The determination of citation, crash and at fault crash rates by medical conditions, functional ability level and the corresponding estimates of relative risk provide useful information for the evaluation of the existing medical conditions program in Utah. The results of this study indicate where the citation and crash risk for the medical condition population exceeds the risk for the general population.
The results of this study suggest that participation in the program does not completely negate the adverse effects of medical conditions on driving. This does not mean that the existing program is not beneficial to public safety. It is important that this study be considered in context, considering both the limitations of the study and of the medical conditions program itself. Conceivably, the existing program could be changed in order to reduce the excess risk of drivers with medical conditions to approximate the risks of the general driving population. Further analyses may be necessary to make specific recommendations for reducing risks by specific functional ability categories. However, general recommendations and observations can be made in order to provide a framework for improving the current system. They include:
Simplification of the existing program, where possible, should be considered to improve evaluation. The need for simplification became clearer when the analysis by specific functional ability level was done. With twelve functional ability levels available for each functional ability category (medical condition), a total of 144 distinct subgroups are created. Although some of these cells were large, others were very small. Many contained less than 25 drivers, and some cells were empty entirely. In these cells, the absolute number of adverse driving events was very few, making useful data analysis impossible. Such a structure only increases the administrative burden without measurable benefit to the program. The problem of small cell size was also encountered in the groups of drivers with multiple medical conditions, which comprises a sizable minority, 20%, of all program drivers. The current program would appear to require a separate functional ability level for each medical condition or functional ability category. This might potentially result in several different restriction levels or licensing periods or both, for the same individual. Restructuring of the program to account for this seamlessly would be worthwhile.
Moving the “restriction line” does not appear to be warranted for most categories. We previously suggested that priority for modification of the system be for those categories where the rate of adverse events seems inordinately high. We further suggested that this might be reasonably defined as a rate that exceeds that of the comparison populations by a factor of 2 or more. Many of the subgroups for which this is true contain small numbers of individuals, such that modification of the program is probably not worthwhile. There are a few subgroups where modification might be considered based on that criterion, however. These are categories containing reasonable numbers of individuals where the relative risk of crashes and/or at-fault crashes exceeds, or at least approaches, 2. Specifically, these are epilepsy, psychiatric, and neurological categories, at functional ability levels 3-6. These are mostly unrestricted drivers, since levels 3-5 do not signify restriction of driving privileges. Since the medical conditions program intervenes on behalf of public safety primarily by restriction of driving activities, the logical action might be increasing restrictions for these drivers. That would have to be done by altering the definitions of the functional ability levels in such a way that some currently unrestricted drivers would have restrictions placed on them. Whether increasing restrictions in this fashion would be acceptable to the public or politically feasible is debatable.
Modification of the restriction pattern may be considered for drivers with more than one medical condition. The relative risk for crashes and at-fault crashes for drivers with more than one medical condition exceeded that of comparison drivers in roughly half of the combinations examined. The ratio exceeded 2 in a fair number of those. This was true for both unrestricted and restricted drivers. This may be interpreted to mean that the present system, wherein the functional ability level is assigned somewhat independently for each medical condition, may underestimate the amount of driving impairment experienced by these drivers with medical conditions. Taking multiple conditions into consideration when assigning functional ability level seems warranted, considering that 1 in 5 (20%) of medical conditions program drivers report multiple conditions.
Any further research into the medical condition program should evaluate these data in the light of exposure data. The main activity of this program is to impose restrictions (by assigning them a functional ability level) on the driving activities of individuals deemed to have diminished driving skills as a result of medical problems. The underlying assumption here is that placing restrictions on drivers compensates for diminished driving competence. It is not certain that this is true, however. Demonstration of this would require that one show that rate of adverse driving events per mile driven be reduced by placement of restrictions. Although we have used age- and location- matching as a proxy for exposure, we cannot be very confident that the exposure rates are, in fact, comparable. It is possible that the existing program could be used to gather exposure data at reasonable expense. Applicants for driver's licenses could be required, as a condition of licensure, to estimate their annual mileage and describe the type of driving they do at the time of license application and renewal.
There is immediate need for research into the medical conditions program to test whether restrictions on driving improve public safety. The main product of the medical conditions driver program is the imposition of restrictions on the driving privileges of certain drivers. The unspoken underlying assumption in the program is that such restrictions improve safety. That is, given that medical conditions impair driving ability and thus adversely affect safety, it is assumed that restriction of driving activities by the program compensates for this impairment so as to make the driver acceptably safe. Whether this is in fact true is unknown; indeed, it appears to be a rather major leap of faith. We believe this is the biggest current need for research in the area of medical condition. Data analysis in this report could not be said to prove that the medical conditions program accomplishes its major goal, that of ensuring safe driving by people with medical programs, at all. Indeed, it could be interpreted as suggesting the opposite.
Efforts to modify the existing program should be prioritized by the Utah Driver License Division and the Utah Medical Advisory Board. Additionally, these agencies should work together to determine the range, scope and order of future research that is necessary to develop the appropriate modifications specific to each functional ability category. Consideration should also be given to this study's existing limitations described in the previous section.
Priorities should be placed on functional ability categories that had smaller estimates of statistically significant risks but larger populations (i.e., vision). As described above, we suggested that a relative risk ratio of 2 or greater would identify a subgroup with sufficiently greater risk such that modification of the program might be useful in that area. However, it may also be useful to direct some attention toward the largest subgroups even if their risk ratios are somewhat smaller. By slightly reducing the risk for a larger number of drivers, the benefit to public safety may be even greater than reducing a large risk for a small number of persons. Functional ability categories that potentially meet this description include diabetes and other metabolic conditions, visual acuity and psychiatric and emotional conditions.
Any modifications to the existing program should be carefully documented. Because of the nature of the medical condition program, changes would be implemented over time. Thus, careful documentation of the date of implementation on an individual level (i.e., the renewal date for the license holder when he/she is affected by the changes) is required in order to evaluate the effects of such changes. Accordingly, the effects of changes implemented on rates of events and estimates of risk should be measured to assure that they are of benefit to public safety.
Any changes to the medical conditions programs should be made cautiously and with consideration of acceptable levels of risk in the real world. Clearly, society at large is willing to accept some risk while using the public roads. Obviously, part of this risk is tolerating some drivers who are less safe than the norm. Teenage males are an obvious example; the general awareness that this age group is more likely to have adverse driving events does not prevent them from being issued driver's licenses. Although medical conditions drivers may have higher rates of some such adverse events, the increased risk posed thereby may be acceptable to society. Some consideration of acceptable levels of risk may be advisable.
Finally, this analysis should not be construed to mean that the medical conditions program is faulty or seriously flawed. Indeed, the results shown here could be as easily interpreted to mean that the program is working well. Adverse driving events in most subgroups of program drivers would seem to be under reasonable control. Wholesale changes to the program cannot be recommended based on data analyzed here.