Naturalistic Studies (Driving In Traffic) continued...
The AD patients performed significantly worse than the healthy elderly controls on all 10 driving behaviors, as rated by the experienced CDRS, who was blinded to the subjects’ cognitive status and group membership. AD patients rated their driving performance better than the evaluator’s ratings on 7 of the 10 items; healthy controls evaluated their performance better than that of the driving evaluator on only one item (manages intersections), and significantly worse on one other item (handles conversational distraction). AD patients’ self-reports of driving ability mirrored healthy elderly drivers’ self-reports of driving ability for 9 of the 10 items, indicating that patients with AD are not cognizant of their diminished driving capabilities. Caregivers’ perceptions of driving ability were similar to the evaluator’s assessments in all but two areas: managing intersections and responding to warning road signs.
Another study of the driving performance of individuals with AD using on-road testing was carried out by Duchek et al. (2003). In this investigation, 41 percent of older subjects with mild dementia of the Alzheimer’s type (CDR = 1) failed a 45-minute in-traffic road test, compared to 14 percent of older subjects with very mild AD (CDR = 0.5) and 3 percent of healthy older subjects (CDR = 0) . The mean age of the 108 participants was 74.9. Participants, who were recruited from the Alzheimer’s Disease Research Center at Washington University School of Medicine, knew that their driving skills would be assessed. They were screened for depression, reversible dementias, and other disorders that could produce cognitive impairment. Almost half of the potential subject pool declined participation in the study (28% with CDR = 0; 44.9% with CDR = 0.5; and 27.1% with CDR = 1). No significant differences were found in age, education, gender, and Short Blessed Test of cognitive abilities (a six-item test of orientation, concentration and memory) between study participants and those who declined participation.
Driving performance in the Duchek et al. (2003) study was assessed using the Washington University Road Test in a standard car with dual brakes. This open-course test was conducted in traffic and assessed the following driving skills: maintaining speed, obeying traffic signs, signaling, turning, yielding right of way, changing lanes, reacting to other drivers, and negotiating intersections. These skills were scored on a 2- or 3-point scale yielding a quantitative driving score between 0 and 108 (perfect performance) by a driving instructor who sat in the front passenger seat and the study investigator who sat in the back seat. In addition, the two raters independently assigned a global rating of safe, marginal, or unsafe as an expert clinical impression of driving performance. The inter-rater reliability of the global rating was high (k=0.85).
Subjects participated in four road tests, administered at 6-month intervals. Two driving behaviors—lane change and using signals—were impaired with increasing dementia severity. Only three driving behaviors showed a significant decline from the first to the second test administration—qualitative judgments (e.g., comprehension of directions, attention to task, awareness), reacts to others (e.g., an awareness of how one’s driving affects others), and speed control. These were independent of CDR group status. The authors note that these behaviors represent the more complex cognitive skills involved in driving (e.g., awareness of the driving environment and decision making) rather than the specific mechanics of driving, such as signaling. These three behaviors were also highly correlated with the global ratings during the first test administration.
Analysis of driving performance over time indicated that subjects without AD (CDR = 0) took significantly longer to receive a rating of not safe than subjects with mild dementia (CDR=1). Subjects with very mild dementia (CDR = 0.5) fell somewhere between the healthy and mild dementia groups. The majority of drivers with mild AD were judged unsafe either at the first test administration, or during follow-up testing within the two-year period. These findings point to aspects of driving performance that should be assessed in any on-road study to measure driver functioning among older people with possible cognitive impairment—including impairment from the use of medications.
De Raedt and Ponjaert-Kristoffersen (2001) used an on-road test to evaluate fitness to drive for 84 licensed, active older drivers (age 65 to 96) who were referred by their insurance company following one or more crashes or by their physician. One of the objectives of this study was to determine whether the road test was predictive of four types of at-fault crashes: crossroad, rear-end, side-swipe, or parking.