Just over 70% of the population hold a driving licence,and in todays society, driving is viewed by themajority as a right. With an ageing population, however,the proportion of people who fail to meet the visualrequirements is increasing and eye care practitionersneed to be in a position to advise patients during the eyeexamination. Although visual input is essential fordriving, it is a highly complex task, making it difficult toprovide rigorous scientific evidence for the level ofvision required for safe driving. The aim of this article isto review the current visual requirements for drivingand consider their limitations, concentrating on therequirements for visual acuity and visual fields.

Visual acuityThe visual acuity requirements to hold a Group 1licence (ordinary car) are specified in terms of thenumber plate test. The exact wording of the requirementis To read in good light (with the aid of glasses orcontact lenses if worn), a registration mark fixed to amotor vehicle and containing characters 79mm high and50mm wide from 20m1. For pre-2001 number plateswith letters of 79mm by 57mm, the distance is 20.5m.

The number plate test is a crude and frequentlycriticised test of vision but does permit members of thepublic to monitor their own ability to meet the standard,in addition to allowing the police to check vision at theroad side. The number plate test is, however,impractical for clinical situations and occupationalvision standards. It suffers from poor repeatability and isvery dependent on lighting levels and the condition of

the licence plate itself. Geometrically, a letter on amodern number plate subtends 13.4 mins of arc,corresponding to a Snellen acuity of 6/15, butdifferences in font design and letter spacing mean thatthere is no exact Snellen equivalent. The number plateavailable on some test charts should not be taken as thelegal standard since such plates often fail to follow thestandard design. They can, however, be used todemonstrate the approximate level of vision requiredand the detrimental effect of low illumination inpatients you may wish to advised to drive duringdaylight hours only.

Drasdo and Haggerty using statistical modelling statedthat a Snellen acuity of 6/9-2 has an equivalent failurerate to the number plate test.2 They found that thosewith a binocular visual acuity of 6/7.5 had a 99% chanceof passing the test, whereas those seeing 6/18binocularly had only a 6% chance of passing. The modelwas based on a population of only 28 subjects, however,and used the pre-2001 number plate design. A morerecent study by Currie and colleagues again using theold number plate design, found that the binocular visualacuity of those who only just read the number platevaried from 6/9+2 to 6/12-2, with an average of 6/10.3They reported that 26% with a binocular VA of 6/9failed to correctly read the number plate and 34% ofthose with a binocular VA of 6/12 passed. Although thisstudy has its limitations, these pass/fail percentagesshould be noted carefully since they highlight theproblem we face advising patients based on theirbinocular Snellen acuity alone. As eye care

By Catharine Chisholm PhD MCOptom, University of Bradford

40

25/0

1/08

DRIVER

SVISION

CLINICAL OVERVIEW

Visualrequirementsfor driving

25/0

1/08

DRIVER

SVISION

41

CLINICAL OVERVIEW

practitioners, we are regularly calledto give our opinion on a patientslikelihood of meeting the visualrequirements for driving. Some helpon this difficult issue has been issuedby the DVLA in the form of a statementindicating their policy is to assumethat an individual with vision in theirbetter eye of 6/12 or more is likely topass the number plate test, whereas anindividual with vision in their bettereye of 6/24 or less is not likely to pass.Those in-between are consideredborderline and should undergo aformal number plate test at a drivingtest centre4. It is only those with good,or conversely, very poor binocularacuity that we can be relativelyconfident in advising with regard totheir likelihood of meeting the visionrequirements for a Group 1 licence.

In this age of evidence-basedstandards, it is pertinent to considerwhether the current visual acuitystandards can be justified. Drivingresearch is fraught with limitationsassociated with the multi-factorialnature of the task and the difficultiesin accessing accurate accident data.Studies have failed to show anythingmore than a weak link between visualacuity and accident rates. A cut-off ofapproximately 6/12 can be derivedfrom the vision required to read anumber plate at the safe stoppingdistance (three car lengths) whentravelling at 30mph5. This justificationis dubious since cars are not driven at30mph the majority of the time andreading the number plate cannot beconsidered a safety-relevant task.

Visual fieldsThe visual field requirements for aGroup 1 licence are A minimumhorizontal field of vision of 120 andno significant defect within 20 offixation. Additional clarification of asignificant defect is given:

The DVLA accepts the followingcentral loss (Esterman protocol): Scattered single missed points A single cluster of up to threecontiguous points

This is a change from the situation adecade ago when no missed pointswere accepted within the central 20.The change originates from a legalchallenge by an individual missingthree contiguous points. Hesuccessfully argued that his scotomawas no bigger than the blind spot of amonocular driver, who was, of course,allowed to hold a licence.

Significant central loss is defined as:

A cluster of four or more contiguouspoints that is either wholly or partlywithin the central 20 area Loss consisting of both a singlecluster of three contiguous missedpoints up to and including 20 fromfixation and any additional separatemissed point(s) within the central 20area Central loss of any size that is anextension of a hemianopia orquadrantanopia

The functional scoring systemdeveloped by Esterman is the currentgold-standard for testing binocularvisual fields and is used by manynational driving authorities6. Nationalvisual field requirements for fitness todrive differ significantly betweencountries7. There is a European-wideminimum requirement of 120 alongthe horizontal but Germany is the onlyother country to require the centralvisual field to be examined. Thebinocular Esterman visual field test ismost commonly run on the HumphreyVisual Field Analyzer. It is asuprathreshold test that presents asingle, very bright stimulus (10dB) ateach of 120 locations within the visualfield. The subject is required to fixatecentrally and press a button when astimulus is detected. The Estermantest has the advantage of beingrelatively quick compared to otherfull field tests (4-5 minutes for anormal subject), and is by far the mostfrequently available binocular fieldtest in ophthalmology and optometry

clinics. The 120 points in the stimulusarray are spread over a large areaextending approximately 75horizontally, 35 superiorly and 55inferiorly (Figure 1).

Occasionally, an asymptomaticbinocular visual field defect is pickedup by the optometrist during a routinevisual field screening when a patientreaches the age of 40. In a fairproportion of the cases with a negativemedical history, the lesion is likely tobe congenital in origin. If theseindividuals have no driving accidentrecord, the DVLA may reissue a Group

Table 1: Visual requirements for a group I (ordinary car) licence set by the United KingdomDriver Vehicle and Licensing Agency (DVLA)

Visual acuity Binocular requirement onlyRead standard number plate in good light with the aid of glasses orcontact lenses if worn, at 20m

Unaided vision No requirement

Visual fields Binocular requirement onlyA minimum horizontal field of vision of 120 degrees and nosignificant defect within 20 degrees of fixation

Colour vision No requirement

Diplopia Cease driving on diagnosis but acceptable if controlled by spectaclesor patching

Monocularity Individuals with one eye must inform the DVLA. Individuals can driveonce adapted if the meet the acuity and visual field standards

Figure 1: A normal binocular Esterman plotshowing the location of 120 points acrossthe visual field. The Group 1 licence visualfield requirements specify that no pointsmust be missed along the horizontalmeridian over an area subtending 120degrees, and no significant defect withinthe central + 20 degrees

25/0

1/08

DRIVER

SVISION

42

CLINICAL OVERVIEW

1 licence after a period of 12 months ifthe patients neurologist orophthalmologist can provide evidencethat the lesion was caused by anisolated event that is unlikely toreoccur, or a non-progressivecondition, and that there is no otherpathology present that is likely toprogress resulting in further visualfield loss. The DVLA also requires theindividual to undergo an on-roadassessment at a specialist centre toprove that full functional adaptation tothe visual field loss has taken place1.

An example I encountered a fewyears ago was a 60 year old patientwhom I shall refer to as patient A. Astroke at the age of 30 resulted inlesions to the left optic radiations andthe left visual cortex and hence apartial upper right quadrantanopia(Figures 2 and 3). At the time therewere no visual field requirement fordriving a car and patient A wasadvised by their GP to cease driving forone month as a matter of caution.Following this period of time, he drovesafely for 20 years before beingcontacted by DVLA during a review ofmedical records, as a result of which,his license was revoked. Patient A hassince regained his license on the basisof 20 years of safe driving and theregularly drives around the busystreets of central London. Figure 4simulates the effect of the visual fieldloss on a sample driving scene.

Providing evidence to justify thevisual field requirements is fraughtwith difficulty. There is no doubt thatin the limit, binocular visual field lossimpacts driving safety since the limitis blindness, but determining the cut-off point has proved impossible. Adriver with normal or near-normalvisual fields would be expected todetect and identify peripheral andparacentral hazards more quickly thansomeone with impaired fields. Northreviewed the literature on this subjectand reported no significantrelationship between visual field lossand driving safety8. She attributed thisfinding in part to the numerousmethodological limitations of many ofthe papers published in this area,which almost exclusively consideredonly the horizontal extent of theperipheral binocular visual field.Studying the effects of isolatedscotomata is complicated since theyvary in terms of size and location, aswell as depth and aetiology.

One option has been to look atmonocular drivers who all show an

identical binocular visual field defectresulting from the physiological blindspot, approximately two metres indiameter at a viewing distance of 20metres. Refixation during drivingoccurs in an area smaller than theblind spot9. Theoretically this couldaffect the perception of, for example, ahand signal from another driver withinthe monocular drivers blind spot9,10.However, no increase in accident ratehas been found for monoculardrivers11-13.

Since real scotomas are so difficultto study, researchers have examinedthe effect of simulated scotomata innormal subjects14-17. Such studiesallow functional performance to beassessed in a sufficiently large group ofsubjects with identical field defects.However, only studies simulatingconstriction of the visual field havebeen undertaken, because of thepractical difficulties in simulatingparacentral visual field loss that moveswith fixation.

Applying the findings of suchexperiments using simulatedscotomata to real visual field loss isproblematic. The main limitation is

the inability to account foradaptation/compensatory eyemovements18, which can develop insome individuals over time and mayresult in safe driving behaviour19.

In addition to the difficulties inproviding justification for the visualfield requirements, the gold standardvisual field test for driving, thebinocular Esterman test, is not reallyup to the job. By testing each of the 120points at a single very extremesuprathreshold level, only the deepestscotomas are revealed. The stimulusarray was not designed to relate todriving, but rather personal mobility,with more locations tested in areasjudged to be of importance fornegotiating obstacles when walking.The sampling density is particularlysparse within the central 20, with 12locations examined above the midlineand 22 below, and no stimuli within7.5 of fixation. The EstermanEfficiency Score (EES) is simply basedon the percentage of stimuli seen andis therefore biased towards the inferiorvisual field due to the greater densityof stimuli within this area.

A typical view when driving isshown in figure 5. It highlights theareas of functional importance, inparticular the central 20 field andthe region along the horizontalmeridian. The lower visual field,beyond about 10, is not particularlyrelevant to driving as it overlays thedashboard. The driver will be aware ofthe presence of the dashboard display,but will only be able to resolve detailwhen fixation is momentarilytransferred to the display area itself.The visual field to the drivers side ispartially obscured by the window A-pillar. The upper visual field isobscured within 30-45 by the rear-view mirror20, although the exactlocation of these obstructions variesbetween cars and with the position of

Figure 2: Monocular HFA 24-2 SITA Standard plots for a patient A showing a bilateralpartial upper right quadrantanopia following a stroke 30 years previously

Figure 3: Binocular Esterman plot forpatient A (also showing ring scotoma fromspectacles)

25/0

1/08

DRIVER

SVISION

43

CLINICAL OVERVIEW

the driver21. An object such as a cyclistor pedestrian moving from the sideinto the drivers path is in danger ofcollision. Early recognition of suchmoving objects is important but only25% of the measured points fall withinthe eloquent area because theEsterman overrates the lower field. Asingle cluster of three contiguouspoints in the upper visual field wouldnot prevent an individual fromholding a Group 1 licence, despite thefact that such a pattern of missedpoints would probably represent asubstantial paracentral visual fieldloss due to the scarcity of points testedin this area.

Examining driversOther than ensuring we are familiarwith the visual requirements fordriving, what more should we bedoing for our patients who drive?Previous surveys have shown thatabout 10% of individuals tested failthe number plate test due to anincorrect prescription. Even a smallincrease in monocular visual acuity inone eye may be enough to allow apatient to meet the visual acuityrequirements under binocularconditions. Checking for binocularproblems is also important, to ensurethat where possible, binocular acuityis at least as good, if not better, thanthe monocular acuity in the better eye.With the latest advice from the DVLAbeing that those with less than 6/12 intheir better eye should undergo thenumber plate test, perhaps we shouldbe giving out leaflets explaining theself-assessment process and listingcontact details for local driving testcentres. Some practitioners may evenconsider mounting a number plateoutside their practices where practical.

With regard to visual fields, shouldwe be performing some form ofperipheral visual field check in alldrivers (the majority of patients)?Binocular visual fields are testedroutinely in drivers in 55% ofEuropean countries. Not all practiceshave access to the binocular Estermantest but any practitioner can perform agood quality gross perimetry orconfrontation test combined with acentral visual field screening. Like anyvisual field test the outcome of theEsterman test is subject to learningeffects, fatigue and concentration.Individuals who fail should firstly begiven the opportunity to repeat the teston a different day.

Monocular visual field loss is

thought to occur in between 3-5% ofthe population with around two-thirdsof such individuals unaware of theirdeficit11,22,23. Binocular visual fieldloss (not necessarily overlapping) wasreported in 1.1% of subjects in onestudy of 20,000 drivers11, but this risesto around 5% in patients with ahistory of cerebrovascular accident24.My own experience of recruitingneurological patients for drivingstudies has taught me that the classicquadrantanopia or hemianopia sooften associated with cerebrovascularaccidents involving the visualpathway, is more often than not, apartial defect and in 10-15% of cases,does not involve the central 30 at all.A central visual field test is therefore

inadequate in such individuals.

Managing patients who fail tomeet the requirementsIt is not a pleasant job explaining to apatient that they no longer meet thevisual requirements for driving. Aspointed out earlier in this article, itmay not be possible to issue firmadvice with respect to visual acuity.Patients should be advised to informthe DVLA, their GP and theirinsurance company. Insurance coverbecomes void in many cases if anindividual fails to meet the medicalrequirements to hold a driving licence.It is the responsibility of the driver toensure they meet all standards,although some insurance companies

Figure 4: Simulated effect of visual field loss on a real driving scene for patient A

Figure 5: A typical scene when driving showing the functionally-important areas of thevisual field

25/0

1/08

DRIVER

SVISION

44

are more lenient than others. A patientwho fails to meet the visualrequirements but makes it clear theywill not inform the DVLA or ceasedriving, should be treated withcaution. Record your discussion withthe patient in their notes and ifpossible, ask them to sign your notes.You are advised to seek the advice ofthe Association of Optometrists beforetaking any further action. Althoughpatient confidentiality is an issue, theCollege of Optometrists code ofconduct indicates that where thepublic are put at risk, you have aresponsibility to report the individual.An occasional, short distance driverwho is likely to self-regulate isprobably less of a risk to public healththan a bus driver. Those of us who aredrivers also have a responsibility toreport individuals if we actually seethem driving unsafely.

The futureHarmonisation at European Union(EU) or even EU/US level is likely atsome point over the next few years.Recent reports by both the EU EyesightWorking Group (May 2005) and theInternational Council ofOphthalmology (ICO) Feb 200625recommended a visual field standardof 120 along the horizontal meridianwith no defect within 20 of fixation.Interestingly, both reports opposebinocular visual field tests such as theEsterman because of the lack offixation monitoring, fusional problemsand the large, bright stimulus, all ofwhich lead to the test being toolenient. The fusing of monocularmeasurements has been recommendedas an alternative26, although this is notwithout its problems27.

Safe driving requires a combinationof good vision, adequate visual field,ability to spread attention over thefield of view and motor skills. There isno doubt that current clinical tests donot in anyway reflect this. It remains tobe seen whether more suitable teststhat more closely relate to the drivingtask, will be developed. Attempts sofar have been relatively fruitless. In themeantime, all we can do as eye carepractitioners is work to the currentvisual requirements and keep ourpatients informed as best we can.

About the authorCatharine Chisholm is a lecturer inOptometry at Bradford University(c.m.chisholm@bradford.ac.uk). Herresearch covers a range of topics, often

employing psychophysical techniquesto investigate visual performance inrelation to occupational issues.

References1. At a glance, DVLA website

http://www.dvla.gov.uk/medical/2. Drasdo N and Haggerty CM

(1981) A comparison of the Britishnumber plate and Snellen vision testsfor car drivers. Ophthalmic andPhysiological Optics; 1: 39-54

3. Currie Z, Bhan A and Pepper I(2000) Reliability of Snellen chartsfor testing visual acuity for driving:prospective study and postalquestionnaire British MedicalJournal; 321: 990-2

4. AOP websitehttp://www.aop.org.uk/uploaded_files/11-drivers_sep_06.pdf

5. Sheedy JE (1980) Police visionstandards. Journal of Police Scienceand Administration; 8: 275-285

6. Esterman B (1982) Functionalscoring of the binocular field.Ophthalmology; 89: 1226-1234

7. Charman WN (1997) Vision anddriving - a literature review andcommentary. Ophthalmic andPhysiological Optics; 17: 371-391

8. North RV (1985) Therelationship between the extent ofvisual field and driving performance -a review. Ophthalmic andPhysiological Optics; 5: 205-210

9. Mourant R and Rockwell T (1970)Mapping eye movement patterns tothe visual scene in driving: anexploratory study. Human Factors;12: 81-87

10. Westlake W (2001) Is a one-eyed racing driver safe to compete:formula one (eye) or two? BritishJournal of Ophthalmology; 18: 619-624

11. Johnson CA and Keltner J (1983)Incidence of visual field loss in20,000 eyes and its relationship todriving performance Archives ofOphthalmology; 101: 371-375

12. Kite CR and King JN (1961) Asurvey of factors limiting the visualfields of motor vehicle drivers inrelation to minimum visual fields andvisibility standards. British Journalof Physiological Optics; 18: 85-107

13. Leismaa M (1973) Theinfluence of drivers vision in relationto his driving ability. Optician; 166:10-13

14. Wood JM, Dique T, andTroutbeck R (1993) The effect ofartificial visual impairment onfunctional visual fields and driving

performance. Clinical VisionScience; 8: 563-575

15. Wood JM and Troutbeck R(1992) Effect of restriction of thebinocular visual field on drivingperformance. Ophthalmic andPhysiological Optics; 12: 291-298

16. Wood JM and Troutbeck R(1994) Effect of visual impairment ondriving. Human Factors; 36: 476-487

17. Wood JM and Troutbeck R(1995) Elderly drivers and simulatedvisual impairment. Optometry andVision Science; 72: 115-124

18. Lovsund P, Hedin A and TornrosJ (1991) Effects on drivingperformance of visual field defects: Adriving simulator study. AccidentAnalysis and Prevention; 23: 331-342

19. Schulte T, Strasburger H,Muller-Oehring E, Kasten E, and SabelBA (1999) Automobile drivingperformance of brain-injured patientswith visual field defects. AmericanJournal of Physical Medicine andRehabilitation: 78: 136-142

20. Schiefer U, Hofer R, Vischer PM,and Wilhelm H (2000)Perimetriebefund undfahrtauglichkeit. Ophthalmologe; 97:491-497

21. Vargas-Martin F and Garcia-Perez MA (2005) Visual fields at thewheel. Optometry and VisionScience; 82: 675-681

22. Bentsson B and Krakau CET(1979) Automatic perimetry in apopulation survey. ActaOphthalmologica; 57: 929-937

23. Keeney AH (1974) Significanceof visual problems in Pennsylvaniadrivers. Pennsylvania Medicine; 77:49-51

24. van Newkirk MR, Weih L,McCarty CA and Taylor HR (2001)Cause-specific prevalence of bilateralvisual impairment in Victoria,Australia - The visual impairmentproject. Ophthalmology; 108: 960-967

25. International Council ofOphthalmology websitewww.icophth.org/standards

26. Crabb DP, Fitzke FW, HitchingsRA and Viswanathan, AC (2004) Apractical approach to measuring thevisual field component of fitness todrive. British Journal ofOphthalmology; 88:1191-1196

27. Chisholm CM, Rauscher, FG,Crabb, DC et al. (2007) Assessingvisual fields for driving in patientswith paracentral scotomata. BritishJournal of Ophthalmology; doi10.1136

CLINICAL OVERVIEW