| Study | Reason for exclusion |
|---|---|
| Adler 2008 | Non‐RCT Review article |
| Akinwuntan 2007 | Stroke cohort, no dementia cases |
| Anderson 2005 | Non‐RCT Case‐control study, 222 participants (70 with mild AD, 152 controls) compared composite neuropsychological test scores and simulated driving performance with prospective crash rate |
| Anstey 2009 | Non‐RCT Only 2 people with probable dementia included |
| ANZSGM 2010 | Non‐RCT Position statement |
| Arai 2009 | Non‐RCT |
| Badenes 2007 | Non‐RCT Case control study, 187 participants, (92 MCI, 55 mild dementia, 40 controls) recruited from hospital population, compared neuropsychological testing with ASDE‐driver test (a Spanish office‐based measure of fitness to drive measuring: estimations of speed, coordination, reaction time, attention and concentration) |
| Badenes‐Guia 2008 | Non‐RCT 92 MCI, 55 mild dementia, 40 controls, tested with RBANS, TMT, Kohs block design compared with ASDE‐driver test and UFOV. No on‐road test |
| Ball 2006 | Non‐RCT Population‐based older driver cohort, no dementia measures included. 2114 older drivers compared initial assessment (measures of vision, gross and fine motor skills and cognitive testing) with 4‐5 years' follow‐up for motor vehicle accidents |
| Barrash 2010 | Non‐RCT Prospective study, self selected, advert respondents, 24 healthy, 26 probable AD, 33 PD. All active drivers with MMSE > 26. People with AD and PD recruited from registry, used neuropsychological tests and on‐road, unable to blind assessors (PD). Differences on neuropsychological tests but not road test failure. Scores did correlate with number of driving errors but effect diminished by demographic adjustments |
| Berndt 2008 | Non‐RCT Road test results in 117 people with dementia stratified by CDR score. Higher road test failure rate in higher CDR score |
| Breen 2007 | Non‐RCT Review article |
| Brown 2005 | Non‐RCT Case‐control study, 75 participants (50 with mild to moderate AD, 25 controls) correlated neuropsychological tests with rating of safe or unsafe in on‐road testing |
| Bylsma 1990 | Duplicate of Rebok 1994 |
| Bylsma 1997 | Non‐RCT Review article |
| Carr 2000 | Non‐RCT Retrospective case‐control study of 121 participants (63 with mild Ad and 58 controls) comparing state‐recorded crash rates in both groups over the previous 5 years |
| Carr 2011 | Non‐RCT 99 people with dementia referred for road test. Washington road test compared with visual, motor and cognitive test. no information on dementia subtype or severity |
| Clark 2005 | Non RCT Cross‐sectional study of 55 participants (41 with Ad, 6 with VaD, 2 with DLB, 4 with FTD and 2 age‐associated memory loss) no controls. Correlated neuropsychological tests with a rating of safe or unsafe in on‐road testing |
| Cooper 1993 | Non‐RCT Case‐control study comparing retrospective crash rates of 165 older people from a dementia clinic with a random sample from the population of drivers in British Columbia |
| Cotrell 1999 | Non‐RCT Cross‐sectional study of 35 people with AD (only 19 still driving), evaluated on MMSE, driving status and carer assessment of driving behaviour |
| Cox 1998 | Non‐RCT Cross‐sectional study of 50 participants (29 with AD) comparing neuropsychological tests with simulated drive |
| Cushman 1992 | Non‐RCT Case‐control study of 17 participants (8 with possible dementia, 9 without suspected dementia) comparing neuropsychological tests with a rating of safe or unsafe in on‐road testing |
| Daiello 2010 | Non‐RCT 24 early AD before and after cholinesterase inhibitor treatment, compared on computerised tests of visual attention and executive function. Also 35 early AD treated vs. matched non‐treated. NO ROAD TEST. Simulated computer‐based drive. Cholinesterase inhibitor improved simulated drive and visual attention |
| Dawson 2009 | Non‐RCT Case control, comparing 40 early AD vs. 115 healthy people on cognitive, visual, motor and driving tests. AD had more driving errors 42 vs. 33. Errors associated with older age, lower cognition status, and poorer performance on BVRT, CFT‐copy, TMT‐A and functional reach test |
| de Simone 2007 | Non‐RCT Case‐control study of 30 participants (15 people with FTD and 15 healthy controls) were administered a driving simulation task. Measures of driving performance and neuropsychiatric symptoms were assessed |
| Devos 2007 | Non‐RCT Case‐control study of 80 participants (40 people with Parkinson's disease and 40 controls). Included 11 people with PD and 3 controls with very mild cognitive decline (CDR 0.5). Participants were assessed using a driving simulator, driving history survey and the CDR. The people with PD also underwent a clinical test battery and an evaluation of fitness to drive performed by an official centre, which included visual, cognitive and on‐road tests |
| Dobbs 1997 | Non‐RCT Case‐control study of 115 older drivers with cognitive decline ‐ mostly AD compared with 35 older and 23 younger controls. Comparisons made with cognitive testing and on‐road driving test |
| Drachman 1993 | Non‐RCT Retrospective questionnaire‐based survey comparing annual rates of occurrence and severity of all crashes based on responses of the carers of 130 people with AD and 112 age‐matched control subjects (no dementia) and their spouses |
| Dubinsky 1992 | Non‐RCT Retrospective questionnaire based survey of 67 people with AD and their families and compared them with 100 elderly, non‐spousal controls focused on driving habits, continued driving and the number of accidents per year for the past 10 years |
| Duchek 2003 | Non‐RCT Prospective case control study following 108 participants (50 with AD and 58 non‐dementia) with sequential on‐road testing |
| Ernst 2010 | Non‐RCT Carer interviews, 30 FTD, 26 AD. Reported differences in driving behaviour: AD: unsteady driving style, FTD: aggressive/ risky/lack of insight. 37% MVA vs. 19% for AD (NS) |
| Fitten 1995 | Non‐RCT Case‐control study of 84 participants (15 with AD, 12 with VaD and 57 non‐dementia subjects). Correlated neuropsychological tests with Sepulveda Road test score and retrospective crash and violation rate |
| Fox 1997 | Non‐RCT Cross‐sectional study of 19 participants with AD, correlating road test with physician and neuropsychologist's assessment of fitness to drive and neuropsychological scores |
| Friedland 1988 | Non‐RCT Case‐control study comparing retrospective crash rate of 30 people with AD with 20 people without dementia |
| Frittelli 2009 | Non‐RCT CCT comparing 20 AD (CDR 1), 20 MCI (CDR 0.5) and 19 healthy on computer‐based driving simulator (STISTIM). All active drivers and not on treatment for AD. Poorer performance on length of run, mean time to collisions, off‐road events in AD vs. MCI and controls. MCI shorter time to collisions than healthy controls. Reaction times longer for AD vs. MCI and healthy control. Driving not correlated with MMSE |
| Gilley 1991 | Non‐RCT Retrospective survey of 522 people and their collateral informants from a dementia clinic. Only 322 licensed to drive at onset of dementia and 93 active drivers at time of questionnaire. Questionnaire on driving cessation and unsafe driving in previous 6 months |
| Grace 2005 | Non‐RCT 21 mild AD, 21 PD without dementia and 21 healthy controls. Compared neuro‐psychological tests and on‐road driving test, miles driven AD < PD < N. Poorer performance on neuro‐psychological testing correlated with unsafe or marginal performance. HVLT for AD and PD, TMT‐A for AD. Hoehn and Yahr correlated with marginal driving performance |
| Hunt 1993 | Non‐RCT Case‐control study of 38 participants (25 with AD, 13 without suspected dementia). Correlated neuropsychological tests with a rating of safe or unsafe in on‐road testing |
| Hunt 1997 | Non‐RCT Case‐control study of 94 participants (56 with AD, 38 without suspected dementia). Correlated neuropsychological tests with a rating of safe or unsafe in on‐road testing |
| Hunt 2010 | Non‐RCT Observational retrospective cohort |
| Innes 2005 | Non‐RCT Cross‐sectional study of 50 drivers with neurological disorders (4 with AD) referred to a Driving and Vehicle Assessment Service comparing a battery of computerised sensory‐motor and cognitive tests and on‐road driving assessment. Developed a model of sensory‐motor and cognitive tests that would predict pass or fail on‐road testing |
| Iverson 2010 | Non‐RCT Consensus statement |
| Johansson 1997 | Non‐RCT Review article |
| Kamimura 2009 | Non‐RCT |
| Kawano 2009 | Non‐RCT Observational retrospective study |
| Kay 2009 | Non‐RCT Prospective multicentre study of 115 driving assessment referrals with functional impairment 96 neurological including 30 AD. Compared 2 standardised office tests with on‐road assessment. Cut‐offs for both scores predicted road test fail |
| Lafont 2010 | Non‐RCT Prospective population study 2104 participants, predictors of driving cessation, including neuro‐psychological testing and dementia diagnosis. Dementia associated with driving cessation but not with crashes. Development of dementia later in the study was associated with crashes (OR 3.4). Poor performance on TMT‐B associated with crashes (OR 7.7) |
| Laks 1999 | Non‐RCT Review article |
| Laks 2000 | Non‐RCT Survey of cognitive performance and driving history (licensing and cessation) in 110 residents of sheltered accommodation for older people |
| Leproust 2008 | Non‐RCT Hypothetical study. Suggested on‐road testing every 3 years for people over 85 years of age as preventing 569 crashes and inducing 270 adverse events. Based on crash risk or 2.0 for drivers with dementia |
| Lincoln 2006 | Non‐RCT Case‐control study of 75 participants (42 with dementia, 33 without suspected dementia). Correlated neuropsychological tests with a rating of safe or unsafe in on‐road testing. An algorithm to predict road‐test failure was generated from the neuropsychological tests and validated on a cohort of 17 people with dementia |
| Lincoln 2010 | Non‐RCT Dementia cohort of 75 people with 34 AD, 14 VaD, 2 DLB, 1 FTD, 1 MCI, 4 mixed ‐ 65 had on‐road test. Dementia severity not recorded. 12 failed road test. Combination cognitive scoring gave sensitivity 80% and specificity 61.5% for predicting road test failure. Validated previous prediction rule correctly classifying 76% of people |
| Love 2007 | Non‐RCT |
| Lucas‐Blaustein 1988 | Non‐RCT Retrospective survey of driving history and crashes in 72 people referred to a dementia clinic |
| Lukas 2009 | Non‐RCT Review article |
| Man‐Son‐Hing 2007 | Non‐RCT Review article |
| Meuser 2006 | Non‐RCT, Trial of education for health professionals |
| Meuser 2009 | Non‐RCT Driving outcomes of reported impaired drivers in Missouri. Reporting resulted in lower crash rate and mortality in reported drivers ‐ 96.5% had license revoked |
| Molnar 2006 | Non‐RCT Review article |
| Molnar 2007 | Non‐RCT Review article |
| Mosimann 2011 | Non‐RCT Consensus statement |
| Neitch 2011 | Non‐RCT Retrospective audit of 43 assessments comparing performance on standardised testing of 27 with and 25 without dementia |
| O'Connor 2010 | Non‐RCT More rapid decline in life space and driving in MCI vs. healthy controls, no driving assessment |
| Odenheimer 1994 | Non‐RCT Cross‐sectional study of 30 licensed drivers with a broad range of cognitive skills (3 with AD, 3 with VaD), over age 60 years, were road tested on a closed course and in traffic by a driving instructor and 2 researchers. Driving instructor scores and cognitive test scores compared with research road test scores |
| Okonkwo 2009 | Non‐RCT Case‐control study comparing 57 MCI vs. 68 healthy controls, on‐road test as part of functional assessment. 70% vs. 50% had some difficulty on road test, non‐significant difference in tendency to overestimate ability. No reported test failures |
| Oswanski 2007 | Non‐RCT Cross‐sectional study of 232 drivers over 55 years referred for license renewal were given the Motor Free Visual Perceptual Test, clock test, and an on‐road driving test. No dementia diagnosis included |
| Ott 2005 | Non‐RCT Cross‐sectional study of 50 drivers with probable or possible AD (very mild to mild), from a longitudinal study of driving and dementia to compare office‐based clinician rating of driving competence with on‐road assessment. Clinician rated the drivers as safe, marginal or unsafe (based on chart review) and compared these ratings with total driving scores on a standardised road test and categorical ratings of driving competence from a professional driving instructor |
| Ott 2008 | Non‐RCT CCT: 65 probable AD, 23 possible AD (53 CDR 0.5, 35 CDR 1) and 45 healthy controls. In AD cases with CDI 0.5‐1 maze test and neuropsychological test were compared to on‐road testing. Road test failure noted in 1/45 controls and 17/88 AD. Road test score correlated with maze and neuropsychological tests |
| Parker 2000 | Non‐RCT Questionnaire survey of 1989 drivers aged 50 years or over (Manchester Driver Behavior Questionnaire) examined factors associated with crashes in older drivers. No dementia diagnosis |
| Patomella 2010 | Non‐RCT 205 impaired drivers referred for on‐road test. 128 had stroke, 43 with MCI and 34 with dementia |
| Rapoport 2007 | Non‐ RCT Review article |
| Rebok 1994 | Non‐RCT Case‐control study of 22 older drivers (10 with AD and 12 controls) compared on neuro‐psychological tests, the Driver Performance test (subjects answer questions based on videotaped driving scenarios) and Driver Advisement System (computer‐based measures of reaction time and coordination using steering wheel and pedals in front of monitor) |
| Reger 2004 | Non‐RCT Review article |
| Regtuijt 2007 | Non‐RCT Case report |
| Rizzo 1997 | Non‐RCT Case‐control study of 39 drivers (21 with AD and 18 controls) comparing neuro‐psychological testing with driving simulator performance and simulator crash rate |
| Rizzo 2001 | Non‐RCT Same design and participants as Rizzo 1997 but with hazards at intersections requiring rapid response to avoid crash |
| Ross 2009 | Non‐RCT Pooled data from 9 longitudinal studies in ageing, 5206 older adults no intervention, no assessment of driving ability |
| Silva 2009 | Non‐RCT Review article |
| Singh 2007 | Non‐RCT Cross‐sectional study of 154 individuals with PD (22 with dementia: diagnosed by team consensus based on MMSE, 15 road sign recognition, visuospatial construction, Trail Making test, forward and reverse digit span, and a story recall) referred to a driving assessment centre. Group included 17 who stated that they had stopped driving by themselves before assessment. Driving ability determined by a combination of clinical tests, reaction times on a static test rig and an in‐car driving test |
| Snellgrove 2005 | Non‐RCT Cross‐sectional study of 115 community‐dwelling older drivers with MCI or early dementia (recruited through a memory clinic). Participants completed the Maze Task, and immediately thereafter, a standardised on‐road driving test. Results also compared with cognitive tests from memory clinic |
| Snyder 2009 | Non‐RCT Retrospective audit of 1664 impaired drivers mandatorily reported, 88% had cognitive impairment, reported by internists and general practitioners. Auto suspension all ‐ 15% requested re‐test, 3% contested and only 10% regained driving privileges. No data on impact on crashes or cessations |
| Soderstrom 2009 | Non‐RCT Retrospective audit of reasons for referral by police to medical advisor boards. 40% of all referrals for confusion disorientation but dementia only in 3% |
| Staplin 2003 | Non‐RCT Same as Staplin 2003a with further 1‐year follow‐up data for crashes |
| Staplin 2003a | Non‐RCT Prospective cohort study of 2500 drivers over 55 years of age comparing cognitive and motor performance with prospective crash rate over 2 years following assessment. No dementia diagnosis |
| Stutts 1998 | Non‐RCT Retrospective cohort study of 3238 drivers aged 65 years and older applying for renewal of their driver's license. Compared cognitive assessments with crash involvement during the 3‐year period prior to testing. No dementia diagnosis |
| Szlyk 2002 | Non‐RCT Initial survey of 292 licensed neuropsychologists on neuropsychological tests currently used to screen people for driving. Followed by development of 12‐item neuropsychological battery used to screen 22 licensed drivers ranging in age from 67 to 91 years. Subjects were administered the neuropsychological battery, a driving simulator test and a Driving Habits Interview. No dementia diagnosis (subject classified as suspected dementia based on MMSE < 24) |
| Taylor 2001 | Non‐RCT Survey of 922 drivers whose licenses had been revoked or suspended due to dementia or suspected dementia. Questioned on the changes in household travel patterns and responsibilities following license revocation and on the psychological effects of these changes on affected household members. The survey was addressed to and in nearly all cases completed by the 'carer' of the former driver |
| Trilling 2001 | Non‐RCT Review article |
| Trobe 1996 | Non‐RCT Retrospective case‐control study of 143 licensed people with AD and 715 licensed comparison subjects matched 5 to 1 in age (± 6 years), sex and county of residence. Compared crashes and violations from state driving records with neuropsychological test scores. A questionnaire‐based inquiry on the influence of physician, family and state interventions on driving cessation was administered |
| Tuokko 1995 | Non‐RCT Retrospective case‐control study of 249 people referred to an outpatient dementia clinic. People were divided into those who met criteria for dementia and those who did not and compared with controls for retrospective crash and traffic violation rate. For each group, control subjects matched on age, gender and location of residence were randomly selected from the records of all drivers in the province |
| Uc 2005 | Non‐RCT Case‐control study of 33 drivers with probable AD of mild severity and 137 neurologically normal older adults comparing a battery of visual and cognitive tests in addition to driving safety and detection of specific landmarks and traffic signs along a segment of an on‐road experimental drive |
| Uc 2006 | Non‐RCT Case‐control study of 61 drivers with mild AD and 115 elderly controls using a driving simulator comparing crash rate and unsafe avoidance behaviour in a driving situation with a high risk of rear end collision |
| Uc 2008 | Non‐RCT Review article |
| Vaux 2010 | Non‐RCT Prospective case control study: 8 people with PD, 6 with AD and 18 healthy controls. Simulated drive on desktop computer tasked to identify whether moving spheres on a collisions course with them |
| Wadley 2009 | Non‐RCT Prospective case control stud: 46 MCI vs. 59 healthy controls compared on driving. MCI had more "less than optimal" performance on global measures but not outright fail |
| Wild 2003 | Non‐RCT Case‐control study of 15 drivers with mild AD and 15 controls comparing questionnaire responses on self reported cognitive function and driving performance in addition to an on‐road test |
| Withaar 2000 | Non‐RCT Review article |
| Woolf 2006 | Non‐RCT Review article |
| Zuin 2002 | Non‐RCT Retrospective case control study of 56 drivers with dementia (43 probable AD, 6 possible AD, 3 VaD, 1 DLB, 2 FTD, 1 unclassified). and 31 elderly controls. Motor vehicle crash and abnormal driving behaviour rates obtained from collateral informants and compared with detailed neurological, psychiatric and neuropsychological examinations |
AD: Alzheimer's dementia; ASDE: Proprietory: Manufacturers name "General ASDE, SA" ; BVRT: Benton Visual Retention Test; CCT: controlled clinical trial; CDR: Clinical Dementia Rating; CFT: Complex Figure Test; DLB: dementia with Lewy body; FTD: fronto‐temporal dementia; HVLT: Hopkins Verbal Learning Test; MCI: mild cognitive impairment; MMSE: Mini‐Mental State Examination; MVA: motor vehicle accident; NS: not significant; OR: odds ratio; PD: Parkinson's disease; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status.; RCT: Randomised Controlled Trial ; TMT: Trail‐Making Test ; UFOV: Useful Field of VIew; VaD: vascular dementia.