British Journal of Ophthalmology 1996;80:799-804

Effect of a cataract simulation on clinical and realworld vision

David B Elliott, Mark A Bullimore, Aftab E Patla, David Whitaker

AbstractAimslbackground-Many reports haveindicated that some patients with cataractcan retain good visual acuity but complainof significant visual problems. This is thefirst in a series of papers trying todetermine what causes these symptomsand whether other clinical tests can pre-dict the real world vision loss.Methods-The effect ofa cataract simula-tion with a similar angular distribution oflight scatter as real cataract on clinical(visual acuity, contrast sensitivity, anddisability glare) and real world vision(face recognition, reading speed, andmobility orientation) was investigated.Results-The simulation had a relativelysmall effect on visual acuity (6/6 with thesimulation), but much larger effects oncontrast sensitivity and low contrast acu-ity with and without glare. The simulationhad no effect on high lunance and highcontrast real world tasks, such as mobilityorientation in room light and optimalreading speed. A small, but significantdeterioration was found for the slightlylower contrast task of face and expressionrecognition. However, under low lumi-nance conditions, substantial defects inmobility orientation were obtained (de-spite 616 acuity).Conclusions-Although the relative effectof the cataract simulation on acuity andcontrast tasks is not typical of the averagecataract, it can be found in those cataractpatients with visual problems despite goodvisual acuity. This corroborates the sug-gestion that it is large amounts of wideangle light scatter (forward andlor back-ward) which are at least partly responsiblefor visual disability in cataract patientswith good visual acuity. A patient's re-ported visual disability may depend on thepercentage oftime he or she spends underlow contrast and/or low luminance condi-tions, such as walking or reading in dimillumination, and walking or driving atnight, in fog, or heavy rain.(BrJ Ophthalmol 1996;80:799-804)

There is a vast body of literature reporting theuse of clinical vision tests to discriminatebetween diseased and normal eyes, to assessdisease progression, and to evaluate the effectsof intervention. Relatively little attention hasbeen paid, however, to the use of clinical visiontesting to predict real world performance.Visual acuity testing, for example, is employedto determine whether an individual is allowed

to drive and to categorise a patient as legallyblind, yet there are little or no data on the rela-tion between a given level of visual acuity andan individual's ability to perform everydayvisual tasks. In essence, all of our clinical visiontests are surrogate measures for real world per-formance but there is a paucity of data relatingthe two. The literature that does exist providesconflicting evidence. For example, three stud-ies have searched for clinical test correlates oforientation and mobility of low visionindividuals.'" All found a strong correlationbetween visual field loss and mobility orienta-tion performance, but Marron and Bailey'found no correlation between mobility orienta-tion performance and visual acuity, unlikeBrown and colleagues2 who reported them tobe closely related. Similar confficting evidenceis found when clinical tests are compared withface perception and reading. Bullimore et al '

found face recognition to be most highly corre-lated with visual acuity (specifically for wordtargets) while others have found the best corre-lation with contrast sensitivity.'7 Peak readingspeed in low vision subjects has, in differentstudies, been found to be correlated best withword reading acuity89 and contrast sensitivity.'0Many of the discrepancies can be accountedfor by differences in test design, the range ofvision tests employed, and the populationstested. For example, if the recognition of facetargets is assessed by varying contrast6 then it ismore likely to correlate with contrast sensitivitymeasures. If the face recognition threshold ismeasured by adjusting face size, then thescores are more likely to correlate with visualacuity.4The relation between real world perform-

ance and clinical tests of vision is particularlygermane in the case of cataract. Cataractextraction is the most commonly performedocular surgery, and it is argued that the needfor cataract extraction is indicated when thequality of the patient's day to day life isimpaired. Because it can be difficult to justifysurgery purely on the basis of patient'ssymptoms, and because there is so littleinformation on the relation between real worldvision and clinical tests in cataract, guidelinesemphasise that surgery is indicated when a cer-tain level of visual acuity has been reached."This is despite substantial evidence that visualacuity is a poor indicator of real world vision insome cataract patients.'2-'7 The AHCPR Clini-cal Practice Guidelines" and the AmericanAcademy of Ophthalmology report on contrastsensitivity and disability glare'6 suggest thatstudies of how real world vision loss compareswith clinical test results in patients withcataract are essential.

Department ofOptometry, UniversityofBradford, Bradford,BD7 lDP, UKD B ElliottD Whitaker

School of Optometry,University ofCalifornia, Berkeley,CA, USAMA Bullimore

Department ofKinesiology, UniversityofWaterloo, CanadaA E Patla

Correspondence to:D B Elliott.

Accepted for publication28 June 1996

799

on Septem

ber 22, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.80.9.799 on 1 Septem

ber 1996. Dow

nloaded from

Elliott, Bullimore, Patla, Whitaker

This is the first in a series of studies to inves-tigate the relation between clinical tests ofvision and real world visual performance incataract. In early cataract, visual disability isprincipally caused by increased forward lightscatter leading to reduced retinal contrast.'1&'8It has been suggested that patients with mediaopacities with good visual acuity who complainof significant visual problems are those withlarge amounts ofwide angle light scatter.'8 19 Toinvestigate this suggestion, we examined theeffects of a wide angle light scattering mediumon clinical and real world vision. Various 'cata-ract simulations' have been used in previousstudies to assess the effect of light scatter onresolution and contrast,20 visual field measure-ments,2" 22 mobility orientation,2" and drivingperformance.24 However, none of these studiesevaluated the angular distribution of the lightscatter produced by their simulations, yet thisdetermines the effect of light scatter on thepoint spread function and thus the contrastsensitivity function.'8 Several light scatteringmedia were assessed and one was chosen whichgave a similar angular distribution of wideangle light scatter as the normal25 and catarac-tous eye.'8 26 The angular distribution of wideangle light scatter has been shown to be similarfor the three main morphological cataracttypes of cortical, nuclear, and posterior sub-capsular.'8 The relatively greater effect ofposterior subcapsular cataract on vision com-pared with the other types is probably due tothe substantial loss of vision which occurs withreduced pupil size with this cataract.

Clinical vision (visual acuity, contrast sensi-tivity, and disability glare) and real world vision(face recognition, reading speed, and mobilityorientation) were then measured in 20 youngsubjects, both with and without the chosenscattering medium.

MethodsSELECTION OF LIGHT SCATTER MEDIUMBeyond 3 degrees of visual angle from its cen-tre, the point spread function declines inamplitude in approximately inverse proportionto the square of the visual angle for bothnormal and cataractous eyes.'8 25 26 To mimicwide angle intraocular light scatter, a similarangular light scatter distribution should beprovided. We assessed the light scatter distribu-tion of the Vistech cataract simulation andvarious optical cells which have been used pre-viously in our laboratory.27 The collimatedbeam of light from a tungsten filamentincandescent lamp was directed onto the frontsurface of the scatter simulation. A computercontrolled spectrophotometer (Bentham In-struments Ltd) was used to measure the lumi-nous flux emanating from a fixed area on therear surface of the illuminated scatter simula-tion. The light source and the scatter mediawere mounted on a rotatable optical bench,enabling photometric measurements to bemade at angles between 5 and 20 degrees fromthe filament. Measurements at angles less than5 degrees were unreliable owing to theextended (non-point source) nature of the fila-

E 3000_ | y = 99266.07*xA(-2.06)> 2500- R = 0.99

X 2000

o 1500

Co\0I 1000

500

-J

0 5 10 15 20 25Scatter angle (degrees)

Figure 1 A plot ofluminance emanatingfrom the Vistechcataract simulation against angle. The -2 exponent of thepower relation which characterises the scatteringdistribution is similar to the wide angle light scattering innormal and cataractous human eyes.

ment. Measurements at angles either side ofzero were averaged, and a graph of luminanceemanating from the scatter cell against anglewas plotted (Fig 1). The Vistech cataract simu-lation was found to scatter light with a similarangular distribution as cataract (light scatterbeing inversely proportional to the square ofthe angle) and was hence chosen for thepresent study (Fig 1). Other features such as itsspectacle mounting and availability are addi-tional useful qualities. The Vistech cataractsimulation is provided with the Vistech con-trast sensitivity system to enable practitionersto illustrate to patients the effect of reducedcontrast on vision. The picture through thesimulation is very 'washed out', although a rea-sonable sharpness to the edges of objects isretained.

SUBJECTSThirty young subjects (mean age 24.5 (SD4.1) years) were recruited. The tenets of theDeclaration of Helsinki were followed and thestudy gained ethical approval from the Univer-sity's Office of Human Research. Informedconsent was obtained after the nature of thestudy had been fully explained. The clinicaland real world vision assessments were madeon 20 subjects with and without the Vistechsimulation. Ten subjects were tested twicewithout the simulation to determine any learn-ing effects. All measurements were madebinocularly with natural pupils and with thesubject's own habitual refractive correction (ifany).

CLINICAL VISUAL ASSESSMENTClinical measurements consisted of binocularBailey-Lovie high contrast visual acuity, lowcontrast visual acuity with and without glare,and Pelli-Robson contrast sensitivity. All ofthese tests have previously been shown to pro-vide reliable measurements.28 Bailey-Loviehigh contrast logMAR visual acuity was meas-ured at 6 metres. LogMAR charts have becomethe standard for clinical research as they have

800

on Septem

ber 22, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.80.9.799 on 1 Septem

ber 1996. Dow

nloaded from

Effect ofa cataract simulation on clinical and real world vision

many advantages over traditional Snellencharts including the provision of non-truncated data down to 6/3, a geometricprogression of letter size that has been shownto provide equal increments in legibility, fiveletters on each line, and letters of similarlegibility. MAR is the minimum angle of reso-lution, so that 6/6 has a MAR of 1 minute ofarc and a logMAR value of 0.0. The chart con-tains 13 lines in 0.1 log unit steps from -0.3logMAR (6/3) to 1.0 logMAR (6/60). A chartluminance of 100 cd/m' and a by letter scoringsystem (0.02 log units per letter) was used. TheBerkeley glare test consists of a reduced lowcontrast Bailey-Lovie chart (Michaelson con-trast = 10%, Weber contrast = 18%) mountedon a triangular opaque panel at the centre of a30 x 27 cm opal Plexiglass panel. The chart isfront illuminated (80 cd/m') and the glaresource provided by transillumination of thePlexiglass panel. At the medium setting theglare source luminance is 750 cd/m'. Measure-ments of low contrast visual acuity were madeat 1 metre with and without the glare source,with credit (0.02 log min arc) given for eachletter read correctly. The Pelli-Robson chart isa 86 x 63 cm chart containing 4.9 x 4.9 cm let-ters arranged in 16 triplets. At a test distance of1 metre these letters correspond to a spatialfrequency of about 1 c/deg. Within each tripletthe letters have the same contrast and the con-trast in each successive triplet decreases by afactor of 0.15 log units. The chart luminancewas 100 cd/m' and a by letter scoring systemwhich gives credit (0.05 log units) for each let-ter read correctly was used.

REAL WORLD VISION ASSESSMENTThree real world tasks were adopted to assessreal world vision:1 mobility orientation2 reading speed3 face recognition.Safe travel relies on our ability to use and inter-pret visual information and select and imple-ment adaptive strategies. Adaptive strategiesused for locomotion over uneven terraininclude obstacle avoidance by appropriate steplength, width and height adjustments, andsteering. Rather than study each of the manyadaptive strategies in isolation, three travelpaths were designed that required most adap-tive strategies to be implemented by thesubjects. These included a variety of obstacles,both on and above ground, of different shapes,sizes, and contrast placed in a variety ofconfigurations. The travel paths were challeng-ing and required continuous visually guidedmodifications to the gait patterns to travelwithout bumping into objects and stayingwithin the boundaries of the path. Thisapproach evaluates the robustness of the visualperception (including visual attention andsearch skills) and action (changes in upper andlower body movements during locomotion)coupling. Mobility performance was assessedby recording the time needed to travel the pathand the number of 'mistakes' made by the sub-ject. A mistake was defined as contact with an

Figure 2 A schematic diagram of one of the travelpathways used to assess mobility orientation.

obstacle, coming to a complete stop, strayingoutside the pathway, and avoidance strategieswhen none were required-for example,changing locomotor patterns when no obsta-cles were present. Patients with cataracts andother media opacities particularly complain ofproblems in dim illumination and in thepresence of glare sources-for example, nightdriving. The illumination for two of thecourses was dim-that is, < 1 lux to simulatetwilight, and two of the three travel pathsincluded appropriately placed glare sources.The third course had illumination in thephotopic range-that is, - 450 lux, consistentwith normal room illumination. A diagram ofone of the courses is shown (Fig 2).Reading speed was measured by having

patients read Bailey-Lovie word charts at 40cm. These are non-continuous text charts withprint ranging in size from 10 M to 0.25 M (80point to 2 point) in 0.1 log unit steps. Subjectsread aloud and were taped for analysis at a latertime. Each subject read three different wordcharts. Reading speed for each print size wascalculated in words per minute, and wasaveraged across the three charts. Consistentwith previous reports,9 reading speed was rela-tively constant for print sizes greater than fivelines above threshold size. Reading perform-ance was determined as the mean of the twopeak speeds.

Face recognition was assessed using themethod developed by Bullimore and col-leagues.4 Black and white photographs of fourmale and four female faces were selected from'Pictures of facial affect' (Consulting Psycholo-gists Press, Palo Alto, CA, USA). For eachindividual, there were four different facialexpressions-happy, sad, angry, and afraid/surprised, giving a total of 32 photographs.The photographs were cropped to remove the

801

on Septem

ber 22, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.80.9.799 on 1 Septem

ber 1996. Dow

nloaded from

Elliott, Bullimore, Patla, Whitaker

Table 1 Mean (SD or range) of the binocular clinical and real world vision test results for 20young subjects.

Normal condition

Clinical tests:LogMAR visual acuityLog contrast sensitivityLogMAR low contrast VALogMAR low contrast VA with glare

Face recognition:LogEVD for identityLogEVD for expression

Reading:Log word acuityPeak reading speed in log words/min

Mobility:Dim illumination with glare

Dim illumination

Bright illumination with glare

MistakesTime (s)MistakesTime (s)MistakesTime (s)

-0.15 (6/4)* (0.07)t2.04 (0.08)0.12 (6/7.5) (0.07)0.15 (6/9) (0.06)

1.54 (35 m) (0.06)1.54 (35 m) (0.04)

-0.06 (6/5) (0.07)1.97 (93 wpm) (0.09)

0.9 (1.0)27 (6)0 (0-1)27 (5)0 (0-1)18 (5)

With Vistech simulation

-0.01 (6/6)* (0.06)t1.36 (0.07)0.48 (6/18) (0.10)0.95 (6/60) (0.02) (n= 1)

1.46 (29 m) (0.06)1.47 (30m) (0.07)

0.04 (6/6) (0.10)1.99 (98 wpm) (0.10)

5.3 (3.1)38 (10)6.5 (2.6)45 (12)0 (0-1)20 (4)

*Approximate equivalent Snellen visual acuity values are given in parentheses.tStandard deviation.

hair outline so that identification was predomi-nantly dependent on the facial features. Thirtyof these photographs were arranged in a letterchart format. There were five faces per linewith each line decreasing in size by 0.15 logunits. The angular size of the faces wasexpressed in terms of the equivalent viewingdistance (EVD), the distance at which a realface would subtend the same angle that thephotograph subtends. The subject's perform-ance was scored in the same way as the visualacuity charts with credit given (0.03 log units)for each correct answer. Two threshold scoreswere obtained: for correct recognition of iden-tity and for correct recognition of expression.During testing subjects could refer to a panelwith large photographs of the eight charactersin neutral facial expressions.

ResultsThere were no significant differences betweenthe test and retest data of the 10 subjects usingtwo tailed t tests. This indicates there were nosignificant learning or fatigue effects. Themean and standard deviations for the clinicaland real world vision tests, with and withoutlight scatter, from 20 subjects are shown inTable 1. The number of mistakes during thetravel paths without the light scatter werehighly skewed for the dim illumination andnormal illumination conditions, with mostsubjects not making any mistakes. Median andrange data are therefore given for these condi-tions rather than mean and standard devia-tions. In addition, nine of the 20 subjects couldnot see any of the letters on the low contrastvisual acuity chart under glare conditions withthe light scatter.The scattering medium induced little change

in visual acuity (0.14 logMAR or one and a halflines), mobility orientation in normal roomillumination (no increase in mistakes, 1 1%increase in time), face recognition (5 metrechange in equivalent viewing distance), wordacuity (0.10 logMAR or one line), and readingspeed (5 words per minute increase). Theslight effect on visual acuity meant that anaverage high contrast visual acuity of 6/6 wasobtained with the cataract simulation. More

substantial deficits were produced, however, incontrast sensitivity (0.68 log contrast sensitiv-ity or four and a half lines on the chart), lowcontrast visual acuity with (at least 8 lines ofvisual acuity) and without glare (three and ahalf lines), mobility orientation in dim illumi-nation (six additional mistakes, 67% increasein time), and in dim illumination with glare(five additional mistakes, 41% increase intime). All these differences were statisticallysignificant using two tailed t tests (p<0.001)except for the time taken and the number ofmistakes in the normal room illumination withglare and the optimal reading speed. Theslightly greater reduction in mobility in dimillumination compared with when a glaresource was added indicate that the glare sourceimproved vision by increasing room illumina-tion, and this effect was greater than anyreduction in vision due to disability glare.

DiscussionThe normal data are similar to previously pub-lished binocular data of visual acuity and Pelli-Robson contrast sensitivity in young subjects.'0The visual acuity data are slightly poorer thanpreviously reported and may reflect the use ofsubjects' habitual rather than optimal refractivecorrections.The cataract simulation produced a contrast

sensitivity:visual acuity loss ratio of 4.5 (Table1). Pardhan and Elliott'0 reported a contrastsensitivity:visual acuity loss ratio of 1.9 be-tween patients with cataract and age-matchedcontrols. The relatively greater effect of thecataract simulation on the contrast sensitivity-:visual acuity ratio compared with data re-ported for cataract may be because the scatter-ing medium is only providing wide angle lightscatter (beyond 5 degrees) of the same angulardistribution as cataract. This suggests that anaverage cataract influences visual acuity by amuch greater extent than the Vistech cataractsimulation. This is most probably due to arelatively greater amount of narrow angle lightscatter in the average real cataract (less thanabout 1 degree), but could be due to effectsother than light scatter in cataract-for exam-ple, irregularities of refractive index or aberra-tions.

802

on Septem

ber 22, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.80.9.799 on 1 Septem

ber 1996. Dow

nloaded from

Effect ofa cataract simulation on clinical and real world vision

However, it is not average cataract patientswe are interested in as their vision is reasonablywell described by traditional visual acuitymeasurements.'2 19 We are interested in the10-20% of cataract patients who have reason-able acuity, but complain of visual problems inthe real world.'2 19 The cataract simulationcaused increased wide angle light scatter withan angular distribution similar to cataract andcaused minimal loss to visual acuity, yet asignificant reduction in walking ability in dimillumination. These findings corroborate thesuggestion that wide angle light scatter is thecause of complaints of significant visualdisability in patients with cataract with goodvisual acuity.'8 9This could be due to increasesin both forward and backward light scatter.The Vistech simulation had two effects onvision: it reduced retinal contrast due toincreased forward light scatter and reducedretinal illumination due to increased backscat-ter. The latter effect is expected to besignificant only at low light levels. The poormobility orientation in dim illumination couldbe due to increased forward scatter reducingretinal contrast or increased backscatter reduc-ing retinal illumination. A contrast sensitivityor glare test could be used to clinically assessforward light scatter and the large reductions incontrast sensitivity and disability glare causedby the cataract simulation are evidence of theirrelatively greater sensitivity to forward lightscatter compared with visual acuity. Manyreports have suggested that low spatial fre-quency contrast sensitivity and/or glare testingshould be used to complement visual acuitytesting in early cataract.""'6 '9 The lack of aneffect of the simulation on mobility orientationin normal illumination suggests that it is thereduction in contrast or illumination of thesimulations which is decreasing mobility orien-tation in dim illumination rather than a restric-tion of the field of view due to the spectaclemounting. Backscatter can be assessed clini-cally in patients with cataract using theslit-lamp. Whether wide angle light scatterreduces real world performance at low illumi-nation levels because of increased forward orbackward light scatter will be investigated insubsequent studies.The fact that reading, face recognition, and

mobility orientation performance in normalillumination are relatively unaffected by thelight scatter implies that these real world visiontasks are relatively immune to reductions inretinal contrast at the light levels used. This isconsistent with previous research." 32Although we found no significant change

with the cataract simulation in mobility orien-tation in normal room illumination, verysignificant deteriorations in performance werefound under twilight conditions. We suggestthat cataract patients (and other patients withlow vision2') may have sufficiently good visionto travel when illumination is good, but do notwhen illumination is poor. Other research sug-gests that cataract and low vision patients areselective about when they go out walking anddriving. Genensky and colleagues" found thatalmost all of their 94 legally blind patients

(caused by a variety of conditions) walked out-doors by themselves (at least in well knownareas), but less than half travelled alone atnight. It is well known that patients withcataract tend to avoid difficult driving condi-tions, such as driving at night.'4 The increase offive to six 'mistakes' in the dim illuminationpathways due to the cataract simulationindicates the seriousness of the reduction inmobility orientation performance in these lightlevels, despite 6/6 visual acuity. This suggeststhat a good acuity may not ensure safe mobilityorientation in dim illumination in patients withcataract and other media opacities. Seriousinjury caused by falls is a significant problem inthe aging adult. It has been suggested that fallscause more than two thirds of the accidentaldeaths in the over 75 age group.35 The 40-66%increases in travel time found in the dimillumination paths should also be considered.The consequences of increases in travel time,such as when crossing a road, can also be serious.

In summary, our results indicate wide anglelight scatter can cause large reductions in con-trast sensitivity and disability glare, but haveminimal effects on visual acuity. Subjects withinduced wide angle light scatter can also beseriously disabled at some real world visiontasks such as walking in dim illumination.These findings corroborate the suggestion thatwide angle light scatter could cause complaintsof significant visual disability in patients withcataract who have reasonable visual acuity. Theresults further suggest that a cataract patientwith good visual acuity should have minimalproblems visually if the illumination conditionsare good. Their reported visual disability willprobably depend on the percentage of timethat they spend under low contrast and/or lowluminance and/or glare conditions, such aswalking or reading in dim illumination, nightdriving, and walking or driving in fog or heavyrain. In particular, a visual acuity as good as 6/6does not ensure that a cataract patient with sig-nificant wide angle light scatter is safe to walkunder poor illumination conditions. Contrastsensitivity and glare tests may be betterrepresentatives of these patients' vision thanvisual acuity. This hypotheses will be furtherexplored in subsequent reports which will dis-cuss results from cataract patients, both beforeand after surgery.

1 Marron JE, Bailey IL. Visual factors and orientation-mobility performance. Am J Optom Physiol Opt1982;59:413-26.

2 Brown B, Brabyn L, Welch L, Haegerstrom-Portnoy G,Collenbrander A. Contribution of vision variables tomobility in age-related maculopathy patients. Am J OptomPhysiol Opt 1986;63:733-9.

3 Lovie-Kitchen J, Mainstone J, Robinson J, Brown B. Whatareas of the visual field are important for mobility in lowvision patients? Clin Vision Sci 1990; 5:249-63.

4 Bullimore MA, Bailey IL, Wacker RT. Face recognition inage-related maculopathy. Invest Ophthalmol Vis Sci 1991;32:2020-9.

5 Ginsburg AP. Specifying relevant spatial information forimage evaluation and display design. Proceedings of the SID1980; 21:219-27.

6 Owsley C, Sekular R, Boldt C. Aging and low contrastvision: face perception. Invest Ophthalmol Vis Sci 1981;21:362-5.

7 Owsley C, Sloane ME. Contrast sensitivity, acuity, and thepreception of 'real-world' targets. BrJ Ophthalmol 1987;17:791-6.

8 Legge GE, Pelli DG, Rubin GS, Schleske MM. Psychophysicsof reading-I. Normal vision. Vision Res 1985;25:239-52.

803

on Septem

ber 22, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.80.9.799 on 1 Septem

ber 1996. Dow

nloaded from

Elliott, Bullimore, Pada, Whitaker

9 Bullimore MA, Bailey IL Reading and eye movements in age-related maculopathy. Optom Vsion Sci 1994;72:125-38.

10 Leat SJ, Woodhouse JM. Reading performance with lowvision aids: relationship with contrast sensitivity. OphthalPhysiol Opt 1993;13:9-16.

11 AHCPR Publication No 93-0542. Cataract ManagementGuideline Panel. Cataract in adults: management offunctional impairment. Clinical Practice Guideline, Number4. Rockville, MD: US Department of Health and HumanServices, Public Health Service, Agency for Health CarePolicy and Research, 1993.

12 Bernth-Petersen P. Visual functioning in cataract patients.Acta Ophthalmol 1981;59: 198-205.

13 Holladay JT, Prager TC, Truillo TC, Ruiz RS. Brighmessacuity test and outdoor visual acuity in cataract patients. JfCataract Refract Surg 1987;13:67-9.

14 Koch DD. Glare and contrast sensitivity testing in cataractpatients. J Cataract Refract Surg 1989;15: 158-64.

15 Elliott DB, Hurst MA, Weatherill J. Comparing clinical testsof visual function in cataract with the patient's perceivedvisual disability. Eye 1990;4:712-7.

16 American Academy of Ophthalmology report. Contrastsensitivity and glare testing in the evaluation of anteriorsegment disease. Ophthalmology 1990;97: 1233-7.

17 Brown NAP. The morphology of cataract and visualperformance. Eye 1993;7:63-7.

18 de Waard PWT, lJspeert JK, van den Berg TJTP, de JongPTVM. Intraocular light scattering in age-related cataracts.Invest Ophthalmol Vis Sci 1992;33:618-25.

19 Elliott DB. Evaluating visual function in cataract. Optom VisSci 1993;70:896-902.

20 Zuckerman JL, Miller D, Dyes W, Keller M. Degradation ofvision through a simulated cataract. Invest Ophthalmol1973;12:213-24.

21 Heuer DK, Anderson DR, Knighton RW, Feuer WJ, GresselMG. The influence of simulated light scattering onautomated perimetric threshold measurements. Arch Oph-thalmol 1986;106: 1247-51.

22 WoodJM,WildJM, Smerdon DL, Crews SJ. Alterations in theshape of the automated perimetric profile arising fromcataract. Graefes Arch Clin Exp Ophthalmol 1989;227:157-61.

23 Pelli DG. The visual requirements of mobility. In: Woo GC,ed. Low vision-principles and applications. Springer; NewYork, 1986:134-46.

24 Wood JM, Dique T, Troutbeck R. The effect of artificialvisual impairment on functional visual fields and drivingperformance. Clin Vis Sci 1993;6:563-75.

25 Vos JJ. Disability glare-a state of the art report. CIEJournal1984;3:39-53.

26 Whitaker D, Steen R, Elliott DB. Light scatter in the normalyoung, elderly and cataractous eye demonstrates littlewavelength dependency. Optom Vis Sci 1993;70:963-8.

27 Steen R, Whitaker D, Elliott DB, Wild JM. Effect of filterson disability glare. Ophthalmic Physiol Opt 1993;13:371-6.

28 Elliott DB, Bullimore MA. Assessing the reliability,discriminative ability, and validity of disability glare tests.Invest Ophthalmol Vis SCi 1993;34:108-19.

29 Mangione CM, Phillips RS, Lawrence MG, Seddon JM,Orav EJ, Goldman L. Improved visual function and attenu-ation of declines in health-related quality of life after cata-ract extraction. Arch Ophthalomol 1994;112: 1419-25.

30 Pardhan S, Elliott DB. Clinical measurements of binocularsummation and inhibition in patients with cataract. Clin VisSci 1991;6:355-9.

31 Peli E, Goldstein RB, Trempe CL, Arend LE. Imageenhancement improves face recognition. Noninvasiveassessment of the visual system. Technical Digest Se-ries.Washington DC: Optical Society of America, 1989;7:64-7.

32 Legge GE, Rubin GS, Luebker A. Psychophysics of reading:V. The role of contrast in normal vision. Vision Res1987;27:1165-77.

33 Genensky SM, Berry SH, Bickson TH, Bickson TK. Visualenvironmental adaptation problems ofthe partially sighted:final report (No CPS-100-HEW). Santa Monica MedicalCenter Center for the partially sighted, CA, January 1979.

34 Ball K, Owsley C. Identifying correlates of accident involve-ment for the older driver. Human Factors 1991;33:583-95.

35 Manchester D, Woolacott M, Zedebauer-Hylton N, Marin0. Visual, vestibular and somatosensory contributions tobalance control in the older adult. J Gerontol 1989;44:Ml 18-27.

804

on Septem

ber 22, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.80.9.799 on 1 Septem

ber 1996. Dow

nloaded from