Journal of Safety Research 42 (2011) 165–169

Contents lists available at ScienceDirect

Journal of Safety Research

j ourna l homepage: www.e lsev ie r.com/ locate / j s r

Road safe seniors: Screening for age-related driving disorders in inpatient andoutpatient settings☆

Linda Hill a,⁎, Jill Rybar a,⁎⁎, Sara Baird c, Susanna Concha-Garcia d, Raul Coimbra b, Kevin Patrick a

a University of California, San Diego, Department of Family and Preventive Medicine, San Diego, CAb University of California, San Diego, Department of Surgery, San Diego, CAc University of Washington, Department of Family Medicined University of California, San Diego, Department of Ophthalmology, San Diego, CA

☆ Funding source: The Office of Traffic Safety, State o⁎ Correspondence to: L. Hill, Department of Family an

9500 Gilman Dr., MS 0811, La Jolla, CA 92093–0811. Tefax: +1 858 5349404.⁎⁎ Correspondence to: J. Rybar, TREDS: Training, ReseaSafety, Department of Family and Preventive Medicine0811, La Jolla, CA 92093–0811. Tel.: +1 858 5349313; f

E-mail addresses: lhillbaird@aol.com, llhill@ucsd.edu(J. Rybar).

0022-4375/$ – see front matter © 2011 National Safetydoi:10.1016/j.jsr.2011.05.005

a b s t r a c t

a r t i c l e i n f o

Available online 29 June 2011

Keywords:Older driversAgingCrashesTraffic safetyCognition

Introduction: Older drivers are increasing in number and they often have health conditions that place them athigh risk for motor-vehicle crashes (MVC). Screening is underutilized, and is rarely done in hospital settings.Methods: A convenience sample of 755 older adults completed age related driving disorders screening atUniversity of California, San Diego inpatient and outpatient health centers. Screening included threestrength/frailty tests, two vision tests (acuity and fields), and two cognitive tests, based on AMArecommendations. The average age of participants was 72.5; 55.5% were male and 94% English-speaking;17.8% of older adults failed at least one aspect of screening. Results: In multivariate analysis, significant

associations of failed status were age, male sex, selfrestrictions of driving, and inpatient screening locations.The screening identified one in six adults to be 'high-risk' for age related driving disorders. Screening waseffective and feasible in both inpatient and outpatient settings. Impact on industry: As the driving populationages, industry, government and health car providers need to plan for the management of driving impairmentsin older adults.

© 2011 National Safety Council and Elsevier Ltd. All rights reserved.

1. Introduction

The population over 65 years of age is the fastest growing group inthe United States, and the number of older adult drivers on the road isexpected to reach 40 million by 2020. However as they age, manyolder adults begin to experience health and functional impairmentsthat interfere with their ability to drive safely. Physical and mentalchanges, including reduced visual acuity, decreased strength, medi-cation side effects, and cognitive impairment, can directly andindirectly result in age-related driving disorders (ARDDs). Conditionsthat adversely affect driving are highly prevalent in older populations,though not limited to that group. Twenty-four percent of adults over80 years old have uncorrectable visual impairment (Congdon et al.,2004), and 34.5% of adults over 85 have dementia, with 37% diagnosedover age 90 (Lindsay et al., 2004). Additionally, medication use,chronic diseases, and frailty all increase with age.

f California.d Preventive Medicine, UCSD,l.: +1 6198406258 (mobile);

rch and Education for Driving, UCSD, 9500 Gilman Dr., MSax: +1 858 5349404.(L. Hill), jrybar@ucsd.edu

Council and Elsevier Ltd. All rights

While older adults bring somepositive attributes to the road, such asreduced risk taking, experience, and greater compliance with the laws,they have more crashes per mile driven and per population than theirmiddle aged counterparts (Evans, 2004). Older drivers and their elderlypassengers are also at higher risk of disabling or life-threatening injurywhen compared to younger drivers and passengers in auto accidents ofsimilar force, due to reducedmusclemass, osteoporosis, and underlyinghealth problems (Meuleners et al., 2006).

Toprovidedirection forphysicians in this area, theAmericanMedicalAssociation (AMA) and the American Occupational Therapy Association(AOTA) organized a task force in 2003 to develop guidelines for ARDDsscreening and intervention for health care providers (AMA, 2010). Thesecond edition of the guidelines, Physician's Guide to Assessing andCounseling Older Drivers, was recently released. The group recom-mended that ARDDs screening include seven exam tasks, addressingvision, frailty, and cognitive function. These recommendations werebased upon relationships between one or more of these individualscreening components and adverse outcomes of concern, such ascrashes, injuries, and death. Cognitive tests have been well studied intheir relationship to driving performance, and have been correlatedwith higher likelihood of crashes and driving cessation (Ball et al., 2006;Friedland et al., 1988; LaFont et al., 2008; Stutts, Stewart, & Martell,1998; Zuin, Ortiz, Boromei, & Lopez, 2002). Freund studied 100 driversand found that the clock-drawing test was highly correlated withmistakes on a driving simulator (Freund et al., 2005).

reserved.

166 L. Hill et al. / Journal of Safety Research 42 (2011) 165–169

In addition to cognitive function, both frailty and its concomitantphysical limitations and vision are correlated with adverse drivingoutcomes. Decreased range of motion has been shown in two studiesto interfere with changing lanes and making left turns (Preusser et al.,1998; Shinar & Scheiber, 1991). Visual impairment affects driving inboth epidemiologic and prospective studies (Lim & Chutka, 2006;Subzwari et al., 2009).

Despite the AMA recommendations, the uptake of ARDDSscreening has been modest, and there is no research addressing theeffects of the AMA recommendations on outcomes such as healthstatus of seniors, others involved in seniors related crashes, or theattendant economic consequences of this public health problem. Fewstudies have addressed the settings best suited for screening or thefeasibility of screening. In addition to the paucity of ARDDs screening,many physicians do not know how to report safety concerns to theDepartment of Motors Vehicles (DMV). A recent report found thatmore than 28% of geriatricians do not know how to report patientswith dementia, yet more than 75% of geriatricians agreed thatphysicians are responsible for reporting patients (Cable et al., 2000).Statutes regarding the reporting of drivers with medical conditionsvary widely by state. For example, in California, physicians arerequired by law to report patients with “disorders characterized bylapses in consciousness,” specifically including dementia, seizures,and loss of consciousness (California Department of Motor Vehicles,2010a). State DMVs across the country also have variance in renewalof license and surveillance of older adults. California prohibits mail-inlicense renewals after the age of 69, and older drivers must undergovision and knowledge tests every five years (California Department ofMotor Vehicles, 2010b). On-road driving tests may be required foranyone exhibiting symptoms (i.e., confusion or tremor) but by law,the driving test may not be required on the basis of advanced agealone (Lafont et al., 2008).

This paper expands on our earlier manuscript that reported thepreliminary results of our project (Baird et al., 2010). The purpose of thisstudy was to reassess the initial analysis based on the larger and final

Table 1Description of the Seven Screening Tests Used in Age-related Driving Disorders ScreeningDrivers.

Test Description of Test Criteria for Pass

Visual Fields Measured using confrontation testing at 3feet distance.

Within normal

Scored as pass/fail by comparison to normallimits.

Visual Acuity Measured using Snellen eye chart. b20/70 in bothcorrectedScored by numerical visual acuity value.

Rapid Pace Walk Measured by time taken to walk 20 feet. ≤9 secondsScored by time in seconds.

Range of Motion Measured by neck rotation, finger curl,shoulder and elbow flexion, ankledorsiflexion and plantar flexion.

Within normalfor all tests

Scored as pass/fail based on normal limits.Motor Strength Measured by shoulder

adduction/abduction/flexion, wristflexion/extension, hand grip strength, hipflexion/extension, and ankledorsiflexion/plantar flexion.

Score of 4/5 orfor all tests

Scored by standard 0/5-5/5 scale.Clock Drawing Measured by asking participant to draw the

face of a clock with the time set at “tenminutes after eleven.”

All criteria met

Scored by: shape of clock, correct amount ofnumbers accurately located and spaced, onlytwo clock hands with one hand pointing tothe 2 and the absence of intrusive marks.

Trail Making B The participant is asked to connect, inalternating order, encircled letters (A-L) andnumbers (1–13). Errors are pointed outwhile the test is being completed.

≤180 seconds

Scored by time required to complete.

numberof participants, and to addanevaluationof the acceptanceof thescreening exam and screening recommendations among older adults.This projectwas funded by theCalifornia Office of Traffic Safety, throughthe National Highway Traffic Safety Administration.

2. Methods

The studywas conducted fromApril 2008 throughSeptember 2009 inthree outpatient centers and two hospitals within the University ofCalifornia, San Diego (UCSD) system. Eligibility criteria included patientsover the age of 60whowere California-licensed drivers andwere Englishor Spanish speaking. Exclusion criteria included medications that mayinterfere with testing (i.e., narcotics), and patients within 24 hours post-op. Permissionwasobtained fromtheUCSD IRB. Research assistantsweretrained to written protocols that included eligibility criteria, recruitmentscript, baseline questions, screening techniques, result interpretation,criteria for consultation with the physician-supervisor, counseling,referrals, and reporting to participants’ physicians. Research assistantscarried the protocols in operation manuals on site. Training includedsupervised practice sessions, and periodic supervision to assess compli-ance with protocols.

The trained research assistants administered the following AMArecommended screening tests, under standing orders, to voluntaryparticipants: (a) visual acuity, (b) visual fields, (c) trail making B,(d) clock drawing, (e) rapid pace walk, (f) range of motion, and(g) motor strength. Detailed descriptions of the screening exams canbe found in Table 1. Patient satisfaction information was obtained bysurvey immediately post-screening.

2.1. Scoring

The participantswere scored based onAMAcriteria (see Table 1). Theparticipantswere thengiven a summary score. ‘Pass’ included individualswho passed all seven tests. ‘Incomplete’ included individuals who passedall measured tests, but were unable to complete all portions of the exam;

(ARDDs), Adapted from the AMA Physician's Guide to Assessing and Counseling Older

ing Criteria for Failing Example Criteria for Incomplete

limits Not within normal limits Inability to understand instructions

eyes, N20/70 in either eye,corrected

Missing corrective lenses

N9 seconds Inability to leave hospital bed area(e.g. Foley catheter)

limits Not within normal limitsfor one or more tests

Temporary reason for failure (e.g. IV,recent orthopedic injury/surgery)

higher Score of ≤3/5 for one ormore tests

Temporary reason for failure(e.g. pain from IV, recent surgery)

Failure to meet any criteria Missing corrective lenses, inability towrite (e.g. tremor)

N180 seconds Missing corrective lenses, Englishilliteracy

167L. Hill et al. / Journal of Safety Research 42 (2011) 165–169

this was most commonly due to issues such as missing glasses, tetheredmedical equipment, or time constraints/conflicting medical procedures.‘Fail’ encompassed two outcome groups: (a) the “Fail-no DMV report”group failedat least onevisionor frailty test butpassed thecognitive tests,and (b) the “Fail- report toDMV”group failed at least one cognitive test orhad a recent unexplained lapse of consciousness (including seizures).Individuals in this latter group were reported to the DMV, unless themedical director of the project or primary care provider felt that furthertestingwas indicated prior to reporting. For discussion purposes, patientswith pass and incomplete scores are referred to as “low risk group,” andall fail scores are considered the “high risk group.”

Data were coded and entered into a database by trained personnelusing a data dictionary and range alerts. Univariate and multivariateanalysis of the dichotomous outcome (Fail vs. Pass/Incomplete) wasperformed using logistic regressionmodels with R Statistical Software(version 2.8.1, 2008). Independent variables included age, gender,language, restriction of driving by self, restriction of driving by others,recent crash (b3 yrs), and screening location. Patient age was groupedin 5-year ranges (60–64, 65–69, 70–74, 75–79, 80+) for comparisonbut treated as a continuous variable when possible; all other variableswere bivariate, including language (English vs. Spanish) and location(inpatient vs. outpatient). The final logistic regression model adjustseach characteristic for all other characteristics. Odds ratios (OR) and95% confidence intervals (CI) were obtained from this analysis; due toa high prevalence (N10%) of ‘Fail’ outcome, OR and CI weresubsequently corrected to better approximate RR using the methodby Zhang and Yu (1998).

3. Results

3.1. Personal characteristics

Table 2 describes the personal characteristics of the 755 partici-pants included for analysis. The majority (74%) of participants were

Table 2Prevalence Data of Patient Characteristics of Patients Screened for ARDDs in Inpatient and O

In

Totals Total 75Pass 55Fail – Report to DMV 85Fail – No Report to DMV 49Incomplete 66

Age 60-64 1265-69 1770-74 1775-79 1180+ 16Mean 72

Gender M 41F 33

Self restriction of driving No 40Yes 35

3 mOther restriction of driving No 58

Yes 163 m

MVC and/or citation in thelast three years

No 66Yes 88

2 mScreening location Inpatient 55

Outpatient 194 m

Use of cell phone while driving No 48Yes 26

4 mLanguage spoken English 71

Spanish 404 m

screened in inpatient settings. Of all participants, 55.5%weremale, 94%English-speaking, and the mean age was 72.5 years (range: 60 – 97).The mean age was higher in the Fail (74.4) than the Pass/Incomplete(72.1) groups.

Several participants reported high-riskmedical histories, includinga current diagnosis of dementia (4, 0.5%), or a ‘lapse of consciousness,’including a history of seizures (19, 2.5%), and history of loss ofconsciousness (LOC; 79, 10%). Twenty-one (2.8%) participantsreported a LOC within 6 months of screening.

The study population was queried about adverse driving situa-tions; a total of 88 (11.7%) reported being in a motor vehicle crash(MVC) in the past 3 years; 5 of these resulted in the current admissionto the hospital. A surprisingly high 263 (34.8%) participants reportedever having used a cell phone while driving.

3.2. Screening outcome

A total of 755 qualified participants were screened; 555 (73.5%)participants received a “Pass” result, 49 (6.5%) were “Fail – no DMVreport,” and 85 (11.3%) participants were “Fail – report to DMV.”Overall,this study yielded a 17.8% Fail screen rate. An additional 66 (8.7%)participants had “Incomplete” screens.

Among the 85 participants who received a “Fail-report to DMV,” 29(34.1%) failed trail-making, 13 (15.3%) failed clock-drawing, and 29(34.1%) failed both portions of the cognitive exam. An additional 14(16.5%) participantswhowould have otherwise passed ARDDs screeningwere designated in this category for other medical reasons and at thediscretion of the provider: 10 due to recent LOC, 1 due to recent seizure,and 3 due a diagnosis of dementia.

Several trends were identified when comparing the low risk group(Pass/Incomplete scores) to the high risk group. Among the 263 personswho reportedusing their cell phonewhile driving, nearly 93%were in thelow risk group; only 7% of cell phone users were in the high risk group,compared to 24% of the non-users. The characteristic could not be

utpatient Settings.

cidence, n (%) Number of Pass/Incomplete, n (%)

Number of Fail (DMVand no DMV), n (%)

55 (73.5%)(11.3)(6.5)(8.7)6 (16.7%) 109 (86.5%) 17 (13.5%)6 (23.3) 151 (85.8) 25 (14.2)7 (23.4) 148 (83.6) 29 (16.4)6 (15.4) 93 (80.2) 23 (19.8)0 (21.2) 120 (75.0) 40 (25.0).5 years 72.1 years 74.4 years9 (55.5%) 338 (80.7%) 81 (19.3%)6 (44.5) 283 (84.2) 53 (15.8)0 (53.0%) 342 (85.5%) 58 (14.5%)2 (46.6) 276 (78.4) 76 (21.6)issing

7 (77.7%) 482 (82.1%) 105 (17.9%)5 (21.9) 136 (82.4) 29 (17.6)issing

5 (88.1%) 548 (82.4%) 117 (17.6%)(11.7) 71 (80.7) 17 (19.3)issing

7 (73.8%) 445 (79.9%) 112 (20.1%)4 (25.7) 174 (89.7) 20 (10.3)issing

8 (64.6%) 373 (76.4%) 115 (23.6%)3 (34.8) 244 (92.8) 19 (7.2)issing

1 (94.2%) 586 (82.4%) 125 (17.6%)(5.3) 31 (77.5) 9 (22.5)issing

168 L. Hill et al. / Journal of Safety Research 42 (2011) 165–169

included inmultivariate analysis due to strong associationwith age, but itis the opinion of the researchers that this finding is related to the drivingconfidence, andoftenyounger age, that is needed to successfullyuse a cellphone while driving. Indeed, the average age of the cell phone-usinggroupwas 69.6 years, compared to the average 74.1 years of the non-cellphone using group. Among inpatients, the number of medications wassignificantly higher among those in the high risk group than the low riskgroup (11.6 vs. 9.3). Similarly, the mean number of outpatientmedications was 8.8 in the high risk group versus 7.2 for the low riskgroup (38 missing values).

3.3. Multivariate analysis

Multivariate analysis identified few patient characteristics thatwere independently associated with a Fail (DMV and no DMV) resultin ARDDs screening (see Table 3). Participants in the highest agecategory (OR 1.87, p=0.025) and screened in inpatient settings (OR2.57, pb0.001) were associated with being in the high risk group.Interestingly, self-restriction of driving (OR 1.61, p=0.006) but notrestriction of driving by others (OR 1.23, p=0.335) was a significantfactor.

Characteristics that were not significantly associated with a Failoutcome included language spoken, and recent (b3 years) crashhistory.

3.4. Participant satisfaction, perceptions of recommendations, futureplans

Of the 755 participants, 537 (71%) completed the post-screeningsurvey immediately after the screening was completed and recom-

Table 3Multivariate Analysis of Patient Characteristics of Patients Screened for ARDDs inInpatient and Outpatient settings.

Incidence, n(%)

CorrectedOR*,**

95% CI p-value

Age 60-64 126 (16.7%) 1.0065-69 176 (23.3) 1.20 0.64 -

2.090.544

70-74 177 (23.4) 1.34 0.75 -2.26

0.305

75-79 116 (15.4) 1.76 0.96 -2.94

0.069

I vote for 80+ 80+ 160 (21.2) 1.87 1.09 -2.97

0.025

continuous 755 1.03 1.01 -1.05

0.002

Gender F 336 (44.5%) 1.00M 419 (55.5) 1.32 0.96 -

1.800.088

Language English 711 (94.2%) 1.00Spanish 40 (5.3) 0.85 0.42 -

1.580.618

4 missingSelf restrictionof driving

No 400 (53.0%) 1.00Yes 352 (46.6) 1.61 1.15 -

2.170.006

3 missingOther restrictionof driving

No 587 (77.7%) 1.00Yes 165 (21.9) 1.23 0.80 -

1.800.335

3 missingCrash and/orcitation inthe lastthree years

No 665 (88.1%) 1.00Yes 88 (11.7) 1.10 0.67 -

1.730.670

2 missingScreeninglocation

Outpatient 194 (25.7%) 1.00Inpatient 557 (73.8) 2.57 1.64 -

3.77b0.001

* for each variable, OR are adjusted for all other listed variables.** The OR were corrected using the method outlined by Zhang and Yu to betterapproximate RR given high outcome prevalence.

mendations weremade; of these, 302 (56%) were men, and 109 (20%)were Fails. Satisfaction with the screening process was high; themajority (382, 71%) stated that screening was “very” useful, with only4% finding it “not at all” useful. Screening took an average of15 minutes, and 442 (82%) participants were “very” satisfied withthe amount of time.

Despite the temporal relationship to the screening advice, 93participants were “not sure” what their screener's recommendationswere – 19 of these participants were failed drivers. Though 109 failed,only 58 thought they should stop driving either temporarily orpermanently. Importantly, less than 20% of participants said theywere “surprised” by their results, and 420 (78%) – including two-thirds of the high risk group - were neutral or not surprised. Finally,75% said they would recommend ARDDs screening to someone else.

4. Discussion

In this population of older adults selected from routinely seenpatients in inpatient and outpatient settings, one in six failed at leastone of the screening tests to be considered high risk for a motor-vehicle related crash. Characteristics of these individuals includedolder age and self-imposed driving restrictions. These numbers mayunderestimate the true percent of high-risk drivers as participation inthis study was voluntary, and those concerned about their skills mayhave been more likely to decline screening.

Nonetheless, the screen was well received with high participationand satisfaction despite the risk of being reported to the DMV, andeven among those who were in the ‘Fail’ group. This study indicatesthat it is feasible to do ARDDs screening in a variety of clinical settingsand that it is generally well accepted by patients and clinical staff.Participants, including those who were screened as “Fail,” werereceptive of the screening and were often not surprised by the resultsand recommendations.

Current recommendations from the AMA promote screening forvision, frailty, and cognitive function; few physicians are capable andcomfortable enough with the screening protocols to implement them.Studies have shown that patients are more likely to cease driving onthe advice of their physicians compared to others, so physicians havethe potential to assume an influential role (Persson, 1993). However,physicians are reluctant to discuss driving issues with their patients(Friedland et al., 2006), and patients may be reluctant to bringconcerns to physicians’ attention. Driving is a sensitive issue for manyolder adults, who depend on driving for independence. Drivingcessation in this population has been associated with a 3-folddecrease in out-of-home activity and a 2.5-fold increase in depressivesymptoms, and physical andmental decline (Berkman & Tinetti, 1997;Fonda, Wallace, & Herzog, 2001; Hebert et al., 2003; Marottoli et al.,1997). Driving cessation has been associated with admission to longterm care facilities, even when controlling for other variables, andwith a reduced network of friends even in those patients withalternative transportation. Thus, ARDDs screening should be con-ducted in a supportive environment, where options for continuedmobility can be given to patients who should no longer be driving. Forexample, in this study, participants who “passed” were providedinformation on classes to improve driving skills, but participants who“failed” received recommendations not to drive and were givenhandouts on alternative transportation.

5. Conclusions

Motor-vehicle crashes and their associated morbidity and mortal-ity are an important public health issue for all age groups. While olderadults represent a small proportion of drivers, the results of this studysuggest that the magnitude of ARDDS is great. Moreover, in the nextfew decades, increasing numbers of older adults will develop theseconditions, yet the systems and incentives needed to identify and

169L. Hill et al. / Journal of Safety Research 42 (2011) 165–169

manage these adults are essentially non-existent. Evidence-basedinterventions are needed to ensure that high-risk older adult driversare identified and successfully offered options to move from behindthe wheel to other forms of continued mobility that reduce thelikelihood that they injure themselves and/or others.

References

AMA (2010). Physician's Guide to Assessing and Counseling Older Drivers AccessedOctober 2010 at: http://www.ama-assn.org/ama/pub/physician-resources/public-health/promoting-healthy-lifestyles/geriatric-health/older-driver-safety/assessing-counseling-older-drivers.shtml.

Baird, S., Hill, L., Rybar, J., Concha-Garcia, S., Coimbra, R., & Patrick, K. (2010). Age-RelatedDriving Disorders: Screening in hospitals and outpatients settings. Geriatrics &Gerontology International, 10(4), 288–294. doi:10.1111/j.1447-0594.2010.00622.x.Epub 2010 May 17.

Ball, K. K., Roenker, D. L., Wadley, V. G., Edwards, J. D., Roth, D. L., McGwin, G., Jr., et al.(2006). Can high-risk older drivers be identified through performance-basedmeasures in a Department of Motor Vehicles setting? Journal of American GeriatricsSociety, 54(1), 77–84.

Berkman, L. F., & Tinetti, M. (1997). Driving Cessation and Increased DepressiveSymptoms: Prospective Evidence from the New Haven EPESE. Journal of AmericanGeriatrics Society, 45, 202–221.

Cable, G., Reisner, M., Gerges, S., & Thirumavalavan, V. (2000). Knowledge, attitudes,and practices of geriatricians regarding patients with dementia who are potentiallydangerous automobile drivers: a national survey. Journal of American GeriatricsSociety, 48(1), 14–17.

California Department of Motor Vehicles (2010a). DMV Health and Safety Code:Reporting Disorders Characterized by Lapses of Consciousness. Sacramento, CA:Author Accessed October 2010: http://www.dmv.ca.gov/pubs/vctop/appndxa/hlthsaf/hs103900.htm.

California Department of Motor Vehicles (2010b). DMV Senior Driver. Sacramento, CA:AuthorAccessedOctober2010.http://www.dmv.ca.gov/about/senior/senior_top.htm.

Congdon, N., O'Colmain, B., Klaver, C. C., Klein, R., Muñoz, B., Friedman, D. S., et al. (2004).Eye Diseases Prevalence Research Group. Causes and prevalence of visual impairmentamong adults in the United States. Archives of Ophthalmology, 22(4), 477–485.

Evans, L. (2004). Science Serving Society. Bloomfield Hills, MI: Traffic Safety.Fonda, S. J., Wallace, R. B., & Herzog, A. R. (2001). Changes in Driving Patterns and

Worsening Depressive Symptoms among Older Adults. Journals of Gerontology,56(6), S343–S351.

Freund, B., Gravenstein, S., Ferris, R., Burke, B. L., & Shaheen, E. (2005). Drawing clocksand driving cars. J Gen Inter Med, 20(3), 240–244.

Friedland, R. P., Koss, E., Kumar, A., Gaine, S., Metzler, D., Haxby, J. V., et al. (1988). Motorvehicle crashes in dementia of the Alzheimer type. Annals of Neurology, 24,782–786.

Friedland, J., Rudman, D. L., Chipman, M., & Steen, A. (2006). Reluctant Regulators:Perspectives of Family Physicians on Monitoring Seniors' Driving. Topics inGeriatric Rehabilitation. Transportation and Mobility, 22(1), 53–60.

Hebert, L. E., Scherr, P. A., Bienias, J. L., Bennett, D. A., & Evans, D. A. (2003). Alzheimerdisease in the US population: prevalence estimates using the 2000 census. Archivesof Neurology, 60(8), 1119–1122.

Lafont, S., Laumon, B., Helmer, C., Dartigues, J. F., & Fabrigoule, C. (2008). DrivingCessation and self-reported car crashes in older drivers: the impact of cognitive

impairment and dementia in a population-based study. Journal of GeriatricPsychiatry and Neurology, 21, 171–182.

Lim, L., & Chutka, D. (2006). Preventive medicine beyond 65. Geriatrics and GerontologyInternational, 6, 73–81.

Lindsay, J., Sykes, E., McDowell, I., Verreault, R., & Laurin, D. (2004). More than theepidemiology of Alzheimer's disease: contributions of the Canadian Study of Healthand Aging. Canadian Journal of Psychiatry, 49(2), 83–91.

Marottoli, R. A., Mendes de Leon, C. F., Glass, T. A., Williams, C. S., Cooney, L. M., Jr.,Berkman, L. F., et al. (1997). Driving cessation and increased depressive symptoms:prospective evidence from the New Haven EPESE. Established Populations forEpidemiologic Studies of the Elderly. Journal of American Geriatrics Society, 45(2),202–206.

Meuleners, L. B., Harding, A., Lee, A. H., & Legge, M. (2006). Fragility and crash over-representation among older drivers in Western Australia. Accident; Analysis andPrevention, 38(5), 1006–1010 Epub 2006 May 22.

Persson, D. (1993). The Elderly Driver: Deciding When to Stop. The Gerontologist, 33(1),88–91.

Preusser, D. F., Williams, A. F., Ferguson, S. A., Ulmer, R. G., & Weinstein, H. B. (1998).Fatal crash risk for older drivers at intersections. Accident; Analysis and Prevention,30(2), 151–159.

Shinar, D., & Scheiber, F. (1991). Visual requirements for safety and mobility of olderdrivers. Human Factors, 33(5), 507–519.

Stutts, J. C., Stewart, J. R., & Martell, C. (1998). Cognitive test performance and crash riskin an older driver population. Accident; Analysis and Prevention, 30(3), 337–346.

Subzwari, S., Desapriya, E., Babul-Wellar, S., Pike, I., Turcotte, K., Rajabali, F., et al.(2009). Vision screening of older drivers for preventing road traffic injuries andfatalities. Cochrane Database of Systematic Reviews, 21(1) CD006252.

Zhang, J., & Yu, K. (1998). What's the Relative Risk? A Method of Correcting the OddsRatio in Cohort Studies of Common Outcomes. JAMA, 280(19), 1690–1691.

Zuin, D., Ortiz, H., Boromei, D., & Lopez, O. L. (2002). Motor vehicle crashes andabnormal driving behaviours in patients with dementia in Mendoza, Argentina.European Journal of Neurology, 9(1), 29–34.

Linda Hill, MD, MPH; Dr. Hill is a Clinical Professor at UCSD. A specialist in preventivemedicine, she has clinical, research and teaching responsibilities. She is the director ofthe UCSD-SDSU General Preventive Medicine Residency.

Jill Rybar, MPH; Ms. Rybar is a public health project coordinator, with an MPH fromSDSU. She has been working in the area of preventive medicine and public health for25 years.

Sara Baird, MD; Dr. Baird is a graduate of Brown University, and is a physician-resident in family medicine at a University of Washington affiliated program in Seattle.

Susanna Concha-Garcia, BS;: Ms. Concha-Garcia is a research technician, with aninterest in geriatrics.

Raul Coimbra, MD, PhD: Dr. Coimbra is Chair of the Division of Trauma at UCSD. He isinvolved in clinical and basic science research at UCSD, and lectures and consultsinternationally in the area of trauma and injury prevention.

Kevin Patrick, MD, MS: Dr. Patrick is a preventive medicine physician and researcherat UCSD. He is the editor of the American Journal of Preventive Medicine.