Cover Image

Not for Sale

View/Hide Left Panel
Click for next page ( 67

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 66
66 5, and this step may be skipped. If the analyst has used in- treatment, and the effectiveness of the third treatment dividual crash types (or crash types by severity and speed applied to the same driver subgroup be reduced to limit) in the above steps, those individual estimates 25 percent of the original level. (Assume that any addi- within each subgroup must be summed to calculate the tional treatments after the third will have no additional total economic benefit for each subgroup. effect.) For example, assume that the first treatment for a 7. Define the annual cost for treating all drivers in each given AR subgroup has an effectiveness level of 20 per- subgroup. cent, the second has an effectiveness level of 15 percent, This will be an estimate of total program cost for and the third has an effectiveness level of 10 percent. The each treatment under study and for each subgroup being estimated combined effectiveness of the three treatments considered. This assessment will require subjective ap- applied to the same segment would be 20% + 15% (.5) + proximations. Some general information on treatment 10% (.25) = 30%. Again, this is only an estimate of the cost is presented for each treatment in the guide, and that true combined effectiveness at best. information should be reviewed by the user. Depending on the treatment, this cost may include start-up cost and Alternative Procedure Choosing cost per driver (i.e., all drivers that would need to be Treatments and Target Subgroups treated without knowing who will subsequently be in- for Alcohol-Related Crash Strategies Based volved in an AR crash). For other treatments, this may On Existing DWI Program Needs simply be an annual cost (e.g., for public information programs). Note that the total cost over the expected life The above two procedures for choosing treatments and of the project will need to be amortized to an "annual subgroups have been based on the size of the alcohol-related cost" basis, since the benefit calculations are in annual crash problem among different target subgroups of drivers numbers [see NCHRP Report 501 (18)]. and, in the second procedure, on estimated treatment effec- 8. Calculate "net benefits" for each treatment by tiveness. Both are based on crash data, which make them the subtracting cost from benefits. recommended procedures to follow. However, in the absence 9. Choose the treatments (and thus treatment subgroups) of crash data, a third alternative procedure that is advocated with the greatest net benefit. in Volume 1, the AR Guide, is to conduct a careful assessment 10. Decide whether to use multiple treatments. of the nature of the jurisdiction's drinking-driving problem and After reviewing the prioritized listing of treatments how the DWI countermeasure system is currently functioning. and estimated costs, the analyst may decide to further de- The choice of AR treatment strategies from those listed in the termine whether multiple treatments would be beneficial. guide (and thus the choice of target subgroups) would be If the treatment combinations being considered affect dif- based primarily on "current AR program needs" strategies ferent driver subgroups (e.g., one affects young AR driv- that are not currently being implemented, or whose imple- ers while the second affects [older] multiple offenders), mentation can be significantly improved. This assessment of then the net benefit of that combination will be the sum current program needs requires a multidisciplinary team, of the individual calculations from Step 8. However, this since the system for dealing with alcohol-impaired driving is not usually the case. Multiple treatment combinations may be the most complex and involve the greatest number of will often affect the same subgroup even if they are aimed disciplines and state agencies of any traffic safety issue. States at different subgroups. In this case, Steps 28 will need to frequently use a task force that represents all the key elements be repeated for each combination under study. Thus the of this system. Without such an approach, a fragmented and potentially affected driver groups will be specified first, incomplete understanding of the problem is likely and then the AR crashes will be calculated, etc. progress will be difficult. Note, however, that one cannot expect that two treat- The National Highway Traffic Safety Administration ments with estimated levels of effectiveness A% and B% (NHTSA) works with highway safety offices within states to will produce a reduction of A% + B% if applied to the facilitate such an assessment procedure using outside experts. same driver group. The combined effect would be ex- A brief description of NHTSA's program assessment process pected to be less. Unfortunately, since we do not have can be found at A more detailed good data on the effectiveness of individual treatments, description of this process for impaired driving and recent we have even less knowledge about the effectiveness of findings from such assessments can be found at http://www. combined treatments. In the absence of such knowledge, it is suggested that the effectiveness level of the second Finally, it is strongly recommended that the findings of an treatment applied to a given driver subgroup be reduced assessment of program needs be combined with crash-based to 50 percent of the level originally estimated for that information on the size of the AR problem attributable to