidence of sensory adaptation to strabismus when the deviation was manifest. Best-corrected visual acuity was 20/ 20 in both eyes in most patients, and no patient had worse than 20/50 visual acuity. There were 21 males and 19 females, ranging in age from 15 to 71 years. Informed consent was obtained from all participants in an institutional review board-approved protocol.
Each patient was seated in a dimly lit examination room wearing trial frames without correction. Testing began with the red Maddox rod placed over the right eye and the clear Maddox rod placed over the left eye. The rods were oriented vertically, producing an image of two horizontal lines, one red and one white. A 4-diopter basedown prism placed in front of one eye ensured that the lines could not be fused. A transilluminator was used as a point source of light, held in the primary position by the examiner, approximately 33 cm from the patient. The patient was asked to note any tilting of the lines from the horizontal, to identify which line appeared to be tilted, and to indicate the direction and magnitude of tilt by rotating the rods in the trial frame. The test then was repeated with the Maddox rods reversed: clear over the right eye, red over the left. The last 18 patients entered into the study also were tested with the same color Maddox rods, red or clear, in front of both eyes under the same conditions, and finally, with two red Maddox rods in a dark room.
To evaluate the optical characteristics of the Maddox rods themselves, two sets of measurements were made. Spot luminance measurements of light transmission through the rods were made by two methods, one of the bright line produced by the standard point source in a dark room, and an additional measurement of diffuse transmission from a distributed source (light box). The spatial frequency bandpass characteristics of the plano meridian of both color Maddox rods also was evaluated. This was done by placing one of the rods with the axis (plano meridian) oriented horizontally in a trial frame in front of one eye of two healthy subjects; the other eye was occluded. A head mount was used to ensure constant head alignment throughout the trial. The subjects then viewed a grating acuity target on a commercial vision testing unit (BV AT II, Mentor 0&0, Norwell, MA) through the Maddox rod. The unit randomly presents successive gratings in a vertical orientation or tilted 30° left or right. The subject's task was to indicate in which trials the grating was vertical. Threshold was defined as the finest grating at which the subject could perform this task correctly on three successive appearances of the grating in the vertical orientation.
Simons et al . Color Dissociation Artifacts
Results
Red/clear and Clear/red Maddox Rod Combinations
Six of 40 patients (marked with an asterisk in Table 1) reported subjective extorsion of the same eye under all conditions tested. But in 33 (83%) of the 40 patients, subjective extorsion could be induced in the nonparetic eye. In every case, this occurred when that eye was viewed through the red Maddox rod (Table 1). In 34 of the 40 patients, apparent extorsion reversed laterality between eyes (33 patients) or appeared and disappeared (patient 3; Table 1) when the red and clear Maddox rods were reversed. The indicated size of the extorsion was identical for approximately half (n = 17) of the 33 patients for either red/clear combination and different in the remainder.
Red/red and Clear/clear Maddox Rod Combinations
Seventeen (94%) of the 18 patients tested localized the extorsion to the paretic eye with all three tests in which the same color Maddox rod was placed before both eyes: double red, double clear, and double red in a dark room (Table 1). There were differences between the three methods in the magnitude of torsion perceived. In ten patients who showed a difference between double clear and double red Maddox rods, the double red rod exhibited a larger degree of torsion in all ten patients (indicated by "<" signs between the measurements in these two categories in Table 1). Four of these patients indicated no perceived cyclodeviation in the double clear Maddox rod condition ("no torsion" in Table 1), although they had exhibited torsion in both red/clear combinations. In all nine patients with a difference between the double reds in the light and dark rooms, the dark room condition produced larger apparent torsion (again indicated by "<" signs in Table 1). Patient 25, who persistently localized extorsion to the non paretic eye in all three conditions of same-color testing, also had shown this persistent laterality (in the nonparetic eye) for both red/clear combinations.
Maddox Rod Luminance and Spatial Frequency Transmission
The clear Maddox rod transmitted more than 7.5 times as much light as the red rod, by either measurement method, and had 1.6 times the grating resolution bandpass of the red rod (Table 2).
Discussion
These results demonstrate clearly that the two-color format of the standard double Maddox rod test, with a red rod placed in front of one eye and a clear Maddox rod in
front of the other, can give rise to an artifactuallocalization of cyclodeviation to one eye in patients with superior oblique paresis. This bias was demonstrated in three different findings: (1) the appearance of subjective extorsion in the nonparetic eyes of33 (83%) of the 40 patients tested, in every case when the nonparetic eye viewed through the red Maddox rod; (2) the reversal of laterality of the subjective extorsion when the clear and red Maddox rods were reversed'between eyes in these patients or, in patient 2, the appearance versus disappearance of torsion; but (3) correct localization of the cyclodeviation to the paretic eye in 17 (94%) of the 18 patients who were retested with the same color lenses before both eyes (Table 1). (Six patients reported cyclodeviation of the same eye, all ofprobable congenital origin, for all conditions tested, in keeping with other evidence of permanent sensory adaptation when the deviation has a childhood onset. 18
)
Cues from the visual environment are known to aid some patients with superior oblique paresis in compensating for cyclotropia, reducing or eliminating perceived tilt.5,6,18,19 While Maddox rods often are considered to pass no usable visual information, this study demonstrated that in fact the plano meridian of the rods passes spatial frequency information at a relatively high visual acuity level-as high as 15 cycles/degree (20/40 equivalent) for the red rod and 24 cycles/degree (20/25 equivalent) for the clear rod (Table 2). The high cylinder power of the Maddox rod segments mean that the plano meridian passes this information over only a narrow range of orientations. However, in being oriented to produce a horizontalline from the point light source, the Maddox rods also are aligned optimally to pass the orientational cues of horizontal edges that typical indoor environments are filled with, such as horizontal shelf edges or junctures between wall and ceiling. The reader can observe this effect by viewing a room through either a red or clear rod and rotating it. Environmental detail will be particularly noticeable when the rod is oriented in the 0° or 90° meridians.
The torsion-reducing effect of visual environment cues on amount of torsion indicated by the double Maddox rod test was well illustrated in the current study. When the two Maddox rod colors were used together, the eye viewing through the clear Maddox rod, with more than 7.5 times the luminance transmission and 1.6 times the spatial frequency bandpass of the red rod (Table 2), showed less (or no) torsion than the eye viewing through the red Maddox rod, irrespective of which eye had the paresis (Table 1). A similar effect was seen when samecolor Maddox rods were used in front of both eyes. Changing from bilateral clear rods to bilateral red rods, both in a lighted room, and then to bilateral red rods in a dark room, resulted in successively dimmer images being presented (i.e., a progressively reduced degree of visible environmental detail). In all nine or ten patients, respectively, where there were within-patient differences between these two steps ("<" signs in Table 1), the condition of less stimulation (i.e., darker images) gave rise to larger perceived amounts of torsion. In patients 26, 27, 34, and 39 (Table 1), viewing through bilateral clear Maddox rods
1899
Tab
le 1
. D
oubl
e M
addo
x R
od
Tes
t R
esul
ts i
n P
atie
nts
wit
h S
uper
ior
Obl
ique
Par
esis
.....
\0
a
Indi
cate
d T
orsi
on w
ith
a A
ge a
t R
edM
R
Bil
ater
al
Pat
ient
S
ympt
om
Ste
reoa
cuit
y I
Fix
atio
n C
lear
MR
B
ilat
eral
Red
B
ilat
eral
Red
N
o.
Dia
gnos
is
Etio
logy
O
nset
F
usio
n P
refe
renc
e aD
as
(li
ght
room
) M
R a
ight
room
) M
R (
dark
roo
m)
1 R
SO
T
raum
a 18
mos
20
" as
5° e
x aD
5°
ex
as
2 R
SO
V
ascu
lar
1 m
o D
iplo
pia
as
10°
ex a
D
7° e
x as
3 L
SO
Con
geni
tal
15 y
rs
140"
aD
N
o to
rsio
n 5°
ex
as
4*
RS
O
Vas
cula
r 1
mo
20"
as
3° e
x aD
3°
ex
aD
5
RS
O
3 yr
s D
iplo
pia
as
10°
ex a
D
8° e
x as
6 L
SO
Tra
uma
lyr
80"
aD
5°
ex
aD
5°
ex
as
0 7
LSO
Pr
esum
ed c
onge
nita
l 10
yrs
40
" aD
5°
ex
aD
5°
ex
as
"d
::r'
8 L
SO
Con
geni
tal
20 y
rs
7CY'
aD
5°
ex
aD
5°
ex
as
.... ::r
9 L
SO
? ?
2CY'
aD
5°
ex
aD
5°
ex
as
~
10
RS
O
Tra
uma
2 m
os
Dip
lopi
a as
10°
ex a
D
8° e
x as
S 11
R
SO
T
raum
a 1
yr
30"
as
5° e
x aD
5°
ex
as
0 -0 12
L
SO
Pres
umed
con
geni
tal
5 yr
s 4C
Y' aD
5°
ex
aD
5°
ex
as
~
13
LSO
C
onge
nita
l 15
yrs
80
" aD
15
° ex
aD
15
° ex
as
14
LSO
N
eopl
asm
4
yrs
80"
aD
5°
ex
aD
5°
ex
as
<
15*
LSO
?
3 yr
s 70
" aD
5°
ex
as
5° e
x as
0
16*
RS
O
Con
geni
tal
20 y
rs
30"
aD
5°
ex
as
5° e
x as
~
17*
LSO
C
onge
nita
l 35
yrs
40
" as
5° e
x aD
2°
ex
aD
S (1
)
18
RS
O
Tra
uma
3 m
os
30"
as
8° e
x aD
8°
ex
as
.....
19*
RS
O
Con
geni
tal
25 y
rs
40"
aD
8°
ex
as
8° e
x as
0 .....
20
RS
O
Vas
cula
r 3
mos
D
iplo
pia
as
2° e
x aD
4°
ex
as
Z
21
LSO
1
yr
Dip
lopi
a as
3° e
x aD
3°
ex
as
~
22
LSO
T
raum
a 1
yr
2CY'
OD
8°
ex
aD
10
° ex
as
S 23
R
SO
V
ascu
lar
6 m
os
Dip
lopi
a as
5° e
x aD
2°
ex
as
3° e
x aD
<
5° e
x aD
5°
ex
aD
C
r"
(1)
24
LSO
C
onge
nita
l 20
yrs
20
" aD
4°
ex
aD
5°
ex
as
5° e
x as
5° e
x as
< 8°
ex
as
~ ......
25*
LSO
C
onge
nita
l 30
yrs
4C
Y' as
5° e
x aD
5°
ex
aD
3°
ex
aD
<
5° e
x aD
5°
ex
aD
~N
26
LSO
T
raum
a 2
mos
70
" aD
1°
ex
aD
3°
ex
as
No
tors
ion
< 3°
ex
as
< 5°
ex
as
tJ
27
RS
O
Con
geni
tal
Chi
ldho
od
20"
OS
5°
ex
aD
5°
ex
as
No
tors
ion
< 3°
ex
aD
<
5 ex
aD
(1
) ('
)
28
RS
O
Tra
uma
4mos
D
iplo
pia
as
6° e
x aD
8°
ex
as
7° e
x aD
<
8° e
x aD
<1
1° e
x aD
(1
) S 29
R
SO
Pr
esum
ed c
onge
nita
l 35
yrs
70
" as
5° e
x aD
2°
ex
as
5° e
x aD
5°
ex
aD
<
8° e
x aD
C
r"
30
RS
O
Tra
uma
2 m
os
Dip
lopi
a as
6° e
x aD
5°
ex
as
6° e
x aD
6°
ex
aD
<1
0° e
x aD
(1
) ~
31
LSO
N
eopl
asm
2
yrs
Dip
lopi
a as
5° e
x aD
10
° ex
as
10°
ex a
s
10°
ex a
s
<15°
ex
as
......
\0
32
LSO
T
raum
a 3
mos
3C
Y' A
ltern
ates
3°
ex
as
3° e
x as
2° e
x O
S <
4° e
x o
s
4° e
x as
\0
33
LSO
T
raum
a 1
yr
4(Y
' aD
5°
ex
aD
5°
ex
as
5° e
x as
5° e
x as
5° e
x as
4>-
34
RS
O
7CY'
aD
3°
ex
aD
3°
ex
as
No
tors
ion
< 3°
ex
aD
3°
ex
aD
35
L
SO
Tra
uma
4mos
D
iplo
pia
Alte
rnat
es
8° e
x aD
6°
ex
as
5° e
x as
< 8°
ex
as
8° e
x as
36
LSO
N
eopl
asm
5
mos
D
iplo
pia
aD
5°
ex
aD
5°
ex
as
5° e
x as
5° e
x as
8° e
x as
37
RS
O
Ane
urys
m
6 m
os
70"
aD
6°
ex
aD
6°
ex
as
5° e
x aD
5°
ex
OD
<
6° e
x O
D
38
LSO
T
raum
a 8m
os
40"
aD
3°
ex
aD
2°
ex
as
3° e
x as
< 5°
ex
as
5° e
x o
s
39
RS
O
? 5
yrs
100"
as
6° e
x O
D
3° e
x as
No
tors
ion
< 5°
ex
aD
<
6° e
x as
40
LSO
Pr
esum
ed c
onge
nita
l 15
yrs
D
iplo
pia
aD
5°
ex
aD
10
° ex
as
10°
ex O
S
10°
ex a
s
10°
ex a
s
MR
= M
addo
x ro
d; O
D =
rig
ht e
ye;
OS
= le
ft e
ye;
RS
O =
rig
ht s
uper
ior
obli
que
mus
cle
pare
sis;
ex
= ex
tors
ion;
LS
O =
left
sup
erio
r ob
liqu
e m
uscl
e pa
resi
s; ?
= u
ndet
erm
ined
ori
gin .
•
Sub
ject
ive
tors
ion
alw
ays
loca
lize
d to
par
etic
eye
.
Simons et al . Color Dissociation Artifacts
Table 2. Maddox Rod Light Transmission and Grating Acuity Thresholds
Transmission Through Through Clear:red (cd/m2
) Clear Rod Red Rod Ratio
Light box (fluorescent) 223 29 7.7:1 Bright line in dark
room (incandescent) 0.53 am 7.6:1
Grating Acuity (cycles/degree)
Subject 1 20 12 1.7:1 Subject 2 24 15 1.6:1 Average 22.0 13.5 1.6:1
eliminated the torsion apparent in the other four test conditions, all of which involved the dimmer red image being presented to at least one eye.
In conclusion, it appears that use of the traditional double Maddox rod test, or even use of a single red Maddox rod with the other eye uncovered, introduces the possibility of the eye viewing through the red rod being artifactually identified as having a cyclodeviation. The results of the current study suggest that the diagnosis of laterality of a cyclodeviation can be made more reliable if Maddox rods of the same color are used before both eyes. To distinguish which eye has torsion, one eye's rod is rotated back and forth a few degrees in the trial frame after the angular measurement has been made, and the patient is asked whether the horizontal or torted line is "rocking." To avoid any bias by environmental visual cues, double Maddox rod testing should be conducted in a completely dark room. If this is not possible, the dimmest available illumination should be used, together with the red rather than white rods.
References
1. von Noorden GK. Binocular Vision and Ocular Motility: Theory and Management of Strabismus, 4th ed. St. Louis: CV Mosby, 1990;184.
2. von Noorden GK, Murray E, Wong SY. Superior oblique paralysis. A review of 270 cases. Arch Ophthalmol 1986; 1 04: 1771-6.
3. Kraft SP, Scott WE. Masked bilateral superior oblique palsy: clinical features and diagnosis. J Pediatr Ophthalmol Strabismus 1986;23:264-72.
4. von Noorden GK. Atlas of Strabismus, 4th ed. St. Louis: CV Mosby, 1983;56.
5. von Noorden GK. Clinical and theoretical aspects of cyclotropia. J Pediatr Ophthalmol Strabismus 1984;21:126-32.
6. Ruttum M, von Noorden GK. The Bagolini striated lens test for cyclotropia. Doc Ophthalmol 1984;58:131-9.
7. Kushner BJ. The diagnosis and treatment of bilateral masked superior oblique palsy. Am J Ophthalmol 1988;105:186-94.
8. Olivier P, von Noorden GK. Excyclotropia of the nonparetic eye in unilateral superior oblique muscle paralysis. Am J Ophthalmol 1982;93:30-3.
9. Slavin ML. Double Maddox rod test in superior oblique muscle palsy [letter]. Am J Ophthalmol 1982;93:807-9.
10. Simons K, Elhatton K. Artifacts in fusion and stereopsis testing based on red/green dichoptic image separation. J Pediatr Ophthalmol Strabismus [in press]
11. Arthur BW, Keech RV. The polarized three-dot test. J Pediatr Ophthalmol Strabismus 1987;24:305-8.
12. Arthur BW, Marshall A, McGillivray D. Worth vs polarized four-dot test. J Pediatr Ophthalmol Strabismus 1992;30:53-5.
13. Lyle TK, Wybar KC. Lyle and Jackson's Practical Orthoptics in the Treatment of Squint: (and Other Anomalies of Binocular Vision), 5th ed rev. Springfield: Charles C Thomas, 1967;478.
14. Parks MM. Ocular Motility and Strabismus. Hagerstown: Harper & Row, 1975;53.
15. Helveston EM. Surgical Management of Strabismus: An Atlas of Strabismus Surgery, 4th ed. St. Louis: CV Mosby, 1993;466.
16. Mein J, Harcourt B. Diagnosis and Management of Ocular Motility Disorders. Oxford: Blackwell Scientific, 1986;96.
17. von Noorden GK. Symposium: Strabismus. Clinical observations in cyclodeviations. Ophthalmology 1979;86:1451-61.
18. Guyton DL, von Noorden GK. Sensory adaptations to cyclodeviations. In: Reinecke R, ed. Strabismus: Proc 3rd Meeting In!'l Strabismological Assoc. New York: Grune & Stratton, 1978:399-403.
19. Ruttum M, von Noorden GK. Adaptation to tilting of the visual environment in cyclotropia. Am J Ophthalmol 1983;96:229-37.
20. Parks MM. Isolated cyclovertical muscle palsy. Arch Ophthalmol 1958;60: 1027-35.
1901
