grading scales in optometry
TRANSCRIPT
39 | October 17 | 2003 OT
Optometric Grading ScalesFor use in everyday practice
Clinical
Richard Pearson describes the Optometric Grading Scales, whichcombine several separate grading schemes already in use.
Richard M. Pearson MPhil, FBOA; HD, FCOptom, DCLP, DOrth
Grading schemes for the assessment ofcontact lens-related complications havebeen available for several years. Thebest-known are those of Efron1 and theCCLRU2. Described as “a useful adjunct topractitioners in gauging and monitoringthe progression of ocular complications incontact lens wear”, they have beenadvocated for routine use3.
Using such scales, it has been shownthat non-optometrists are capable ofmaking similar, but less reliable, estimatesas optometrists4 and that gradingreliability improves with experience butnot with training5. In view of the provenbenefit of grading scales in the contactlens field, it seems anomalous that asimilar series has not been available foruse in routine optometric practice. TheOptometric Grading Scales presented hereare an attempt to correct this omission bybringing together for the first time severalseparate grading schemes already in use.
The Optometric Grading Scales arearranged under the following fourheadings:• Limbal anterior chamber depth• Age-related cataract• Vertical cup/disc ratio• Retinal vessels
Limbal anteriorchamber depthAnterior chamber depth should beassessed routinely, particularly in patientssuspected of having glaucoma and thosewho require pupil dilation. Thecircumstances which warrant the use of amydriatic are familiar to everyoptometrist6. The van Herick method,which requires the use of the slit lampbiomicroscope, has been widely adopted
by optometrists as a method of assessinglimbal anterior chamber depth7.
A pen torch can also be used in what isknown as the penlight, flashlight oroblique illumination test. The penlight isheld parallel to the plane of the iris andtemporally in order to direct light nasally.If the entire iris appears to be illuminated,the inference is that the angle of theanterior chamber is wide open (Grade 4).The presence of a shadow nasally signifiesshallowness of the anterior chamber, thedegree depending upon the extent of theshadow (Grades 3 to 1). The presence of acrescentic shadow in shallow anteriorchambers has prompted some authors todescribe this as the ‘eclipse test’8. vanHerick et al considered that this method ofestimating angle width was reliable in themajority of cases7. Since this test takes onlyseconds to complete, it can easily beundertaken together with the van Hericktest.
Both the van Herick and penlight testsassess limbal anterior chamber depth incontrast to axial anterior chamber depth,which can be measured with a simple slitlamp method9-11, by means of apachometer (such as the Haag-Streit modelII) or using A-scan ultrasonography10,11.Table 1 shows the results of several studiesof the sensitivity and specificity of the vanHerick and penlight tests. The variation inresults may be due to the different patientpopulations (numbers, ethnicity, etc) andthe use as the ‘gold standard’ ofpachymetry in some studies12,13 andgonioscopy in others14-17.
Grading schemes generally use anincreasing numerical value to correspondwith greater levels of severity of acondition, and these are usually illustrated
from left to right. The van Herick andpenlight tests are unusual in that the‘least severe’ situation is described asGrade 4 and most severe situation asGrade 1. In an attempt to reconcile thissituation on the Optometric GradingScales, decreasing depth of the anteriorchamber is arranged from left to rightwhile retaining the numerical valuesassigned by van Herick et al7.
It is relevant to note that van Herick etal found that only 1.64% of 2,185subjects examined by slit lamp andgonioscopy were classified as havingnarrow angles (Grades 1 or 2)7. It is alsoreassuring to note that none of the 4,870Caucasian and Afro-Caribbean subjectswho were dilated developed angle-closureglaucoma, despite the fact that 38 ofthem were judged on the basis ofgonioscopy to have occludable angles18.Nevertheless, the authors of this studyacknowledged that their findings mightnot apply to an Asian population, inwhich angle-closure glaucoma isreportedly more common.
Reference to the illustrations of thevan Herick and penlight proceduresshould encourage consistency inestimation of limbal anterior chamberdepth.
Age-related cataractOptometrists monitor various forms ofcataract over long periods of time and ithas been standard practice to make asketch of the condition at eachexamination. It is possible to gradecataract, and reference to the literaturesuggests that the scheme which currentlyhas the widest acceptance is the LensOpacities Classification System (LOCS)introduced by Chylack et al in 198819.
Initially, the LOCS used black andwhite photographs for the classificationof cortical and posterior sub-capsularcataracts and a coloured photograph forthe classification of nuclear colour andopalescence. This system subsequentlyevolved into LOCS II20 and LOCS III21 inwhich each type of cataract is illustratedwith colour photographs.
The drawings used for theclassification of age-related cataract in theOptometric Grading Scales are basedupon LOCS III photographs and are notintended to replace sketches made by thepractitioner, but to complement them byallowing them to be graded. Those
Table 1 Sensitivity and specificity of the van Herick and penlight tests
SSeennssiittiivviittyy
82
91
81.8
61.9
84*
SSppeecciiffiicciittyy
84.6
53
95.8
89.3
86*
SSeennssiittiivviittyy
89
90.9
45.5
96.67
SSppeecciiffiicciittyy
88
76.2
82.7
74.53
12
13
14
15
16
17
vvaann HHeerriicckk tteesstt PPeennlliigghhtt tteesstt
* Using a scale of six grades instead of four
RReeffeerreennccee
Text continued on page 42
Clinical Richard M. Pearson MPhil, FBOA; HD, FCOptom, DCLP, DOrth
40 | October 17 | 2003 OT
OPTOMETRIC GRADING SCALES
LIMBAL ANTERIOR CHAMBER DEPTH
Grade 4 Grade 3 Grade 2 Grade 1
Dilation: angle closure unlikely Dilation: involves risk Dilation: contra-indicated
Van
Her
ick
danger
Chamber ≥ section Chamber = 1/4 to 1/2ti
Chamber = 1/4 section Chamber < 1/4 section
Pen
ligh
t
NO shadow > 2/3 illuminated 1/3 to 2/3 illuminated < 1/3 illuminated
AGE-RELATED CATARACT
Grade 1 Grade 2 Grade 3 Grade 4 Grade 5
Nu
clea
r
1
23
4
Co
rtic
alP
ost
erio
rS
ub
-cap
sula
r
© Richard M Pearson
Clinical
41 | October 17 | 2003 OT
OPTOMETRIC GRADING SCALES
VERTICAL CUP/DISC RATIO
0.1 0.2 0.3 0.4 0.5
0.6 0.7 0.8 0.9 95% confidence limit ofnormality
Vert disc Vertdiam (mm) C/D ratio
1.2 0.5 1.5 0.6 1.8 0.7 2.2 0.8
(Garway-Heath et al 1998)
RETINAL VESSELS
100% 80% 60% 40% 20%
A/V
rati
o
1 2 3 4
0% 25% 50% 75% 100%
Art
eri
al re
flex
0 1 2 3 4
To
rtu
osit
y
© Richard M Pearson
Clinical Richard M. Pearson MPhil, FBOA; HD, FCOptom, DCLP, DOrth
42 | October 17 | 2003 OT
recognised as having the shape of a prolateellipse. At higher ratios, the cupincreasingly assumes an oblate shape.Vertical CDRs are depicted since the cuptends to enlarge in that meridian inglaucoma. One study has demonstrated asignificant correlation between the extentof cupping and the size of the visual field,and that the vertical CDR is the mostuseful quantitative index of cupping30.
The margin of the cup is depicted as adotted line to acknowledge the fact that itis sometimes less easily distinguished thanthat of the disc. It should be emphasisedthat the cup margin represents adiscernible change in contour, not pallor.
Jonas et al provided data on thevariation in width of the neuroretinal rim,reporting that it was greatest inferiorlyfollowed, in decreasing order, by superior,nasal and temporal regions28. This findinghas been referred to as the ISNT rule,which may be an inappropriatedesignation in view of the significantstandard deviations for the data on rimwidth.
In subjects with normal optic discs, themean value of the maximum difference inneuroretinal rim width was 0.2mm. Theseauthors examined the relationshipbetween the total neuroretinal rim areaand optic disc area but they provided noinformation about meridional variation ofneuroretinal rim width for differentcup/disc ratios. Accordingly, this featurehas been omitted in the CDR diagrams inthe Optometric Grading Scales. It is,nonetheless, important to assess theneuroretinal rim and to note notching orchanges in shape.
When recording the CDR, it is pertinentto recall that Afro-Caribbean subjects havebeen found to have higher values and agreater prevalence of ratios of 0.5, or more,than Caucasians31,32.
Due to the variation in cup size in thenormal population, the CDR alone haslimited value in the identification of aglaucomatous optic disc. Indeed, thevertical CDR can be as high as 0.85 innormal eyes28. Ideally, the CDR should beconsidered in relation to optic disc sizebecause when this is larger, the cup iscorrespondingly larger. Optic disc size canbe measured using indirect slit lampbiomicroscopy with a high plus-poweredcondensing lens (e.g. 78 or 90D)33.
The CDR chart includes a table showingthe 95% confidence limits of normality forseveral vertical ratios and disc sizes, whichwas compiled using the linear regressionequation of Garway-Heath et al.
Retinal vesselsIt has been common practice to record therelative calibre of retinal arteries to veins asan A/V ratio expressed as a fraction. Aslong ago as 1890, Gowers remarked that,
depicting nuclear cataract illustrateincreasing density of opacification but donot, of course, allow grading of its colour.
Colour slides and photographs can bepurchased from the principal author ofLOCS III, Dr Leo Chylack, and are clearlyessential for use in cross-sectional orlongitudinal research studies.
Vertical cup-to-disc ratioIn 1921, Pickard22 was perhaps the first tomention comparison of the size of theoptic disc and cup in relation to glaucoma.It has been said that one of the bestestablished methods of screening for anddetecting the early signs of open-angleglaucoma is examination of the opticdisc23. More recently, it has been stated thatclinical estimation of the size of the cupremains the simplest and most frequentlyperformed assessment of the disc in thediagnosis and follow-up of the glaucomasuspect24.
If the recorded cup/disc ratio (CDR) isto have clinical value, it is essential that itbe estimated consistently both by differentindividuals and by each individual overperiods of time.
Intra-observer agreement has beenfound to be higher than inter-observeragreement in the assessment of CDR fromstereoscopic optic disc photographs, andthe inter-observer agreement ofophthalmologists was higher than that ofoptometrists25.
In order to facilitate estimation of thesize of the optic disc and cup, the use of achart comprised of concentric circles torepresent cup/disc ratios from 0.1 to 0.9 inintervals of 0.1 was advocated as long agoas 196026. A template of similarconstruction, but in the form of atransparent sheet, has been used tomeasure the CDR from colourphotographs of the optic disc27.
In 1971, Allergan Pharmaceuticalspublished a Cup Disc Ratio Guide whichillustrated values from 0.0 to 0.8, in 0.1intervals. Despite the use of colourphotographs, the potential value of thischart was undermined by the fact that eachsmall figure was not confined to the areaof the disc but included an extensive areaof normal retina.
An alternative approach has been theincorporation of a graticule for theevaluation of CDR in directophthalmoscopes, such as the KeelerPractitioner and Professional models23.
The CDR chart in the OptometricGrading Scales is intended to achieve amore accurate representation than thatprovided by diagrams of concentric circles,and is novel since it is based upon the dataof Jonas et al28 and constructed using theBurek algorithm29. Accordingly, the dischas the shape of an oblate ellipse and forCDR values from 0.1 to 0.5, the cup can be
“As a rule the width of the artery is abouttwo-thirds or three quarters that of thevein” and optometric case records arefrequently annotated with ‘A/V 2/3’34. Thepreponderance of this recorded value maybe due to the fact that the majority ofoptometric patients exhibit ‘normal’retinal vasculature. There may, however, besome difficulty in envisaging alternativefractions – a problem which can beobviated by reference to a range ofillustrations.
In order to improve description ofretinal vasculature, Wolffsohn et al useddiagrams of the artery-to-vein ratioexpressed as a percentage35. Although therange of their grades was from 120% to40%, the upper limit encountered in theirsubjects was 100%. In view of this finding,the Optometric Grading Scales illustrateA/V grades from 100% to 20%.
Grades depicting the width of thearteriolar reflex and arteriolar tortuosityhave also been included, since these haveshown by the same authors to beparticularly useful indicators of systemichypertension. Photographs of normal andcongenitally ‘abnormal’ vessels served asmodels for the two extremes shown in thediagrams depicting tortuosity. Intermediategrades were derived by manipulation ofthe images in digital form.
ConclusionAs with contact lens grading scales, it isof course possible to optimise gradingsensitivity by interpolating between thegrades illustrated36. The OptometricGrading Scales are presented in thehope that they will minimise intra andinter-observer variation in theassessment of various clinical features.
AcknowledgementThe author is indebted to Henry Burek,who devised the algorithm for theconstruction of the CDR chart.
About the authorHaving worked in multiple,independent and hospital practice,Richard Pearson has spent most of hiscareer in optometric education.
ReferencesFor a full set of references, email [email protected].