Identifying crash involvement among older drivers: agreement between self-report and state records

Abstract

Older drivers have a high crash rate per vehicle mile of travel. Coupled with the growth of the number of older drivers on the road, this has generated interest in the identification of factors which place older drivers at increased risk. However, much of the existing research on medical and functional risk factors for crash involvement has generally been inconsistent. Methodological differences between studies have been hypothesized as being partly responsible for such inconsistencies. The source of information used to identify crash-involved drivers has been identified as one such difference. This paper reports on the agreement between self-report and state record for identifying crash involved-older drivers. We also sought to determine whether the prevalence of visual and cognitive impairment differs across crash-involved drivers identified by either or both sources. Finally, we assessed whether risk factors for crash involvement differed when crash-involved drivers were identified by either self-report or state records. Results indicated that there was a moderate level of agreement between self-reported and state-recorded crash involvement (kappa=0.45). However, we did find significant differences between crash-involved drivers identified via state records and/or self-report with respect to demographic (age, race), driving (annual mileage, days per week driven), and vision impairment (acuity, contrast sensitivity, peripheral visual field sensitivity, useful field of view). We also found that the possibility for biased measures of association is real. Useful field of view impairment was associated with both self-reported and state-recorded crash involvement; however, the magnitude of the associations was disparate. Moreover, glaucoma was identified as a significant risk factor when considering state-recorded crashes but not self-reported crashes. While validation of these findings is required, research designed to identify risk factors for crash involvement among older drivers should carefully consider the issue of case definition, particularly if self-report is used to identify crash-involved older drivers.

Introduction

Driving is a complex task requiring visual, cognitive and physical skills. Impairment in these skills is increasingly common in the later decades of life (Katz, 1983; Tielsch et al., 1990; Whitehouse, 1993), and has been associated with increased risk of crash involvement in older drivers (National Highway Traffic Safety Administration, 1989). Chronic diseases (e.g. diabetes, cardiovascular disease) which engender these types of functional impairments have also been identified as increasing crash risk for older adults (National Highway Traffic Safety Administration, 1989). After 16–25-year-olds, persons aged 70 and older have the second highest rates of motor vehicle collisions and fatal and non-fatal crash injuries per vehicle mile of travel (National Safety Council, 1993). With the aging of the U.S. population, there is a pressing need to understand the causes of crashes by older drivers so that preventive strategies can be developed.

Although relatively sparse, much of the research on medical and functional risk factors for motor vehicle crashes in the elderly has not produced consistent results. For example, three studies have assessed diabetes as a risk factor for crash involvement among older drivers (Foley et al., 1994; Gresset and Meyer, 1994a; Koepsell et al., 1994). Among them, one (Koepsell et al., 1994) found a significantly elevated, almost 3-fold risk, while the remaining studies (Foley et al., 1994; Gresset and Meyer, 1994a) found no such significant increase. Studies investigating medication use and crash risk among older drivers have also been equivocal. Although a number of studies have found elevated crash rates for drivers using benzodiazepines (Hemmelgarn et al., 1997; Honkanen et al., 1980; Neutel, 1995; Ray et al., 1992; Skegg et al., 1979), other studies have produced null results (Benzodiazepine/Driving Collaborative Group, 1993; Jick et al., 1981; Leveille et al., 1994). It has been suggested that these discrepancies may be partly methodologic in origin (Koepsell et al., 1994; Ray, 1992). Differences in study populations, data collection techniques (e.g. interview, medical record), and availability of information on potentially important confounding variables (e.g. driving exposure) have been identified as such methodological discrepancies. Another methodological issue relates to case definition (i.e. how crash-involved drivers are identified). Among previously conducted studies, case definitions have included self-reported crashes (Hoffstetter, 1976; Marottoli et al., 1994; Shinar, 1977), crashes involving injury to the driver (Koepsell et al., 1994; Leveille et al., 1994; McCloskey et al., 1994), and state-recorded crashes where the driver was found to be at-fault (Ball et al., 1993; Owsley et al., 1991). Several studies have also utilized a mixture of injury, fault and/or self-report when defining cases (Gresset and Meyer, 1994a, Gresset and Meyer, 1994b; Marottoli et al., 1994). In the majority of studies conducted to date, the impact of case definition on study results has not been given due consideration (see Marottoli, 1997; Owsley, 1997for discussion). When attempting to identify risk factors for any health problem, including cataract, neural tube defects and hypertension, heterogeneity among cases can potentially obscure or mask important determinants. Therefore, it is possible that discrepancies between prior studies utilizing different case definitions are attributable to the fact that risk factors are not uniform across different case definitions. For example, the use of self-report rather than state records for identifying crashes may only identify a fraction of all crashes. If this fraction is somehow different with respect to risk factors of interest, the end result will be measures of association that are biased.

Marottoli et al. (1997)recently discussed the issue of the agreement between self-report versus state records for the identification of crashes among older drivers. These authors concluded that self-report and state records provide complementary information and that the former may provide a reasonable alternative to state records. Others have been less sanguine about the agreement between self-report and state records (Ball et al., 1993; Owsley et al., 1991). For example, in their study of older drivers in Alabama, Ball et al. (1993)found a low correlation (r=0.11) between the raw number of self-reported and state-recorded crashes during a 5-year period. However, when subjects were categorized into groups according to none versus one or more self-reported crashes and none versus one or more state-recorded aggregate crashes, the level of agreement between self-report and state records was moderate (kappa=0.40). Thus, the apprehension on the part of some researchers may be partly related to the manner in which sources of information on crash involvement are compared. Nevertheless, agreement between sources is only part of the issue. Of greater concern is the potential for important risk factors to be over- or under-represented among crash-involved subjects identified via self-report or state records. To our knowledge, no studies to date have directly addressed this issue.

The objective of this report is three-fold. First, we estimate the level of agreement between self-reported and state-recorded crashes among a sample of older drivers. Second, we evaluate whether the prevalence of visual and cognitive impairment differs across three groups of older crash-involved drivers: those with crashes that were both self-reported and state-recorded, self-reported but not state-recorded, and not self-reported but state-recorded. Third, we assess whether risk factors for crash involvement differed when crash-involved drivers were identified by either self-report or state records. We focus on visual and cognitive factors, because driving is a complex task obviously involving visual and cognitive skills for its successful execution, and impairment of these skills has been associated with increased risk of crash involvement in older drivers (Ball et al., 1993; Cooper et al., 1993; Decina and Staplin, 1993; Johansson et al., 1996; Johnson and Keltner, 1983; Kahneman et al., 1973; Marottoli et al., 1994; Owsley et al., 1991; Transportation Research Board, 1988).