Practical Experience-Based Approaches to Assessing Fitness to Drive in Dementia

Practical Experience-Based Approaches to Assessing Fitness to Drive in Dementia



Note: If you are a practicing Canadian physician, you can request a free subscription to Geriatrics & Aging. You may request your subscription by following this link.

Frank J. Molnar, MSc, MDCM, FRCPC, Canadian Institutes of Health Research (CIHR) CanDRIVE Research Team, Clinical Epidemiology Program, University of Ottawa Health Research Institute; Division of Geriatric Medicine, Department of Internal Medicine, University of Ottawa; Division of Geriatric Medicine, the Ottawa Hospital; REVTAR Research Group and CT Lamont Centre for Primary Care Research, Élisabeth-Bruyère Research Institute, Ottawa, ON.
Anna M. Byszewski, MD, FRCPC, CIHR CanDRIVE Research Team; Division of Geriatric Medicine, Department of Internal Medicine, University of Ottawa; Division of Geriatric Medicine, the Ottawa Hospital, Ottawa, ON.
Mark Rapoport, MD, FRCPC, CIHR CanDRIVE Research Team; Department of Psychiatry,
University of Toronto; Sunnybrook Health Sciences Centre, Toronto, ON.
William B. Dalziel, MD, FRCPC, Division of Geriatric Medicine, Department of Internal Medicine, University of Ottawa; Division of Geriatric Medicine, the Ottawa Hospital; the Regional Geriatric Program of Eastern Ontario, Ottawa, ON.

There may be up to 1.5 million persons with dementia who are driving in North America. In many jurisdictions, physicians are mandated to assess and report fitness to drive in such patients. Lack of knowledge of patients’ driving status does not protect physicians from lawsuits. There is a paucity of research to aid physicians in the assessment of fitness to drive in persons with dementia. Guidelines recommend the Mini-Mental State Examination, the clock-drawing test, and Trails A and B but lack evidence-based instructions regarding how to interpret such tests. This article provides experience-based approaches to the assessment of fitness to drive in dementia as well as an approach to disclosure of the findings to patients.
Key words: dementia, Alzheimer, driving, family physicians, cognitive testing.

While the majority of older drivers remain safe drivers, a subset experience the cumulative functional effects of medical conditions (e.g., dementia, strokes, arthritis, Parkinson’s disease) and medications (i.e., those with sedating properties) that impact on their fitness to drive.1

In North America, there are estimated to be 3.4 million people with dementia; if the published estimated proportion of persons with dementia who are driving2 is correct, this suggests that there are more than 1.5 million drivers with dementia. In Canada, there are now an estimated 500,000 people with dementia, with an expected 250,000 new cases to be diagnosed over the next 5 years. As our population ages, the number of persons with dementia who are driving is also expected to escalate.2

In many jurisdictions front-line physicians are responsible for reporting patients who have medical conditions that may impact on fitness to drive. These legal reporting duties vary by province and territory and can be found in the Canadian Medical Association’s driving guidelines (available at ).3 What is less clear is how to determine which patients are unsafe to drive during assessments in front-line clinical settings (e.g., physicians’ offices).4

This is particularly true in the field of dementia. A recent systematic review revealed that no cognitive tests have cut-off scores that are validated to determine fitness to drive status in dementia.5 Consequently, the Canadian Institutes of Health Research has funded a 5 year longitudinal prospective cohort study to develop and validate screening tools for fitness to drive that can be employed by physicians in their offices ( The study will begin recruiting this year and results can be expected in 5-7 years. When such validated screening tests are available they will still need to be employed within the framework of clinically sensible approaches such as those that will be presented in this article.

Pending the results of such research, we are left to refer to consensus guidelines that, due to a lack of evidence, are largely based on individual expert opinion or the consensus of small groups of experts.3,6 Such guidelines tend to recommend tests such as the Mini-Mental State Examination (MMSE),7-16 the clock-drawing test, and the Trail Making Test (Trails A and B),7,16-19 none of which have well validated cut-off scores predicting fitness to drive in dementia, and some of which have conflicting published data.5 Consequently, the guidelines cannot provide evidence-based information regarding how to interpret the cognitive tests recommended (i.e., what would represent fatal errors on these tests or which validated cut-off scores to employ).5

This article presents the practical approaches that we developed for the in-office screening and assessment of medical fitness to drive in persons with dementia.4,20-22 The approaches presented in this article are based on a combination of clinical guidelines and clinical acumen and experience. They represent the attempts of seasoned clinicians to incorporate clinical guidelines into approaches that can be employed in busy clinical practices. The approaches have been refined via an ongoing iterative process of discussion and debate among us and our many clinical and research colleagues. The approaches represent our current opinions regarding the best approach to employ in this evidence-based vacuum. Consequently, readers must use their own judgment to decide how to use the approaches described in their own clinical practices.

Assessment of Fitness to Drive in Dementia
When caring for persons with dementia, it is necessary to ask if they drive. A lack of knowledge of patients’ driving status does not legally protect physicians should these patients become involved in at-fault motor vehicle crashes. To the contrary, a precedent has been set as physicians have been successfully sued when their patients were involved in crashes due to neurological conditions, even when the physicians were unaware that the patients were active drivers.23,24

Moderate-to-Severe Dementia
When cognitive impairment is so severe or obvious that it is clearly unsafe for the patient to continue driving, in-depth testing is not needed.

Mild-to-Moderate Dementia
The diagnosis of dementia does not, however, automatically mean that a person cannot drive. Some people with mild dementia may still be able to drive safely for a limited period of time, but require individualized assessment and periodic follow-up.3,6 Attempts to mandate that all persons with dementia should be forced to cease driving regardless of whether they are still safe or not, aside from being legally unsupportable, could inadvertently increase the risk to the general public. Such draconian measures could result in more people with dementia avoiding a diagnostic assessment which might thereby result in more people with undiagnosed dementia continuing to drive (i.e., patients whose unfitness to drive might have been detected during the diagnostic assessment).

For less severe cases, clinicians need to decide if they have enough information to make a clinical decision regarding fitness to drive. The Canadian Medical Association driving guidelines3 and the Canadian Consensus Conference on Dementia guidelines25 indicate that persons with moderate to severe dementia should not drive, and they employ an opinion-based definition of moderate to severe dementia as demonstrating new impairments (relative to the patient’s baseline) due to cognition in one or more personal activities of daily living and/or two or more instrumental activities of daily living (see Table 1).

The assessment of fitness to drive in persons with mild dementia is complex and should take into account not only cognitive issues but also other medical and physical reasons indicating that they are unfit to drive. Driving cessation is often more acceptable or palatable to such patients if the decision is also based on physical (i.e., noncognitive) findings. We propose two different methods to organize the complex array of factors impacting on driving (see Tables 2 and 3). The approaches are not as lengthy to apply as they may first appear. Primary care physicians with an in-depth longitudinal knowledge of a patient will be able to answer many of the questions listed in these approaches before meeting with the patient for a more focused examination of fitness to drive. The initial elements of such a focused examination, for example, points 1-5 in Table 3, may answer the question of fitness to drive; in this case, further assessment (e.g., points 6-10, Table 3) may not be necessary. In many instances, the approach suggested in Table 3 may only take 10 minutes to complete.


These approaches are heavily based on history and physical examination. Many clinicians may prefer to start with cognitive tests. When physicians employ cognitive tests such as the MMSE, clock-drawing test and/or Trails A and B, they should keep in mind that none of these tests have well-validated cut-off scores for persons with dementia (and when validated, such cut-off scores will likely be averages and may vary by individual). It is, therefore, recommended that clinicians use their judgment to trichotomize the results of these tests into categories of “clearly safe,” “unclear--needs more testing,” or “clearly unsafe” by asking themselves if they would get into or allow a loved one in a car that the patient is driving, given the tests results.5 As presented in point 8 of Table 3 (Trails B) and Figures 1 and 2, the unclear category may be further evaluated by considering qualitative dynamic information regarding how the test was performed (e.g., observations such as slowness, hesitation, multiple corrections, anxiety, impulsive or perseverative behaviour, lack of focus, forgetting instructions, inability to understand test, etc., may facilitate more precise judgment of this category). Given the lack of research on validated cut-off scores, and on trichotomization in general, where to set the cut-off scores remains dependent on physician judgement pending further research.5 The trichotomization approach essentially asks, “Which patients are obviously unfit to drive, which are clearly safe, and which require further evaluation?”

What to Do if Fitness to Drive Remains Unclear
If fitness to drive remains unclear after performing assessments such as those described in Tables 2-3 and Figures 1-2, then physicians should refer for further evaluation. Referral to a centre specializing in the diagnosis and treatment of dementia should be considered if there are dementia-related issues other than driving to also consider (i.e., there are insufficient resources in dementia clinics to handle large numbers of referrals purely for assessment of fitness to drive). If fitness to drive is the only issue to be addressed then referral to a centre providing specialized on-road testing would be more appropriate (in regions where such centres exist).



This recommendation comes with a caveat. In some provinces the ministry of transportation will not accept their own on-road tests as being sufficient to assess persons with cognitive impairment. Rather, the ministry of transportation requires that a more comprehensive on-road evaluation be performed at specialized ministry certified centers that are often run by occupational therapists. The high costs of these specialized comprehensive on-road tests ($500-800 to be paid by the patient in some provinces) create a barrier to the assessment and reporting of fitness to drive as they place physicians in the position of presenting patients with an ultimatum; pay for such expensive on-road tests or stop driving. This type of interaction is destructive to the physician-patient relationship and is unfair to patients of limited financial means. Systems in which patients have to pay for on-road testing discourage physicians from assessing and reporting fitness to drive and may thereby unintentionally create a risk to public safety. Some provinces such as British Columbia have addressed this by funding comprehensive on-road testing for patients with dementia if the physician recommends such on-road testing to the ministry of transportation and the ministry agrees with this recommendation. In Quebec on-road testing only costs patients $80. Ideally all provincial and territorial ministries of transportation should fund comprehensive on-road testing for persons with dementia in the way British Columbia and Quebec do. Regrettably, most ministries of transportation are not themselves adequately funded by their province to undertake this responsibility. If we, as a society, want to have safer roads then we must ask our provincial governments to better fund our ministries of transportation so they, in turn, can fund comprehensive on-road testing.

Another approach would be to consider which organizations would benefit financially from better funded comprehensive on-road testing. When people are involved in car crashes (as drivers, passengers, pedestrians, or drivers and passengers of other cars), it is the ministries of health and the insurance industry that pay the extremely high immediate and long-term costs of care and disability. The health care system and the insurance industry could potentially save tax payers and investors millions of dollars by funding comprehensive on-road testing or by sharing the costs with the ministries of transportation (i.e., a tripartite payer system including the insurance industry, ministries of health, and ministries of transportation). Such forward thinking could save both lives and money.



After the Assessment: Approaching a Person with Mild Dementia who Is Still Temporarily Safe to Drive
If a person with mild dementia is found to be able to continue to drive safely, physicians should still broach the subject of eventual driving cessation when the dementia progresses (as it inevitably will). Fitness to drive must then be re-evaluated every 6-12 months.3,29 If the clinician is concerned that the patient may not return for re-evaluation, then it would be prudent to report the patient to the ministry of transportation as “having mild dementia, but being deemed still safe to drive with re-evaluation required in 6-12 months (period for re-evaluation dependent on physician judgment).” The physician also has the option of specifying the type of follow-up required (e.g., in the physician’s office, by a specialist, or via comprehensive on-road assessment) when completing this form.

After the Assessment: Disclosing That a Person Is Unsafe to Drive
Once fitness to drive has been assessed, if the findings suggest an unacceptable risk, they must be acted on. Many clinicians find the disclosure of unfitness to drive to be a difficult, if not painful, task that fundamentally alters the physician-patient relationship. They understandably express a desire to avoid this potentially confrontational situation as they fear it will emotionally harm patients and may result in these patients, and their families, leaving their practice.27,28 As outlined in the Canadian Medical Association guidelines, physicians in most provinces are legally required to assess and report persons with dementia who are unsafe to drive.3 Even in jurisdictions where reporting is not mandated, it is still possible for physicians to be sued if their patient with dementia injures others in a car crash. Disclosure becomes unavoidable. However, as in many areas of medicine, the manner in which bad news is disclosed can moderate the negative impact on patients and families. Table 2 presents an approach that has been employed clinically by one of authors (F.M.) and that has formed the basis for presentations given on behalf of the Ontario Alzheimer Knowledge Exchange (accessible on the Exchange’s dementia and driving resource webpage at ). Once a physician has disclosed a finding of unfitness to drive, it is generally prudent to also provide the finding in writing to the patient and family as the patient may forget the conversation. A sample letter is provided in Figure 3. For legal reasons, the disclosure meeting (including the date and participants’ names) should be documented in the patient’s chart.


By employing approaches such as those presented in Tables 2 and 3, clinicians with baseline knowledge of a patient can assess fitness to drive in a relatively short period of time and can appropriately select only those patients who truly need referral for further in-depth assessment of fitness to drive. By not referring patients whose fitness to drive can be determined in the primary physician’s office, our system will be able to better adapt to the rapidly growing numbers of older drivers who truly require specialized assessment of fitness to drive. To preserve public safety, provinces must better fund their ministries of transportation to allow these ministries to, in turn, fund comprehensive on-road testing for the escalating number of persons with mild dementia whose fitness to drive cannot be determined without an on-road test. To do otherwise will perpetuate the disincentives to physician assessment and reporting of fitness to drive described above and will place the general public at unnecessary risk.

For those interested in learning more regarding the evaluation of fitness to drive in dementia, we recommend the Ontario Alzheimer Knowledge Exchange dementia and driving resources available at, and the Dementia and Driving Toolkit, available on the Regional Geriatric Program of Eastern Ontario website at

No competing financial interests declared.




  1. 1. Parker D, McDonald L, Rabbitt P, et al. Elderly drivers and their accidents: the Aging Driver Questionnaire. Accid Anal Prev 2000;32:751-9.
  2. Hopkins RW, Kilik L, Day DJA, et al. Driving and dementia in Ontario: a quantitative assessment of the problem. Can J Psychiatry 2004;49:434-8.
  3. Canadian Medical Association. Determining Medical Fitness to Operate Motor Vehicles. CMA Driver’s Guide, 7th edition. (available at
  4. Molnar FJ, Byszewski AM, Marshall SC, et al. In-office evaluation of medical fitness-to-drive: practical approaches for assessing older people. Can Fam Physician 2005;51:372-9.
  5. Molnar FJ, Patel A, Marshall S, et al. Clinical utility of office-based predictors of fitness to drive in persons with dementia: a systematic review. J Am Geriatr Soc 2006;54:1809-24.
  6. American Medical Association, U.S. Department of Transportation, and National Highway Traffic Safety Administration. Physician’s guide to assessing and counseling older drivers. Washington (DC): National Highway Traffic Safety Administration, 2003; Accessed February 3, 2009.
  7. Friedland RP, Koss E, Kumar A, et al. Motor vehicle crashes in dementia of the Alzheimer type. Ann Neurol 1988;24:782-6.
  8. Lucas-Blaustein MJ, Filipp L, Dungan C. Driving in patients with dementia. J Am Geriatr Soc 1988;36:1087-91.
  9. Gilley DW, Wilson RS, Bennett DA, et al. Cessation of driving and unsafe motor vehicle operation by dementia patients. Arch Intern Med 1991;151:941-6.
  10. Trobe JD, Waller PF, Cook-Flannagan CA, et al. Crashes and violations among drivers with Alzheimer disease. Arch Neurol 1996;53:411-6.
  11. Rebok GW, Keyl PM, Blysma FW, et al. The effects of Alzheimer disease on driving-related abilities. Alzheimer Dis Assoc Disord 1994;8:228-40.
  12. Harvey R, Fraser D, Bonner D, et al. Dementia and driving: results of a semi-realistic simulator study. Int J Geriatr Psychiatry 1995;10:859-64.
  13. Cox DJ, Quillian WC, Thorndike FP, et al. Evaluating driving performance of outpatients with Alzheimer disease. J Am Board Fam Pract 1998;11:264-71.
  14. Fitten LJ, Perryman KM, Wilkinson CJ, et al. Alzheimer and vascular dementias and driving. JAMA 1995;273:1360-5.
  15. Bieliauskas LA, Roper BR, Trobe J, et al. Cognitive measures, driving safety, and Alzheimer ’s disease. Clin Neuropsychol 1998;12:206-12.
  16. Fox GK, Bowden SC, Bashford GM, et al. Alzheimer’s disease and driving: prediction and assessment of driving performance. J Am Geriatr Soc 1997;45:949-53.
  17. Hunt L, Morris JC, Edwards D, et al. Driving performance in persons with mild senile dementia of the Alzheimer type. J Am Geriatr Soc 1993;41:747-53.
  18. Duchek JM, Hunt L, Ball K, et al. Attention and driving performance in Alzheimer’s disease. J Gerontol B Psychol Sci Soc Sci 1998;53B(2):130-41.
  19. Rizzo M, Reinach S, McGehee D, et al. Simulated car crashes and crash predictors in drivers with Alzheimer’s disease. Arch Neurol 1997;54:545-51.
  20. Byszewski A, Molnar F, Aminzadeh F. The impact of disclosure of unfitness to drive in persons with newly diagnosed dementia: patient and caregiver experiences Clin Gerontol 2009. In press.
  21. Byszewski AM, Graham ID, Amos S, et al. A continuing medical education initiative for Canadian primary care physicians: the Driving and Dementia toolkit: a pre and post evaluation of knowledge, confidence gained and satisfaction. J Am Geriatr Soc 2003;51:1484-9.
  22. Rapoport M, Zucchero Sarracini C, et al. Driving with dementia: how to assess safety behind the wheel. Curr Psychiatr 2008;7:37-48.
  23. Capen K. New court ruling on fitness-to-drive issues will likely carry “considerable weight” across country. CMAJ 1994;151:667.
  24. Capen K. Are your patients fit to drive? CMAJ 1994;150:988-90
  25. Hogan DB, Bailey P, Carswell A, et al. Management of mild to moderate Alzheimer’s disease and dementia. Alzheimers Dement 2007;3:355-84.
  26. Rapoport MJ, Herrmann N, Molnar FJ, et al. Psychotropic medications and motor vehicle collisions in patients with dementia. J Am Geriatric Soc 2008;56:1968-70.
  27. Marshall SC, Gilbert N. Saskatchewan physicians’ attitudes and knowledge regarding medical fitness to drive. CMAJ 1999;160:1701-4.
  28. Jang RW, Man-Son-Hing M, Molnar FJ, et al. Family physicians’ attitudes and practices regarding assessments of medical fitness to drive in older persons. J Gen Intern Med 2007;22:531-43.
  29. Molnar FJ, Patel A, Marshall SC, et al. Systematic review of the optimal frequency of follow-up in persons with mild dementia who continue to drive. Alzheimer Dis Assoc Disord 2006;20:295-7.

Practical Experience-Based Approaches to Assessing Fitness to Drive in Dementia

Practical Experience-Based Approaches to Assessing Fitness to Drive in Dementia


Frank J. Molnar, MSc, MDCM, FRCPC, Canadian Institutes of Health Research (CIHR) CanDRIVE Research Team, Clinical Epidemiology Program, University of Ottawa Health Research Institute; Division of Geriatric Medicine, Department of Internal Medicine, University of Ottawa; Division of Geriatric Medicine, the Ottawa Hospital; REVTAR Research Group and CT Lamont Centre for Primary Care Research, Élisabeth-Bruyère Research Institute, Ottawa, ON.
Anna M. Byszewski, MD, FRCPC, CIHR CanDRIVE Research Team; Division of Geriatric Medicine, Department of Internal Medicine, University of Ottawa; Division of Geriatric Medicine, the Ottawa Hospital, Ottawa, ON.
Mark Rapoport, MD, FRCPC, CIHR CanDRIVE Research Team; Department of Psychiatry,
University of Toronto; Sunnybrook Health Sciences Centre, Toronto, ON.
William B. Dalziel, MD, FRCPC, Division of Geriatric Medicine, Department of Internal Medicine, University of Ottawa; Division of Geriatric Medicine, the Ottawa Hospital; the Regional Geriatric Program of Eastern Ontario, Ottawa, ON.

There may be up to 1.5 million persons with dementia who are driving in North America. In many jurisdictions, physicians are mandated to assess and report fitness to drive in such patients. Lack of knowledge of patients’ driving status does not protect physicians from lawsuits. There is a paucity of research to aid physicians in the assessment of fitness to drive in persons with dementia. Guidelines recommend the Mini-Mental State Examination, the clock-drawing test, and Trails A and B but lack evidence-based instructions regarding how to interpret such tests. This article provides experience-based approaches to the assessment of fitness to drive in dementia as well as an approach to disclosure of the findings to patients.
Key words: dementia, Alzheimer, driving, family physicians, cognitive testing.

Évaluation de l’aptitude à conduire chez les personnes atteintes de démence

Évaluation de l’aptitude à conduire chez les personnes atteintes de démence


Évaluation de l’aptitude à conduire chez les personnes atteintes de démence

Conférencier : Frank Molnar, M. Sc., MDCM, FRCPC, membre du personnel, service de gériatrie, Hôpital d’Ottawa; professeur agrégé, département de médecine, Université d’Ottawa; chercheur affilié, Institut de recherche en santé d’Ottawa; scientifique, Institut de recherche Élisabeth-Bruyère, Ottawa (Ontario).

Le Dr Frank Molnar, membre du réseau de chercheurs interdisciplinaires du programme CanDRIVE (The Canadian Driving Research Initiative for Vehicular Safety in the Elderly), a présenté les approches pratiques permettant d’évaluer l’aptitude à conduire un véhicule dans le cadre d’un diagnostic de démence.

La recherche du programme CanDRIVE

L’équipe de recherche du programme CanDRIVE, subventionné par les Instituts de recherche en santé du Canada (IRSC), a effectué des recherches substantielles et proposé d’importantes recommandations sur l’aptitude à conduire, en s’appuyant sur une double approche. Le premier objectif du groupe fut de trouver et de valider des outils de dépistage pour cette population de patients. Ce travail fondamental a nécessité la constitution d’une équipe nationale de recherche pour examiner les aspects médicaux de l’aptitude à conduire. Cette collaboration entre divers professionnels de la santé, qui a abouti à la mise en œuvre de tests comportant des valeurs seuils en fonction des données de groupe, a permis de passer au second volet de leur objectif : faciliter le réseautage et la transmission des savoirs (aboutissant à l’ajustement des valeurs seuils et à l’utilisation de résultats spécifiques pour éva-luer chaque patient). CanDRIVE va effectuer une grande étude prospective de cohortes pour suivre les aptitudes à la conduite d’adultes atteints de démence.

Alors que la recherche primaire en est à ses débuts, le Dr Molnar a précisé que les équipes de recherche de CanDRIVE s’attacheront à transmettre les connaissances acquises aux médecins. De plus, CanDRIVE aspire à tenir compte des opinions des cliniciens sur les recherches à effectuer.

L’étendue du problème
Bien que les conducteurs âgés soient généralement plus prudents que les cohortes plus jeunes, le taux d’accidents de véhicules motorisés par kilomètre parcouru en fonction de l’âge du conducteur augmente après l’âge de 80 ans. Ce décalage, principalement dû à l’accumulation de problèmes médicaux à un âge avancé, signifie que le nombre de collisions et de victimes va prendre de l’ampleur avec le vieillissement de la population. Les salles d’urgence auront à traiter un nombre croissant de personnes âgées victimes d’accidents de la route.

Les projections de Transport Canada jusqu’en 2026 indiquent que les accidents vont surtout augmenter chez les personnes âgées. Impliquées dans un accident, ces dernières ont une probabilité quatre fois plus élevée d’être grièvement blessées ou hospitalisées, et sont plus susceptibles d’être atteintes d’une incapacité permanente ou de décéder. Elles se réta-blissent également plus lentement. Les études ont montré que la majorité des personnes âgées victimes d’accidents de la route conduisaient le véhicule, et que la plupart des accidents impliquant des conducteurs âgés mettent en cause plusieurs véhicules et touchent des innocents. Le Dr Molnar a invité ses auditeurs à considérer que ce n’est pas une faveur que de permettre aux conducteurs âgés de conduire lorsqu’ils atteignent la limite de leur aptitude à le faire sans danger.

Ce n’est pas seulement un problème d’âge
La vaste majorité des conducteurs âgés sont des conducteurs prudents, a insisté le Dr Molnar, et CanDRIVE est sensible au fait que son travail pourrait contribuer à l’âgisme ou à l’alarmisme concernant les aînés au volant. Il a souligné que les troubles médicaux et les traitements médicamenteux sont la première cause de l’incompétence des conducteurs âgés, et que tout médicament peut contribuer au risque de collision. Les personnes âgées sont affectées de façon disproportionnée en raison de la polypharmacie.

Le Dr Moldar a déclaré explicitement qu’il est impossible d’associer de manière catégorique un danger à une maladie ou un trouble chronique précis. Ce ne sont pas les troubles qui sont dangereux, mais leur gravité ou leur instabilité, auxquelles s’ajoutent des doses élevées ou les changements de doses de médicaments. Les médecins ne peuvent pas prévenir chaque accident, mais ils sont bien placés pour identifier de nombreuses personnes susceptibles d’être dangereuses au volant. Les troubles médicaux ou les médicaments qui sont les plus corrélés à une diminution de l’aptitude à la conduite sont ceux qui modifient les capacités physiques, sensorielles, mentales ou émotionnelles.

La conduite met en jeu un ensemble complexe de capacités et d’attitudes cognitives, notamment des catégories d’action opérationnelles, tactiques et stratégiques. Les médecins ne peuvent pas évaluer correctement à 100 % une fonction altérée, en raison des limites de l’examen physique (conçu principalement pour détecter la présence de maladies, et non pour évaluer la fonction ou la sécurité) et du manque de temps adéquat dans un cadre clinique de première ligne. Par exemple, les décisions prises au volant reposent sur l’aptitude des conducteurs à faire des choix tactiques, assistant les décisions contextuelles en temps réel. Si des déficiences existent dans cette catégorie, elles sont souvent difficiles à dépister par le médecin. Les médecins peuvent difficilement éva-luer la capacité stratégique, à savoir les décisions prises avant de se mettre au volant. Aucun outil de dépistage ne sera jamais entièrement efficace pour prévenir tous les accidents de véhicules motorisés. La plupart des protocoles d’évaluation ne testent que des caractéristiques intrinsèques stables de l’aptitude à conduire. Les médecins peuvent ne pas détecter une maladie récente ou fluctuante, ou anticiper le discernement de leurs patients sur des facteurs extrinsèques comme la météo, les autres conducteurs, les conditions routières ou la sécurité d’une voiture.

Évaluation clinique : survol des enjeux

On peut cependant améliorer l’évaluation, et CanDRIVE cherche à attirer l’attention sur la question, car il est bien établi qu’une altération de la fonction cognitive augmente le risque d’accident avec responsabilité pour les conducteurs. Une étude de 2004 a révélé qu’il y a actuellement des dizaines de milliers de conducteurs âgés atteints de troubles démentiels en Ontario; en 2028, ils approcheront les 100 000.

Un diagnostic de démence ne signifie pas automatiquement une interdiction de conduire, a déclaré le Dr Molnar. Cependant, un tel diagnostic signifie que le clinicien doit demander si la personne conduit encore, et la sécurité de sa conduite doit être évaluée et établie. Les exigences provinciales en matière de déclaration varient, mais elles spécifient invariablement que le trouble doit être évalué et déclaré.

Les recommandations de l’Association médicale canadienne (AMC), « Évaluation médicale de l’aptitude à conduire : Guide du médecin » (7e éd.) rejoignent les déclarations consensuelles internationales qui reconnaissent les limites des données disponibles sur l’évaluation, mais recommandent les directives suivantes : 1) interdiction de conduite pour les personnes atteintes de démence modérée ou grave (AMC : démence modérée = 1 AVQ ou 2 AIVQ altérées par un problème cognitif); 2) évaluation individuelle des personnes qui présentent une démence légère; 3) suivi périodique obligatoire (tous les 6 à 9 mois); 4) évaluation complète au volant (la norme de référence en matière d’évaluation). Le Dr Molnar a émis l’avis que les recommandations de l’AMC devraient aller plus loin en ce qui concerne les AVQ : toute altération des AIVQ dans le domaine cognitif devrait entraîner automatiquement une évaluation de l’aptitude à conduire. En outre, il trouve que la règle d’un suivi tous les 6 à 9 mois n’est pas assez flexible, et conseille une approche personnalisée (p. ex. : une évaluation tous les 3 mois dans le cadre d’une maladie évoluant rapidement).

Le Dr Molnar a attiré l’attention sur le fait que, malgré l’existence de protocoles d’évaluation clairs (comment utiliser le MMSE, le test de l’horloge, la partie B du test des tracés, etc.), aucune indication n’est fournie au médecin sur la façon d’appliquer de tels tests. Par exemple, comment réagir à différents scores, quelles valeurs seuils utiliser, et quelles erreurs constituent automatiquement un échec? Ces questions font encore l’objet de débats. CanDRIVE a examiné des douzaines d’articles sur la démence et la conduite, mais n’a pas trouvé un seul test cognitif dont l’analyse s’appuie sur une valeur seuil validée. Les cliniciens travaillent dans un vide de données prouvées.

Une approche pour évaluer l’aptitude à la conduite
Les cliniciens doivent d’abord poser la question : « Est-ce que vous conduisez? » L’absence d’une telle vérification n’a pas protégé les cliniciens en procès, a averti le Dr Molnar. Ensuite, souvenez-vous que l’aptitude à la conduite repose sur un tableau clinique global, comprenant la cognition, le bilan fonctionnel, les capa-cités physiques, l’état de santé, le comportement et le dossier de conducteur du patient. Et puis, après les questions générales, réalisez des tests cognitifs spécifiques. Les renseignements corroborés par la famille peuvent être utiles, et le Dr Molnar a suggéré plusieurs pistes d’enquête qu’il vaut mieux aborder en l’absence du patient (Tableau 1). De plus, prenez en compte les états pathologiques qui, lorsqu’ils sont graves, mal contrôlés ou en évolution rapide, peuvent compromettre l’aptitude à la conduite (il a invité les cliniciens à se poser la question : « Est-ce que je monterais dans une voiture conduite par cette personne sur la foi de ces résultats? »). Le Dr Molnar a recommandé d’attacher la plus haute importance aux « 3 D » (démence, délire et dépression). Il a ensuite passé en revue les médicaments pouvant affecter la conduite (Tableau 2).

Il est essentiel de tester des domaines cognitifs spécifiques, comme le font les protocoles susmentionnés. Le jugement est apprécié par la réponse à des tests éva-luant la réaction à des situations dangereuses (comme le feu); les capacités visuo-spatiales sont testées grâce au MMSE et au test de l’horloge; la fonction exécutive est évaluée par les parties A et B du test des tracés, le test de l’horloge et le test de fluence verbale (nommer des animaux pendant une minute); et le temps de réaction peut être vérifié par le test de rattrapage de la règle lâchée. Dans le cas où les scores cognitifs se chevauchent ou ne sont pas clairs, le Dr Molnar a plaidé pour une trichotomisation sérielle (p. ex., conducteur vraiment dangereux, résultat incertain nécessitant d’autres tests et aucun souci pour la sécurité; figure 1).

Il a recommandé de commencer par le MMSE : les patients obtenant un score inférieur à 20 représentent probablement un danger au volant. Tous les tests mentionnés apportent des données précieuses; le test de rattrapage de la règle lâchée, bien que non validé, est important pour évaluer les réactions. De tels tests sont très utiles parce que les décréments du temps de réaction (qui n’est pas évalué lors de l’examen physique) ne deviennent souvent apparents qu’en dehors des tests, où les laps de temps se comptent en secondes. Néanmoins, la conduite implique des temps de réaction de l’ordre de la milliseconde.

Le Dr Molnar a terminé en insistant sur le fait qu’en cas de diagnostic de démence, il convient de s’interroger sur l’aptitude à conduire, et en faire une évaluation formelle et bien documentée. Les médecins peuvent effectuer une évaluation clinique approfondie de la sécurité au volant en 15 à 20 minutes. Si les cliniciens ne sont pas convaincus d’une telle sécurité, il convient d’orienter le patient vers des évaluations spécialisées ou des tests spécialisés sur route. En cas de démence, il faut réévaluer la sécurité au volant tous les 6 à 9 mois. Finalement, il a invité toute personne ayant des idées concernant l’évaluation de la conduite de les porter à l’attention du personnel de CanDRIVE par l’intermédiaire du site Web du programme (

Assessment of Fitness-to-Drive in Persons with Dementia

Assessment of Fitness-to-Drive in Persons with Dementia


Click here to view the entire report from the 28th Annual Scientific Meeting of the Canadian Geriatrics Society

Assessment of Fitness-to-Drive in Persons with Dementia

Speaker: Frank Molnar, MSC, MDCM, FRCP(C), Staff, Division of Geriatric Medicine, The Ottawa Hospital; Associate Professor, Department of Medicine, University of Ottawa; Affiliate Investigator, The Ottawa Health Research Institute; Scientist, The Elisabeth Bruyere Research Institute, Ottawa, ON.

Dr. Frank Molnar, a member of the network of interdisciplinary investigators for the Canadian Driving Research Initiative for Vehicular Safety in the Elderly (CanDRIVE), reviewed practical approaches to assessing fitness to drive in the setting of a dementia diagnosis.

CanDRIVE Research

The Canadian Institutes of Health Research (CIHR)-funded CanDRIVE research team has worked to produce substantive research and recommendations on fitness to drive via a two-pronged approach. First, the group has aimed at deriving and validating screening tools for this patient segment. This pillar of their work has involved building a national research team to examine medical aspects of fitness to drive in conjunction with an array of health professionals (leading to the development of tests with cut-offs based on group data), in order to pursue the second pillar of their purpose: facilitating networking and knowledge translation (leading to adjusting cut-offs and using specific findings to assess individual patients). CanDRIVE will conduct a large prospective cohort study that tracks fitness to drive among adults with dementing illness.

While the primary research is in its infancy, Dr. Molnar explained that the CanDRIVE research teams will focus on disseminating their acquired knowledge to physicians. Further, CanDRIVE aims to incorporate input from clinicians on what they should be looking at.

The Scope of the Problem
While older drivers are generally safer when compared to younger cohorts, the rate of motor vehicle crashes per km driven according to driver’s age increases beyond age 80. This shift, due primarily to the accumulation of medical illnesses in late age, means the net number of collisions and casualties will soar with an aging populace. Emergency rooms will treat an increasing number of older crash victims.

Projections through 2026 from Transport Canada show that crashes will rise primarily in older groups. An older person involved in a crash has a fourfold higher likelihood of being seriously injured or hospitalized; has a higher risk of becoming permanently disabled or dying; and takes longer to recover. Studies have shown that the majority of crash-injured seniors were driving the vehicle, and that most of the crashes involving older drivers are multivehicle and involve innocents. Dr. Molnar asked listeners to consider that they are not doing older drivers a favour by letting them drive when they reach the limit of their ability to do so safely.

The Problem Is not Age Alone

The vast majority of older drivers are safe drivers, Dr. Molnar insisted, and CanDRIVE is sensitive to the concern that their work contributes to ageism or alarmism about seniors at the wheel. He emphasized that medical conditions and medications are the primary cause of older drivers’ incompetence, and any medication can contribute to collision risk. Older people are affected disproportionately due to polypharmacy.

Dr. Molnar explicitly stated that no disease or chronic condition can be isolated as categorically risky. It is not the presence but severity and/or instability of conditions, plus high doses and/or changing doses of medications, that are perilous. While physicians cannot prevent every accident, they are well-placed to detect many persons who are at risk for unsafe driving. Qualities of medical conditions or medications most correlated with impaired driving capacity are those that alter physical, sensory, mental, or emotional abilities.

Driving recruits a complex set of cognitive capacities and behaviours, including operational, tactical, and strategic categories of action. Doctors cannot correctly assess impaired function 100% due to limitations of the physical exam (which is primarily designed to detect presence or absence of disease, not to assess function or safety) and the inadequate time available in front-line clinical settings. For example, tactical maneuvering is involved in decisions drivers make on the road—it supports real-time contextual decisions. Impairments in this category are often hard for doctors to catch. Strategic capacity refers to decisions made before getting on the road, which is difficult for doctors to assess. No screening tool will ever be completely effective for screening for all motor vehicle crashes. Most assessment protocols only test stable intrinsic features of driving ability. Doctors may miss new or fluctuating illness. Further, physicians cannot anticipate patients’ judgment of extrinsic factors such as weather, other drivers, road conditions, or a car’s safety.

Clinical Assessment: An Overview of the Issues

Assessment can be improved, however CanDRIVE seeks to galvanize attention on this issue as it is well-documented that cognitive impairment puts drivers at increased risk of at-fault crashes. A 2004 study found that currently there are tens of thousands of older drivers with dementing illnesses in Ontario; by 2028, the figure will approach 100,000.

A diagnosis of dementia does not automatically mean no driving, Dr. Molnar stated; however, a diagnosis of dementia means that the clinician must ask if the person is still driving, and driving safety must be assessed and documented. Provincial reporting requirements vary but uniformly state that the condition must be assessed and reported.

The Canadian Medical Association’s (CMA) guidelines, “Determining Medical Fitness to Operate Motor Vehicles” (7th ed.), joins international consensus statements that recognize the limitations of available data on assessment but recommend that: one, those with moderate to severe dementia should not drive (CMA: Moderate = 1 ADL or 2 IADLs impaired due to cognition); two, that individual assessment should be performed for those with mild dementia; three, that periodic follow-up is required (every 6-9 months); and four, the “gold standard” is comprehensive on-road assessment. Dr. Molnar opined that the CMA guidelines should go further in terms of ADLs—any single IADL impairment due to cognition should trigger an assessment of fitness to drive. Further, he finds the 6-9 month follow-up rule insensitive and advises an individualized approach (e.g., assessing every 3 months in the setting of rapidly progressing disease).

Dr. Molnar cautioned that while clear assessment protocols are given (e.g., for using the MMSE, Clock Drawing, Trails B), no guidance is provided as to how physicians should apply such tests. For example, how to respond to different scores, what cut-offs to use, and which errors equal automatic failure remain under debate. CanDRIVE has examined dozens of dementia and driving articles, and was unable to find one cognitive test that was analyzed via a validated cut-off. Clinicians are working in an evidence-based vacuum, Dr. Molnar stated.

An Approach to Assessing Fitness to Drive
Clinicians must inquire, “Do you drive?” Failure to verify has not protected clinicians in litigation, Dr. Molnar advised. Two, recall that driving capacity depends on a global clinical picture, including the patient’s cognition, function, physical abilities, medical conditions, behaviour, and driving record. Then, follow general questions with specific cognitive tests. Corroborative information from the family can help, and Dr. Molnar suggested several areas of inquiry that are best asked when the patient is not in the room (Table 1). Further, review medical conditions that when severe, poorly controlled, or changing rapidly can compromise capacity to drive (he suggested clinicians ask themselves, “Would I get in a car with this person based on these findings?”). Dr. Molnar recommended that the “3 Ds”— dementia, delirium, and depression— are most important to consider. He then reviewed medications that could affect driving (Table 2).

The key intervention is to test specific cognitive domains, as with the aforementioned protocols. Judgment is assessed by the test response to dangerous situations (e.g., fire); visuospatial ability is tested with the MMSE and clock drawing; executive function is assessed with Trails A and B, clock drawing, and 1-minute animal naming; and reaction time can be verified with the ruler drop test. In the case of overlapping/unclear cognitive scores, he argued for serial trichotomization (e.g., clearly unsafe, uncertain with further testing required, no concerns regarding safety), as shown in Figure 1.

He advised that the MMSE is the best place to start; patients scoring under 20 are likely unsafe to drive. All areas of testing mentioned yield valuable data; the ruler drop test, while not validated, is important in assessing reaction. Such tests are valuable because decrements of reaction time (which is not tested in the physical examination) often only become apparent outside of testing when lapses involve seconds. However, driving involves the need for reactions on a millisecond scale.

Dr. Molnar closed with emphasizing that if dementia is diagnosed, driving must be asked about, formally assessed, and documented. Physicians can perform a comprehensive driving safety clinical evaluation in approximately 15 to 20 minutes. If clinicians are unsure of safety, refer to specialized assessment or specialized on-road testing. In dementia, reassess driving safety every 6 to 9 months. Finally, he encouraged those with any ideas about driving assessment to bring them to the attention of CanDRIVE staff via their website (

Driving, Cancer and Discrimination

Driving, Cancer and Discrimination


At the time of the writing of this editorial, there is a 'high profile' inquest going on in Toronto concerning driving and the elderly. Two years ago, an elderly woman making a right hand turn struck and killed a young woman. The young woman was then dragged under the car for almost a kilometre with the driver apparently unaware. There was no suggestion that the elderly driver had any physical or cognitive impairment that affected her driving. However, despite the absence of cognitive impairment, this was felt to be a case that could raise the profile of cognitive impairment and the aging driver. The inquest has not concluded, but fortunately initial testimony has stressed that most elderly drivers are competent to drive.

The same day that my testimony at this inquest was reported in the papers, another story was reported, more gruesome than the first. A 25-year-old Texas woman struck a homeless man, impaled him on her windshield, and then locked him and the car in the garage while he slowly bled to death over two or three days. She and her friends then removed the body and 'dumped' it in a garbage bin. For some reason, the first case has sparked an intense interest in whether or not the elderly should drive, but I have not read or heard any musing about restricting the driving privileges of 25-year-olds. Perhaps all young people should have random drug testing to maintain their driving privileges (a presumed factor in the Texas incident)!

Clearly, the difference in the two cases from a geriatrician's perspective is as follows: The incident with the elderly driver is immediately generalized to reflect all the elderly, whereas the incident with the young driver is a reflection of her actions, and her actions alone. In the first case, the trial judge last year pronounced that the woman's ability to drive was 'impaired by age.' I have yet to identify any evidence that shows age is an independent risk factor for driving. Rather, it is the morbidity that accompanies aging that impairs driving. I suspect that any slowing of reaction time and reflexes in the elderly is more than compensated by better judgment and increased caution. Even though we know that a large number of the over 80 population has cognitive impairment, we do not have accurate information on how many still drive, vital information to have if any screening endeavours are considered.

This issue focuses on cancer and the elderly and, as I have discussed in the past, the presumption is often made that the elderly should be treated less aggressively than should younger patients, even though comorbidity is a more important factor than age alone. The lesson, brought home once again by this inquest, is that management must be tailored to the individual and based on comprehensive assessment, not just a single factor such as age.

Fortunately, in this issue we feature articles by experts who do not fall prey to age bias. Dr. Townsley and Dr. Hedley discuss pancreatic cancer in the elderly, and other articles address the issues of cardiac tumours (Desai and Butany), ovarian cancer (Gould and McMeekin), male breast cancer (Glück and Friedenreich), and screening for colorectal cancer (Rossos and Yeung). As well, we have our usual assortment of other articles, including a special piece on estrogen and the aging brain by Elise Levinoff and Dr. Howard Chertkow, one of Canada's leading investigators in the field of cognitive impairment.

Enjoy this issue.

Alberta Researcher Develops Competence Screen for Drivers with Dementia

Alberta Researcher Develops Competence Screen for Drivers with Dementia


Michelle Durkin, BSc

How to determine when people suffering from Alzheimer's disease and other dementias should stop operating a motor vehicle is a complex and controversial issue. The decision can affect the patient (by impairing independence and mobility), his or her family, and the safety of the general public. As a greater percentage of the population ages, the number of functionally impaired drivers only increases, further complicating the issue.

The Canadian Medical Association's Guide for Physicians in Determining Fitness to Drive (5th edition) states that physicians should monitor the driving competence of a patient with dementia. Until now this has been difficult because an appropriate and accurate evaluation tool has not been available. Physicians relied solely on their own judgement. Now, however, the research of Dr. Allen Dobbs of the University of Alberta may provide this needed, effective evaluation tool with the development of a computer-based test called the Competence Screen.

In an interview, Dr.