Driver Screening and Evaluation Program
Volume I: Project Summary and Model Program Recommendations
This chapter presents an overview of the project activity having the greatest impact on Model Program development, the Maryland Pilot Older Driver Study. Pilot Study development and data collection; analyses and results; research products; and cost-benefit considerations are covered below. For in-depth discussion of these and related topics, see Volume 2.
The Pilot Study was carried out through the combined efforts of the Maryland Medical Advisory Board, the MVA's Office of Driver Safety Research, the NHTSA research team, and numerous partners in the Maryland Research Consortium. Key individuals who contributed most significantly to the success of the Pilot Study are recognized in the Acknowledgements section. The material below highlights our principal considerations in selecting the included screening measures; training the staff who would administer the functional tests; choosing the physical locations for data collection; and meeting technical support needs to compile and check the quality of data analyzed during the Pilot Study.
Two goals were established during planning for the Pilot Study--to examine the validity and to evaluate the administrative feasibility of measuring drivers' functional status to help detect individuals at higher risk of driving impairments and crashes. Construct validity for a group of ten measures spanning designated perceptual-cognitive and physical abilities linked to safe driving was established through a synthesis of prior research, plus expert opinion solicited as described in the preceding chapter. Empirical validity was to be examined through predictor-criterion relationships denoted by odds ratios, calculated using functional status and driving history data collected and analyzed in the Pilot Study. The odds ratio technique was selected because of its ability to illustrate how the predictive value of a given functional measure changes when different pass-fail "cutpoints" are applied, something of clear interest for any broader implementations that might emerge from this research.
The potential for broader implementation of research findings also depends strongly upon the judged feasibility of recommended screening activities. Additional criteria in selecting measures for the Pilot Study thus included brevity; low cost; and the ability to be administered by non-professionals (i.e., general office staff), with limited training, in diverse settings. A field test of candidate measures was carried out in a "pre-pilot" study, using subjects and research assistants participating in the Johns Hopkins University Salisbury Eye Evaluation (JHU/SEE) project.1 This effort focused on test instructions, materials, measurement methods, scoring procedures, and data entry requirements, as well as other problems or concerns on the part of the subjects or the JHU research assistants trained as test administrators.
Ultimately, validity and feasibility considerations were balanced to define a battery of "gross impairments screening" (GRIMPS) measures in the two domains indicated below:
The physical measures concentrated on domains of functional ability including lower limb strength and mobility, upper limb strength and mobility, and flexibility of the neck and upper torso. The perceptual-cognitive measures addressed domains of functional ability that included directed visual search, divided attention, information processing speed, working memory, visuospatial abilities, and the organization of drivers' scanning patterns.
Given the prior existence of established protocols for vision testing by the MVA, no visual performance measures were included in the Pilot Study.
After the GRIMPS battery was narrowed to the procedures listed above, two perceptual-cognitive measures were added. A PC-based test labeled "Dynamic Trails" was incorporated into the screening battery added after data collection with the other measures had begun; it automated and modified the "Trail-making, Part B" procedure. Also, one component of the Useful Field of View test (Subtest 2, Information Processing Speed and Divided Attention) was included in the Pilot Study, under the sponsorship of the National Institute on Aging.2 Thus, functional status data were collected and analyzed using a total of ten procedures--six perceptual-cognitive measures and four physical measures--in the Pilot Study, which together could be administered by trained data collectors in 20 to 30 minutes.
A "Mobility Questionnaire" addressing the nature and extent of the study participants' driving habits was developed for administration in conjunction with the functional status measure. Subjective estimates of miles driven on a weekly basis were obtained, plus categorical responses indicating annual miles driven as well as the frequency with which specified driving situations were avoided (nighttime, adverse weather, etc.).
The personnel who collected functional status data in the Pilot Study were employees of the MVA. At the outset of the study, in November, 1998, data collection responsibilities were met by "line personnel" assigned to the project by the MVA. Over a 2-day period, these individuals received training that included classroom lecture on general research principles (e.g., the importance of consistency in test administration), video tape demonstration of methods, hand's-on demonstration of methods, and practice with observation and feedback. Continued observation at random intervals during the following weeks reinforced training lessons.
During the final year (2000) of data collection, responsibilities were shifted to staff designated by the MVA as Driver License Examiners. Similar training was provided to these individuals. This change was prompted by a less-than-desired level of consistency in test administration methods that persisted among the initial group of data collectors; the latter group, already accustomed to performing a range of examination functions, in fact demonstrated greater diligence and attention to detail as hoped.
The physical locations in which Pilot Study data were collected reflect the different sampling strategies pursued in this research. A License Renewal sample (n = 1,876) and a Medical Referral sample (n = 366) were tested in MVA field offices; 11 separate locations were utilized Statewide, each with a private conference/training room that was dedicated to screening activities. A Residential Community sample (n = 266) was tested on-site at Leisure World in Montgomery County, MD.
The randomly selected License Renewal sample was deemed sufficiently representative of its age cohort to permit generalization to the broad population of older drivers, with respect to crash and violation experience; it served as the test bed for project data analyses examining the relationship between functional ability and crash and violation safety outcome measures. The Medical Referral sample, as the name suggests, was comprised of individuals already identified as likely to be impaired. The Residential Community sample, by contrast, included a more affluent, self-selected group of "well elderly" drivers.
A fourth location for data collection was planned in a social service setting: Senior Centers operated by the Howard County, Maryland, Office on Aging. Early experience at this site determined that driver functional screening as required to meet the objectives of this research could not feasibly be completed. The greatest, though not the only difficulty was the very low response from older citizens in the county; concern that test results would be shared with the MVA--despite explicit assurances to the contrary--was the apparent reason. Data collection activities were curtailed in the Senior Centers in the spring of 1999.
Also deserving mention in this chapter is the necessary work to prepare analysis files, an essential and quite involved step en route to the numerous summary tables, graphs and figures, and statistical test results emerging from this research. In partnership with the MVA Office of Driver Safety Research, safety data were extracted from Maryland State Highway Authority (SHA) databases and filtered as described below to create a primary project database for analyses relating drivers' functional status to crashes and moving violations. This was keyed to a unique period of time--relative to each individual's test date--during which driving history variables would be analyzed.
Original data tables as received from the MVA were imported into MS Access 97 for analysis, using the driver record (Soundex) number as the linking variable. Data sorting using MVA system codes defined specific outcome variables of interest--at-fault crashes, unknown-fault crashes, and all crashes; as well as moving violations with and without speeding and occupant restraint violations. Before proceeding with summary analyses and statistical tests, the data were further examined to confirm that variability in driving experience observation intervals inherent in the pilot study design (because of test dates that differed by driver, coupled with a common study end date for sampling crash and violation data) was random with respect to crash-involved versus non-crash-involved populations. Having addressed concerns about this potential bias in the data, the Access tables were imported into SPSS SYSTAT (v. 9.01) using an Open Database Connectivity (ODBC) feature, and the analysis of Pilot Study results could proceed as described in Volume 2.
Pilot Study analyses initially compared the age distributions for each study sample, plus functional performance distributions for each included screening measure. The License Renewal sample was approximately 10 years younger (mean age = 68.3) than the Medical Referral sample (mean age = 76.8) and Residential Community sample (mean age = 77.1). In terms of functional ability, however, the Residential Community sample mirrored the population-based License Renewal sample very closely, especially with respect to perceptual-cognitive tests, while the Medical Referral sample was consistently skewed toward greater functional loss. This result was not surprising, and only reinforced the premise that functional status, not age per se, is of primary importance.
By contrast, the Residential Community and Medical Referral samples were more alike in terms of self-reported mobility restrictions. It thus appears that drivers of similar age but differing in functional ability may make similar behavioral adaptations in the ways they limit their driving habits to compensate for a perceived increase in driving risk.
The nature and strength of relationships between the various included screening measures and safety outcomes were assessed using odds ratio (OR) analyses. As noted earlier, three levels of crashes (at-fault only, at-fault plus unknown fault, and all crashes) and three levels of moving violations (all moving violations, all except speeding, and all except speeding and occupant restraint violations) were examined in the Pilot Study analyses. These safety outcomes were tabulated for analysis for each driver, bracketing his/her test date with one year of prior data plus as much later driving history data as available. In addition, crash analyses were repeated using prospective data only.
The OR calculations were performed in SPSS/SYSTAT with significance tests (chi-square) applied at functional performance levels where peak valid OR values were obtained. As explained in more detail in Volume 2, this analysis technique expresses the odds of experiencing a given outcome (e.g., crash involvement) if a person fails a test than if the person passes the test. By noting where the maximum OR value is attained, a candidate cutpoint for pass-fail decisions may be identified for each measure where an OR greater than 1.0 is demonstrated; OR values below 1.0 indicate that a test has no predictive value.
Although quite popular in studies of this kind, the calculation of an odds ratio--along with the similar analysis technique "relative risk"--is subject to strict limitations on its validity. With reference to the four cell matrix defined by the combinations of "pass" and "fail" versus "crash" and "no crash" outcomes, OR cannot be calculated when any of the cell values are zero. Paradoxically, this includes instances where the measure is a perfect predictor, i.e., where there are no "misses" (where a driver passes the test but still has a crash) or "false alarms" (where a driver fails the test but remains crash-free). Also, an OR calculation when there are fewer than 5 observations in any cell in the aforementioned matrix is statistically unreliable and easily susceptible to misinterpretation. This requirement for valid OR calculations was applied in the analysis and interpretation of Pilot Study results without exception.
The analysis results demonstrated significant predictor-criterion relationships in all domains of functional ability studied except upper limb strength/mobility and the organization of drivers' visual scanning patterns. The most prominent example relates to the detection of functional decline in visuospatial abilities, specifically, the ability to visualize whole objects or patterns when there are missing elements and only partial information is available. Performance on the Motor-Free Visual Perception Test/Visual Closure Subtest easily evidenced the strongest relationships to safety outcomes, at the highest levels of statistical significance. Moreover, this finding was consistent across not only crash analyses, but also in relation to convictions for moving violations.
Next, the importance of detecting losses in drivers' divided attention abilities is highlighted by the significant results obtained for the Trail-making, Part B and Useful Field of View Subtest 2 measures in the (at-fault) crash analyses. Trail-making, Part B also demonstrated statistically significant results in the analyses of moving violations; and again there was close agreement between analysis results using crash and violation outcome measures.
The solid results demonstrated for Trail-making, Part B in this research also highlights directed visual search as a functional ability that, when compromised, significantly impairs safe driving.
The significant result for Useful Field of View Subtest 2 in the crash analyses, noted above, focuses attention on information processing speed as another functional ability where decline may be reliably associated with increased risk of driving impairment. None of the results obtained in the analyses of moving violations were significant for this procedure, however. It also may be noted that variation in the actual size of the "useful field of view" was not evaluated as a safety predictor in this research.
A decline in working memory was shown to be a significant predictor of impaired driving through the analyses relating performance on the Delayed Recall measure to at-fault crashes. But, significant results were not demonstrated for this measure in the analyses of moving violations.
With respect to physical measures, the importance of lower limb strength and mobility was indicated by significant results for the Rapid Pace Walk measure in the crash analyses; results for this measure failed to reach significance for the analyses of moving violations, however, and the related Foot Tap measure approached but failed to reach significance for both sets of analyses. A significant outcome in the crash analyses for the Head/Neck Rotation measure indicated that flexibility of the neck and upper torso can be a useful predictor of driving impairment; unfortunately, a valid based on moving violation experience was not permitted as too few drivers failed this test in the study sample.
Other measures which were found to have negligible value as predictors of driving risk in the Pilot Study included the Arm Reach and Scan Test. In both cases, virtually all drivers were able to perform the measure without error or deficit; without any variance in functional ability, a measure cannot discriminate between different levels of crash or conviction experience.
The analysis results, beyond providing evidence that functional capacity screening can yield scientifically valid predictions about the risk of driving impairment, also supported the identification of preliminary cutpoints for pass-fail decisions for selected measures. However, in a number of cases, the rationale for selecting cutpoints depended less on an isolated, peak OR value than on broader trends in the distributions of crash-involved versus non-crash-involved drivers. In particular, candidate cutpoints were chosen where there was a clear performance-versus-safety transition, signified by a level of functional loss where the proportion of drivers involved in crashes began to consistently exceed the proportion of drivers remaining crash-free. In fact, the analysis results provided evidence that supports the notion of not one, but two cutpoints for each functional measure adopted for use in a screening program--one keyed to prevention, at a modest level of decline, and another keyed to intervention, where gross impair-ments in functional ability are detected. Volume 2 presents additional details about analysis methods and outcomes, while the importance of these findings for Model Program development is discussed in the concluding chapter of this report.
This section identifies and describes research products developed to support Maryland Pilot Older Driver Study activities which, like the Safe Mobility for Older Persons Notebook, have demonstrated a sustained value in applications extending well beyond the present investigation. These include educational materials; software and materials for conducting functional screening; and a database proposed as the primary test bed for longitudinal study of the relationship between functional status and traffic safety.
A key objective in the Pilot Study was to help older drivers in Maryland gain a better understanding of their functional abilities, the changes in abilities to expect with normal aging, and how these changes relate to safe driving. Educational materials were developed in this project with these specific goals in mind, most notably the "How Is Your Driving Health?" brochure distributed to its customers by the MVA.
This brochure, reproduced in appendix C, contained contact information tailored to the particular jurisdiction (Maryland) of the Pilot Study. But it also contained general information to raise awareness that safe driving depends upon intact functional abilities; examples of declining abilities common among older persons; and suggestions for changes in driving habits to help compensate for certain kinds of diminished capabilities. By design, the self knowledge that an older driver gains from thoughtful consideration of the brochure's contents will lead to a more frank discussion about medical fitness-to-drive with the individual's physician, and with his or her spouse and family members as well.
Much of the material developed for the "How Is Your Driving Health?" brochure subsequently was adopted by NHTSA for a publication produced in cooperation with the USAA insurance company, Driving Safely While Aging Gracefully.
The functional screening data for the Pilot Study was, for the large majority of included measures, collected manually by MVA staff using materials and procedures developed for this purpose. Nine of the ten measures obtained in the Pilot Study were elements of the Gross Impairments Screening (GRIMPS) battery emerging from earlier work in this project.3 GRIMPS data collection was supported by test kits including:
After initial development of the contents listed above, accomplished under this research contract, materials could be reproduced and assembled into kits at a cost of approximately $100 each. GRIMPS test kits were distributed to all MVA staff involved in data collection at field offices Statewide, following preliminary training exercises to familiarize them with their use.
Following its introduction in Maryland, the GRIMPS battery has also been implemented for research purposes and in driver evaluation and rehabilitation settings that have no formal linkage to the Pilot Study. Per request from the responsible parties, GRIMPS test kits were provided at no charge or at cost, with an understanding that data sharing to support the develop-ment of population norms for the included procedures would be allowed at a future point in time. The venues in which the GRIMPS battery has been applied, using test kits supplied by project staff, include:
A novel software application was also developed in this project, spurred by Maryland MVA desires to automate screening procedures wherever feasible--especially those that are most time-consuming and prone to test administration errors that threaten data quality. According to these criteria, the Trail-making test measuring visual search and sequencing and divided attention abilities was identified as the priority for automation. Software subsequently developed with this goal in mind resulted in a derivative procedure, labeled "Dynamic Trails."
The Dynamic Trails procedure is a PC-based test that maintains the mixed letter and digit stimuli used in the traditional paper-and-pencil Trail-making procedures; however, instead of the blank, white background used in the traditional protocol, Dynamic Trails presents a compressed video image of a freeway driving scenario, in color. This approach was selected to incorporate an additional element of distraction into the test procedure. At the same time, the overall number of stimuli (letters and digits) superimposed on the moving traffic background was reduced. A shorter but more challenging measure of perceptual-cognitive abilities related to safe driving, with high face validity to examinees, was the intended result.
As reported in Volume 2, the analysis of Dynamic Trails data collected in the Pilot Study was complicated by difficulties in test administration, and because of a reduced sample size resulting from the introduction of this procedure after data collection with the rest of the battery had already been underway for several months. Nevertheless, valid odds ratios greater than 1.0 were calculated when examining how well this measure could predict at-fault crashes among the License Renewal sample. Also, significant methodological improvements were undertaken after data collection was concluded; these included software refinements for better data capture, the use of a touch-screen interface instead of a light pen, and the addition of audio as well as text instructions to standardize an element of test administration that was often inconsistent during the Pilot Study.
Apart from the research in Maryland, the enhanced Dynamic Trails protocol was applied in a 2001-02 study of functional impairment and driving safety by the Florida Aging Driver Council. It has also been selected for use by a Continuing Care Retirement Community (CCRC) in the Peninsula United Methodist Homes (PUMH) network in the state of Delaware, which is offering a functional screening service with counseling about safe driving habits as an educational benefit for its residents.
The Pilot Study analyses, detailed in Volume 2, were performed upon a MS Access 97 database comprised of functional screening data, and crash and moving violation data provided by Maryland DOT officials. Coordination between the State Highway Authority (SHA) and the Motor Vehicle Administration (MVA) permitted the aggregation of the raw data files; extensive error checking and filtering of these data to determine the period of time--relative to each of more than 2,000 individuals' test dates--for which driving history variables should be examined resulted in the final analysis database for the Maryland Pilot Older Driver Study (MaryPODS).
The MaryPODS database itself is a valuable resource for continuing study of the relationships presently under investigation. Currently shared among all research partners participating in the Pilot Study, including NHTSA and Maryland MVA, this Access database establishes a baseline against which future changes in functional status and crash and violation experience can be compared. Relationships documented through longitudinal study with the same Maryland drivers that are consistent with the cross-sectional analyses reported herein, would provide a compelling argument for functional capacity screening to detect high-risk drivers. In addition, the performance distributions (for the License Renewal sample) that are recorded in this database--supplemented to the greatest extent practical by data collected in other venues employing a common methodology--may be fairly considered as a starting point for the development of population age norms for each included measure of functional ability.
It was found in the Survey of State Licensing Officials conducted in this project, that implementation of the types of screening and evaluation activities envisioned under the Model Program would depend, to a large extent, on a Department's ability to offset the costs associated with such activities. This section of the report compares the estimated costs for conducting screening activities in a "production" environment by a licensing agency, derived through consultation with the Maryland Motor Vehicle Administration (MVA), to benefits (cost savings) that are realized through increased efficiencies in the performance of certain, indispensable components of a medical determination of fitness to drive.
Fitness-to-drive determinations are required for drivers referred to a motor vehicle agency, through means that can vary greatly from jurisdiction to jurisdiction. Using Maryland as our example, drivers suspected of (functional) impairment of one sort or another may be referred by physicians, occupational therapists, and other health care providers; law enforcement officers or the courts; social service providers, including those who perform geriatric assessments for the State; by MVA personnel (e.g., counter staff) based on in-person observations of particular behaviors associated with possible impairment; by family, friends, or other citizens; and by the motorist himself/herself via acknowledgements of one or more medical conditions (stroke, cardiovascular conditions, diabetes, visual problems, seizure disorders, etc.) that are included on checklists attached to License Renewal forms and Learner's Permits.
Under the Model Program, it is anticipated--though not specifically advocated--that jurisdictions may, in the future, require some type of assurance that all individuals applying for license renewal are free of any gross functional impairments. Alternately, individuals above a designated age only may be subject to such a requirement, based on evidence showing that the incidence of functional loss resulting in driving impairment increases sharply somewhere between the mid-60's to mid-70's, depending on the individual; and that, accordingly, it is extremely inefficient to broadly screen for functional deficits when they are so rarely detected in young and middle-aged drivers. If such a policy were implemented in Maryland for individuals age 65 and over, 452,591 drivers or 12.7 percent of the licensed population, apportioned according to the (5-year) renewal cycle in that State, would have been affected in the year 2000. By comparison, the number of drivers age 75 and older in Maryland in 2000 was 182,530; on a 5-year renewal cycle, approximately 36,500 individuals would be affected annually.
Whatever mechanisms drive the number of persons for whom a motor vehicle agency makes fitness-to-drive determinations each year, each of those customers must be evaluated in terms of criteria including, at a minimum, health history information provided by the individual plus a current physician's report. A case review file containing this and any additional information deemed important by a jurisdiction is forwarded to the professional--an agency employee (e.g., Medical Advisory Board) or outside consultant--who ultimately provides a recommendation for disposition of the matter. Generalizing from the Maryland experience, this discussion assumes that three outcomes are possible at this stage, i.e., a determination of (1) OK to drive (with or without restrictions); (2) NOT OK to drive (license suspended or revoked); or (3) HOLD pending further information.
An evaluation was performed in the Maryland Pilot Study to see what impact, if any, on the disposition of cases referred for review by a Medical Advisory Board (MAB) physician would result from providing functional capacity screening data in addition to the other information (driver's self-reported medical history and personal physician's report) contained in the traditional case review file. MAB physicians initially reviewed each of 450 cases without access to the screening data; then, blind to their earlier recommendations for disposition, conducted another review where the results for all of the Pilot Study screening measures were included in the driver's file.
The functional screening data for this evaluation were obtained from the Medical Referral sample using procedures described in Volume 2 of this report. The consequences of including the screening data on the dispositions of the MAB physicians are shown in figure 1.
As indicated, these Pilot Study findings indicate an increase in the proportion of cases with a disposition of "OK" (from 38 to 55 percent) and "NOT OK" (from 22 to 29 percent), coupled with a decrease from 40 percent to 15 percent in the proportion of "HOLD" outcomes. In real numbers, including functional screening data in the MAB review process for medical determination of fitness to drive reduced the number of drivers placed on "HOLD" status from 180 to 68, a 38 percent decline.
The benefit that may be attached to this shift is gauged in terms of the relative costs of the screening procedures, applied in a production setting, versus the procedures that historically would be applied in cases where a disposition is on "HOLD" pending further information. In Maryland, such information is obtained through one or more of the following:
While the costs of lab tests, medical specialty reviews, and OT/CDRS evaluations are typically borne by the driver, the MVA bears the costs of interviews and drive tests. Excluding equipment, facilities, and miscellaneous overhead expenses, an estimate of the cost-per-driver to administer these procedures calculated just in terms of the associated labor costs of agency personnel may be derived.
According to the MVA,4 two Driver License Examiner-level staff at fully loaded rates of $20/hr and one physician at a rate of $100/hr are involved in each driver interview. The DLE's perform scheduling, coordination, and record keeping activities that require one-third hour each per driver, while one-half hour of the physician's time is engaged in preparations for and interaction with each driver interviewed. Together, these labor costs total an estimated $63.20 per driver. Each closed course drive test, by comparison, can be performed by a single DLE; this activity requires one-third hour, at an estimated cost of $6.67 per driver.
It is important to note that the closed-course drive test used by the Maryland MVA was developed to provide assurance that novice drivers could demonstrate basic competency in handling a vehicle. It may therefore be suited to assessing maneuvering skills, but insufficient to determine whether a driver is able to meet the attentional and perceptual-cognitive demands experienced across a range of traffic conditions encountered in everyday driving. For this reason, the MVA in some instances applies a road test, in traffic, to reach a fitness-to-drive determination; this requires between 45 to 60 minutes of DLE labor, or $15 to $20 per driver tested.4
To calculate the overall cost experienced by the MVA for interviews and (closed-course) drive tests to reach a disposition of cases initially placed on "HOLD" pending further inform-ation, both the number of "HOLD" cases and the percentage of cases that receive an interview alone, versus an interview-plus-drive test, must be specified. Historically, roughly 30 percent of all cases referred for medical review result in a driver interview, and approximately one-third of those interviewed, or 10 percent of cases, also require a drive test before a disposition can be reached.5 Per 1,000 drivers who are referred under current agency practices for a medical determination of fitness-to-drive, then, it is expected that 300 will be interviewed at an aggregate (labor) cost of $18,960, and that 100 of these will also receive a (closed-course) drive test at an aggregate (labor) cost of $667. A conservative estimate of the supplemental cost to the MVA to reach a disposition of cases placed on "HOLD" status after their initial review by the MAB, which does not include any costs for on-road assessments, is therefore $19,627.
As reported earlier, Pilot Study results point to an expected 38 percent reduction in the number of drivers referred for medical evaluation that will be placed on "HOLD" status if physicians are provided with functional capacity screening data at the time of initial review. This translates to a savings of $7,458 using the figures developed in the paragraph above. According to the cost analysis described in Volume 2 of this report, it was concluded that the cost-per-driver to conduct functional screening in a production environment could be brought down to a five dollar ($5.00) range allowing for automation of certain test procedures performed manually in the Pilot Study. Per 1,000 drivers, an aggregate cost of $5,000 to perform functional screening therefore yields an estimated net reduction in costs experienced by the MVA of nearly $2,500.
It is understandable, given these findings, that the MVA will extend functional screening beyond the term of the Pilot Study for all drivers referred for medical determination of fitness to drive. But can a cost-benefit analysis justify functional screening for all renewing drivers?
Again relying on Maryland Motor Vehicle Administration data, to perform functional capacity screening for all 36,500 annually renewing customers age 75 and older would cost the agency--using optimistic but defensible projections--something approaching $200,000, in round numbers. Obviously, only a portion of this sum would be offset by increased efficiencies of the nature described above; this amount would depend, at least in part, on the percentage of this population cohort who would be expected to be referred to the agency for medical evaluation through the various existing mechanisms. The greater problem, though, is that this entire line of inquiry is constrained by adherence to an intervention model governing policy and practices in the area of medical fitness to drive. All of the drivers who were subjects of the preceding analysis have already manifested problems of sufficient magnitude, that one or more referring parties judged the individuals' safety and the safety of others to be at immediate risk.
When screening is performed "across the board," for a designated cohort of drivers, the most profound benefits are foreseen within the context of a prevention model. Functional screening not only improves the detection of impairments signifying immediate risk, but also provides individuals and their health care professionals with early warning of functional decline in the abilities needed to drive safely. This can only enhance the potential for remediation of a wide range of deficits, resulting in more older persons driving safely longer, if they choose to do so. According to recent NHTSA estimates, the overall cost to society of a single traffic fatality or critical injury approximates $1 million (cf. Blincoe, Seay, Zaloshnja, Romano, Luchter, and Spicer, 2002), which significantly exceeds the projected expense of implementing a driver screening and evaluation program in all but the largest states. Additional and very substantial benefits to society will accrue from the lower levels of assistance that must be provided to an elderly population that remains independently mobile. Benefits to individuals from meeting their own transportation needs with dignity, meanwhile, are incalculable.
Finally, the considerations detailed above do not speak to the costs or benefits of conducting screening activities in health care, social service or other settings, with results submitted to the DMV according to an established protocol with safeguards to ensure data quality, confidentiality, etc. These and other ways in which prevention and intervention components could be integrated into a Model Driver Screening and Evaluation Program are examined in the following discussion.
BACK | CONTENTS | NEXT