ARTICLE

Effects of a blue ligh

t–filtering intraocularlens on driving safety in glare conditions

Rob Gray, PhD,WarrenHill, MD, Brooke Neuman,MS, Diane Houtman, OD,MBA, Richard Potvin, OD

Q

P

816

2012 A

ublished

PURPOSE: To evaluate whether the previously established benefit of blue light–filtering intraocularlenses (IOLs) when driving in glare conditions is maintained in patients previously implanted witha blue light–filtering toric IOL.

SETTING: Department of Applied Psychology, Arizona State University, Mesa, Arizona, USA.

DESIGN: Comparative case series.

METHODS: The study comprised patients with a blue light–filtering toric IOL (test IOL) or an ultra-violet (UV)-only filtering nontoric IOL (control IOL). All patients had good visual acuity and a validdriver’s license. While wearing best spherocylindrical correction, patients performed left-turnmaneuvers in front of oncoming traffic in a driving simulator. The safety margin was defined asthe time to collision less the time taken to turn at an intersection with oncoming traffic.Measures were repeated with a glare source simulating low-angle sun conditions (daytime driving).

RESULTS: Of the 33 evaluable patients, 18 had a test IOL and 15 had a control IOL. In the presenceof glare, patients with test IOLs had significantly greater safety margins (mean 2.676 seconds G0.438 [SD]) than patients with control IOLs (mean 2.179G 0.343 seconds) and significantly lowerglare susceptibility (P<.05). In no-glare and glare conditions, patients with test IOLs hadsignificantly lower glare susceptibility than patients with control IOLs.

CONCLUSION: The blue light–filtering toric IOL produced a significantly greater reduction in glaredisability than the UV-only filtering nontoric IOL and increased the ability of drivers to safely executeleft turns in low-sun conditions.

Financial Disclosure: Dr. Houtman is an employee of Alcon Laboratories, Inc. No other author hasa financial or proprietary interest in any material or method mentioned.

J Cataract Refract Surg 2012; 38:816–822 Q 2012 ASCRS and ESCRS

Gray et al.1 recently reported that intraocular lenses(IOLs) with a blue light–filtering chromophore(Acrysof Natural IOL, model SN60AT, AlconLaboratories, Inc.) significantly reduced driving im-pairment in glare conditions. This was based on differ-ences in the decision to turn left in the face of oncomingtraffic during a driving simulation. Patients with thechromophore had statistically significantly highersafety margins under glare conditions than patientswith an ultraviolet (UV)-only filtering IOL.

The 2 IOLs tested in this previous study were of thesame material and design (Acrysof spherical correc-tion nonaspheric single-piece IOLs), with the only dif-ference being the blue-filtering chromophore. The testmethods usedwere previously documented in a studythat did not involve IOLs2; results in this earlier studyare the first to objectively show a relationship betweendisability glare and driving performance.

SCRS and ESCRS

by Elsevier Inc.

Establishing the effectiveness of the blue-filteringchromophore with spherical IOLs in improving driversafety in this previous study was important. However,more patients are now having implantation ofaspheric IOLs to reduce spherical aberration and oftoric IOLs to reduce postoperative astigmatism. It isuseful, therefore, to establish whether the benefitsidentified in the earlier study are maintained whenthese alternative IOL designs are implanted. It isconceivable that the effect of the difference betweenIOL designs changes the relative importance of theblue-filtering chromophore.

More than 40% of eyes are estimated to have 0.75 di-opter or more of corneal astigmatism.3,4 With the cur-rently available range of astigmatic correction in toricIOLs, most of these eyes would benefit from implanta-tion of a toric IOL, although at present the percentageof toric IOLs implanted is significantly lower than 40%.

0886-3350/$ - see front matter

doi:10.1016/j.jcrs.2011.11.047

817BLUE LIGHT–FILTERING TORIC IOLS AND DRIVING SAFETY

The current study was designed to determinewhether the significant increase in safety marginfrom the blue light–filtering chromophore in theAcrysof material observed previously with sphericalIOLs was evident with toric IOLs (aspheric or nona-spheric) as well. It was decided that testing would beperformed in the best-corrected state (ie, includingany required spherocylindrical correction over the to-ric IOL). This removes the element of surgical variabil-ity and helps establish that any effects on drivingsafety are related to the chromophore rather than touncorrected refractive error.

The study was designed to include 2 groups ofpatients. The first was a control group of patientswith a UV-only filtering spherical nonaspheric IOL(Acrysof SA60AT), similar to that used in an earlierstudy.1 The second was a test group of patients witha toric IOL that has a blue-filtering chromophore;this group included toric IOLs with and withoutaspheric correction. The hypothesis was that the con-trol group and test group in the current study wouldperform similarly to the control group and test groupin the previous study. Stated another way, it wasexpected that (1) the toric IOLs with the blue light–filtering chromophore would provide a statisticallysignificantly higher safety margin than the UV-onlyfiltering spherical IOLs in this study and that (2)the performance of the toric IOLs with the bluelight–filtering chromophore would be similar to theperformance of the spherical IOLs with the same chro-mophore in the previous study.1

In this study, we evaluated the benefit of blue-lightfiltering and astigmatic correction technology on real-world tasks by studying the ability of 2 groups of pseu-dophakic patients to drive safely in the presence ofsimulated glare. Performancewas assessed in a drivingsimulator by measuring the safety margin, defined asthe difference in the time to complete a left turn and

Submitted: June 10, 2011.Final revision submitted: November 14, 2011.Accepted: November 21, 2011.

From the School of Sport and Exercise Sciences (Gray), Universityof Birmingham, Birmingham, United Kingdom; East Valley Ophthal-mology (Hill) and the Department of Applied Psychology (Neuman),Arizona State University, Mesa, Arizona, USA; Alcon Laboratories,Inc, (Houtman), Fort Worth, and Science in Vision (Potvin),Burleson, Texas, USA.

Sponsored by Alcon Laboratories, Inc., Fort Worth, Texas, USA.

Corresponding author: Rob Gray, PhD, School of Sport andExercise Sciences, College of Life and Environmental Sciences, Uni-versity of Birmingham, Edgbaston, Birmingham B15 2TT, UnitedKingdom. E-mail: r.gray.2@bham.ac.uk.

J CATARACT REFRACT SURG

the time before colliding with oncoming traffic. Mea-surements were repeated in the presence of a glaresource (halogen light) simulating low-angle sun condi-tions. The safety margin has been reported1,2 and pro-vides a valuable assessment of driving performancebecause the measure has been shown to correlate wellwith simple clinical tests of glare susceptibility (eg,visual acuity measurements in the presence of glare).

PATIENTS AND METHODS

Study Population

This cross-sectional prospective study comprised patientswho had unilateral or bilateral implantation of the blue-lightfiltering Acrysof Natural toric IOL (Models SN6ATT orModel SN60TT, Alcon Laboratories, Inc.) (test group) orthe UV-only filtering Acrysof IOL without astigmatic correc-tion (model SA60AT, Alcon Laboratories, Inc.) (controlgroup) from January 2008 through April 2010. The IOLmodels were selected for comparison because they are basedon the same Acrysof platform: the Acrysof UV-absorbingfoldable single-piece posterior chamber IOL. The key differ-ences between the IOLs are the presence of a blue light–filter-ing chromophore and the correction for corneal astigmatismin the test group. In addition, some of the toric IOLs includedcorrection for asphericity. Driving performance testing wasperformed monocularly so patients with bilaterally toricIOLs or spherical IOLs were tested using the eye that hadthe best logMAR distance-corrected visual acuity. This wasthe same methodology used in the earlier studies.1,2 Themajority of driving simulators use monocular test condi-tions; studies5,6 have established that there is no significantdifference in driving performance between binocular andmonocular conditions.

A power analysis was performed to determine the neces-sary sample size to detect a statistically significantly highersafety margin in patients with the IOL with the blue light–filtering chromophore, based on previously collected data.With an a of 0.05, a 1-tailed test of differences and an ex-pected difference in safetymargin of 0.5 seconds (with a stan-dard deviation of 0.5 seconds), a sample size of 13 patientswas determined sufficient to provide a power of 0.80. Actualsample sizes below were somewhat higher because dropoutis expected in driving simulation studies due to the potentialfor simulator motion sickness.

Study patients were identified by chart reviews. Allpatients were required to possess a valid driver’s license atthe time of the study. The studywas approved by a central in-stitutional review board and the Arizona State University Re-search Ethics Committee. Participants provided informedconsent after receiving an explanation of the nature andpossible consequences of the study. Patients knew what thestudy involved but were effectively masked because theywere not aware of any grouping (test IOL versus controlIOL). All participants were paid a fixed amount for their time.

Patients with vestibular pathology (vertigo, motion sick-ness); ocular pathology, including diabetic retinopathy, opticatrophy, retinal degeneration, or corneal abnormality; a his-tory of ocular surgery or amblyopia; or a corrected distancevisual acuity worse than 0.30 logMAR were excluded fromthe study. Good post-cataract visual acuity was required,eliminating the potential that secondary cataract would in-fluence the results. Outside of the noted physical andocular health screening and the possession of a valid license,

- VOL 38, MAY 2012

818 BLUE LIGHT–FILTERING TORIC IOLS AND DRIVING SAFETY

no other criteria were applied. The recruitment criteria weredesigned to reduce the likelihood of systemic effects (eg,dementia) or ocular effects (eg, neuropathy) from affectingresults.

Study patients attended a screening visit and 1 follow-upstudy visit. Screening visits involved routine clinical testing,including measures of monocular and binocular correcteddistance (4 m), intermediate (60 cm), and near (40 cm) log-MAR visual acuity; manifest refraction at 4 m (distance)and 60 cm (intermediate); and a slitlamp and dilated fundusexamination. The screening visit ensured patients met eligi-bility criteria. Only patients who met study criteria werescheduled for a follow-up study visit, performed on a differ-ent day, for driving performance assessment.

The variables for the assessment of driving performanceincluded safety margin, glare susceptibility ratio, and per-centage of turns made in front of the oncoming vehicle.The comparison between the 2 groups (test and control)was performed separately for measures obtained under theglare condition, under no-glare condition, and for the differ-ences in the measures between the glare condition and theno-glare conditions (calculated as glare minus no-glare).The latter provided an understanding of the effect of glarebecause it captured the difference in driving performancemeasures between the 2 conditions (glare and no glare).These comparisons, as well as the comparison with respectto the glare susceptibility ratio,were performedusing a linearmodel that included a fixed-factor treatment.Wilcoxon rank-sum tests were used to compare the number of collisionsexperienced under each driving condition (no glare, glare)between the test group and the control group. A McNemartest was used to test for differences in the relative proportionof patients who had 1 or more collisions under glare or no-glare conditions. All statistical tests were 2 sided and useda level of significance of a Z 0.05. The statistical analyseswere performed using SPPS software (version 17, SPSS,Inc.). To allow further comparison with the previous study,1

the data were also converted into glare susceptibility ratios,defined as the ratio of performance with glare absent to per-formance with glare present. More specifically, glare suscep-tibility ratio was defined as follows:

No glare: Mean TTC TTCT at turn onsetWith glare: Mean TTC TTCT at turn onset

Driving Simulator

The driving simulator was composed of 2 main compo-nents: (1) a steeringwheel mounted on a table top and pedals(Wingman Formula Force GP, Logitech) and (2) 3 Dell liquidcrystal display monitors (19 inch). The monitors wereviewed from a distance of 57 cm. The 3 monitors were posi-tioned side by side to create a driving scene that subtendeda total of 130 degrees horizontal � 30 degrees vertical ofvisual angle. The visual scene was rendered and updatedat a rate of 60 Hz by DriveSafety driving simulator softwarerunning on 4 computers (Dell Optiplex GX270). Driver be-havior in this model of simulator has been highly correlatedwith behavior during real-world driving.A

The system simulated glare from low-angle sun condi-tions using a large beam splitter (18 cm � 18 cm) orientedat a 45-degree angle to the participant’s line of sight. Thisbeam splitter was used to superimpose a 0.8-degree diameterlight source on the driving scene at a height of 20 degreesabove the simulated road. The light source was a SonnemanLibra desk lamp (model 1650.55) equipped with a bi-pin

J CATARACT REFRACT SURG

50-watt halogen bulb with a power of approximately150 lux (920 lumens) measured at the eye point. For safetyreasons, no attempt was made to accurately simulate the ac-tual luminance level of the sun. As in the previous study,2 thelight source was chosen to approximate the high relativecontrast between the road surface and the sun. The meanroad luminance in the simulator was approximately 3.1 lux(19.0 lumens). This setup allowed real-world simulation ofglare during daytime driving from low-angle sun conditions(as in sunset orwhen driving toward the sun). In the no-glarecondition, the light source was turned off. Testing was uni-lateral, and all participants wore an eye patch over the eyenot being tested. In addition, all participants were correctedfor best intermediate distance (distance of the computermonitors of the driving simulator) with spectacle lensesand wore these lenses for all conditions in the driving study.Intermediate correction included both sphere and cylinder,as required.

Procedure

At the study visit, the patients performed a left-turnmaneuver at a simulated intersection in the presence of anoncoming car. Judging that a left turn can be completedsafely requires a driver to subtract the estimated time to com-plete the turn (TTCT) from the estimated time to collision(TTC) with the oncoming vehicle (the safety margin). Asafe left turn requires that the driver initiate the turn whenthe TTC is comfortably larger than the TTCT or wait untilthe oncoming vehicle has passed through the intersection.Thus, the variable TTC� TTCTwas used as the main depen-dent measure for left-turn performance. This variable givesthe amount of time that would elapse between the instantthe left-turning driver vehicle cleared the intersection andthe oncoming vehicle entered the intersection; that is, thesafety margin.2 A similar variable has been used previouslyfor the design of left-turn warnings.B The speed and initialdistance of the oncoming vehicle was varied to presentdrivers with different TTC � TTCT values ranging frompositive (ie, turn can be completed safely in front of theoncoming car) to negative.

A rural 2-lane driving environment was simulated. Theroad was textured, and each trial involved a 4400 m simu-lated drive. Speed limit signs of 55 miles per hour (mph)were posted throughout the drive, and participants were in-structed to obey the speed limit at all times. The roadwaywas made up of 5 4-way intersections with traffic lightsand 1 T-junction intersection with traffic lights connectedwith straight road segments. Participants were instructedto make a left-hand turn at each of the 4-way intersectionsand a right-hand turn at the T-junction. The right-handturn was included to create a driving route that did not dou-ble back on itself after 3 left turns. Traffic lights were alwaysgreen as the participant’s car approached the intersection. Atmost intersections, there was an oncoming car. The initialdistance and speed of the approaching car was varied to cre-ate different values of TTC � TTCT. All oncoming cars weredark graywith a contrast of 9% relative to the gray road. Thisparadigm is similar to the low-contrast vehicle conditions re-ported in the previous study by Gray and Regan2 and wasidentical to that used by Gray et al.1

The value of TTCT for each participant was determined asfollows: Before the experimental trials began, each partici-pant completed 10 practice left turns with no oncomingcars. Participants were instructed to execute the turn asquickly as possible without losing control of the simulated

- VOL 38, MAY 2012

819BLUE LIGHT–FILTERING TORIC IOLS AND DRIVING SAFETY

vehicle. Time to complete the turn was defined as the timebetween when the steering movement began and when therear bumper of the participant’s vehicle was completelythrough the intersection. For each participant, the minimumTTCT of these 10 practice trials was used; thus, the marginscalculated would slightly overestimate the safety margin un-der most conditions.

The no-glare and glare driving conditions each comprised4 trials of 5 left turns for a total of 20 left turns per condition.A total of 6 values of TTC � TTCT were used ranging from3.0 to 1.0 seconds. These are approximate TTC � TTCTvalues. Because the participants’ approach speed to the inter-section varied, there was some variability in the actualTTC � TTCT values. The safety margin (TTC � TTCT) wasmeasuredwhen the participant was 3 seconds from the inter-section because it has been shown that this is roughly the in-stant in time when a driver decides whether it is safe to turnin front of an oncoming vehicle.7 Each of these 6 TTC� TTCTvalues was repeated 3 times. There were also 2 left turns inwhich there was no oncoming car. Immediately beforeeach driving session, participants completed 2 practice trialswith no glare. The order of presentation of the differentTTC � TTCT values was randomized. For both the glareand no-glare conditions, participants were instructed to“execute a left turn in front of any oncoming vehicles ifthey felt it was safe to do so.” Participants altered betweenglare and no-glare runs, with the order of these 2 conditionscounterbalanced across participants. In addition to the safetymargin, the glare susceptibility ratio (expressed as a ratio ofsafety margin without glare to the safety margin with glare)was assessed.

RESULTS

Participant Demographics

Table 1. Age,MRSE, and time since implantation by IOL group.

Parameter

Treatment Group

Test Control

Mean SD Mean SD

Age (y) 67.39 9.91 68.24 7.94MRSE (D) �0.43 0.90 �0.44 0.67Time since implantation (y) 1.29 0.61 1.42 0.64

MRSE Z manifest refraction spherical equivalent

Of the 35 patients screened, 2 did not complete thestudy because they had simulator motion sicknessduring the driving performance testing. Therefore,evaluable data were available for 33 patients. Of these,18 had blue light–filtering IOLs, including 12 patients(9 men, 3 women) with a mean age of 67.8 years G11.5 (SD) who had previous implantation of theaspheric toric IOL and 6 patients (2 men, 4 women)with a mean age of 66.6G 6.3 years who had previousimplantation of the spherical toric IOL. Fifteen patients(6 men, 9 women) with a mean age of 68.0G 7.9 yearshad previous implantation of the UV-only filteringnontoric IOL.

Preliminary analyses of safety margins underno-glare and glare conditions showed no statisticallysignificant difference between the aspheric and nona-spheric toric IOLs (PO.23) so they were combinedinto 1 test group for reporting purposes. Given thephotopic conditions of the test setup, it is not surpris-ing that asphericity (a benefit in dim-light large-pupilsituations) was not a significant factor. There was alsono statistically significant difference by magnitude ofcylinder corrected (T3, T4, or T5). Thus, the 2 groupsof data analyzed were for the test group (blue light–filtering chromophore toric IOL) and the control group

J CATARACT REFRACT SURG

(UV-only filtering nontoric IOL). Table 1 shows theage, manifest refraction spherical equivalent, andtime since implantation by IOL group. There wereno significant differences in any of the 3 parameters be-tween the 2 groups.

Driving Performance Assessment

Figure 1 shows themean left-turn safetymargins (ie,TTC� TTCT at the instant of initiating a turn) for eachparticipant under glare and no-glare conditions. Thesemeans only include data from turns in which thedriver started the left turn before the oncoming carhad passed through the intersection. Table 2 showsthe safety margin results under the no-glare and glareconditions. This analysis found a significant maineffect of glare presence because overall, patients inboth groups (test and control) initiated left turns ata significantly lower safety margin with glare thanwhen the glare source was switched off. The differencewas not statistically significant in the test group(PO.05) but was statistically significant in the controlgroup (P!.05). In addition, although there was no dif-ference in the safety margin between groups under theno-glare condition, in the presence of glare, patients inthe test group had a significantly greater safetymarginthan patients in the control group (P!.05). Theseresults indicate that the test group was significantlyless affected by glare than the control group.

Figure 2 shows the glare susceptibility ratio results.Patients in the test group had a significantly lowerglare susceptibility ratio (P!.001), and hence wereless susceptible to the effect of glare, than the patientsin the control group. Table 3 shows the results for theglare susceptibility ratio in the 2 study groups.

Also included in Figure 2 are the data for theAcrysof SN60AT (blue light–filtering spherical nona-spheric) and Acrysof SA60AT (UV-only filteringspherical nonaspheric) IOLs from the previous study.Results in the test group in the present study and theAcrysof SN60AT IOL in the previous study showedthat patients with toric IOLs had a significantly lowerglare susceptibility ratio (P!.05), and hence were less

- VOL 38, MAY 2012

Figure 1.Mean left-turn safety margin (ie, TTC� TTCT) in the glareand no-glare conditions in the 2 IOL groups. Error bars show meanG 1 SD. Closed figures in toric data are nonaspheric IOLs; open fig-ures are aspheric IOLs (as can be seen results were not significantlyaffected by the presence or lack of asphericity) (IOL Z intraocularlens; TTC Z estimated time to collision; TTCT time to completethe turn).

820 BLUE LIGHT–FILTERING TORIC IOLS AND DRIVING SAFETY

susceptible to the effect of glare, than patients with theAcrysof SN60AT IOLs (Figure 2 and Tables 3 and 4).Table 4 shows the glare susceptibility ratio results forthe test IOL in the present study and the AcrysofSN60AT IOL in the previous study. The difference be-tween the 2 blue light–filtering IOLs, while statisticallysignificant, was lower than the difference between theblue light–filtering IOLs and UV-only filtering IOLs.

DISCUSSION

Consistent with previous research,2 we found that thedrivers in the present study executed left turns witha significantly smaller safety margin in the presenceof glare (by 0.50 seconds on average). However, thiseffect was significantly mediated by the type of IOLimplanted. For drivers with the Acrysof Natural toricIOL, the reduction in the left-turn safety margin asso-ciated with glare (mean 0.24 sec) was significantly less

Table 2. Safety margin for the no glare, glare, and the difference betwe

Group

Safety Marg

No Glare Glare

Mean SD Mean

Test (n Z 18) 2.916 0.473 2.676Control (n Z 15) 2.971 0.218 2.179P value† .68 .001

*Between conditions†Between groups

J CATARACT REFRACT SURG

than the reduction in the left-turn safety margin fordrivers with the UV-only filtering nontoric IOL(mean 0.79 seconds). Because the Natural toric IOLshave a blue light–filtering chromophore that is notpresent in the UV-only filtering nontoric IOL, theresults in the present study are consistent with thehypothesis that filtering blue light reduces the effectof disability glare.1 These results are also consistentwith the psychophysical experiments of Hammondet al.,8 in which glare disability was reduced in eyesimplanted with an IOL with a blue light–filteringchromophore.

A comparison of glare susceptibility ratios in thepatients with the toric IOLs in the present study andpatients with the Acrysof SN60AT IOL in our previousstudy1 showed that the effect of glare on left-turnsafety margin was significantly lower with the toricIOLs. In addition, 2 patients with the toric IOLs inthe present study had glare susceptibility ratio valuesless than 1.0 (ie, drove more safely under glareconditions). In our 2 previous studies,1,2 all patients(including patients with normal vision who were 19to 29 years of age) had glare susceptibility ratio valuesgreater than 1.0. Because both the Acrysof toric IOLand the Acrysof SN60AT IOL have the blue light–fil-tering chromophore, the difference was not due toblue-light filtering. Instead, the present findings sug-gest that the astigmatic correction in the toric IOLs pro-vides additional driving safety benefits to the bluelight–filtering chromophore. The patients had bothsphere and cylinder corrected for best intermediatevision; therefore, the difference might be related tothe internal (IOL) versus external correction of astig-matism, the geometry of toric versus spherical IOLs,differences in the IOL power implanted, or differencesin corneal astigmatism between groups. The data inthe current study are insufficient to draw a definitiveconclusion. To our knowledge, no previous researchhas examined the effects of astigmatism on theprocessing of motion-in-depth and time to collision;further research into this topic appears warranted.

en the glare and no-glare conditions by IOL group.

in (Seconds)

P Value

Difference(Glare – No Glare)

SD Mean SD

0.438 �0.240 0.294 .12*0.344 �0.792 0.301 !.01*

!.001 d

- VOL 38, MAY 2012

Figure 2. Glare susceptibility ratios for drivers in the 2 IOL groups.Error bars shows mean G 1 SD. Closed circles in toric data denoteaspheric IOLs, open circles are nonaspheric. Closed and open trian-gles show comparable data for the Acrysof SN60AT (blue light–filtering) and Acrysof SA60AT (UV-only filtering) IOLs from Grayet al.1 (IOL Z intraocular lens; UV Z ultraviolet light).

Table 4. Glare susceptibility ratio (in the natural toric IOLgroup) in the current study and in the SN60AT IOL groupfrom Gray et al.1 (between-group P!.05).

Group

Glare Susceptibility Ratio

Mean SD Min Max

Test (n Z 18) 1.093 0.092 0.971 1.330SN60AT (n Z 17) 1.196 0.159 1.050 1.713

821BLUE LIGHT–FILTERING TORIC IOLS AND DRIVING SAFETY

As noted earlier, the difference between the AcrysofNatural toric IOL and the Acrysof SN60AT IOL (thedifference being astigmatism correction) was muchlower than the difference between the blue light–filter-ing and UV-only filtering IOLs (the difference beingthe chromophore). This supports the notion that theblue-light filtration is the primary factor in the im-provement in driving safety under glare conditions.

The pattern of results for the other driving variablesanalyzed shows some effects that were consistent withthe analysis of left-turn safety margin and glare sus-ceptibility ratio. In the presence of glare, the likelihoodof a collision between drivers with the test IOL wassignificantly lower than for drivers with the controlIOL.

What do the results of the present study mean foron-road driving safety? In the glare condition, thesafety margin used for left turns was, on average.0.55 second longer for patients with the blue light–filtering toric IOL than for patients with the UV-onlyfiltering nontoric IOL and was 0.2 second longer forpatients with the blue light–filtering toric IOL thanfor patients with the Acrysof SN60AT IOL. Although

Table 3. Glare susceptibility ratio by IOL group (between-groupP!.001).

Group

Glare Susceptibility Ratio

Mean SD Min Max

Test (n Z 18) 1.093 0.092 0.971 1.330Control (n Z 15) 1.392 0.222 1.134 1.975

J CATARACT REFRACT SURG

on the surface these may seem like relatively small dif-ferences, this amount of time is substantial in driving.If a driver executes a left turnwith too short a temporalmargin, he or she leaves less time/distance to respondto unpredictable events, such as the oncoming vehiclespeeding up, a pedestrian stepping into the road, or anunexpected wet or icy patch. When driving at 55 mph,the driver’s vehicle will travel a distance of roughly13.5 m (44.3 feet) in 0.55 second and 4.9 m (16 feet) in0.2 second. This extra time/distance to respond canmean the difference between an accident and safe pas-sage. Consistent with this idea, in-car warning systemsdesigned to reduce collisions typically reduce driver’sbrake reaction time by roughly 0.2 to 0.4 second.9–11

The importance of fractions of a second in driverresponse can be further appreciated in the context of2 recent and significant developments in drivingsafety. The first is Center High Mounted Stop Lamps(CHMSL), the new standard for brake lights on vehi-cles in the United States, which provide a high, centralthird brake light to warn drivers. In 1988, the U.S.National Highway Traffic Safety Administrationperformed extensive tests of the reaction times ofvolunteers to simulated light trucks with CHMSL orwith conventional brake lights. The reaction time fordrivers following a truckwithCHMSLwas 0.09 secondshorter than for drivers following a truck withoutCHMSL. That is just a slightly lower benefit than inpassenger cars, where the reduction in reaction timewith CHMSL was 0.11 second. The safety data forCHMSL are related, of course, to braking events.Several analytical studies of subsequent accident dataindicate that CHMSL was associated with a 5.1%reduction of rear-impact crashes.12

A second relevant issue is the use of cell phoneswhile driving. A recent metaanalysis of cell phonestudies was performed; 18 studies measured reactiontime in relation to cell phone use.C The average reduc-tion in reaction time reported was 0.25 second. Cellphone use while driving has been banned in manystates and provinces because of the negative effect ondriving.

The 2 examples above show the importance of smalltime intervals when driving. The effects of the blue

- VOL 38, MAY 2012

822 BLUE LIGHT–FILTERING TORIC IOLS AND DRIVING SAFETY

light–filtering chromophore, with a mean safety mar-gin increase of 0.5 second, are therefore not only statis-tically significant but also functionally significant interms of the potential to reduce accidents.

Heinrich13 developed an accident classification sys-tem with 3 events: major injury, minor injury, and noinjury. He proposed that for every major injury therewere 29 minor injuries and 300 no-injury events. Inthe context of driving, a major injury refers to an eventinvolving bodily harm or death to the driver, a minorinjury refers to an event involving minor injury or ve-hicle damage, and a no-injury accident is an event inwhich the driver narrowly avoids a crash (eg, due toluck or a reaction by the other driver). Although the ex-act values in the model are still under debate (eg, Birdand Germain14), the concept of Heinrich’s pyramid iswell accepted by safety researchers in a variety of do-mains. It has been applied to the analysis and predic-tion of driving15 and aviation accidents.16 The modelpredicts that for every major accident (ie, car crash)there are numerous situations in which a major crashcould have occurred had the circumstances been dif-ferent. Executing a left turn with a short safety marginis an example of a no-injury accident in this model. Ifa driver repeatedly performs this behavior, it is likelythat on some occasion a major accident will occur(due to the unpredictable circumstances describedabove).

Any intervention that can reduce the occurrence ofno-injury accidents is expected to decrease the likeli-hood of a minor or major accident in Heinrich’s model.Blue-light filtration appears to be such an intervention;the results in thepresent study showthat ablue-filteringtoric IOL can significantly increase the safety marginduring left turns and improve the glare susceptibilityratio of patients relative to a UV-only filtering IOL.

REFERENCES1. Gray R, Perkins SA, Suryakumar R, Neuman B, Maxwell AW.

Reduced effect of glare disability on driving performance in

patients with blue light–filtering intraocular lenses. J Cataract

Refract Surg 2011; 37:38–44

2. Gray R, Regan D. Glare susceptibility test results correlate with

temporal safety margin when executing turns across approach-

ing vehicles in simulated low-sun conditions.Ophthalmic Physiol

Optics 2007; 27:440–450

3. Hoffer KJ. Biometry of 7,500 cataractous eyes. AmJOphthalmol

1980; 90:360–368; correction, 890

4. Ferrer-Blasco T, Mont�es-Mic�o R, Peixoto-de-Matos SC,

Gonz�alez-M�eijome JM, Cervi~no A. Prevalence of corneal astig-

matism before cataract surgery. J Cataract Refract Surg 2009;

35:70–75

5. Johnson CA, Keltner JL. Incidence of visual field loss in 20,000

eyes and its relationship to driving performance. Arch Ophthal-

mol 1983; 101:371–375

J CATARACT REFRACT SURG

6. McKnight AJ, Shinar D, Hilburn B. The visual and driving per-

formance of monocular and binocular heavy-duty truck drivers.

Accid Anal Prev 1991; 23:225–237

7. Cody D, Tan S, Garcia A. Human Driver Model Development.

Berkeley, Institute of Transportation Studies, University of Cali-

fornia, 2005; 13–25

8. Hammond BR, Bernstein B, Dong J. The effect of the AcrySof

Natural lens on glare disability and photostress. Am J Ophthal-

mol 2009; 148:272–276

9. Scott JJ, Gray R. A comparison of tactile, visual, and auditory

warnings for rear-end collision prevention in driving. Hum

Factors 2008; 50:264–275

10. Mohebbi R, Gray R, Tan HZ. Driver reaction time to tactile and

auditory rear-end collision warnings while talking on a cell

phone. Hum Factors 2009; 51:102–110. Available at: http://

cobweb.ecn.purdue.edu/whongtan/pubs/PDFfiles/J41_Mohebbi

GrayTan_HF2009.pdf. Accessed December 30, 2011

11. Straughn SM, Gray R, Tan HZ. To go or not to go: stimulus-

response compatibility and pedestrian collision warnings in driv-

ing. IEEE Transactions on Haptics 2009; 2:111–117

12. Kahane CJ, Hertz E. The Long-Term Effectiveness of Center

High Mounted Stop Lamps in Passenger Cars and Light Trucks.

NHTSA Technical Report Number DOT HS 808 696, March

1998. Available at: http://www.nhtsa.gov/cars/problems/Equip

ment/CHMSL.html. Accessed December 30, 2011

13. Heinrich HW. Industrial Accident Prevention: A Scientific

Approach. New York, NY, McGraw-Hill, 1931

14. Bird FE Jr, Germain GL. Loss Control Management: Practical

Loss Control Leadership, rev ed. Loganville, GA, Det Norske

Veritas (USA), 1996

15. Tijerina L. Issues in the evaluation of driver distraction associated

with in-vehicle information and telecommunication systems.

Washington,DC,NationalHighwayTrafficSafetyAdministration,

June 23, 2000; Available at: http://www-nrd.nhtsa.dot.gov/

departments/Human%20Factors/driver-distraction/PDF/3.PDF.

Accessed December 30, 2011

16. Federal Aviation Administration. System Safety Handbook,

Washington DC, FAA, December 30, 2000; chapter 16: Opera-

tional safety in aviation. Available at: http://www.faa.gov/

library/manuals/aviation/risk_management/ss_handbook/media/

Chap16_1200.pdf. Accessed December 30, 2011

OTHER CITED MATERIALA. Slick RF, Evans DF, Kim E, Steele JP, “Using Simulators to

Train Novice Teen Drivers: Assessing Psychological Fidelity

as a Precursor to Transfer of Training,” presented at the Driving

Simulation Conference (DSC-Asia/Pacific), Ibaraki, Japan, May

2006

B. Nowakowski C, “A Preliminary Evaluation of a Driver-

Infrastructure Interface for Left-Turn Warning,” presented at

the 50th annual meeting of the Human Factors And Ergonomics

Society, San Francisco, California, USA, October 2006

C. Caird JK, Scialfa CT, Ho G, Smiley A, “A Meta-Analysis of

Driving Performance and Crash Risk Associated with the Use

ofCellular TelephonesWhileDriving,” presentedat the 3rd Inter-

national Driving Symposium on Human Factors in Driver

Assessment, Training, and Vehicle Design, Rockport, Maine,

USA, June 2005. Available at: http://www.nsc.org/news_

resources/Resources/Documents/A%20meta-analysis%20of

%20driving%20performance%20and%20crash%20risk%20

associated%20with%20the%20use%20of%20cellular%20

telephones%20while%20driving.pdf. Accessed December

30, 2011

- VOL 38, MAY 2012