Elsevier

Accident Analysis & Prevention

Volume 45, March 2012, Pages 634-638
Accident Analysis & Prevention

Cognitive screening of older drivers does not produce safety benefits

https://doi.org/10.1016/j.aap.2011.09.032Get rights and content

Abstract

Although screening policies for older drivers based on chronological age are widely used in many countries, previous research has shown that increasing age does not cause higher crash rates and that consequently, chronological age per se is at best only a weak predictor of safe driving performance. Previous research on age-based mandatory screening of older drivers has not been able to demonstrate any safety benefits from screening measures.

The present study is a population-based evaluation of the safety effects that the introduction of the cognitive test as an age-based screening tool has had in Denmark. The primary data used came from the Danish road accident register. The present study compared the number of fatal accidents before and after the implementation of screening for cognitive impairment.

There were two main findings. First, there was no statistically significant difference in the number of older drivers involved in fatal accidents before and after the implementation of the screening process, indicating that the screening had no effect on the safety of older drivers. Second, there was a significant increase in the number of unprotected older (but not younger) road users who were killed between the two periods of observation, suggesting that the screening process produced a modal shift among older persons from driving to unprotected, significantly less safe modes of transportation. As a consequence, the number of fatalities in this group increased.

Older driver screening is an example of a political measure that intuitively makes sense, but fails to produce the desired benefits. On the contrary, on a system level, it decreases the overall safety and is connected to various direct and indirect costs.

Highlights

► Cognitive test was introduced as an age-based driver screening tool in Denmark. ► Cognitive screening had no effect on the safety of older drivers. ► After the screening fatalities among unprotected, older road users increased. ► Screening a measure that intuitively makes sense, but fails to produce the desired benefits.

Introduction

As the size of the older population is increasing, managing older road users’ safe mobility has become a topic of interest in most Western industrialised countries. Previous research has found mobility and older individuals’ ability to use the transportation system independently to be closely related to different aspects of well-being and health (Farquhar, 1995, Marottoli et al., 1997, Fillenbaum, 1985). Providing satisfactory opportunities for independent travel and mobility helps the older population to maintain an independent lifestyle and their own well-being. As the safest and most convenient mode of transportation for older persons, private cars are often seen as the best option for independent mobility (OECD, 2001). The question of the fitness and safety of older drivers has, however, also been widely discussed, resulting in a debate about the meaningfulness of screening as well as various measures the aim of which is to identify those older drivers who “no longer are fit to drive” (e.g. Fain, 2003, Fitten, 2003, White and O’Neill, 2000).

Already Hakamies-Blomqvist et al. (2002) demonstrated in their seminal paper that an increase in age does not cause higher crash rates per exposure. Since then, this finding has been repeatedly confirmed by independent studies (Fontaine, 2003, Keall and Frith, 2006, Langford et al., 2006), thus challenging the traditional conception of the age-related deterioration of safety-relevant driving skills. Chronological age per se seems to be, in the case of mature drivers, at best only a weak predictor of safe driving performance. However, screening policies based on chronological age are widely used in most European countries and many US and Australian states (e.g., Insurance Institute for Highway Safety, 2011, Langford et al., 2004b, Meuser, 2008, Mitchell, 2008, White and O’Neill, 2000).

The safety effects of screening older drivers have been studied in many different contexts and with different study designs. Rock (1998) compared the accident rates in the state of Illinois before and after the license renewal rules were revised. For the younger age group (69–74 years old), the rules were eased by removing the earlier mandatory driving test from the license renewal procedure. For the older age group (81+ years old), the rules were tightened and the license renewal was set to every second year (every year for drivers aged 87+), where it had previously been every 4 years. The conclusion was that no safety effect could be observed; the new rules did not lead to an increase in the number of accidents caused by younger drivers, nor did it contribute to a decrease in the number of accidents involving older drivers.

Three American studies (Grabowski et al., 2004, Lange and McKnight, 1996, Levy et al., 1995) compared accident rates in different states with different age-based license renewal policies. Levy et al. (1995) compared 50 states analysing Fatality Analysis Reporting System (FARS) accident statistics, and found only testing for visual acuity to be somewhat related to lower accident rates, while adding theory tests to the renewal procedure had no effect. In a study by Lange and McKnight (1996), per-driver accident rates in states with age-based license renewal systems were compared with neighbouring states without such tests, and the results showed that the rates of accidents involving older drivers were actually higher in those states with age-based testing. A more recent study by Grabowski et al. (2004) investigated a number of factors including in-person renewal, vision tests, road tests and the frequency of license renewal (which vary in different states) as predictors of older driver safety. The study showed that the only predictor of a lower accident rate was in-person renewal (as opposed to renewal by post), and that this effect could be seen only for those aged 85 years and older. Additional tests, regardless of whether they were medical or tests of practical driving skills, had no effects on safety.

A Finnish-Swedish study (Hakamies-Blomqvist et al., 1996) compared accident rates in Sweden, where there is no age-related screening, and Finland, where all drivers from the age of 70 onwards have to go through a medical check in order to renew their license. The results did not show any safety benefits resulting from the Finnish system, and, on the contrary, showed that Finland had a higher pedestrian fatality rate after the age of 70. The authors argued that by providing a modal shift from being a car driver to a unprotected road user (pedestrian, cyclist, moped rider), the screening indirectly caused an increase in the number of unprotected road users who were killed, and concluded that the age-based mandatory screening thus produced an overall negative safety effect.

An Australian study (Langford et al., 2004a) compared older (80+ years old) drivers in Melbourne, where no age-based screening is used, and Sydney, where drivers have to undergo a medical check and a driving test at the age of 80. No safety benefit could be observed for the seniors in Sydney in this study. Another Australian study (Langford et al., 2004b) compared older driver accident statistics in six Australian states with different licensing policies. The results showed that the older driver accident rates were lowest in Victoria, the only state without age-based mandatory screening. Furthermore, Langford et al. (2008) studied the effects that screening policies for older drivers had on other road users’ risk of being killed by an older driver. They found that age-based mandatory assessment programmes did not have demonstrable safety benefits, in terms of either the total number of fatalities or the number of deaths of other road users.

In the European context, Mitchell (2008) has compared seven EU-member states with different older driver licensing policies. He found that the accident rates for older drivers were lower in countries that had fewer requirements for license renewal in old age.

All in all, previous research indicates that mandatory age-based screening has no positive safety effects. On the contrary, the research suggests that stopping older people from driving may have negative safety effects, as they tend to shift from driving a car to more high-risk modes of transport. To our knowledge, no studies have evaluated the mobility effects of screening policies directly, but we know that screening policies contribute to premature driving cessation, especially among older women (Hakamies-Blomqvist and Wahlström, 1998, Siren et al., 2004).

In Denmark, a driving license is valid until the driver reaches the age of 70. Thereafter, it needs to be renewed at the age of 70, 74, 76, 78 and 80. After the age of 80, the license has to be renewed every year. In order to renew their license, a person has to go through a medical check performed by a physician. If the physician is undecided about the applicant's fitness to drive, the applicant undergoes a practical driving test. As of May 1st 2006, a short version of the mini-mental examination (m-MMSE) and the clock-drawing test were added to the medical check in order to identify drivers with cognitive impairments. The costs and benefits of the implementation of these tests have not been evaluated. Before its implementation, the new screening test was piloted in one area in Denmark (Sønderjylland, an area with ca. 250,000 inhabitants), resulting in an increase in the number of drivers who did not renew their license. It was assumed that those who did not renew their licenses were those people who were likely to be an elevated risk in traffic. Thus, the overall conclusion was that the measure was likely to succeed and produce safety benefits (Hansen and Hansen, 2002).

The aim of the present study is to evaluate the safety effects, which the implementation of the cognitive test as a screening tool has had in Denmark since it was introduced. The introduction of a new measure has provided a unique quasi-experimental setting, enabling us to carry out a population-based evaluation of the effects.

Section snippets

Accidents

The primary data for the present study were collected from the Danish accident database. This database holds information on all police-reported road accidents in Denmark. Representativeness is best for fatal accidents, as practically all of these are reported to the police and consequently recorded in the database. Several details of the accidents are recorded, such as the time and date, the location of the crash, the type of crash, the characteristics of the parties involved and the types of

Results

During the first period of observation, fewer drivers were involved in fatal accidents than during the second period (Table 2). This applied to both age groups; the increase was ca. 7% for drivers aged 18–69 and ca. 14% for drivers aged 70 and older. However, this increase was not statistically significant for the younger (χ2 = 2.8601, df = 1, p < 0.10) or the older (χ2 = 1.1095, df = 1, p < 0.25) group.

The number of unprotected road users who were killed in road accidents increased in both age groups (

Discussion

The purpose of the present study was to evaluate the safety effects of a new type of driver screening programme, which was introduced in Denmark in 2006. There were two main findings. First, there was no statistically significant difference in the number of drivers involved in fatal accidents in the periods before and after the implementation of the programme. This applied to both older and younger drivers, and indicates that the screening programme had no effect on the safety of the group of

Conclusion

Ideally, the implementation of political measures should be evidence-based. It is important to assess, when possible, the costs and benefits of different measures for society. Screening programmes for older drivers are an example of a political measure that seems to make sense, but that fails to produce the desired benefits. On the contrary, on a traffic systems level, this measure decreases overall safety and is connected to various direct and indirect costs. The focus on managing older road

Acknowledgements

Financial support from the TrygFonden foundation is gratefully appreciated. The authors would like to thank data analyst Hjalmar Christiansen from DTU Transport Model Centre for his valuable comments on the data analysis, and Professor Liisa Hakamies-Blomqvist for her valuable and constructive comments on a previous version of the manuscript.

References (33)

  • P-A. Rimmö et al.

    Older drivers’ aberrant driving behaviour, impaired activity, and health as reasons for self-imposed driving limitations

    Transportation Research Part F: Traffic Psychology and Behaviour

    (2002)
  • S.M. Rock

    Impact from changes in Illinois driver's license renewal requirements for older drivers

    Accident Analysis and Prevention

    (1998)
  • L.K.M. Donorfio et al.

    A qualitative exploration of self-regulation behaviors among older drivers

    Journal of Aging and Social Policy

    (2008)
  • M.J. Fain

    Should older drivers have to prove that they are able to drive?

    Archives of Internal Medicine

    (2003)
  • J.E. Fillenbaum

    Screening the elderly: a brief instrumental activities of daily living measure

    Journal of the American Geriatrics

    (1985)
  • L.J. Fitten

    Driver screening for older adults

    Archives of Internal Medicine

    (2003)
  • Cited by (63)

    • Transport Modes and an Aging Society

      2021, International Encyclopedia of Transportation: Volume 1-7
    View all citing articles on Scopus
    1

    Tel.: +45 35324895.

    View full text