- Email: [email protected]
Impact of an educational program on the safety of high-risk, visually impaired, older drivers
Impact of an Educational Program on the Safety of High-Risk, Visually Impaired, Older Drivers Cynthia Owsley, MSPH, PhD, Gerald McGwin, Jr, MS, PhD, Janice M. Phillips, BS, Sandre F. McNeal, MPH, Beth T. Stalvey, MPH, PhD, CHES Background: Older drivers (licensed drivers aged 60 years and older) have among the highest rates of motor vehicle collision involvement per mile driven of all age groups. Educational programs that promote safe driving strategies among seniors are a popular approach for addressing this problem, but their safety benefit has yet to be demonstrated. The objective of this study was to determine whether an individualized educational program that promoted strategies to enhance driver safety reduces the crash rate of high-risk older drivers. Design/ Setting:
Randomized, controlled, single-masked intervention evaluation at an ophthalmology clinic.
Participants: A total of 403 older drivers with visual acuity deficit or slowed visual processing speed or both who were crash-involved in the previous year, drove at least 5 days or 100 miles per week or both, and were at least 60 years old. Intervention: Patients were randomly assigned to usual care (comprehensive eye examination) or usual care plus an individually tailored and administered educational intervention promoting safe-driving strategies. Main outcome Police-reported vehicle collision rate, expressed both in terms of person-years of follow-up measure: and person-miles of travel for 2 years postintervention. Results:
The intervention group did not differ significantly from the usual care only group in crash rate per 100 person-years of driving (relative risk [RR], 1.08; 95% confidence interval [CI], 0.71–1.64) and per 1 million person-miles of travel (RR, 1.40; 95% CI, 0.92–2.12). The intervention group reported more avoidance of challenging driving maneuvers and self-regulatory behaviors during follow-up than did the usual care only group (p⬍0.0001).
Conclusions: An educational intervention that promoted safe-driving strategies among visually impaired, high-risk older drivers did not enhance driver safety, although it was associated with increased self-regulation and avoidance of challenging driving situations and decreased driving exposure by self-report. (Am J Prev Med 2004;26(3):222⫺229) © 2004 American Journal of Preventive Medicine
Introduction
W
hen measured on a per-mile-driven basis, older drivers (licensed drivers aged ⱖ60 years) have a crash rate nearly equivalent to that of drivers younger than aged 25 years, whose crash rate is the highest among all age groups (approximately
From the Department of Ophthalmology, School of Medicine (Owsley, McGwin, Phillips, McNeal); Department of Epidemiology and International Health, School of Public Health (McGwin); Division of Trauma, Burns, and Surgical Critical Care, Division of General Surgery, Department of Surgery (McGwin), University of Alabama at Birmingham, Birmingham, Alabama; Office of Aging Policy and Information, Texas Department on Aging (Stalvey), Austin Texas Address correspondence and reprint requests to: Cynthia Owsley, MSPH, PhD, Department of Ophthalmology, School of Medicine, University of Alabama at Birmingham, 700 South 18th Street, Suite 609, Birmingham, AL 35294-0009. E-mail: [email protected].
222
15 crashes per million miles of travel).1,2 There are more older drivers on the road than in previous decades who have a higher mileage record than previous cohorts, with these trends increasing. Older drivers in a crash are more likely to be injured or die than are younger adults.3,4 Driving a vehicle is the preferred mode of travel by older adults in the United States.5,6 When older adults stop driving, they are at increased risk of depression and decreased quality of life.7–9 These findings stimulate a need to identify ways to enhance older driver safety. Organizations have developed educational programs to promote safe driving practices among seniors,10 –13 and these programs have widespread popularity. For example, in 1999 the American Association of Retired Persons (AARP) Driver Safety program (formerly called
Am J Prev Med 2004;26(3) © 2004 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/04/$–see front matter doi:10.1016/j.amepre.2003.012.005
55-Alive) had nearly 650,000 enrollees in 31,000 classes held nationwide.12 Programs for older drivers have several delivery designs, including classroom, videos, and workbooks. Only two programs have had their effectiveness evaluated in terms of reducing crash rate,10,14 demonstrating that the enrollees gained knowledge about safe driving and how aging affects skills important for driving. Neither program reduced crash rate compared with nonintervention groups (i.e., no demonstrated safety benefit). These findings are reminiscent of research on the effectiveness of driver education programs for both novice and postlicense drivers in that knowledge is successfully imparted, but there is no evidence that collision rate is reduced.15–19 It is premature to abandon the idea that educational interventions are effective in enhancing older driver safety. First, established risk factors for crash involvement for senior drivers—functional impairments because of medical conditions20,21—are different from those for young drivers in which impulsiveness, risk taking, and alcohol are contributory. These personalityrelated characteristics can be less responsive to educational intervention. However, it might be possible to teach an older driver motivated to remain behind the wheel about how his or her functional deficits affect driving skills and to facilitate the acquisition of compensatory strategies like self-regulation. Studies on health education illustrate that older adults are able to adopt health-protective behaviors and attitudes,22,23 implying that their use of new driving strategies is possible. Earlier work suggests high-risk older drivers acknowledge high self-efficacy in their ability to carry out compensatory driving strategies,24 and high selfefficacy in older adults is associated with behavior change.25 Another reason why the effectiveness of older driver education deserves further scrutiny is because previous evaluation studies10,14 did not focus on older drivers who were at a high risk for crash involvement. Most drivers had no crashes on their records in recent years and did not have significant functional impairments. The curriculum was group administered with standardized curriculum content for all enrollees, rather than individually tailored for the functional characteristics, driving needs, and attitudes of each driver. The aim of this randomized controlled study was to determine whether an individually administered and tailored health education program that promoted strategies to enhance driver safety reduces the crash rate of older drivers at high risk for collision involvement. High risk in this study was defined in terms of characteristics that are established risk factors for collisions by older drivers,2 namely older drivers who had visual impairment, had at least moderate levels of driving exposure, and had a recent crash.
Methods The protocol was approved by the institutional review board at the University of Alabama at Birmingham on March 11, 1997. The study was funded by General Motors Corporation pursuant to an agreement with the U. S. Department of Transportation (no. G.2). Informed consent procedures were followed. The recruitment population was licensed drivers aged 60 years and older in the Birmingham, Alabama, area who had been the driver in a crash in the prior year, according to the Alabama Department of Public Safety. Potential participants (N⫽8057) were contacted by letter and telephone to determine interest in participation and to confirm that they were legally licensed and active drivers. If so, they were invited to visit the Clinical Research Unit in the University of Alabama at Birmingham Department of Ophthalmology for further determination of eligibility. Other eligibility criteria consisted of being a driver at least 5 days per week or at least for 100 miles per week or both by self-report as determined by the Driving Habits Questionnaire,26 had a Mini-Mental State Examination (MMSE) score of 23 or greater,27 and had visual impairment. Visual impairment was defined as (1) visual acuity (habitual, binocular) between 20/30 and 20/60 (lower limit for licensure in Alabama) as measured by the Early Treatment of Diabetic Retinopathy (ETDRS) chart,28 or (2) a restriction in the useful field of view (score ⱖ40% reduction) as measured by the useful field of view test (measure of visual processing speed and attention)29,30 or both. General health was estimated through a questionnaire that asked about medical problems in 17 areas.31,32 Depressive symptoms were assessed by the Center for Epidemiological Studies Depression Scale (CES-D)33; participants rated 20 items on the basis of how often they felt in the past week the way described in each item (responses on a four-point scale from “none of the time” to “all of the time”). Total scores ranged from 0 to 60 (higher numbers represent more depressive symptoms). Social desirability (tendency to respond in a manner person believes will present himself or herself in a favorable light) was estimated by the Marlowe-Crowne scale,34 consisting of 20 statements that the participant indicated were true versus false. Total score was the number of items in which true was the response (higher scores indicate more social desirability).
Study Design Patients were randomly assigned to one of two groups by fixed randomization35 (55% to usual care plus intervention, 45% to usual care only). A target enrollment of 400 was based on a hypothesized 50% reduction of crash rate in the usual care plus intervention group with 80% power. The flow of the study protocol is depicted in Figure 1. All enrollees received usual care consisting of a free comprehensive eye examination by an optometrist to diagnose and treat eye problems and any relevant discussion with the patient about the implications of his or her conditions for everyday life.36 The intervention group consisted of usual care plus an educational intervention to promote driver safety (description follows). Because the screening process for inclusion had the potential for identifying vision impairment that was treatable, it was ethically important to send participants to usual care. If the eye examination resulted in the diagnosis of visual acuity
Am J Prev Med 2004;26(3)
223
Figure 1. Flow of participants through study. deficit or slowed processing speed that was reversed by subsequent treatment, then the person was no longer considered to be eligible for the study. Randomization to usual care versus usual care plus educational intervention occurred after the eye examination. The educational curriculum (KEYS, Knowledge Enhances Your Safety) was the basis of the intervention24,37,38 and has been presented in detail in a separate published report.37 Its main elements and approach are described here. The intervention was led by a health educator specializing in driver safety and gerontology, who created a tone of an encouraging yet frank atmosphere to create rapport and to facilitate communication. The curriculum was motivated by models of health behavior change and health promotion.39 – 41 It was administered one on one in two sessions in an eye clinic conference room, and it was individually tailored to the participant’s own driving needs, lifestyle, and visual problems as follows. During the first, 2-hour session, the educator led a discussion according to the results of his or her eye examination and the visual impairment identified at the study participant’s screening visit. The discussion began by promoting the participant’s self-awareness about his or her visual impairment and results of the eye examination. Information was imparted about how his or her type of visual impairment
224
poses problems for safe driving, causes driving performance problems, and elevates crash risk. This discussion was evidence based in that it focused on visual risk factors for crash involvement described in the scientific literature. A slide presentation of situations challenging for older drivers (e.g., driving in rain, at night, making left turns across on-coming traffic)1 was shown, enhancing discussion about how the participant’s visual skills are challenged in these situations. The discussion focused on maneuvers and settings that each participant found challenging or frustrating from his or her own perspective and then moved to strategies that the participant could implement to avoid these driving situations. Such strategies included structuring the day so that the need to run errands and travel by car was carried out during daylight rather than at dusk or night, driving the less popular routes, driving at non–rush hour times, postponing trips during inclement weather, seeking out routes in which three right turns could be substituted for a left turn at busy intersections, and selecting routes with protected left-turn arrows. Drivers were encouraged to identify strategies that could realistically be incorporated into their own routine driving. Before the session ended, participants set specific goals in a personal written contract that listed the ways they would try to modify their driving behavior (using strategies like those mentioned earlier). Session 2 of the educational intervention was administered 1 month after the first session in a 1-hour visit to the clinic. It was designed to be a booster session to review information from session 1 and to discuss progress toward achieving the contract’s goals. Driving exposure information was collected at the initial visit and then by telephone at 6-month intervals for the 2 years subsequent to randomization. The driving exposure section of the Driving Habits Questionnaire,26,42 a valid and reliable instrument for estimating driving exposure, was used for this purpose. The amount of driving done in a typical week was estimated by each participant through a structured interview that asked about the places driven to in a typical week as well as their distance from home. From this interview an estimate of annual mileage for each participant was computed. At baseline screening and at follow-up, questionnaires were administered that evaluated the subject’s strategy usage (discussed in the intervention curriculum) in terms of whether these strategies were incorporated into driving practices. A subscale of the Driving Perception and Practice Questionnaire (DPPQ)24 assessed the frequency with which the participant performed eight self-regulatory strategies at 6 and 18 months’ follow-up (e.g., driving at times other than rush hour, scheduling outings during daylight hours, making right turns around the block to avoid left turns across traffic). Participants indicated the frequency of performance by using a four-point scale (1⫽never, 2⫽rarely, 3⫽sometimes, 4⫽often). Total score is the sum of all eight items (range, 8 –24). A subscale of the Driving Habits Questionnaire26,38 assessed the self-reported avoidance of challenging driving situations (e.g., at night, in rain, at rush hour, heavy traffic roads) at the 6-, 12-, 18-, and 24-month follow-ups. Responses were on a five-point scale (1⫽never, 2⫽rarely, 3⫽sometimes, 4⫽often, 5⫽always). Total score is the sum of all eight items (range, 8 – 40). The examiner administering questionnaires was unaware of group assignment, demographics, screening characteristics, and questionnaire responses at the initial visit.
American Journal of Preventive Medicine, Volume 26, Number 3
Outcomes The primary outcome of interest was crash involvement during the 2-year follow-up, expressed as a rate, either person-years of follow-up or person-miles of travel (see Statistical Methods). Information on collisions in which the participant was the driver was obtained from accident reports provided by the Alabama Department of Public Safety. Secondary outcomes consisted of average weekly mileage and average days, trips, and places driven per week. Scores were evaluated at baseline and at follow-up on self-reported use of self-regulatory driving practices and avoidance of challenging driving situations.
Statistical Methods The intervention and usual care groups were compared according to demographic, behavioral, and clinical characteristics by using t tests and chi-square statistics for continuous and categorical variables, respectively. Crash rate for each group was computed by first enumerating the total number of collisions sustained by each group. Only collisions that occurred between each subject’s randomization date and 2-years hence were considered. Two separate denominators for crash rates were considered. The first denominator, person-years, was calculated as the amount of chronologic time elapsed between the date of randomization and 2 years subsequent, the date of death, or the date of driving cessation, whichever came first. The second denominator, personmiles, was calculated as the total amount of driving done by each subject between the same dates used to calculate personyears. Information on average weekly mileage was obtained at 6-, 12-, 18-, and 24-month follow-up interviews. Average weekly mileage reported at each follow-up was considered to be constant over the entire previous 6-month period. Thus, total person-miles of travel were calculated by summing each estimate of weekly mileage and multiplying this value by 26. For subjects with incomplete follow-up information but known to be alive, mileage information was carried forward from the previous follow-up. If a subject died or stopped driving, mileage was prorated until the date of death or driving cessation. Both sets of crash rates were statistically compared by using Poisson regression. For the secondary outcomes (average weekly mileage, days, trips, and places driven per week; self-regulatory driving practices; avoidance of challenging driving situations), the intervention and usual care groups were compared by using repeated measures analysis of variance. All data analyses were conducted on an intention-to-treat basis. pⱕ0.05 (two-sided) was considered statistically significant.
Results There were 987 persons screened for the study (Figure 1). Of these, 408 met the preliminary eligibility criteria and agreed to participate and underwent comprehensive eye examinations. Of the 408 participants, 5 had acuity impairment because of correctable refractive error; they were prescribed spectacles that improved vision to better than 20/30 in at least one eye; thus, they no longer met the vision impairment eligibility require-
ment for enrollment. The remaining 403 participants were randomly assigned to the two arms of the study, 227 to intervention and 176 to usual care only. After randomization, 30 subjects assigned to intervention never attended the educational sessions. For intervention, 197 attended the first session; 162 of these persons attended the booster second session. Figure 1 lists the number of subjects available for telephone follow-up questionnaires in both groups over the 2-year period. At 24 months, 87.2% of the intervention group and 85.8% of the usual care only group participated in the telephone interview. Table 1 presents baseline demographic, behavioral, and clinical characteristics of the two groups. Subjects in both groups averaged 73 years of age and were more likely to be men, to be white, and to have at least a high school education. MMSE scores for both groups were within normal limits with respect to the age and educational composition for adults aged 60 years and older. Depressive symptoms (CES-D scores), health status, visual function, social desirability, and driving exposure were comparable in the two groups. Figure 2 presents average mileage per week over the course of the follow-up period for the two study groups. Both the intervention and usual care groups reported declines in mileage; however, the rate of decline in the intervention group was significantly greater (p⫽0.02). Although both groups also reported declines in average days driving per week, they occurred similarly, yielding no significant difference between the groups (p⫽0.98). The intervention group reported a significant decline in average places driven per week after baseline (p⫽0.01) (Figure 3). A similar pattern was observed for average trips per week, but the difference between the groups did not achieve statistical significance (p⫽0.07). For both driving avoidance (Figure 4) and self-regulation scores (Figure 5), after baseline equivalence, the intervention group had significantly higher scores than the usual care group at each follow-up visit (both p⬍0.0001). Table 2 presents the crash outcomes over the follow-up period. Whether measured as person-years or person-miles, there was no significant difference in the rates of crash involvement between intervention and usual care groups (relative risk [RR], 1.08; 95% confidence interval [CI], 0.71–1.64 and RR, 1.40; 95% CI, 0.92–2.12, respectively). When the intervention group was limited to those drivers who participated in both sessions, the results still indicated no association (RR, 1.03; 95% CI, 0.52–2.02 for person-years and RR, 0.99; 95% CI, 0.51–1.88 for person-miles).
Discussion An educational intervention to promote driver safety in visually impaired older drivers did not reduce their rate of crash involvement in the following 2 years, even Am J Prev Med 2004;26(3)
225
Table 1. Baseline demographic, behavioral, and clinical characteristics among study groups
Characteristic Age, years Gender, % (n) Female Male Race, % (n) White African-American Years of education, mean (SD) Cognitive statusa, mean (SD) Social desirability, mean (SD) Depressive symptomsb, mean (SD) Comorbidity scorec, mean (SD) Visual acuity, mean (SD) Weekly driving, mean (SD) Mileage Days Places Trips
Intervention ⴙ usual care (n ⴝ 227) Mean (SD)
Usual care only (n ⴝ 176) Mean (SD)
p
73.7 (6.0)
73.3 (6.2)
0.5136
34.2 (77) 65.8 (148)
27.3 (48) 72.7 (128)
0.1360
76.9 (173) 23.1 (52) 12.8 (3.1) 27.4 (1.8) 13.7 (3.5) 6.7 (6.0) 6.5 (4.4) 0.09 (0.14)
77.3 (136) 22.2 (39) 12.8 (3.1) 27.3 (1.8) 13.9 (3.5) 6.7 (5.5) 6.5 (4.7) 0.08 (0.15)
0.5169 0.9325 0.5845 0.6626 0.9624 0.9530 0.5401
244.7 (309.4) 6.5 (0.8) 6.0 (1.7) 15.6 (9.6)
263.2 (332.3) 6.4 (1.0) 6.1 (1.9) 15.5 (8.7)
0.5654 0.2367 0.5119 0.8797
a
Mini-Mental State Examination (MMSE) score. Center for Epidemiological Studies Depression Scale (CES-D) score. c Score from the general health questionnaire. SD, standard deviation. b
though the study focused on a high-risk population and used an individually tailored and interactive curriculum. Older drivers participating in the program had a crash rate similar to drivers who did not, whether characterized by person-years of driving or personmiles of driving. Drivers experiencing the intervention did report lower driving exposure as compared with the usual care only group both in terms of miles driven and the number of places driven per week, a trend that was
maintained 2 years after the intervention. Given the self-report nature of the driving exposure measure, there may be some question as to whether these drivers did indeed drive less after the intervention, or whether social desirability was operative. The latter explanation is unlikely because baseline social desirability scores between the groups were similar. Regardless of one’s view on the validity of the self-report exposure data, when the relative risk is expressed in terms of person-
Figure 2. Self-reported annual mileage at baseline and at follow-up for intervention plus usual care and usual care only groups.
Figure 3. Self-reported places driven at baseline and at follow-up for intervention plus usual care and usual care only groups.
226
American Journal of Preventive Medicine, Volume 26, Number 3
Figure 4. Driving avoidance score at baseline and at follow-up for intervention plus usual care and usual care only groups.
years as a driver, which is not based on self-report, it is clear that the intervention had no safety benefit in terms of collision rate reduction. Previous studies suggest that driver improvement programs can sometimes backfire by increasing collision rates.10,15 Some analyses in the present study showed a tendency in this direction, which was not statistically significant. Nevertheless, it raises the question as to whether some driver improvement programs could generate overconfidence on the road. Given
Figure 5. Self-regulation score at baseline and at follow-up for intervention plus usual care and usual care only groups.
their current popularity in our society coupled with the absolute increase in the number of older drivers projected for future decades, the safety effect of these programs needs to be clearly understood if they are to be promoted. It is still possible that our intervention, and other programs, have some subtle, as yet unmeasured, safety benefit to older drivers. However, in terms of standard definitions of safety (crash rates) evidence that educational programs enhance driver safety for any age group remains unavailable.10,14,19 Why did the intervention fail to enhance older driver safety? Cognitive impairment, prevalent among older adults, might prevent the comprehension of the educational program’s content. By study design, the mental status of participants in this sample was good, so this explanation for the lack of effect is unlikely. The intervention would be expected to fail to enhance driver safety if the older drivers did not learn the curriculum’s lessons (increased self-regulatory practices) and incorporate them into their behavior. However, results imply that older drivers in the intervention group, as compared with the usual care only group, had a greater use of self-regulatory strategies and were more likely to avoid challenging driving situations, maintained up to 2 years postintervention. However, these behavior changes may not have been frequent enough or sufficiently large to generate a measurable protective effect against crash involvement. Could the content, structure, or delivery of the curriculum itself have been the source of its ineffectiveness in affecting safety? The intervention36 was built on knowledge domains widely agreed on to have high relevance to older driver safety10 –14; was individualized for each person’s driving practices and attitudes, functional impairments, and lifestyle; and was based on the latest health education and behavior models of how to engender behavior changes that facilitate health,39 – 41 models that have been useful in facilitating behavior change in other health domains.43 A process evaluation of the curriculum on the basis of health education models of behavior change37 demonstrated that those older drivers taking part in it increased their awareness of their vision impairment and its effect on driving, perceived increased benefits to self-regulation when driving, and moved closer to the preparation and action/maintenance stage of readiness in adopting new driving behaviors. Thus, the intervention’s content and structure was comprehensive, and its delivery had its intended effect in terms of theoretical models. It is difficult to identify how such an intervention could be improved on. Crash involvement is influenced by a multiplicity of factors (e.g., functional and medical characteristics of driver, highway design, other vehicles, fatigue, personality, weather), and it might be overly simplistic to expect that educational programs by themselves, how ever thorough and individualized, would affect crash rate. Am J Prev Med 2004;26(3)
227
Table 2. Crash involvement among study groups Group Intervention ⫹ usual care Usual care only
Total subjects
Collisions
Personyears
Personmiles
Ratea
Rateb
RR (95% CI)a
RR (95% CI)b
227
53
448.4
3,459,032
11.8
15.3
1.08 (0.71–1.64)
1.40 (0.92–2.12)
176
38
341.7
3,529,777
11.1
10.8
Reference
Reference
a
Crash rate per 100 person-years. b Crash rate per 1,000,000 person-miles of travel. CI, confidence interval.
Driving involves a skill set that is habitual, overlearned, and viewed as a basic life necessity regardless of adult age. These factors create obstacles to changing driver attitudes and practices. The failure to find evidence that educational programs enhance driver safety has led to an emphasis on “passive-type” driver-safety approaches in which standards for highway and vehicle design are enacted (e.g., speed limit laws, reducing the risk of injury during a crash through airbag and restraint systems).15,44 Another strategy might be to focus on the aggressive treatment of chronic medical conditions to reverse or slow functional decline in older adults. Medical conditions and functional impairment are the primary risk factors for crash involvement among older drivers.5,20,21 This approach not only improves public health in general but also enhances public safety in a society in which the older driver population is growing. This study agrees with earlier work indicating that there is no empiric support that educational programs enhance older driver safety (i.e., reduce crash rates) despite their widespread popularity. Given the convergence of findings here and from earlier studies,10,14,19 future public health initiatives should focus resources on identifying and implementing evidence-based strategies to improve older driver safety. This research was supported by General Motors Corporation pursuant to an agreement with the U.S. Department of Transportation. Supplemental funds were provided by Research to Prevent Blindness, Inc., and the EyeSight Foundation of Alabama. Cynthia Owsley is a Research to Prevent Blindness Senior Scientific Investigator.
References 1. National Highway Traffic Safety Administration. Traffic safety facts 1997. Washington DC: U.S. Department of Transportation, 1998. DOT HS 808 806. 2. National Highway Traffic Safety. Statistical relationships between vehicle crash rates, driving cessation and age-related physical or mental limitations: final summary report. Washington DC: U.S. Department of Transportation, November 1997. DTNH22-92-X-05475. 3. Evans L. Risk of fatality from physical trauma versus sex and age. J Trauma 1988;28:368 –78. 4. Barancik JI, Chatterjee BF, Greene-Cadden YC, Michenzi EM. Motor vehicle trauma in northeastern Ohio. I. Incidence and outcome by age, sex, road-use category. Am J Epidemiol 1986;74:473–8. 5. Transportation Research Board. Transportation in an aging society, Vol 1. Washington DC: National Research Council, 1988.
228
6. Jette AM, Branch LG. A ten-year follow-up of driving patterns among the community-dwelling elderly. Hum Factors 1992;34:25–31. 7. Fonda SJ, Wallace RB, Herzog AR. Changes in driving patterns and worsening depressive symptoms among older adults. J Gerontol B Psychol Sci Soc Sci 2001;56B:S343–51. 8. Marottoli RA, de Leon CFM, Glass TA, et al. Driving cessation and increased depressive symptoms: prospective evidence from the New Haven EPESE. J Am Geriatr Soc 1997;45:202–6. 9. DeCarlo DK, Scilley K, Wells J, Owsley C. Driving habits and health-related quality of life in patients with age-related maculopathy. Optom Vis Sci 2003;80:207–13. 10. Janke MK. Mature driver improvement program in California. Transportation Res Rec 1994;1438:77–83. 11. AAA Foundation for Traffic Safety. The older and wiser driver. Available at: www.aaafoundation.org and www.seniordrivers.org. Accessed August 6, 2003. 12. AARP. AARP driver safety program. Available at: www.aarp.org/drive/ and www.aarp.org/ar99/practice/. Accessed August 6, 2003. 13. National Safety Council. Coaching the mature driver. Available at: http:// secure.nsc.org/train/course.cfm?id⫽88. Accessed August 6, 2003. 14. McKnight AJ, Simone GA, Weidman JR. Elderly driver retraining. Alexandria VA: National Public Services Research Institute, September 1982. Publication U.S. Department of Transportation HS-806 336. 15. Insurance Institute For Highway Safety. Education alone won’t make drivers safer. It won’t reduce crashes. Status Rep 2001;36:1–8. 16. Vernick JS, Li G, Ogaitis S, MacKenzie EJ, Baker SP, Gielen AC. Effects of high school driver education on motor vehicle crashes, violations, and licensure. Am J Prev Med 1999;16(suppl 1):40 –6. 17. Struckman-Johnson DL, Lund AK, Williams AF, Osborne DW. Comparative effects of driver improvement programs on crashes and violations. Accid Anal Prev 1989;21:203–15. 18. Insurance Institute for Highway Safety. Driver education does not equal safe drivers. Status Rep 1997;32:1–8. 19. Ker K, Roberts I, Collier T, Renton F, Bunn F. Post-license driver education for the prevention of road traffic crashes (Cochrane Review). The Cochrane Library, Issue 3, 2003. Oxford: Update Software. 20. Retchin SM. Clinics in geriatric medicine: medical considerations of the older driver. Philadelphia: Saunders, 1993. 21. Owsley C. Driver capabilities. In: Conference Proceedings 27: Transportation in an aging society: a decade of experience. Washington DC: Transportation Research Board, National Research Council, 2004. In press. 22. Wagner EH, LaCroix AZ, Grothaus L, et al. Preventing disability and falls in older adults: a population-based randomized trial. Am J Public Health 1994;84:1800 –6. 23. Clark NM, Janz NK, Dodge JA, Sharpe PA. Self regulation of health behavior: The ‘take PRIDE’ program. Health Educ Q 1992;19:341–54. 24. Stalvey BT, Owsley C. Self-perceptions and current practices of high-risk older drivers: implications for driver safety interventions. J Health Psychol 2000;5:441–56. 25. Mendes de Leon CF, Seeman TE, Baker DI, Richardson ED, Tinetti ME. Self-efficacy, physical decline, and change in functioning in communityliving elders: a prospective study. J Gerontol B Psychol Sci Soc Sci 1996;51B:S183–90. 26. Owsley C, Stalvey B, Wells J, Sloane ME. Older drivers and cataract: driving habits and crash risk. J Gerontol A Biol Sci Med Sci 1999;54A:M203–11. 27. Folstein MF, Folstein SW, McHugh PR. “Mini-mental state” a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189 –98. 28. Ferris FL III, Kassoff A, Bresnick GH, Bailey I. New visual acuity charts for clinical research. Am J Ophthalmol 1982;94:91–6.
American Journal of Preventive Medicine, Volume 26, Number 3
29. Ball KK, Roenker DL, Bruni JR. Developmental changes in attention and visual search throughout adulthood. In: Enns JT, ed. The development of attention: research and theory. North-Holland: Elsevier Science Publishers, 1990;489 –507. 30. Ball K, Owsley C, Sloane ME, Roenker DL, Bruni JR. Visual attention problems as a predictor of vehicle crashes in older drivers. Invest Ophthalmol Vis Sci 1993;34:3110 –23. 31. Owsley C, Stalvey BT, Wells J, Sloane ME, McGwin GJ. Visual risk factors for crash involvement in older drivers with cataract. Arch Ophthalmol 2001; 119:881–7. 32. Owsley C, McGwin GJ, Sloane ME, Wells J, Stalvey BT, Gauthreaux S. Impact of cataract surgery on motor vehicle crash involvement by older adults. JAMA 2002;288:841–9. 33. Radloff LS, Teri L. Use of the Center for Epidemiological Studies– Depression Scale with older adults. In: Brink TL, ed. Clinical gerontology: a guide to assessment and intervention. New York: Haworth Press, 1986; 119 –36. 34. Strahan R, Gerbasi KC. Short, homogeneous versions of the MarloweCrowne Social Desirability Scale. J Clin Psychol 1972;28:191–3. 35. Meinert CL. Clinical trials. Design, conduct, and analysis. New York: Oxford University Press, 1986. 36. American Optometric Association. Optometric clinical practice guideline, comprehensive adult eye and vision examination, reference guide for clinicians. 2nd ed. St. Louis, MO: American Optometric Association, 1997.
37. Stalvey BT, Owsley C. The development of a theory-based educational curriculum to promote self-regulation among high-risk older drivers. Health Promotion Pract 2003;4:109 –19. 38. Owsley C, Stalvey BT, Phillips JM. The efficacy of an educational intervention in promoting self-regulation among high-risk older drivers. Accid Anal Prev 2003;35:393–400. 39. Bandura A. Social foundations of thought and action: a social cognitive theory. Englewood Cliffs, NJ: Prentice Hall, 1986. 40. Rosenstock IM. The health belief model: explaining health behavior through expectancies. In: Glanz K, Lewis FM, Rimer BK, eds. Health behavior and heath education: theory, research, and practice. San Francisco: Jossey-Bass Publishers, 1990;39 –62. 41. Prochaska JO, DiClemente CC. Stages of change in the modification of problem behaviors. In: Hershen M, Eisler RM, Miller PM, eds. Progress in behavior modification, Vol 28. Sycamore, IL: Sycamore Publishing Company, 1992;184 –18. 42. Murakami E, Wagner DP. Comparison between computer-assisted selfinterviewing using GPS with retrospective trip reporting using telephone interviews. Washington DC: Federal Highway Administration, U.S. Department of Transportation, 1997. 43. Raczinski JM, DiClemente RJ. Handbook of health promotion and disease prevention. New York: Kluwer Academic/Plenum Publishers, 1999. 44. Haddon WJ. A logical framework for categorizing highway safety phenomena and activity. J Trauma 1972;12:193–207.
Am J Prev Med 2004;26(3)
229
Randomized, controlled, single-masked intervention evaluation at an ophthalmology clinic.
Participants: A total of 403 older drivers with visual acuity deficit or slowed visual processing speed or both who were crash-involved in the previous year, drove at least 5 days or 100 miles per week or both, and were at least 60 years old. Intervention: Patients were randomly assigned to usual care (comprehensive eye examination) or usual care plus an individually tailored and administered educational intervention promoting safe-driving strategies. Main outcome Police-reported vehicle collision rate, expressed both in terms of person-years of follow-up measure: and person-miles of travel for 2 years postintervention. Results:
The intervention group did not differ significantly from the usual care only group in crash rate per 100 person-years of driving (relative risk [RR], 1.08; 95% confidence interval [CI], 0.71–1.64) and per 1 million person-miles of travel (RR, 1.40; 95% CI, 0.92–2.12). The intervention group reported more avoidance of challenging driving maneuvers and self-regulatory behaviors during follow-up than did the usual care only group (p⬍0.0001).
Conclusions: An educational intervention that promoted safe-driving strategies among visually impaired, high-risk older drivers did not enhance driver safety, although it was associated with increased self-regulation and avoidance of challenging driving situations and decreased driving exposure by self-report. (Am J Prev Med 2004;26(3):222⫺229) © 2004 American Journal of Preventive Medicine
Introduction
W
hen measured on a per-mile-driven basis, older drivers (licensed drivers aged ⱖ60 years) have a crash rate nearly equivalent to that of drivers younger than aged 25 years, whose crash rate is the highest among all age groups (approximately
From the Department of Ophthalmology, School of Medicine (Owsley, McGwin, Phillips, McNeal); Department of Epidemiology and International Health, School of Public Health (McGwin); Division of Trauma, Burns, and Surgical Critical Care, Division of General Surgery, Department of Surgery (McGwin), University of Alabama at Birmingham, Birmingham, Alabama; Office of Aging Policy and Information, Texas Department on Aging (Stalvey), Austin Texas Address correspondence and reprint requests to: Cynthia Owsley, MSPH, PhD, Department of Ophthalmology, School of Medicine, University of Alabama at Birmingham, 700 South 18th Street, Suite 609, Birmingham, AL 35294-0009. E-mail: [email protected].
222
15 crashes per million miles of travel).1,2 There are more older drivers on the road than in previous decades who have a higher mileage record than previous cohorts, with these trends increasing. Older drivers in a crash are more likely to be injured or die than are younger adults.3,4 Driving a vehicle is the preferred mode of travel by older adults in the United States.5,6 When older adults stop driving, they are at increased risk of depression and decreased quality of life.7–9 These findings stimulate a need to identify ways to enhance older driver safety. Organizations have developed educational programs to promote safe driving practices among seniors,10 –13 and these programs have widespread popularity. For example, in 1999 the American Association of Retired Persons (AARP) Driver Safety program (formerly called
Am J Prev Med 2004;26(3) © 2004 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/04/$–see front matter doi:10.1016/j.amepre.2003.012.005
55-Alive) had nearly 650,000 enrollees in 31,000 classes held nationwide.12 Programs for older drivers have several delivery designs, including classroom, videos, and workbooks. Only two programs have had their effectiveness evaluated in terms of reducing crash rate,10,14 demonstrating that the enrollees gained knowledge about safe driving and how aging affects skills important for driving. Neither program reduced crash rate compared with nonintervention groups (i.e., no demonstrated safety benefit). These findings are reminiscent of research on the effectiveness of driver education programs for both novice and postlicense drivers in that knowledge is successfully imparted, but there is no evidence that collision rate is reduced.15–19 It is premature to abandon the idea that educational interventions are effective in enhancing older driver safety. First, established risk factors for crash involvement for senior drivers—functional impairments because of medical conditions20,21—are different from those for young drivers in which impulsiveness, risk taking, and alcohol are contributory. These personalityrelated characteristics can be less responsive to educational intervention. However, it might be possible to teach an older driver motivated to remain behind the wheel about how his or her functional deficits affect driving skills and to facilitate the acquisition of compensatory strategies like self-regulation. Studies on health education illustrate that older adults are able to adopt health-protective behaviors and attitudes,22,23 implying that their use of new driving strategies is possible. Earlier work suggests high-risk older drivers acknowledge high self-efficacy in their ability to carry out compensatory driving strategies,24 and high selfefficacy in older adults is associated with behavior change.25 Another reason why the effectiveness of older driver education deserves further scrutiny is because previous evaluation studies10,14 did not focus on older drivers who were at a high risk for crash involvement. Most drivers had no crashes on their records in recent years and did not have significant functional impairments. The curriculum was group administered with standardized curriculum content for all enrollees, rather than individually tailored for the functional characteristics, driving needs, and attitudes of each driver. The aim of this randomized controlled study was to determine whether an individually administered and tailored health education program that promoted strategies to enhance driver safety reduces the crash rate of older drivers at high risk for collision involvement. High risk in this study was defined in terms of characteristics that are established risk factors for collisions by older drivers,2 namely older drivers who had visual impairment, had at least moderate levels of driving exposure, and had a recent crash.
Methods The protocol was approved by the institutional review board at the University of Alabama at Birmingham on March 11, 1997. The study was funded by General Motors Corporation pursuant to an agreement with the U. S. Department of Transportation (no. G.2). Informed consent procedures were followed. The recruitment population was licensed drivers aged 60 years and older in the Birmingham, Alabama, area who had been the driver in a crash in the prior year, according to the Alabama Department of Public Safety. Potential participants (N⫽8057) were contacted by letter and telephone to determine interest in participation and to confirm that they were legally licensed and active drivers. If so, they were invited to visit the Clinical Research Unit in the University of Alabama at Birmingham Department of Ophthalmology for further determination of eligibility. Other eligibility criteria consisted of being a driver at least 5 days per week or at least for 100 miles per week or both by self-report as determined by the Driving Habits Questionnaire,26 had a Mini-Mental State Examination (MMSE) score of 23 or greater,27 and had visual impairment. Visual impairment was defined as (1) visual acuity (habitual, binocular) between 20/30 and 20/60 (lower limit for licensure in Alabama) as measured by the Early Treatment of Diabetic Retinopathy (ETDRS) chart,28 or (2) a restriction in the useful field of view (score ⱖ40% reduction) as measured by the useful field of view test (measure of visual processing speed and attention)29,30 or both. General health was estimated through a questionnaire that asked about medical problems in 17 areas.31,32 Depressive symptoms were assessed by the Center for Epidemiological Studies Depression Scale (CES-D)33; participants rated 20 items on the basis of how often they felt in the past week the way described in each item (responses on a four-point scale from “none of the time” to “all of the time”). Total scores ranged from 0 to 60 (higher numbers represent more depressive symptoms). Social desirability (tendency to respond in a manner person believes will present himself or herself in a favorable light) was estimated by the Marlowe-Crowne scale,34 consisting of 20 statements that the participant indicated were true versus false. Total score was the number of items in which true was the response (higher scores indicate more social desirability).
Study Design Patients were randomly assigned to one of two groups by fixed randomization35 (55% to usual care plus intervention, 45% to usual care only). A target enrollment of 400 was based on a hypothesized 50% reduction of crash rate in the usual care plus intervention group with 80% power. The flow of the study protocol is depicted in Figure 1. All enrollees received usual care consisting of a free comprehensive eye examination by an optometrist to diagnose and treat eye problems and any relevant discussion with the patient about the implications of his or her conditions for everyday life.36 The intervention group consisted of usual care plus an educational intervention to promote driver safety (description follows). Because the screening process for inclusion had the potential for identifying vision impairment that was treatable, it was ethically important to send participants to usual care. If the eye examination resulted in the diagnosis of visual acuity
Am J Prev Med 2004;26(3)
223
Figure 1. Flow of participants through study. deficit or slowed processing speed that was reversed by subsequent treatment, then the person was no longer considered to be eligible for the study. Randomization to usual care versus usual care plus educational intervention occurred after the eye examination. The educational curriculum (KEYS, Knowledge Enhances Your Safety) was the basis of the intervention24,37,38 and has been presented in detail in a separate published report.37 Its main elements and approach are described here. The intervention was led by a health educator specializing in driver safety and gerontology, who created a tone of an encouraging yet frank atmosphere to create rapport and to facilitate communication. The curriculum was motivated by models of health behavior change and health promotion.39 – 41 It was administered one on one in two sessions in an eye clinic conference room, and it was individually tailored to the participant’s own driving needs, lifestyle, and visual problems as follows. During the first, 2-hour session, the educator led a discussion according to the results of his or her eye examination and the visual impairment identified at the study participant’s screening visit. The discussion began by promoting the participant’s self-awareness about his or her visual impairment and results of the eye examination. Information was imparted about how his or her type of visual impairment
224
poses problems for safe driving, causes driving performance problems, and elevates crash risk. This discussion was evidence based in that it focused on visual risk factors for crash involvement described in the scientific literature. A slide presentation of situations challenging for older drivers (e.g., driving in rain, at night, making left turns across on-coming traffic)1 was shown, enhancing discussion about how the participant’s visual skills are challenged in these situations. The discussion focused on maneuvers and settings that each participant found challenging or frustrating from his or her own perspective and then moved to strategies that the participant could implement to avoid these driving situations. Such strategies included structuring the day so that the need to run errands and travel by car was carried out during daylight rather than at dusk or night, driving the less popular routes, driving at non–rush hour times, postponing trips during inclement weather, seeking out routes in which three right turns could be substituted for a left turn at busy intersections, and selecting routes with protected left-turn arrows. Drivers were encouraged to identify strategies that could realistically be incorporated into their own routine driving. Before the session ended, participants set specific goals in a personal written contract that listed the ways they would try to modify their driving behavior (using strategies like those mentioned earlier). Session 2 of the educational intervention was administered 1 month after the first session in a 1-hour visit to the clinic. It was designed to be a booster session to review information from session 1 and to discuss progress toward achieving the contract’s goals. Driving exposure information was collected at the initial visit and then by telephone at 6-month intervals for the 2 years subsequent to randomization. The driving exposure section of the Driving Habits Questionnaire,26,42 a valid and reliable instrument for estimating driving exposure, was used for this purpose. The amount of driving done in a typical week was estimated by each participant through a structured interview that asked about the places driven to in a typical week as well as their distance from home. From this interview an estimate of annual mileage for each participant was computed. At baseline screening and at follow-up, questionnaires were administered that evaluated the subject’s strategy usage (discussed in the intervention curriculum) in terms of whether these strategies were incorporated into driving practices. A subscale of the Driving Perception and Practice Questionnaire (DPPQ)24 assessed the frequency with which the participant performed eight self-regulatory strategies at 6 and 18 months’ follow-up (e.g., driving at times other than rush hour, scheduling outings during daylight hours, making right turns around the block to avoid left turns across traffic). Participants indicated the frequency of performance by using a four-point scale (1⫽never, 2⫽rarely, 3⫽sometimes, 4⫽often). Total score is the sum of all eight items (range, 8 –24). A subscale of the Driving Habits Questionnaire26,38 assessed the self-reported avoidance of challenging driving situations (e.g., at night, in rain, at rush hour, heavy traffic roads) at the 6-, 12-, 18-, and 24-month follow-ups. Responses were on a five-point scale (1⫽never, 2⫽rarely, 3⫽sometimes, 4⫽often, 5⫽always). Total score is the sum of all eight items (range, 8 – 40). The examiner administering questionnaires was unaware of group assignment, demographics, screening characteristics, and questionnaire responses at the initial visit.
American Journal of Preventive Medicine, Volume 26, Number 3
Outcomes The primary outcome of interest was crash involvement during the 2-year follow-up, expressed as a rate, either person-years of follow-up or person-miles of travel (see Statistical Methods). Information on collisions in which the participant was the driver was obtained from accident reports provided by the Alabama Department of Public Safety. Secondary outcomes consisted of average weekly mileage and average days, trips, and places driven per week. Scores were evaluated at baseline and at follow-up on self-reported use of self-regulatory driving practices and avoidance of challenging driving situations.
Statistical Methods The intervention and usual care groups were compared according to demographic, behavioral, and clinical characteristics by using t tests and chi-square statistics for continuous and categorical variables, respectively. Crash rate for each group was computed by first enumerating the total number of collisions sustained by each group. Only collisions that occurred between each subject’s randomization date and 2-years hence were considered. Two separate denominators for crash rates were considered. The first denominator, person-years, was calculated as the amount of chronologic time elapsed between the date of randomization and 2 years subsequent, the date of death, or the date of driving cessation, whichever came first. The second denominator, personmiles, was calculated as the total amount of driving done by each subject between the same dates used to calculate personyears. Information on average weekly mileage was obtained at 6-, 12-, 18-, and 24-month follow-up interviews. Average weekly mileage reported at each follow-up was considered to be constant over the entire previous 6-month period. Thus, total person-miles of travel were calculated by summing each estimate of weekly mileage and multiplying this value by 26. For subjects with incomplete follow-up information but known to be alive, mileage information was carried forward from the previous follow-up. If a subject died or stopped driving, mileage was prorated until the date of death or driving cessation. Both sets of crash rates were statistically compared by using Poisson regression. For the secondary outcomes (average weekly mileage, days, trips, and places driven per week; self-regulatory driving practices; avoidance of challenging driving situations), the intervention and usual care groups were compared by using repeated measures analysis of variance. All data analyses were conducted on an intention-to-treat basis. pⱕ0.05 (two-sided) was considered statistically significant.
Results There were 987 persons screened for the study (Figure 1). Of these, 408 met the preliminary eligibility criteria and agreed to participate and underwent comprehensive eye examinations. Of the 408 participants, 5 had acuity impairment because of correctable refractive error; they were prescribed spectacles that improved vision to better than 20/30 in at least one eye; thus, they no longer met the vision impairment eligibility require-
ment for enrollment. The remaining 403 participants were randomly assigned to the two arms of the study, 227 to intervention and 176 to usual care only. After randomization, 30 subjects assigned to intervention never attended the educational sessions. For intervention, 197 attended the first session; 162 of these persons attended the booster second session. Figure 1 lists the number of subjects available for telephone follow-up questionnaires in both groups over the 2-year period. At 24 months, 87.2% of the intervention group and 85.8% of the usual care only group participated in the telephone interview. Table 1 presents baseline demographic, behavioral, and clinical characteristics of the two groups. Subjects in both groups averaged 73 years of age and were more likely to be men, to be white, and to have at least a high school education. MMSE scores for both groups were within normal limits with respect to the age and educational composition for adults aged 60 years and older. Depressive symptoms (CES-D scores), health status, visual function, social desirability, and driving exposure were comparable in the two groups. Figure 2 presents average mileage per week over the course of the follow-up period for the two study groups. Both the intervention and usual care groups reported declines in mileage; however, the rate of decline in the intervention group was significantly greater (p⫽0.02). Although both groups also reported declines in average days driving per week, they occurred similarly, yielding no significant difference between the groups (p⫽0.98). The intervention group reported a significant decline in average places driven per week after baseline (p⫽0.01) (Figure 3). A similar pattern was observed for average trips per week, but the difference between the groups did not achieve statistical significance (p⫽0.07). For both driving avoidance (Figure 4) and self-regulation scores (Figure 5), after baseline equivalence, the intervention group had significantly higher scores than the usual care group at each follow-up visit (both p⬍0.0001). Table 2 presents the crash outcomes over the follow-up period. Whether measured as person-years or person-miles, there was no significant difference in the rates of crash involvement between intervention and usual care groups (relative risk [RR], 1.08; 95% confidence interval [CI], 0.71–1.64 and RR, 1.40; 95% CI, 0.92–2.12, respectively). When the intervention group was limited to those drivers who participated in both sessions, the results still indicated no association (RR, 1.03; 95% CI, 0.52–2.02 for person-years and RR, 0.99; 95% CI, 0.51–1.88 for person-miles).
Discussion An educational intervention to promote driver safety in visually impaired older drivers did not reduce their rate of crash involvement in the following 2 years, even Am J Prev Med 2004;26(3)
225
Table 1. Baseline demographic, behavioral, and clinical characteristics among study groups
Characteristic Age, years Gender, % (n) Female Male Race, % (n) White African-American Years of education, mean (SD) Cognitive statusa, mean (SD) Social desirability, mean (SD) Depressive symptomsb, mean (SD) Comorbidity scorec, mean (SD) Visual acuity, mean (SD) Weekly driving, mean (SD) Mileage Days Places Trips
Intervention ⴙ usual care (n ⴝ 227) Mean (SD)
Usual care only (n ⴝ 176) Mean (SD)
p
73.7 (6.0)
73.3 (6.2)
0.5136
34.2 (77) 65.8 (148)
27.3 (48) 72.7 (128)
0.1360
76.9 (173) 23.1 (52) 12.8 (3.1) 27.4 (1.8) 13.7 (3.5) 6.7 (6.0) 6.5 (4.4) 0.09 (0.14)
77.3 (136) 22.2 (39) 12.8 (3.1) 27.3 (1.8) 13.9 (3.5) 6.7 (5.5) 6.5 (4.7) 0.08 (0.15)
0.5169 0.9325 0.5845 0.6626 0.9624 0.9530 0.5401
244.7 (309.4) 6.5 (0.8) 6.0 (1.7) 15.6 (9.6)
263.2 (332.3) 6.4 (1.0) 6.1 (1.9) 15.5 (8.7)
0.5654 0.2367 0.5119 0.8797
a
Mini-Mental State Examination (MMSE) score. Center for Epidemiological Studies Depression Scale (CES-D) score. c Score from the general health questionnaire. SD, standard deviation. b
though the study focused on a high-risk population and used an individually tailored and interactive curriculum. Older drivers participating in the program had a crash rate similar to drivers who did not, whether characterized by person-years of driving or personmiles of driving. Drivers experiencing the intervention did report lower driving exposure as compared with the usual care only group both in terms of miles driven and the number of places driven per week, a trend that was
maintained 2 years after the intervention. Given the self-report nature of the driving exposure measure, there may be some question as to whether these drivers did indeed drive less after the intervention, or whether social desirability was operative. The latter explanation is unlikely because baseline social desirability scores between the groups were similar. Regardless of one’s view on the validity of the self-report exposure data, when the relative risk is expressed in terms of person-
Figure 2. Self-reported annual mileage at baseline and at follow-up for intervention plus usual care and usual care only groups.
Figure 3. Self-reported places driven at baseline and at follow-up for intervention plus usual care and usual care only groups.
226
American Journal of Preventive Medicine, Volume 26, Number 3
Figure 4. Driving avoidance score at baseline and at follow-up for intervention plus usual care and usual care only groups.
years as a driver, which is not based on self-report, it is clear that the intervention had no safety benefit in terms of collision rate reduction. Previous studies suggest that driver improvement programs can sometimes backfire by increasing collision rates.10,15 Some analyses in the present study showed a tendency in this direction, which was not statistically significant. Nevertheless, it raises the question as to whether some driver improvement programs could generate overconfidence on the road. Given
Figure 5. Self-regulation score at baseline and at follow-up for intervention plus usual care and usual care only groups.
their current popularity in our society coupled with the absolute increase in the number of older drivers projected for future decades, the safety effect of these programs needs to be clearly understood if they are to be promoted. It is still possible that our intervention, and other programs, have some subtle, as yet unmeasured, safety benefit to older drivers. However, in terms of standard definitions of safety (crash rates) evidence that educational programs enhance driver safety for any age group remains unavailable.10,14,19 Why did the intervention fail to enhance older driver safety? Cognitive impairment, prevalent among older adults, might prevent the comprehension of the educational program’s content. By study design, the mental status of participants in this sample was good, so this explanation for the lack of effect is unlikely. The intervention would be expected to fail to enhance driver safety if the older drivers did not learn the curriculum’s lessons (increased self-regulatory practices) and incorporate them into their behavior. However, results imply that older drivers in the intervention group, as compared with the usual care only group, had a greater use of self-regulatory strategies and were more likely to avoid challenging driving situations, maintained up to 2 years postintervention. However, these behavior changes may not have been frequent enough or sufficiently large to generate a measurable protective effect against crash involvement. Could the content, structure, or delivery of the curriculum itself have been the source of its ineffectiveness in affecting safety? The intervention36 was built on knowledge domains widely agreed on to have high relevance to older driver safety10 –14; was individualized for each person’s driving practices and attitudes, functional impairments, and lifestyle; and was based on the latest health education and behavior models of how to engender behavior changes that facilitate health,39 – 41 models that have been useful in facilitating behavior change in other health domains.43 A process evaluation of the curriculum on the basis of health education models of behavior change37 demonstrated that those older drivers taking part in it increased their awareness of their vision impairment and its effect on driving, perceived increased benefits to self-regulation when driving, and moved closer to the preparation and action/maintenance stage of readiness in adopting new driving behaviors. Thus, the intervention’s content and structure was comprehensive, and its delivery had its intended effect in terms of theoretical models. It is difficult to identify how such an intervention could be improved on. Crash involvement is influenced by a multiplicity of factors (e.g., functional and medical characteristics of driver, highway design, other vehicles, fatigue, personality, weather), and it might be overly simplistic to expect that educational programs by themselves, how ever thorough and individualized, would affect crash rate. Am J Prev Med 2004;26(3)
227
Table 2. Crash involvement among study groups Group Intervention ⫹ usual care Usual care only
Total subjects
Collisions
Personyears
Personmiles
Ratea
Rateb
RR (95% CI)a
RR (95% CI)b
227
53
448.4
3,459,032
11.8
15.3
1.08 (0.71–1.64)
1.40 (0.92–2.12)
176
38
341.7
3,529,777
11.1
10.8
Reference
Reference
a
Crash rate per 100 person-years. b Crash rate per 1,000,000 person-miles of travel. CI, confidence interval.
Driving involves a skill set that is habitual, overlearned, and viewed as a basic life necessity regardless of adult age. These factors create obstacles to changing driver attitudes and practices. The failure to find evidence that educational programs enhance driver safety has led to an emphasis on “passive-type” driver-safety approaches in which standards for highway and vehicle design are enacted (e.g., speed limit laws, reducing the risk of injury during a crash through airbag and restraint systems).15,44 Another strategy might be to focus on the aggressive treatment of chronic medical conditions to reverse or slow functional decline in older adults. Medical conditions and functional impairment are the primary risk factors for crash involvement among older drivers.5,20,21 This approach not only improves public health in general but also enhances public safety in a society in which the older driver population is growing. This study agrees with earlier work indicating that there is no empiric support that educational programs enhance older driver safety (i.e., reduce crash rates) despite their widespread popularity. Given the convergence of findings here and from earlier studies,10,14,19 future public health initiatives should focus resources on identifying and implementing evidence-based strategies to improve older driver safety. This research was supported by General Motors Corporation pursuant to an agreement with the U.S. Department of Transportation. Supplemental funds were provided by Research to Prevent Blindness, Inc., and the EyeSight Foundation of Alabama. Cynthia Owsley is a Research to Prevent Blindness Senior Scientific Investigator.
References 1. National Highway Traffic Safety Administration. Traffic safety facts 1997. Washington DC: U.S. Department of Transportation, 1998. DOT HS 808 806. 2. National Highway Traffic Safety. Statistical relationships between vehicle crash rates, driving cessation and age-related physical or mental limitations: final summary report. Washington DC: U.S. Department of Transportation, November 1997. DTNH22-92-X-05475. 3. Evans L. Risk of fatality from physical trauma versus sex and age. J Trauma 1988;28:368 –78. 4. Barancik JI, Chatterjee BF, Greene-Cadden YC, Michenzi EM. Motor vehicle trauma in northeastern Ohio. I. Incidence and outcome by age, sex, road-use category. Am J Epidemiol 1986;74:473–8. 5. Transportation Research Board. Transportation in an aging society, Vol 1. Washington DC: National Research Council, 1988.
228
6. Jette AM, Branch LG. A ten-year follow-up of driving patterns among the community-dwelling elderly. Hum Factors 1992;34:25–31. 7. Fonda SJ, Wallace RB, Herzog AR. Changes in driving patterns and worsening depressive symptoms among older adults. J Gerontol B Psychol Sci Soc Sci 2001;56B:S343–51. 8. Marottoli RA, de Leon CFM, Glass TA, et al. Driving cessation and increased depressive symptoms: prospective evidence from the New Haven EPESE. J Am Geriatr Soc 1997;45:202–6. 9. DeCarlo DK, Scilley K, Wells J, Owsley C. Driving habits and health-related quality of life in patients with age-related maculopathy. Optom Vis Sci 2003;80:207–13. 10. Janke MK. Mature driver improvement program in California. Transportation Res Rec 1994;1438:77–83. 11. AAA Foundation for Traffic Safety. The older and wiser driver. Available at: www.aaafoundation.org and www.seniordrivers.org. Accessed August 6, 2003. 12. AARP. AARP driver safety program. Available at: www.aarp.org/drive/ and www.aarp.org/ar99/practice/. Accessed August 6, 2003. 13. National Safety Council. Coaching the mature driver. Available at: http:// secure.nsc.org/train/course.cfm?id⫽88. Accessed August 6, 2003. 14. McKnight AJ, Simone GA, Weidman JR. Elderly driver retraining. Alexandria VA: National Public Services Research Institute, September 1982. Publication U.S. Department of Transportation HS-806 336. 15. Insurance Institute For Highway Safety. Education alone won’t make drivers safer. It won’t reduce crashes. Status Rep 2001;36:1–8. 16. Vernick JS, Li G, Ogaitis S, MacKenzie EJ, Baker SP, Gielen AC. Effects of high school driver education on motor vehicle crashes, violations, and licensure. Am J Prev Med 1999;16(suppl 1):40 –6. 17. Struckman-Johnson DL, Lund AK, Williams AF, Osborne DW. Comparative effects of driver improvement programs on crashes and violations. Accid Anal Prev 1989;21:203–15. 18. Insurance Institute for Highway Safety. Driver education does not equal safe drivers. Status Rep 1997;32:1–8. 19. Ker K, Roberts I, Collier T, Renton F, Bunn F. Post-license driver education for the prevention of road traffic crashes (Cochrane Review). The Cochrane Library, Issue 3, 2003. Oxford: Update Software. 20. Retchin SM. Clinics in geriatric medicine: medical considerations of the older driver. Philadelphia: Saunders, 1993. 21. Owsley C. Driver capabilities. In: Conference Proceedings 27: Transportation in an aging society: a decade of experience. Washington DC: Transportation Research Board, National Research Council, 2004. In press. 22. Wagner EH, LaCroix AZ, Grothaus L, et al. Preventing disability and falls in older adults: a population-based randomized trial. Am J Public Health 1994;84:1800 –6. 23. Clark NM, Janz NK, Dodge JA, Sharpe PA. Self regulation of health behavior: The ‘take PRIDE’ program. Health Educ Q 1992;19:341–54. 24. Stalvey BT, Owsley C. Self-perceptions and current practices of high-risk older drivers: implications for driver safety interventions. J Health Psychol 2000;5:441–56. 25. Mendes de Leon CF, Seeman TE, Baker DI, Richardson ED, Tinetti ME. Self-efficacy, physical decline, and change in functioning in communityliving elders: a prospective study. J Gerontol B Psychol Sci Soc Sci 1996;51B:S183–90. 26. Owsley C, Stalvey B, Wells J, Sloane ME. Older drivers and cataract: driving habits and crash risk. J Gerontol A Biol Sci Med Sci 1999;54A:M203–11. 27. Folstein MF, Folstein SW, McHugh PR. “Mini-mental state” a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189 –98. 28. Ferris FL III, Kassoff A, Bresnick GH, Bailey I. New visual acuity charts for clinical research. Am J Ophthalmol 1982;94:91–6.
American Journal of Preventive Medicine, Volume 26, Number 3
29. Ball KK, Roenker DL, Bruni JR. Developmental changes in attention and visual search throughout adulthood. In: Enns JT, ed. The development of attention: research and theory. North-Holland: Elsevier Science Publishers, 1990;489 –507. 30. Ball K, Owsley C, Sloane ME, Roenker DL, Bruni JR. Visual attention problems as a predictor of vehicle crashes in older drivers. Invest Ophthalmol Vis Sci 1993;34:3110 –23. 31. Owsley C, Stalvey BT, Wells J, Sloane ME, McGwin GJ. Visual risk factors for crash involvement in older drivers with cataract. Arch Ophthalmol 2001; 119:881–7. 32. Owsley C, McGwin GJ, Sloane ME, Wells J, Stalvey BT, Gauthreaux S. Impact of cataract surgery on motor vehicle crash involvement by older adults. JAMA 2002;288:841–9. 33. Radloff LS, Teri L. Use of the Center for Epidemiological Studies– Depression Scale with older adults. In: Brink TL, ed. Clinical gerontology: a guide to assessment and intervention. New York: Haworth Press, 1986; 119 –36. 34. Strahan R, Gerbasi KC. Short, homogeneous versions of the MarloweCrowne Social Desirability Scale. J Clin Psychol 1972;28:191–3. 35. Meinert CL. Clinical trials. Design, conduct, and analysis. New York: Oxford University Press, 1986. 36. American Optometric Association. Optometric clinical practice guideline, comprehensive adult eye and vision examination, reference guide for clinicians. 2nd ed. St. Louis, MO: American Optometric Association, 1997.
37. Stalvey BT, Owsley C. The development of a theory-based educational curriculum to promote self-regulation among high-risk older drivers. Health Promotion Pract 2003;4:109 –19. 38. Owsley C, Stalvey BT, Phillips JM. The efficacy of an educational intervention in promoting self-regulation among high-risk older drivers. Accid Anal Prev 2003;35:393–400. 39. Bandura A. Social foundations of thought and action: a social cognitive theory. Englewood Cliffs, NJ: Prentice Hall, 1986. 40. Rosenstock IM. The health belief model: explaining health behavior through expectancies. In: Glanz K, Lewis FM, Rimer BK, eds. Health behavior and heath education: theory, research, and practice. San Francisco: Jossey-Bass Publishers, 1990;39 –62. 41. Prochaska JO, DiClemente CC. Stages of change in the modification of problem behaviors. In: Hershen M, Eisler RM, Miller PM, eds. Progress in behavior modification, Vol 28. Sycamore, IL: Sycamore Publishing Company, 1992;184 –18. 42. Murakami E, Wagner DP. Comparison between computer-assisted selfinterviewing using GPS with retrospective trip reporting using telephone interviews. Washington DC: Federal Highway Administration, U.S. Department of Transportation, 1997. 43. Raczinski JM, DiClemente RJ. Handbook of health promotion and disease prevention. New York: Kluwer Academic/Plenum Publishers, 1999. 44. Haddon WJ. A logical framework for categorizing highway safety phenomena and activity. J Trauma 1972;12:193–207.
Am J Prev Med 2004;26(3)
229