Overview

Introduction

The United States has made substantial progress in reducing alcohol-related traffic fatalities in recent decades. Since the National Highway Traffic Safety Administration (NHTSA) began keeping records on alcohol involvement in fatal crashes in 1982, the proportion of all traffic fatalities that are alcohol-related has declined steadily from 57% to 38%.1 Despite this progress, alcohol-related motor vehicle crashes continue to be a major public health problem, resulting in 15,786 deaths and more than 300,000 injuries in 1999.1

Since 1970, individual states and communities have implemented a broad range of strategies to reduce alcohol-impaired driving. Laws to deter alcohol-impaired driving and to control the sale or public consumption of alcohol are among the most widely used strategies. By 1987, all states had enacted a minimum legal drinking age of 21 years. As of May 1, 2001, a total of 24 states, the District of Columbia, and Puerto Rico had lowered the illegal blood alcohol concentration (BAC) for drivers aged 21 years and older from 0.10 g/dL to 0.08 g/dL. Community-based interventions, including sobriety checkpoints, enhanced enforcement of alcohol control policies, and training programs for servers of alcoholic beverages, have also been implemented in some states.

Data provided by NHTSA and the U.S. Bureau of the Census in 1999 indicate that the United States nearly met the Healthy People 2000 objective for alcohol-related motor vehicle deaths of no more than 5.5 deaths per 100,000 persons, with a rate of 5.8 per 100,000 persons.1-3 The Healthy People 20104 target for alcohol-related motor vehicle fatalities is 4.0 per 100,000 persons or fewer. Meeting the 2010 objective will require a further decrease of 31% in the rate of alcohol-related motor vehicle fatalities. The recommendations of the Task Force on Community Preventive Services (the Task Force)5 are intended to help communities, working with public health and traffic safety professionals, identify and implement effective programs and policies.

Some of the laws evaluated in the systematic review have already been widely enacted in the United States. Information about the effectiveness of these laws will be useful in evaluating any future proposals to repeal or to revise them. In addition, the systematic review highlights important unanswered questions about the effectiveness of these laws in various settings (e.g., urban vs rural).

Conceptual Approach

This systematic review was undertaken to assess the effectiveness of a number of laws and other community-based interventions in reducing alcohol-impaired driving and alcohol-related motor vehicle crash fatalities in the United States and other Established Market Economies.^a Of the 76 studies included in the review, 55 (72%) were conducted in the United States. Other studies were conducted in Australia, Canada, New Zealand, France, and The Netherlands.

We focused on interventions for which the primary goal was to reduce alcohol-impaired driving. We did not review interventions intended primarily to restrict access to alcohol (e.g., alcohol taxation, alcohol outlet zoning restrictions) or to address health outcomes of alcohol abuse or misuse other than alcohol-impaired driving. Those topics will be included in the Task Force's review of interventions to prevent alcohol abuse and misuse as part of the Guide to Community Preventive Services (the Community Guide).

The logic framework shown in Figure 1 depicts the conceptual approach that guided the review process. This figure illustrates the hypothesized links between interventions to reduce alcohol-impaired driving and the outcomes of interest. These interventions are thought to work through three separate pathways: They may reduce alcohol-impaired driving by increasing the perceived risk of detection and punishment; They may reduce alcohol consumption in high-risk settings or among high-risk groups; and they may foster a social norm that reduces the acceptable amount of alcohol to consume before driving.

Figure

Figure 1. Logic framework for reviews of interventions to reduce alcohol-impaired driving. BAC, blood alcohol concentration.

Methods

The Community Guide's methods for systematic reviews and for linking evidence to recommendations have been published elsewhere.6 An overview of the general methods used in the systematic reviews of interventions to reduce motor vehicle occupant injury appears in the supplement.7 This discussion is limited to topics that apply specifically to interventions to reduce alcohol-impaired driving.

Selecting Interventions

The consultation team (see Acknowledgments) generated a comprehensive list of interventions to reduce alcohol-impaired driving and created a priority list of interventions to be reviewed after surveying consultants and other experts. Those individuals were asked to rank interventions as priorities for systematic review, considering whether each intervention is (1) thought to be effective but underused; (2) thought to be ineffective but overused; (3) popular, but its effectiveness is not well established; (4) costly, but its effectiveness is not well established; (5) targeted to a specific population of interest (e.g., youth); or (6) broad reaching, and could achieve large reductions in alcohol-impaired driving if found to be effective. Rankings were compiled, and the 12 interventions with the most votes were selected as priorities for this review (Table 1). Resource limitations prevented us from completing reviews of all of the priority interventions in time for this publication. Additional reviews will be published as they are completed.

Table

Table 1. Priority interventions selected for review.

Selecting Summary Effect Measures

The primary outcomes assessed in this literature are fatal and nonfatal injuries resulting from alcohol-related motor vehicle crashes. This information is primarily derived from police incident reports. In the United States, information about all fatal crashes that occur on public roads is available in electronic form through NHTSA's Fatality Analysis Reporting System (FARS).8 There is no comparable single source of electronic information about all nonfatal crashes. Of the 69 studies that examined crash data, 35 (51%) examined only fatal crashes.

Differences in how "alcohol-relatedness" of crashes is operationally defined from study to study contribute to the variability in the effect measures in this review. Until recent decades, the BACs of drivers involved in fatal crashes were measured too sporadically to be useful in evaluating interventions to reduce alcohol-impaired driving. Objective measures of alcohol involvement in nonfatal crashes continue to be collected only sporadically. Given the limited availability of BAC data, many studies have used proxy variables for alcohol-related crashes. Commonly used proxy variables, and their estimated level of association with alcohol involvement, are listed in Table 2. Using proxy variables for alcohol involvement produces effect estimates that are biased toward the null, with the degree of bias being more pronounced for proxies with weaker association with alcohol involvement.

Table

Table 2. Estimated probability of alcohol involvement for various crash types in the United States, 1999.

Important differences exist in the operational definitions of "alcohol-relatedness" of crashes even among studies that use driver BAC data. For fatal crashes in the United States, the FARS system uses a statistical model to estimate the BAC of drivers for whom BACs were not obtained.9 Some of the studies in this review used the estimated values for BAC provided by the FARS statistical model. Other studies considered as alcohol-related only those crashes involving drivers with measured BACs above an established level. Additionally, studies defined alcohol-relatedness using various BAC cutpoints (i.e., >0.01 g/dL, >0.08 g/dL, or >0.10 g/dL).

We often had to select from several possible effect measures. We established and consistently applied rules for identifying the outcome measure that most adequately reflected alcohol-related crashes and addressed potential confounding variables. Briefly, we considered BAC data to be the most objective measure of alcohol-relatedness of crashes. For studies that reported results using more than one cutpoint for BAC, we chose the results based on the BAC cutpoint closest to 0.10 g/dL.

When available, we selected effect measures that compared alcohol-related fatalities with non-alcohol-related fatalities (e.g., proportion of all fatal crashes involving drivers with BACs of >0.10 g/dL; ratio of single-vehicle nighttime fatal crashes to multi-vehicle daytime fatal crashes) over the absolute number of alcohol-related fatalities. These effect measures help control for both the long-term downward trend in total fatal crashes and factors that influence the total number of crashes, such as weather, economic conditions, vehicle miles traveled, and safety characteristics of vehicles and highways.10 When available, we also selected effect measures that incorporated a concurrent comparison group such as drivers in adjacent states or drivers within the same state who were unaffected by the intervention. For those studies, results were reported in the form of the net change, reflecting the difference between the percent change for the intervention group and the comparison group. Net change was calculated by using the formula

(I_post - I_pre)/I_pre - (C_post - C_pre)/C_pre

where: I refers to the group exposed to the intervention,C refers to the group not exposed to the intervention (the comparison group),post refers to outcome measurements after implementation of the intervention, andpre refers to outcome measurements before implementation of the intervention.

For studies using interrupted time series or other regression-based analyses, results were reported in terms of the percent change estimated from the model.

The other outcomes assessed in this review were BACs of drivers at roadside surveys, as well as measured and estimated BACs of people leaving bars or other licensed establishments. Net changes in these outcomes were calculated by using the same formula as for the crash outcomes.

Effect measures from individual studies are displayed in figures, and a median effect measure and range for each outcome of interest is reported. For median effect measures based on seven or more studies, the interquartile range is reported. For interventions with a large number of studies, we also evaluated whether the intervention's effect varied by follow-up time.

Intervention Effectiveness and Economic Efficiency

.08 BAC Laws

In the United States, states have two basic types of alcohol-impaired driving laws. The first type prohibits a person from driving while intoxicated (DWI). Originally, these laws did not require evidence of a specific BAC. The second type of law, which came later, made it illegal "per se" to operate a motor vehicle at or above a specified BAC. These laws, referred to as per se laws, were usually enacted in addition to the existing DWI laws. Originally, most per se laws specified a BAC of 0.10 g/dL or 0.15 g/dL as being illegal. In 1983, Utah and Oregon lowered the illegal BAC from 0.10 g/dL to 0.08 g/dL. By May 1, 2001, a total of 24 states, the District of Columbia, and Puerto Rico had enacted laws lowering the illegal BAC to 0.08 g/dL. These laws, referred to as .08 BAC laws, are the subject of this review.

In the United States, per se laws apply to all drivers, but they target primarily drivers aged 21 years and older. This target is because, as of July 1998, all states had enacted per se laws for drivers aged 20 years and younger that establish BAC limits of 0.02 g/dL or less.

Reviews of evidence

Effectiveness

The evidence base for .08 BAC laws included published journal articles, technical reports, and conference papers. Our search identified nine studies, all of which were of sufficient design quality and execution to be included in the review.10-18 Descriptive information about the quality, study design, and outcome measures from these studies is presented in Table 3. Details of the nine qualifying studies are provided in the Appendix and at the website (www.thecommunityguide.org).

Table

Table 3. .08 BAC laws: descriptive information about included studies.

All nine studies analyzed data from police incident reports of crashes occurring on public roadways. Post-law follow-up times for individual state laws ranged from 1 to 14 years (median, 5).

Eight of the nine studies reported the percent change in alcohol-related fatalities (post-law period vs pre-law period) or provided the data needed to calculate the measure.10,12-18 Seven studies provided state-specific percent change values, and the remaining study provided a summary percent change value for all 16 states that enacted .08 BAC laws before January 1, 1998 (Figure 2). The median post-law percent change in alcohol-related motor vehicle fatalities was -7% (interquartile range, -15% to -4%). Results were generally consistent in direction and size across the studies.

Figure

Figure 2. Percent change in measures of alcohol-related motor vehicle fatalities after .08 BAC laws were enacted, by state. * Numbers correspond to reference numbers of cited studies. † Median percent change calculated by using the median value (more...)

One study reported a 45% net increase (95% confidence interval [CI], -13% to +144%) in fatalities after enactment of the .08 BAC law in Vermont.10 This study compared alcohol-related fatalities in Vermont with those in New Hampshire. The result of this comparison is imprecise because there were fewer than 100 alcohol-related motor vehicle fatalities in each state during the 3-year study period.

Evaluations of .08 BAC laws in the states of California, Maine, Oregon, Utah, and Vermont10,11,14-17 were limited in their ability to separate the effect of .08 BAC laws from that of administrative license revocation (ALR) laws, which were enacted at about the same time. ALR laws allow the arresting officer, judge, or magistrate to seize the license of a driver who refuses or fails a BAC test. Two recently published studies provide summary estimates of the effect of .08 BAC laws independent of ALR laws. In a secondary analysis, Hingson et al.13 reported an overall post-law decline in alcohol-related fatal crashes of 5% in four states that had long-standing ALR laws. Voas et al.18 estimated the separate effects of BAC laws and ALR laws by using multivariate regression analysis. They reported an 8% decline in fatally injured alcohol-impaired drivers attributable to .08 BAC laws.

Applicability

The states represented in the evidence base are geographically diverse with varying population densities. Because all of the studies analyzed data from statewide police incident reports of fatal crashes, the evidence of effectiveness should be applicable to all drivers affected by .08 BAC laws. None of the studies, however, provided data to assess differences in effectiveness for various subgroups of the driving population.

In support of .08 BAC laws, the U.S. Congress included a provision in the Fiscal Year 2001 Department of Transportation and Related Agencies Appropriations Act19 that requires states to implement .08 BAC laws by October 2003 or risk losing federal highway construction funds.

Other positive or negative effects

Three studies measured outcomes other than motor vehicle crashes, including public knowledge and perception of impaired driving laws, self-reported impaired driving, and impaired driving arrests.12,15,16 Information about these other potential effects was not summarized in this review.

Economic

No studies were found that met the requirements for inclusion in a Community Guide review.7

Barriers to intervention implementation

One potential barrier to implementation of .08 BAC laws is the view that the laws discourage "social drinkers" from driving after drinking small amounts of alcohol but do not deter "hard-core" drinking drivers. Results of the systematic review provide some evidence to counter this view. Five of the nine studies measured fatalities involving drivers with BACs of 0.10 g/dL or higher, and these studies reported post-law reductions for most states.11-14,18

Conclusion

According to the Community Guide's rules of evidence, available studies provide strong evidence that .08 BAC laws are effective in reducing alcohol-related crash fatalities.

Lower BAC Laws for Young or Inexperienced Drivers

Lower BAC laws for young or inexperienced drivers establish a lower illegal BAC for these drivers than for older or more experienced drivers. Although these laws are commonly referred to as "zero tolerance" laws, in many jurisdictions the BAC limit for affected drivers is slightly above zero (e.g., 0.02 g/dL).

Young people who drive after drinking alcohol pose an inordinate risk to themselves, their passengers, and other road users. A recent U.S. study20 estimated that male drivers aged 16 to 20 years with BACs in the range of 0.08 to <0.10 g/dL were 24 times more likely to die in a motor vehicle crash than those with BACs of zero.

In the United States, lower BAC laws have typically applied to all drivers younger than the minimum legal drinking age of 21 years. In Austria, Australia, New Zealand, and some Canadian provinces, lower BAC laws apply either to all newly licensed drivers or to newly licensed drivers younger than a specified age.21 The first U.S. laws lowering the illegal BAC for underage drivers were enacted in 1983 in Maine and North Carolina. By December 1994, 27 states and the District of Columbia had enacted lower BAC laws, with BAC limits ranging from any detectable level of BAC to 0.07 g/dL.22 In support of lower BAC laws, the U.S. Congress included a provision in the National Highway Systems Designation Act of 1995 that required states to implement a BAC limit of 0.02 g/dL or less for all drivers younger than the age of 21 years by October 1998 or risk losing federal highway construction funds. By July 1998, all 50 states had enacted lower BAC laws.

Reviews of evidence

Effectiveness

The evidence base for this intervention included published journal articles, technical reports, and conference proceedings. We found nine publications22-30 that reported on 14 separate studies of the effectiveness of lower BAC laws. Two additional papers provided more information about an already-included study.31,32 Descriptive information about the quality, study design, and outcome measures from these studies is presented in Table 4. Details of the six qualifying studies are provided at the website (www.thecommunityguide.org).

Table

Table 4. Lower BAC laws for young or inexperienced drivers: descriptive information about included studies.

Four of the six studies were conducted in the United States,22,23,29,30 and the remaining two were conducted in Australia.24,26 Two of the U.S. studies evaluated lower BAC laws in multiple states.22,30 All six studies analyzed data from police incident reports of motor vehicle crashes occurring on public roadways. Post-law follow-up times for individual state laws ranged from less than 1 year to 15 years. The median post-law follow-up time for the six studies was 22 months.

Each of the six studies reported a post-law reduction in crashes. The three studies that examined fatal crash outcomes reported declines of 24%,30 17%,22 and 9%.29 The two studies that examined fatal and nonfatal injury crashes reported declines of 17%26 and 3.8%.24 The study that examined crashes in which the investigating police officer believed that the driver had been drinking alcohol reported a decline of 11%.23

Applicability

The same body of evidence used to assess effectiveness was used to assess the applicability of these interventions in various settings. The states studied are geographically diverse and have both urban and rural populations. Because all of the studies analyzed data from the statewide files of police-reported crashes, the evidence of effectiveness should be applicable to all drivers affected by these lower BAC laws. None of the studies, however, provided data to assess differences in effectiveness for various subgroups of the affected population.

Lower BAC laws have been enacted for other defined populations not addressed in this review, including commercial truck drivers and people convicted of driving while impaired. The Centers for Disease Control and Prevention has recommended that states consider enacting lower BAC laws for all drivers who transport children.33

Other positive or negative effects

It is possible that drivers younger than the age of 21 years with high BACs could receive "zero tolerance" citations for violating the lower BAC law, whereas adults with the same BAC would be arrested for the more serious offense of driving under the influence of alcohol (DUI). Voas et al.29 explored this potential negative effect in an evaluation of California's 1994 lower BAC law. They reported that the combined rate of alcohol-related license suspensions for zero tolerance citations and DUI arrests among underage drivers increased only slightly after enactment of the 1994 lower BAC law. Furthermore, 57% of underage drivers who received zero tolerance citations had BACs above 0.08 g/dL. The investigators concluded that California's 1994 lower BAC law resulted in about half of the potential DUI arrests among underage drivers being converted to less serious zero tolerance citations.

Economic

One study34 met the criteria for inclusion6,7 in the review of lower BAC laws for young or inexperienced drivers. The study applied previously published crash costs and used effectiveness data from other previously published studies to illustrate how these costs could be applied to lower BAC laws in the United States. The benefits from a reduction in alcohol-related crashes were estimated on the basis of the assumption that lower BAC laws reduce young drivers' alcohol-related crashes by 20%. Monetary benefits and costs were reported in dollars per mile driven.

The study conducted a cost-benefit analysis. The estimated benefit-to-cost ratio^b for lower BAC laws was $11 per dollar invested when violators receive a 6-month license suspension. Costs included the cost of trials and sanctions imposed and compliance costs to young drivers (i.e., cost of the loss of mobility).

The study was classified as satisfactory, based on the quality assessment criteria for economic evaluations used in the Community Guide.7,35 Study details, adjusted results, and quality scoring are provided in the economic evaluation summary tables at the website (www.thecommunityguide.org).

Barriers to intervention implementation

All U.S. states currently have lower BAC laws for drivers younger than age 21 years. Voas et al.29 discussed several potential barriers to full enforcement of these laws. Because young people are less likely than adults to drink in bars, police patrols that target bar neighborhoods are likely to miss underage drinking drivers. Also, officers may have difficulty identifying underage drinking drivers with low BACs who do not show signs of impairment. Finally, because of ambiguities, some state laws do not authorize officers to test the BAC of an underage driver unless the officer has probable cause to believe that the driver's BAC is above the legal limit for adults.

Conclusion

According to the Community Guide's rules of evidence, there is sufficient evidence that lower BAC laws are effective in reducing alcohol-related crashes among young or inexperienced drivers.

Minimum Legal Drinking Age Laws

Minimum legal drinking age (MLDA) laws specify an age below which the purchase or public consumption of alcoholic beverages is illegal. Studies included in this review assessed the effect of raising or lowering the MLDA on crashes and related fatal and nonfatal injury outcomes.

In the United States, several states lowered their MLDA during the early 1970s. Shortly thereafter, in response to an increase in motor vehicle fatalities among young people, some of these states raised their MLDA. To address continuing concerns about youth drinking and driving, federal legislation requiring states to adopt a minimum drinking age of 21 years or lose highway funds was passed in 1984. By 1987, all U.S. states had adopted an MLDA of 21.

Reviews of evidence

Effectiveness

The evidence base for this intervention included only published journal articles. We reviewed three bodies of evidence that evaluated the effect of MLDA changes: studies of the effect of raising the MLDA, studies of the effect of lowering the MLDA, and studies that used multiple regression to evaluate the effect of MLDA changes. The regression-based studies are reported separately because they cover all U.S. states during overlapping time periods, and their results are not independent of each other. Most studies in the review assessed the effect of changes in the MLDA from 18 to 21 years or vice versa. Outcomes were typically assessed in the age groups affected by the law change.

Forty-nine studies reported in 46 papers met the inclusion criteria for this review: 17 studies of the effect of raising the MLDA,36-52 11 studies of the effect of lowering the MLDA contained in eight reports,53-60 and 18 regression-based studies of the effect of changing the MLDA.30,61-77 Three papers provided additional information about already-included studies.78-80 Descriptive information about the quality, study design, and outcome measures from all MLDA studies is presented in Table 5. The effects of changes in the MLDA on crash outcomes likely to involve alcohol are summarized in Table 6. Details of the 33 qualifying studies are provided at the website (www.thecommunityguide.org).

Table

Table 5. Minumum legal drinking age laws: descriptive information about included studies.

Table

Table 6. Effects of changing the mimiumum legal drinking age: summary effects from the body of evidence on crash outcomes likely to involve alcohol.

Figure 3 presents the findings from the evidence base, aggregated across all crash outcomes. These results suggest that changes in the MLDA result in changes of roughly 10% to 16% in alcohol-related crash outcomes for the targeted age groups, decreasing when the MLDA is raised, and increasing when it is lowered. These effects were consistent over follow-up times ranging from 7 to 108 months.

Figure

Figure 3. Percent change in aggregated crash outcomes after changes in the MLDA. The "a" and "b" in study names refer to the first or second study by the same author in that year, included in this review. Listed studies for which the author name and (more...)

In some studies, the age group directly affected by the change in the MLDA was not identical to the age group on which outcomes were evaluated. These discrepancies usually arose when crash data pertaining only to the affected age group were not available, or when young people who were of legal drinking age before the law change were allowed to continue to purchase or consume alcohol (i.e., were grandfathered). For 15 studies with perfect overlap between the age group targeted by the law and the age group analyzed,30,40-43,45,46,48,51-55,58,67 the median change in crashes was 19%. For 17 studies (in 14 reports) with misclassification into the outcome group of subjects not affected by the MLDA change,36,37,39,44,50,56,60,61,65,66,68,73-75 the median change was only 12%.

Applicability

All of the changes in MLDA assessed in this review affected drivers aged 18 to 20 years. All studies assessed changes in state or provincial laws. Of the 33 studies included in the review, 27 were conducted in the United States,30,36,37,39-48,50-52,54-56,61,65-68,73-75 one was conducted in the United States and Canada,53 and the remaining studies were conducted in Australia (4 studies in one report)60 or Canada.58 The generalizability of these findings to other countries may be limited by differences in patterns of alcohol consumption and driving among 18- to 20-year-olds.

Other positive or negative effects

Several studies reported that raising the MLDA was associated with decreased alcohol consumption.36,45,47,51 We did not review this literature systematically but present relevant findings in the evidence tables (available at the website www.thecommunityguide.org). Nine studies also investigated the effect of raising the MLDA on crashes involving adolescent drivers who were younger than the MLDA both before and after it was raised.37,42,44,46,47,51,66,68,74 Although these studies indicated that raising the MLDA was associated with a median decline in crashes of 6% (interquartile range, -18% to 5%), the size of this effect was inconsistent across studies, with several showing no effect.

Some investigators have postulated that when drivers who have not been legally allowed to drink reach the MLDA, their risk of alcohol-related crash involvement will dramatically increase because of their inexperience in drinking, thus partly or completely offsetting the benefits of MLDA laws.44,72 Studies that attempt to directly estimate the "drinking experience" effect have produced inconsistent results because of problems in statistically disentangling it from the effect of the MLDA itself.44,67,72,81 In one study of a cohort that would be affected by both the MLDA and drinking experience effects, raising the MLDA from 18 to 21 years was estimated to decrease nighttime fatal crashes by 15% (95% CI, 3% to 27%).67 This result is similar to the median effect estimate for MLDA alone, suggesting that if the putative drinking experience effect exists, it does not substantially diminish the benefits of raising the MLDA.

Economic

No studies were found that met the requirements for inclusion in a Community Guide review.7

Barriers to intervention implementation

Currently, the MLDA is 21 years throughout the United States and 18 years in many other countries (e.g., Australia). The belief among some opponents of MLDA laws that prohibition of drinking among young adults unjustly punishes them for the irresponsible behavior of the subgroup that drives after drinking poses a potential barrier to the strengthening or maintenance of MLDA laws.82

Conclusion

According to the Community Guide's rules of evidence, there is strong evidence that MLDA laws, particularly those that set the MLDA at age 21, are effective in preventing alcohol-related crashes and associated injuries.

Sobriety Checkpoints

At sobriety checkpoints, law enforcement officers systematically stop drivers to assess their degree of alcohol impairment. There are two types of sobriety checkpoints. At random breath testing (RBT) checkpoints, all drivers stopped are given breath tests for BACs. RBT checkpoints are used in Australia and several European countries. Issues about the violation of constitutional protections against unreasonable search and seizure83 have prevented the use of RBT checkpoints in the United States. At selective breath testing (SBT) checkpoints, used in many U.S. states, police must have reason to suspect the driver stopped at a checkpoint has been drinking before a breath test can be demanded. Both types of sobriety checkpoint programs generally include media efforts to publicize the enforcement activity and the consequences of driving with a BAC above the legal limit.

The rationale for the use of checkpoints is based on deterrence theory. Although checkpoints may remove some drinking drivers from the road, their primary goal is to reduce driving after drinking by increasing the perceived risk of arrest. This perceived risk may be influenced by the level of publicity accompanying the enforcement effort, visibility of the checkpoint operations themselves, the likelihood of detection, and drivers' beliefs about their ability to avoid detection.84

Reviews of evidence

Effectiveness

The evidence base for this intervention included published journal articles, technical reports, and Association for the Advancement of Automotive Medicine proceedings. We found 15 studies of the effectiveness of SBT checkpoints.85-99 Two additional papers provided more information about an already-included study,100,101 and one presented data in a form that could not be converted to our summary effect measure.99 Descriptive information about the quality, study design, and outcome measures from these studies is presented in Table 7. Our search identified 17 studies of the effectiveness of RBT checkpoints (four of these studies were reported in one paper).84,102-114 Two additional papers provided information about an already-included study.115,116 Descriptive information about the quality, study design, and outcome measures from these studies is presented in Table 8.

Table

Table 7. Selective breath testing sobriety checkpoints: descriptive information about included studies.

Table

Table 8. Random breath testing sobriety checkpoints: descriptive information about included studies.

Details of the 23 qualifying studies are provided at the website (www.thecommunityguide.org). Summaries of study outcomes are reported in Tables 9 (SBT) and 10 (RBT). Outcomes from studies reporting crash-related outcomes are also provided in Figures 4 (SBT) and 5 (RBT). Both SBT and RBT checkpoints consistently resulted in decreased crashes. Length of follow-up time ranged from 1 to 120 months (median, 14) and did not appear to influence the size of the declines.

Table

Table 9. Effects of selective breath testing checkpoints on crash outcomes likely to involve alcohol: summary effects from the body of evidence.

Table

Table 10. Effects of random breath testing checkpoints on various outcomes: summary effects from the body of evidence.

Figure

Figure 4. Percent change in crashes likely to involve alcohol after implementing selective breath testing checkpoint programs.

Figure

Figure 5. Percent change in crashes likely to involve alcohol after implementing random breath testing checkpoint programs. Listed studies for which the author name and year are identical come from a single paper.110.

One study assessed the effect of RBT checkpoints on the observed incidence of drinking and driving. This study found that during an RBT checkpoint program, the proportion of drivers with any detectable BAC level declined 13% and the proportion of drivers with BACs above 0.08 g/dL declined 24% from prior levels.111

Although RBT checkpoints have greater sensitivity in detecting drinking drivers than SBT checkpoints, this review found no evidence that their effectiveness for reducing alcohol-related crashes differed. None of the studies reviewed was designed to directly compare the effectiveness of RBT and SBT checkpoints, however, so these results should be interpreted with caution. Sensors that allow police to passively sample air in the car for alcohol vapors (passive alcohol sensors) can improve the detection rate at SBT checkpoints by approximately 50%.99 If such technology becomes more widely used, the sensitivity in detecting drinking drivers at SBT checkpoints may approach that of RBT checkpoints.

Applicability

The same body of evidence used to assess effectiveness was used to assess the applicability of these interventions. Studies that met our quality criteria involved a somewhat larger scale of enforcement and publicity activity than studies that were excluded because of quality limitations. Thus, the reported results may be most generalizable to these larger-scale interventions. The studies were conducted on interventions implemented at the city,86,87,95,97,98,102,111 county,91 state,84,85,90,93,94,104,105,108,110,112 and national level114 and were evaluated in rural areas,106,112 in urban areas,86,87,95,97,98,102,111,112 and in mixed rural and urban areas.84,85,90,93,94,104,105,108,110

Other positive or negative effects

Several studies report the arrest of drivers stopped at sobriety checkpoints for other offenses, such as driving with a suspended license or carrying weapons, as an added benefit.90,94,95,97

One negative effect of stopping drivers at checkpoints is the resulting inconvenience and intrusion on driver privacy. According to the U.S. Supreme Court, the brief intrusion of a properly conducted sobriety checkpoint is justified in the interest of reducing alcohol-impaired driving.117 Some civil libertarian groups have also endorsed this position.84

In the United States, checkpoints use established protocols to ensure that they are conducted properly.83 Common components of the protocols include selecting checkpoint locations on the basis of objective criteria (e.g., the incidence of alcohol-related crashes in the area) and stopping cars according to a predetermined system (e.g., every third car that approaches the checkpoint).90,94,97

Economic

Four studies95,104,118,119 were included in the review of sobriety checkpoints. Two studies95,118 evaluated SBT checkpoints and two studies104,119 evaluated RBT checkpoints. All studies conducted cost-benefit analyses. Three studies104,118,119 reported annual net benefits (the stream of benefits minus the stream of costs incurred in 1 year), and the fourth study95 reported net benefits for the length of the intervention (9 months). Study details, adjusted results, and quality scoring for all four studies are provided in the economic evaluation summary tables at the website (www.thecommunityguide.org).

Selective breath testing checkpoints. The first study118 modeled a 1-year campaign conducted in a hypothetical community of 100,000 licensed drivers in the United States. The modeled campaign consisted of 156 checkpoints (4 hours each) per year. The effect size assumed in the analysis was a 15% reduction in alcohol-related crashes. Program costs included in the analysis were personnel, equipment, travel delay, trial, punishment, and mobility loss associated with sanctioning (e.g., loss of driver's license). The estimated annual total benefit from alcohol-related crashes averted was $9.2 million (in 1997 U.S. dollars). Benefits were estimated by accounting for medical care, property damage, and public costs averted plus future earnings and quality of life gained. Estimated annual total costs of the intervention were $1.6 million. The estimated annual net benefit was $7.6 million (in 1997 U.S. dollars), resulting in a benefit-to-cost ratiob of $6 per dollar invested. This study was classified as very good, based on the quality assessment criteria used in the Community Guide.

The second study95 evaluated a 9-month campaign conducted in four communities in California with checkpoint sites (a fifth community was a comparison group and a sixth implemented roving DWI patrols). The program consisted of 18 checkpoints per community plus publicity campaigns and education programs. Net reductions in alcohol-related crash injuries and deaths ranged from 17.5% to 31.6%. Total aggregated benefits of $3.86 million (in 1997 U.S. dollars) came from societal savings realized through injuries and fatalities avoided. Costs included personnel and equipment. Cost of the publicity campaigns and education programs were not included in the analysis. Total costs of the intervention (aggregated for four communities) were $164,552. The aggregated net benefit was $3.7 million, resulting in a benefit-to-cost ratio of $23 per dollar invested. This study was classified as satisfactory, based on the quality assessment criteria used in the Community Guide.

In summary, both studies show positive net benefits and, therefore, from a societal viewpoint, economic benefits of these interventions exceed costs. The hypothetical study reported net benefits almost double those reported by the California study although greater disparity was observed in the benefit-to-cost ratio ($6 in the hypothetical study vs $23 in the California study). The high benefit-to-cost ratio reported by the California study is due, in part, to the underestimation of total costs because the analysis did not include the cost of publicity and education.

Random breath testing checkpoints The first study104 was conducted 3 years after statewide RBT checkpoints were introduced in New South Wales, Australia. The program included police operations, media publicity, and revised drunk-driving penalties. Annual total benefits were $228 million (in 1997 U.S. dollars) and were calculated on the basis of the assumption that 75% of the reduction in fatal crashes, serious injuries, minor injuries, and tow-away crashes was attributable to the checkpoints. Annual total program costs were $4 million (in 1997 U.S. dollars) and included personnel, equipment, publicity, and transportation. The net annual benefit reported in the study was $224 million. This study was classified as satisfactory, based on the quality assessment criteria used in the Community Guide.

The second study119 evaluated a proposed nationwide RBT checkpoint intervention in The Netherlands. The proposed intervention included a publicity component and incorporated a more efficient method of transporting offenders to police stations. The checkpoint program was assumed to result in a 25% reduction in alcohol-related injury or property damage on weekend nights. Annual total benefits from cost-savings in the reduction in alcohol-related injury and property damage were estimated at $31.4 million (in 1997 U.S. dollars). The investigators did not specify the value of statistical life used to calculate the cost-savings from averted death or the procedure and assumptions to calculate cost-savings from averted injury. Annual total costs, including materials and publicity, were estimated at $15.6 million. The annual net benefit of the intervention was estimated to be $15.8 million, resulting in a benefit-to-cost ratio of $2 per dollar invested. This study was classified as good, based on the quality assessment criteria used in the Community Guide.

In summary, both studies showed positive net benefits (i.e., the economic benefits of the interventions are greater than the economic costs). The Australian intervention, which was more intensive and reached one in three drivers, showed larger net benefits than the modeled Netherlands intervention, which was designed to reach one in nine drivers.

Barriers to intervention implementation

Although the U.S. Supreme Court has determined that SBT checkpoints are permissible,117 some state courts prohibit them. Where checkpoints are permitted, police concern about low arrest rates can be an important barrier.89 Informing police officers about the general deterrence benefit of their efforts and providing them with regular feedback that links these efforts to crash prevention may decrease this frustration.85,109

Conclusion

According to the Community Guide's rules of evidence, available studies provide strong evidence that both SBT and RBT sobriety checkpoints are effective in preventing alcohol-impaired driving, alcohol-related crashes, and associated fatal and nonfatal injuries.

Research Issues

Discussion

Interventions to prevent alcohol-impaired driving are implemented within the social and legal context of a community. Although these reviews evaluate each intervention as an independent activity, effective prevention of impaired driving requires a comprehensive and systematic approach that addresses various individual and ecologic influences on drinking and driving behavior.143-145 These reviews can help decision makers identify and implement effective interventions that fit within an overall prevention strategy.

We thank the following individuals for their contributions to this review: Erin Finley, Krista Hopkins, and Patryce Young-Curtis, Research Assistants; Sandra Bonzo and Joanna Taliano, Research Librarians; Kate W. Harris, Editor; our Consultation Team—J. C. Bolen, PhD, MPH, National Center for Chronic Disease Prevention and Health Promotion, CDC, Atlanta, GA; R. D. Brewer, MD, MSPH, Nebraska Department of Health, Lincoln; S. D. Bryn, MPH, Health Resources Services Administration, Rockville, MD; F. M. Council, PhD, University of North Carolina, Chapel Hill; R. W. Denniston, MA, Substance Abuse and Mental Health Services Administration, Rockville, MD; A. C. Gielen, ScD, ScM, Johns Hopkins University, Baltimore, MD; S. Gorcowski, MA, National Highway Traffic Safety Administration, Washington, DC; C. A. Hurley, BA, National Safety Council, Washington, DC; B. H. Jones, MD, MPH, National Center for Injury Prevention and Control, CDC, Atlanta, GA; T. A. Karlson, PhD, University of Wisconsin, Madison; M. R. Kinde, MPH, Minnesota Department of Health, Minneapolis; D. W. Lawrence, MPH, RN, San Diego State University, CA; S. E. Martin, PhD, National Institute for Alcohol Abuse and Alcoholism, Rockville, MD; J. A. McKnight, PhD, National Public Service Research Institute, Landover, MD; A. D. Mickalide, PhD, CHES, National SAFE KIDS Campaign, Washington, DC; L. F. Novick, MD, MPH, Onandaga County Department of Health, Syracuse, NY; F. P. Rivara, MD, MPH, University of Washington, Seattle; C. W. Runyan, PhD, MPH, University of North Carolina, Chapel Hill; R. J. Smith, MS, Health Resources and Services Administration, Rockville, MD; P. F. Waller, PhD, University of Michigan, Ann Arbor; A. F. Williams, PhD, Insurance Institute for Highway Safety, Arlington, VA; and our Abstraction Team -- J. L. Alongi, MPH, Los Angeles County Department of Health Services, Los Angeles, CA; P. Bellas, MD, MPH, South Bay Family Healthcare Center, Redondo Beach, CA; S. Corbett-Perez, MHSE, University of Florida, Gainesville; R. A. Eidman, MPH, Triple Hayz Corporation, Los Angeles, CA; J. Kattapong, MA, MD, MPH, Social Security Administration, Tucson, AZ; S. Levesque, MD, MPH, Inscription House Health Center, Navajo/Indian Health Service, Shoto, AZ; D. Pallin, MD, MPH, Mt. Sinai School of Medicine, New York, NY; M. Patterson, MS, PhD, University of Washington and Harborview Injury Prevention Research Center, Seattle; S. L. Usdan, PhD, Mississippi State University, Mississippi State; C. Ward, Johns Hopkins School of Public Health, Dr. PH candidate, Baltimore, MD.

References and Notes

[1]

Discussion National Highway Traffic Safety Administration. Traffic safety facts, 1999: a compilation of motor vehicle crash data from the Fatality Analysis Reporting System and the General Estimates System. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 2000. DOT HS 809 100.

[2]

National Center for Health Statistics. Healthy People 2000 Review, 1998-99. Hyattsville, MD: Public Health Service, 1999. DHHS Pub No (PHS) 99-1256.

[3]

Bureau of the Census, Economics and Statistics Administration, U.S. Department of Commerce. Census data to calculate the rate of alcohol-related motor vehicle crash fatalities. 2000. Available at: www​.census.gov/population​/www/estimates/popest.html. Accessed November 1, 2000.

[4]

U.S. Department of Health and Human Services, With understanding and improving health and objectives for improving health, 2 vols. (2000) Healthy People 2010, 2nd ed. Washington, DC: U.S. Government Printing Office.

[5]

Task Force on Community Preventive Services, Recommendations to reduce injuries to motor vehicle occupants: increasing child safety seat use, increasing safety belt use, and reducing alcohol-impaired driving Am J Prev Med 2001,21suppl 4:16–22. [PubMed: 11691558]

[6]

Briss P.A., Zaza S. and Pappaioanou M. et al. Task Force on Community Preventive Services; Developing an evidence-based Guide to Community Preventive Services—methods Am J Prev Med 2000,18suppl 1:35–43.

[7]

Task Force on Community Preventive Services, Zaza S., Carande-Kulis V.G. and Sleet D.A. et al. Methods for conducting systematic reviews of the evidence of effectiveness and economic efficiency of interventions to reduce injuries to motor vehicle occupants Am J Prev Med 2001,21suppl 4:23–30. [PubMed: 11691559]

[8]

National Highway Traffic Safety Administration, National Center for Statistics and Analysis Fatality Analysis Reporting System (FARS). 2001. Available at: www​.nhtsa.dot.gov/people/ncsa/fars.html. Accessed February 21, 2001.

[9]

Klein T. A method for estimating posterior BAC distributions for persons involved in fatal traffic accidents: final report. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1986. DOT HS 807 094.

[10]

Hingson R., Heeren T. and Winter M. Lowering state legal blood alcohol limits to 0.08%: the effect on fatal motor vehicle crashes. Am J Public Health. 1996,;86:1297–1299. [PMC free article: PMC1380594] [PubMed: 8806383]

[11]

Apsler R, Char AR, Harding WM, Klein TM. The effects of .08 BAC laws. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, National Center for Statistics and Analysis, 1999. DOT HS 808 892.

[12]

Foss RD, Stewart JR, Reinfurt DW. Evaluation of the effects of North Carolina's 0.08% BAC law. Chapel Hill, NC: University of North Carolina Highway Safety Research Center, 1998. DOT HS 808 893. Available at: www​.nhtsa.dot.gov/people/ncsa/nc08.html. Accessed September 6, 2001.

[13]

Hingson R., Heeren T. and Winter M. Effects of recent 0.08% legal blood alcohol limits on fatal crash involvement. Inj Prev. 2000,;6:109–114. [PMC free article: PMC1730627] [PubMed: 10875666]

[14]

Johnson D, Fell J. The impact of lowering the illegal BAC limit to .08 in five states. 39th Annual Proceedings, Association for the Advancement of Automotive Medicine, October 16-18, 1995, Chicago, Illinois.

[15]

Research and Evaluation Associates. The effects following the implementation of an 0.08 BAC limit and administrative per se law in California. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, National Center for Statistics and Analysis, 1991. DOT HS 807 777.

[16]

Rogers PN. The general deterrent impact of California's 0.08% blood alcohol concentration limit and administrative per se license suspension laws, Volume 1. Sacramento, CA: California Department of Motor Vehicles, Research and Development Section, 1995. No. CAL-DMV-RSS-95-158.

[17]

Scopatz, R.A., Methodological study of between-states comparisons, with particular application to .08% BAC law evaluation. Paper presented at the Transportation Research Board 77th annual meeting, January 11-15, 1998, Washington, DC.

[18]

Voas R.B., Tippetts A.S. and Fell J. The relationship of alcohol safety laws to drinking drivers in fatal crashes. Accid Anal Prev. 2000,;32:483–492. [PubMed: 10868751]

[19]

Department of Transportation and Related Agencies Appropriations Act, 2001. Public law no. 106-346. 2000.

[20]

Zador P.L., Krawchuk S.A. and Voas R.B. Alcohol-related relative risk of driver fatalities and driver involvement in fatal crashes in relation to driver age and gender: an update using 1996 data. J Stud Alcohol. 2000,;61:387–395. [PubMed: 10807209]

[21]

National Highway Traffic Safety Administration. On DWI laws in other countries. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 2000. DOT HS 809 037.

[22]

Hingson R., Heeren T. and Winter M. Lower legal blood alcohol limits for young drivers. Public Health Rep. 1994,;109:738–744. [PMC free article: PMC1403574] [PubMed: 7800781]

[23]

Blomberg RD. Lower BAC limits for youth: evaluation of the Maryland .02 law. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1992. DOT HS 807 860.

[24]

Haque M.O. and Cameron M. Effect of the Victorian Zero BAC legislation on serious casualty accidents: July 1984-December 1985. J Safety Res. 1989,;20:129–137.

[25]

Lacey JH, Jones RK, Wiliszowski CH. Zero tolerance laws for youth: four states' experience. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 2000. DOT HS 809 053.

[26]

Maisey GE. The effect of lowering the statutory alcohol limit for first year drivers from 0.08 to 0.02 gm/100ml (monograph). Perth, Western Australia: Western Australia Police Department, Research and Statistics Section, 1984. Research Report 84/2.

[27]

Smith D.I. Effect of low proscribed blood alcohol levels (BALs) on traffic accidents among newly-licensed drivers. Med Sci Law. 1986,;26:144–148. [PubMed: 3724396]

[28]

Streff FM, Hopp ML. Evaluation of Michigan's under age 21 zero-tolerance alcohol-impaired driving law. Ann Arbor, MI: University of Michigan Transportation Research Institute, 1997. UMTRI 97-50.

[29]

Voas RB, Lange JE, Tippetts AS. Enforcement of the zero tolerance law in California: a missed opportunity? 42nd Annual Proceedings: Association for the Advancement of Automotive Medicine, October 5-7, 1998. Charlottesville, VA. pp. 369-383.

[30]

Voas RB, Tippetts AS, Fell J. The United States limits drinking by youth under age 21: Does this reduce fatal crash involvements? 43rd Annual Proceedings Association for the Advancement of Automotive Medicine, September 20-21, 1999. Barcelona (Sitges), Spain. pp. 265-78.

[31]

Hingson R., Heeren T. and Morelock S. Effects of Maine's 1982 .02 law to reduce teenage driving after drinking. Alcohol Drugs Driving. 1989,;5:25–36.

[32]

Hingson R., Heeren T., Howland J. and Winter M. Reduced BAC limits for young people (impact on night fatal crashes). Alcohol Drugs Driving. 1991,;7:117–127.

[33]

Quinlan K.P., Brewer R.D., Sleet D.A. and Dellinger A.M. Child passenger deaths and injuries involving drinking drivers. JAMA. 2000,;283:2249–2252. [PubMed: 10807383]

[34]

Miller T.R., Lestina D.C. and Spicer R.S. Highway crash costs in the United States by driver age, blood alcohol level, victim age, and restraint use. Accid Anal Prev. 1998,;30:137–150. [PubMed: 9450118]

[35]

Task Force on Community Preventive Services, Carande-Kulis V.G., Maciosek M.V. and Briss P.A. et al. Methods for systematic reviews of economic evaluations for the Guide to Community Preventive Services Am J Prev Med 2000,18suppl 1:75–91. [PubMed: 10806980]

[36]

O'Malley P.M. and Wagenaar A.C. Effects of minimum drinking age laws on alcohol use, related behaviors and traffic crash involvement among American youth: 1976-1987. J Stud Alcohol. 1991,;52:478–491. [PubMed: 1943105]

[37]

Decker M.D., Graitcer P.L. and Schaffner W. Reduction in motor vehicle fatalities associated with an increase in the minimum drinking age. JAMA. 1988,;260:3604–3610. [PubMed: 3193591]

[38]

MacKinnon D.P. and Woodward J.A. The impact of raising the minimum drinking age on driver fatalities. Int J Addict. 1986,;21:1331–1338. [PubMed: 3804518]

[39]

Legge J. Jr Reforming highway safety in New York State: an evaluation of alternative policy interventions. Soc Sci Q. 1990,;71:373–382. [PubMed: 11613969]

[40]

Figlio D.N. Effect of drinking age laws and alcohol-related crashes: time-series evidence from Wisconsin. J Policy Anal Manage. 1995,;14:555–566.

[41]

Womble K. Impact of minimum drinking age laws on fatal crash involvements: an update of the NHTSA analysis. J Traffic Safety Educ. 1989,;37:4–5.

[42]

Wagenaar A.C. and Maybee R. Legal minimum drinking age in Texas: effects of an increase from 18 to 19. J Safety Res. 1986,;17:165–178.

[43]

Hoskin A.F., Yalung Mathews D. and Carraro B.A. Effect of raising the legal minimum drinking age on fatal crashes in 10 states. J Safety Res. 1986,;17:117–121.

[44]

Males M. Minimum purchase age for alcohol and young-driver fatal crashes: a long-term view. J Legal Stud. 1986,;15:181–211.

[45]

Hingson R., Scotch N. and Mangione T. et al. Impact of legislation raising the legal drinking age in Massachusetts from 18 to 20. Am J Public Health. 1983,;73:163–169. [PMC free article: PMC1650505] [PubMed: 6849474]

[46]

Williams A.F., Zador P.L., Harris S.S. and Karpf R.S. The effect of raising the legal minimum drinking age on involvement in fatal crashes. J Legal Stud. 1983,;12:169–179.

[47]

Smith R.A., Hingson R.W. and Morelock S. et al. Legislation raising the legal drinking age in Massachusetts from 18 to 20: effect on 16 and 17 year olds. J Stud Alcohol. 1984,;45:534–539. [PubMed: 6521479]

[48]

Wagenaar A.C. Preventing highway crashes by raising the legal minimum age for drinking: the Michigan experience 6 years later. J Safety Res. 1986,;17:101–109.

[49]

Vingilis E. and Smart R.G. Effects of raising the legal drinking age in Ontario. Br J Addict. 1981,;76:415–425. [PubMed: 6947816]

[50]

Durant R. and Legge J.S. Policy design, social regulation and theory building: lessons from the traffic safety policy arena. Political Res Q. 1993,;46:641–656.

[51]

Lillis R, Williams T, Williford W. The impact of the 19-year-old drinking age in New York.Advances in Substance Abuse 1987;Suppl 1, Control Issues in Alcohol Abuse Prevention: Strategies for States and Communities:133-46.

[52]

Wagenaar A.C. Raising the legal drinking age in Maine: impact on traffic accidents among young drivers. Int J Addict. 1983,;18:365–377. [PubMed: 6874158]

[53]

Williams A.F., Rich R.F. and Zador P.L. The legal minimum drinking age and fatal motor vehicle crashes. J Legal Stud. 1975,;4:219–239.

[54]

Naor E.M. and Nashold R.D. Teenage driver fatalities following reduction in the legal drinking age. J Safety Res. 1975,;7:74–79.

[55]

Ferreira J. and Sickerman A. The impact of Massachusetts reduced drinking age on auto accidents. Accid Anal Prev. 1976,;8:229–239.

[56]

Brown D.B. and Maghsoodloo S.A. A study of alcohol involvement in young driver accidents with the lowering of the legal age of drinking in Alabama. Accid Anal Prev. 1981,;13:319–322.

[57]

Zylman R. Fatal crashes among Michigan youth following reduction of legal drinking age. Q J Stud Alcohol. 1974,;35:283–286. [PubMed: 4827289]

[58]

Whitehead P.C., Craig J., Langford N., MacArthur C., Stanton B. and Ferrence R.G. Collision behavior of young drivers: impact of the change in the age of majority. J Stud Alcohol. 1975,;36:1208–1223. [PubMed: 240974]

[59]

Bako G., Mackenzie W.C. and Smith E.S.O. The effect of legislated lowering of the drinking age on fatal highway accidents among young drivers in Alberta, 1970-1972. Can J Public Health. 1976,;67:161–163. [PubMed: 1260633]

[60]

Smith D.I. and Burvill P.W. Effect on traffic safety of lowering the drinking age in three Australian states. J Drug Issues. 1986,;16:183–198.

[61]

Ruhm C.J. Alcohol policies and highway vehicle fatalities. J Health Econ. 1996,;15:435–454. [PubMed: 10164038]

[62]

Asch P. and Levy D.T. Does the minimum drinking age affect traffic fatalities? J Policy Anal Manage. 1987,;6:180–192.

[63]

Wilkinson J.T. Reducing drunken driving: Which policies are most effective? South Econ J. 1987,;54:322–334.

[64]

Legge J.S. Jr and Park J. Policies to reduce alcohol-impaired driving: evaluating elements of deterrence. Soc Sci Q. 1994,;75:594–606.

[65]

Saffer H. and Chaloupka F. Breath testing and highway fatality rates. Appl Econ. 1989,;21:901–912.

[66]

Saffer H. and Grossman M. Drinking age laws and highway mortality rates: cause and effect. Econ Inquiry. 1987,;25:403–417.

[67]

DuMouchel W., Williams A.F. and Zador P. Raising the alcohol purchase age: its effects on fatal motor vehicle crashes in twenty-six states. J Legal Stud. 1987,;16:249–266.

[68]

Cook P.J. and Tauchen G. The effect of minimum drinking age legislation on youthful auto fatalities. J Legal Stud. 1984,;13:169–190.

[69]

Robertson L.S. Blood alcohol in fatally injured drivers and the minimum legal drinking age. J Health Polit Policy Law. 1989,;14:817–825. [PubMed: 2614044]

[70]

Colon I. and Cutter H.S.G. The relationship of beer consumption and state alcohol and motor vehicle policies to fatal accidents. J Safety Res. 1983,;14:84–89.

[71]

Colon I. The alcohol beverage purchase age and single-vehicle highway fatalities. J Safety Res. 1984,;15:159–162.

[72]

Asch P. and Levy D.T. Young driver fatalities: the roles of drinking age and drinking experience. South Econ J. 1990,;57:512–520.

[73]

Chaloupka F.J., Saffer H. and Grossman M. Alcohol control policies and motor vehicle fatalities. J Legal Stud. 1993,;22:161–186.

[74]

Saffer H. and Grossman M. Beer taxes, the legal drinking age, and youth motor vehicle fatalities. J Legal Stud. 1987,;16:351–374.

[75]

Dee T.S. State alcohol policies, teen drinking and traffic fatalities. J Public Econ. 1999,;72:289–315.

[76]

Houston D.J., Richardson L.E. and Neeley G.W. Legislating traffic safety: a pooled time series analysis. Soc Sci Q. 1995,;76:328–345.

[77]

Houston D.J., Richardson L.E. and Neeley G.W. Mandatory seat belt laws in the states: a study of fatal and severe occupant injuries. Eval Rev. 1996,;20:146–159. [PubMed: 10182201]

[78]

Arnold R. Effect of raising the legal drinking age on driver involvement in fatal crashes: the experience of thirteen states. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1985. DOT HS 806 902.

[79]

Wagenaar A.C. Effects of an increase in the legal minimum drinking age. Public Health Policy. 1981,;2:206–224. [PubMed: 7298839]

[80]

Wagenaar A.C. Effects of the raised legal drinking age on motor vehicle accidents in Michigan. HSRI Res Rev. 1981,;11:1–8.

[81]

Skinner D, Hoxie P. Effects of minimum drinking age laws on highway fatalities. Cambridge, MA: National Highway Traffic Safety Administration, Transportation Systems Center, 1989. DOT-TSC-HS970-PM-89-1.

[82]

Ross H. Drunk driving: social policy for saving lives. (1992) New Haven: Yale University Press.

[83]

National Highway Traffic Safety Administration. The use of sobriety checkpoints for impaired driving enforcement. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1990. DOT HS 807 656.

[84]

Homel R., Carseldine D. and Kearns I. Drink-driving countermeasures in Australia. Alcohol Drugs Driving. 1988,;4:113–144.

[85]

Castle SP, Thompson JD, Spataro JA, et al. Early evaluation of a statewide sobriety checkpoint program. 39th annual proceedings, Association for the Advancement of Automotive Medicine, October 16-18, 1995, Chicago, IL. pp. 65-78.

[86]

Jones R, Joksch H, Lacey J, Wiliszowski C, Marchetti L. Site report: Wichita, Kansas field test of combined speed, alcohol, and safety belt enforcement strategies. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1995. DOT HS 808 244.

[87]

Lacey J.H., Stewart J.R., Marchetti L.M., Popkin C.K. and Murphy P.V. Enforcement and public information strategies for DWI (driving-while-intoxicated) general deterrence: ARREST DRUNK DRIVING—the Clearwater and Largo, Florida experience. (1986) Chapel Hill, NC: University of North Carolina Highway Safety Research Center.

[88]

Lacey JH, Marchetti LM, Stewart JR, et al. Enforcement and public information strategies for DWI general deterrence: the Indianapolis, Indiana experience. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1988. DOT HS 807 434.

[89]

Lacey JH, Jones RK. Assessment of changes in DWI enforcement/level: final report. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1991. DOT HS 807 690.

[90]

Lacey JH, Jones RK, Smith RG. Evaluation of checkpoint Tennessee: Tennessee's statewide sobriety checkpoint program. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Adminis-tration, 1999. DOT HS 808 841.

[91]

Levy D., Shea D. and Asch P. Traffic safety effects of sobriety checkpoints and other local DWI programs in New Jersey. Am J Public Health. 1989,;79:291–293. [PMC free article: PMC1349549] [PubMed: 2916713]

[92]

Levy D.T., Asch P. and Shea D. An assessment of county programs to reduce driving while intoxicated. Health Educ Res. 1990,;5:247–256.

[93]

Mercer G.W. The relationships among driving while impaired charges, police drinking-driving roadcheck activity, media coverage and alcohol-related casualty traffic accidents. Accid Anal Prev. 1985,;17:467–474. [PubMed: 4096805]

[94]

Mercer GW, Cooper PJ, Kristiansen LA. A cost/benefit analysis of a 5-month intensive alcohol-impaired driving road check campaign. 40th annual proceedings of the Association for the Advancement of Automotive Medicine, October 7-9, 1996, Vancouver, British Columbia. pp. 283-92.

[95]

Stuster JW, Blowers PA. Experimental evaluation of sobriety checkpoint programs. Washington, DC: U.S. Department of Transportation, National Highway Safety Traffic Administration, 1995. DOT HS 808 287.

[96]

Vingilis E. and Salutin L. A prevention programme for drinking and driving. Accid Anal Prev. 1980,;12:267–274.

[97]

Voas RB, Rhodenzer E, Lynn C. Evaluation of Charlottesville checkpoint operation: Final report, December 30, 1983 to December 31, 1984. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1985. US DOT Contract no. DTNH 22-83-C-05088.

[98]

Wells J.K., Preusser D.F. and Williams A.F. Enforcing alcohol-impaired driving and seat belt use laws, Binghamton, NY. J Safety Res. 1992,;23:63–71.

[99]

Voas R.B., Holder H.D. and Gruenewald P.J. The effect of drinking and driving interventions on alcohol-involved traffic crashes within a comprehensive community trial. Addiction. 1997,;92:S221–S236. [PubMed: 9231446]

[100]

Levy D. Methodologies for the evaluation of local traffic safety programs: with an application to New Jersey DWI programs. Eval Program Plann. 1988,;11:255–266.

[101]

Vingilis E., Salutin L. and Chan G. R.I.D.E. (Reduce Impaired Driving in Etobicoke): a driving-while-impaired countermeasure programme, one-year evaluation. (1979) Toronto: Addiction Research Foundation.

[102]

Armour M, Monk K, South D, Chomiak G. Evaluation of the 1983 Melbourne random breath testing campaign: interim report, casualty accident analysis. Melbourne, Australia: Victoria Road Traffic Authority, 1985. N8-85.

[103]

Armstrong B.K. and Howell C.M. Trends in injury and death in motor vehicle accidents in Australia in relation to the introduction of random breath testing. Aust Drug Alcohol Rev. 1988,;7:251–259.

[104]

Arthurson RM. Evaluation of random breath testing. Sydney: Traffic Authority of New South Wales, 1985. Research Note RN 10/85.

[105]

Cameron MH, Cavallo A, Sullivan G. Evaluation of the random breath testing initiative in Victoria, 1989-1991: multivariate time series approach. Melbourne, Australia: Monash University Accident Research Centre, 1992. Report 38.

[106]

Cameron M, Diamantopolou K, Mullan N, Dyte D, Gantzer S. Evaluation of the country random breath testing and publicity program in Victoria, 1993-1994. Melbourne: Monash University Accident Research Center, 1997. Report 126.

[107]

Dunbar J.A., Penttila A. and Pikkarainen J. Drinking and driving: success of random breath testing in Finland. Br Med J. 1987,;295:101–103. [PMC free article: PMC1246975] [PubMed: 3113620]

[108]

Hardes G., Gibberd R.W., Lam P., Callcott R., Dobson A.J. and Leeder S.R. Effects of random breath testing on hospital admissions of traffic-accident casualties in the Hunter Health Region. Med J Aust. 1985,;142:625–626. [PubMed: 4000035]

[109]

Hendrie D, Cooper L, Ryan G, Kirov C. Review of the random breath testing program in Western Australia in 1996/1997. Nedlands, WA 6907: Road Accident Prevention Research Unit, Department of Public Health, the University of Western Australia, 1998. Research Report RR67.

[110]

Henstridge J, Homel R, Mackay P. The long-term effects of random breath testing in four Australian states: a time series analysis. Canberra, Australia: Federal Office of Road Safety, 1997. No. CR 162.

[111]

McCaul K.A. and McLean A.J. Publicity, police resources and the effectiveness of random breath testing. Med J Aust. 1990,;152:284–286. [PubMed: 2314329]

[112]

McLean AJ, Clark MS, Dorsch MM, Holubowycz OT, McCaul KA. Random breath testing in South Australia: effects on drink-driving. Adelaide, South Australia: NHMRC Road Accident Research Unit, University of Adelaide, 1984. HS 038 357.

[113]

Paciullo G. Random breath testing in New South Wales. Med J Aust. 1983,;1:620–621. [PubMed: 6855684]

[114]

Ross H.L., McCleary R. and Epperlein T. Deterrence of drinking and driving in France: an evaluation of the law of July 12, 1978. Law Soc Rev. 1981,;16:345–374.

[115]

Homel R. The impact of random breath testing in New South Wales, December 1982 to February 1983. Med J Aust 1983;616-9. [PubMed: 6855683]

[116]

Homel R. Drink-driving law enforcement and the legal blood alcohol limit in New South Wales. Accid Anal Prev. 1994,;26:147–155. [PubMed: 8198683]

[117]

Michigan Department of State Police v Sitz, 496 U.S. 444, 110 L. Ed. 2d 412, 1990 U.S. LEXIS 3144, 110 S. Ct. 2481, 58 U.S.L.W. 4781 (1990).

[118]

Miller T.R., Galbraith M.S. and Lawrence B.A. Costs and benefits of a community sobriety checkpoint program. J Stud Alcohol. 1998,;59:462–468. [PubMed: 9647429]

[119]

Wesemann P. (1990) Costs and benefits of police enforcement in the Netherlands. In: Perrine M.W. (Eds.) Alcohol, drugs, & traffic safety: proceedings of the 11th International Conference on Alcohol, Drugs & Traffic Safety, Chicago, IL, October 24-27, 1989. Chicago, IL: National Safety Council.

[120]

O'Donnell M.A. Research on drinking locations of alcohol-impaired drivers: implications for prevention policies. J Public Health Policy. 1985,;6:510–525. [PubMed: 3912407]

[121]

Lang E. and Stockwell T.R. Drinking locations of drink-drivers: an analysis of accident and non-accident cases. Accid Anal Prev. 1991,;23:573–584. [PubMed: 1772559]

[122]

Alcohol Epidemiology Program, Alcohol policies in the United States: highlights from the 50 states. (2000) Minneapolis: University of Minnesota.

[123]

Toomey T.L., Kilian G.R., Gehan J.P., Perry C.L., Jones-Webb R. and Wagenaar A.C. Qualitative assessment of training programs for alcohol servers and establishment managers. Public Health Rep. 1998,;113:162–169. [PMC free article: PMC1308656] [PubMed: 9719818]

[124]

Ward W. and Lewis F.M. (Eds.) (1991) Advances in health education and promotion. (Vol. 2147483647) London, England: Jessica Kingsley.

[125]

McKnight A.J. and Streff F.M. The effect of enforcement upon service of alcohol to intoxicated patrons of bars and restaurants. Accid Anal Prev. 1994,;26:79–88. [PubMed: 8110359]

[126]

Holder H, Wagenaar A, Saltz RF, Mosher J, Janes K. Alcoholic beverage server liability and the reduction of alcohol-related problems: evaluation of dram shop laws (final report). Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1990. DOT HS 807 628.

[127]

Hauritz M., Homel R., McIlwain G., Burrows T. and Townsley M. Reducing violence in licensed venues through community safety action projects: the Queensland experience. Contemp Drug Problems. 1998,;25:511–551.

[128]

Single E. Server intervention: a new approach to the prevention of impaired driving. Health Educ Res. 1990,;5:237–245.

[129]

Saltz R.F. The role of bars and restaurants in preventing alcohol-impaired driving: an evaluation of server intervention. Eval Health Professions. 1987,;10:5–27.

[130]

Saltz R.F. Research needs and opportunities in server intervention programs. Health Educ Q. 1989,;16:429–438. [PubMed: 2793495]

[131]

Holder H.D. and Wagenaar A.C. Mandated server training and reduced alcohol-involved traffic crashes: a time series analysis of the Oregon experience. Accid Anal Prev. 1994,;26:89–97. [PubMed: 8110360]

[132]

Gliksman L., McKenzie D., Single E., Douglas R., Brunet S. and Moffatt K. The role of alcohol providers in prevention: an evaluation of a server intervention programme. Addiction. 1993,;88:1195–1203. [PubMed: 8241919]

[133]

McKnight A.J. Factors influencing the effectiveness of server-intervention education. J Stud Alcohol. 1991,;52:389–397. [PubMed: 1943093]

[134]

Howard-Pitney B., Johnson M.D., Altman D.G., Hopkins R. and Hammond N. Responsible alcohol service: a study of server, manager, and environmental impact. Am J Public Health. 1991,;81:197–199. [PMC free article: PMC1404951] [PubMed: 1990858]

[135]

Russ N.W. and Geller E.S. Training bar personnel to prevent drunken driving: a field evaluation. Am J Public Health. 1987,;77:952–954. [PMC free article: PMC1647259] [PubMed: 3605473]

[136]

Saltz R.F. and Stanghetta P. A community-wide responsible beverage service program in three communities: early findings Addiction 1997,92suppl 2:s237–s249. [PubMed: 9231447]

[137]

Lang E., Stockwell T., Rydon P. and Beel A. Can training bar staff in responsible serving practices reduce alcohol-related harm? Drug Alcohol Rev. 1998,;17:39–50. [PubMed: 16203467]

[138]

Geller E.S., Russ N.W. and Delphos W.A. Does server intervention training make a difference? Alcohol Health Res World. 1987,;11:64–69.

[139]

McKnight AJ. Development and field test of a responsible alcohol service program. Volume III: Final results. Washington, DC: U.S. Department of Transportation, National Highway Traffic Safety Administration, 1989. DOT HS 807 449.

[140]

Saltz R.F. Server intervention: Will it work? Alcohol Health Res World. 1986,;10:12–19.

[141]

Stockwell TR, Rydon P, Lang E, Beel AC. An evaluation of the "Freo Respects You" responsible alcohol service project. Perth, Western Australia: National Centre for Research into the Prevention of Drug Abuse, Division of Health Sciences, Curtin University of Technology, 1993. NDRI Technical Report No. T40.

[142]

Peters J. Beyond server training: an examination of future issues. Alcohol Health Res World. 1986,;10:24–27.

[143]

DeJong W. and Hingson R. Strategies to reduce driving under the influence of alcohol. Annu Rev Public Health. 1998,;19:359–378. [PubMed: 9611624]

[144]

Holder H.D., Saltz R.F., Grube J.W., Voas R.B., Gruenewald P.J. and Treno A.J. A community prevention trial to reduce alcohol-involved accidental injury and death: overview. Addiction. 1997,;92:S155–S171. [PubMed: 9231442]

[145]

Liu S., Siegel P.Z., Brewer R.D., Mokdad A.H., Sleet D.A. and Serdula M. Prevalence of alcohol-impaired driving. JAMA. 1997,;277:122–125. [PubMed: 8990336]

Appendix: Studies Measuring the Effectiveness of 0.08% Blood Alcohol Concentration (BAC) Laws

Footnotes

a

Established Market Economics as defined by the World Bank include Andorra, Australia, Austria, Belgium, Bermuda, Canada, Channel Islands, Denmark, Faeroe Islands, Finland, France, Germany, Gibraltar, Greece, Greenland, Holy See, Iceland, Ireland, Isle of Man, Italy, Japan, Liechtenstein, Luxembourg, Monaco, The Netherlands, New Zealand, Norway, Portugal, San Marino, Spain, St. Pierre and Miquelon, Sweden, Switzerland, the United Kingdom, and the United States.

b

A benefit-to-cost ratio is provided as a stand-alone piece of information and should not be used to rank interventions unless (1) there is a known budget constraint, (2) the interventions are mutually independent, or (3) interventions exhibit constant returns to scale.

Endnote: In 1993, the Council on Linkages was composed of representatives of the Association of State and Territorial Health Officials (ASTHO), the National Association of County Health Officials (NACCHO), American Public Health Association (APHA), Association of Schools of Public Health (ASPH), American College of Preventive Medicine (ACPM), Centers for Disease Control and Prevention (CDC), and the Health Resources Services Administration (HRSA). One of the authors (LN) was Chair of the Council (1993-1996) and has been a consultant to the Task Force since 1996.

aWe use the Community Guide's definition of "time series study," which includes any study that obtains multiple measurements before, during, or after an intervention, as well as those using traditional time series analysis. Multiple measurements are equated with a better accounting for trend and are thus given a "moderate" rating in study quality (compared with a "least" rating for before-after studies).

Published in: American Journal of Preventive Medicine, 21:4(Supplement 1):66-88.

The research described in this article was awarded the Department of Health and Human Services Secretary's Award for Distinguished Service.