DoctorFinder | Join/Renew | MyAMA | Site Map | Contact Us

Drivers impaired by alcohol

Note: This report presents information and AMA policy on this subject as of June 1997.

Full text

In response to Resolution 423 (I-95), the American Medical Association (AMA) Board of Trustees (BOT) directed the Council on Scientific Affairs (CSA) to re-examine the "serious public health problem of chemically impaired drivers" with particular emphasis on (1) legislation that makes it illegal to drive at any age with a blood alcohol level greater than 0.02%, (2) the collection of injury data from countries that stipulate driver alcohol levels of 0.02% or less, and (3) assistance to American physicians in educating the public on the hazards of chemically impaired driving. This report addresses these issues.

In 1968, the American Medical Association (AMA) published Alcohol and the Impaired Driver. A Manual on the Medicolegal Aspects of Chemical Tests for Intoxication,1 developed by the AMA Committee on Medicolegal Problems. This manual and review of the scientific literature concluded (pp. 58-59):

Analysis of the epidemiological studies indicates that a clear relationship exists between the consumption of alcoholic beverages and the frequency of involvement in vehicular accidents.. The frequency with which drinking drivers are involved in accidents in comparison with nondrinking drivers indicates a definite relationship between the consumption of alcoholic beverages and the deterioration of driving ability to the point that an undesirable event occurs.

All studies indicate a proportionality in the probability of being involved in an auto accident with the increase in blood-alcohol level. It is apparent that impairment in some drivers commences at a low level and 0.04 percent w/v (40 mg/100 ml) seems to be the threshold. Despite...limitations all of the results of the tests, both real and simulated, have led to the conclusion that driving skill deteriorates with a relatively low blood-alcohol level, certainly less than 0.05 percent w/v (50 mg/100 ml).

...In viewing all of this experimental data, and evaluating the epidemiological findings, there is no evidence presented that alcohol improves driving or that it does not decrease driving skills in all individuals. The levels at which these occur vary somewhat with drinking habits, driving skills and other personal attributes.1

These findings have been substantially confirmed since that report, but the issue of alcohol-impaired drivers continues to be a significant public health concern. In recent years, national morbidity and mortality statistics document an overall decline in alcohol-related motor vehicle fatalities.2-4 This trend suggests that scientific research, policy initiatives, and legislative efforts to keep alcohol-impaired drivers from driving have helped reduce alcohol-related motor vehicle deaths.

The extent and cost of alcohol-impaired driving in the United States

In 1995, alcohol-related motor vehicle crashes resulted in 17,274 deaths in the United States with intoxication rates in fatal crashes highest for persons aged 21 to 24. These deaths represent an average of one alcohol-related fatality every 30 minutes. In addition, more than 300,000 persons were injured in crashes where police reported the presence of alcohol. That represents one person injured approximately every 2 minutes. Over their lifetimes, about 2 in every 5 Americans will be involved in an alcohol-related crash.5

The economic costs of drinking and driving also are quite high. One researcher estimated that in 1990 alcohol-related motor vehicle crashes cost the nation $3.87 billion, exclusive of the economic costs of medical services, morbidity, and mortality.6 This amounted to 3.9% of total economic costs of alcohol abuse. Reviewing the motor vehicle crash data for 1994, the National Highway Traffic Safety Administration (NHTSA)7 estimated that alcohol-involved crashes caused $45 billion or 30% of all crash costs, including 17% of property damage-only crash costs, 29% of nonfatal injury crash costs and 47% of fatal injury crash costs. Of these crashes, 78% of the costs occurred in crashes where a driver or pedestrian was legally intoxicated at 0.10% blood alcohol concentration (BAC) or higher. Each crash fatality (due to alcohol use, speeding, or other cause) cost the equivalent of $144 in added taxes for each US household, resulted in an average lifetime cost of $830,000, and cost each critically injured survivor an average of $706,000 (of which 84% was for medical expenses and lost productivity). Medical costs accounted for 22% of non-fatal injury crash costs. The medical costs for all motor vehicle crashes were paid for as follows: 9% from public revenue (6% federal, 3% state and local), 55% from private insurers, and 29% by crash victims.7

Overall, the alcohol-attributable medical spending for driving under the influence (DUI) is estimated at $5.4 billion per year, with each victim bearing an average cost of $36,000. Every drink, on the average, generates $1.00 in public expenditures, including $.35 attributable to DUI. Insurers bear an average cost of $10,000 for each DUI crash. By contrast, the use of a proven prevention strategy such as a sobriety checkpoint (used to temporarily stop and identify intoxicated drivers while signaling to all drivers that they should not drink and drive) is only approximately $6,600 each time one is set up for several hours and may save numerous lives.8

American Medical Association Policy
[Note: The discussion below reflects AMA policy at the time this report was written—June 1997]

Review of AMA policy on alcohol-impaired driving reveals a multiple-decade commitment to lowered BAC limits. In addition, AMA policy has supported physician-based public health efforts to educate, warn, identify, and counsel individuals about the risks of drinking alcohol and driving.

Adult drivers: In 1985, the AMA (Policy H-30.986, AMA Policy Compendium) supported a 0.05 g/dL blood-alcohol level as the per se legal limit for driving. In 1986 and 1987, AMA Policies H-30.982 and H-30.980, respectively, reaffirmed the 0.05% BAC and urged the states and federal government to incorporate this limit in their drunk driving laws.

In 1990, Policy H- 30.964 broadened the AMA position by calling for "all efforts to remove drunken drivers from the nation's highways." Policies H-30.962, 30.963, and 30.967 urged the states to "retitle the crime of Driving Under the Influence (DUI) since one is under the influence of however much one has consumed regardless of outward signs of intoxication." In addition, the AMA called for development of reliable surveillance and testing mechanisms and rehabilitation programs to augment penalty-focused efforts. AMA policy calling for a 0.05% BAC was again reaffirmed in 1995.

Underage Drivers: The AMA position on underage drivers was first addressed in the early 1980s with Policies H-15.990 and 30.989, which endorsed state-wide legislation to raise the legal drinking age to 21. In 1990, Policy H-15.972 advocated revocation of licenses for underage drivers convicted of driving while under the influence of alcohol. In 1991, Policy H-30.959 endorsed "zero tolerance" legislation mandating penalties for underage drivers having a BAC greater than 0.02%.

Current state laws on alcohol impaired driving

All states have identified BAC levels or thresholds that indicate alcohol concentrations that are used to define the alcohol element of traffic offenses. A comparison of AMA policy with current legislation in the 50 states and the District of Columbia shows the following trends:9

0.05% BAC: Thirty-eight states and the District of Columbia hold to a 0.10% BAC threshold, although there have been active efforts to pass such legislation in almost every state.9 Thirteen states (including Hawaii) have lowered the BAC threshold to 0.08% (only 3 states since 1993 and none in 1996). At present, no states subscribe to the AMA policy of 0.05% BAC as the legal standard for intoxication for the general population of adult drivers. All states conform to federal requirements that commercial drivers convicted of driving with a BAC of 0.04% or more will have their license revoked for a specified time. Although some states have 10-year reinstatement clauses, most revoke commercial drivers' licenses for life following the second offense.

"Zero tolerance" for drivers under age 21: Many states endorse the AMA's "zero tolerance" policy for alcohol-impaired adolescent drivers in legislation that stipulates a BAC of no greater than 0.02% (although states vary in the specific low level BAC they identify to implement "zero tolerance"). Thirty-five states and the District of Columbia have legally established a standard of 0.02% or below for drivers under 21 years of age. The number of states incorporating this policy into law is growing rapidly. "Zero tolerance" laws became effective in 4 states in 1996, and will become effective in 3 more in 1997.9

Enforcement efforts: At present, state legislation imposing penalties for alcohol-impaired drivers of all ages includes: mandatory blood-alcohol testing in 33 states and the District of Columbia; portable breath-alcohol analyzer testing (PBT) in 30 states and the District of Columbia; and sobriety checkpoints in 30 states and the District of Columbia. Twenty states have made driving while intoxicated a felony offense. Forty-six stipulate a mandatory jail sentence for repeat offenders, and 16 require jail sentences for first-time offenders.9

BAC testing and results

BAC testing is an important law enforcement tool to regulate drivers who drink. Test devices known as breath-alcohol analyzers (made in stationary and portable models) are used by all 50 states and the District of Columbia. These devices measure breath alcohol levels that can then be used to estimate a driver's BAC. All states have established specific impairment levels based on BAC as presumptive evidence of intoxication and have legislation deeming that operation of a vehicle on state highways implies consent for breath-alcohol analyzer testing if a driver is arrested for DUI. (See Figure 1.)

Although the BAC also can be measured in serum after drawing a blood sample, the accuracy and reliability of breath-alcohol analyzer devices are generally accepted, even at alcohol concentrations as low as 0.01%. The states have regulations to calibrate and approve breath-alcohol analyzer units and to train and supervise breath test operators. Breath-alcohol analyzers provide an inexpensive, accurate, and swift measurement method that enables the United States to avoid many problems experienced by countries that require only blood alcohol tests (i.e., lengthy delays because a physician is required to draw a blood sample).10 Law enforcement officers often stop suspected drivers and administer standard field sobriety tests at the roadside (to assess balance, speech, and perceptual skills). Persons who demonstrate impairment are then taken into custody, and breath-alcohol analyzer testing is performed prior to charging them under state DUI laws.

Recent reviews of the scientific literature on BAC levels and driver skill and performance impairment have reaffirmed the findings of the AMA study cited above and are summarized in Figures 1, 2, and 3 [Editor's note: Figures 2 and 3 are not available online.]. A 1988 literature review for the NHTSA11 examined 177 experimental studies of the effects of alcohol on reaction time, tracking, concentrated and divided attention, information processing capabilities, visual function, perception, psychomotor performance, and driver performance. Of these studies, 158 reported impairment of one or more behavioral skills at one or more BAC level,19 showed no impairment at the particular BAC level studied, and the majority found impairment below 0.07% and 35 at 0.04% or less. The NHTSA review concluded that "Impairment in all areas was significant by BACs of 0.05% and first appeared in many areas by 0.02% to 0.03%. Thus scientific evidence suggests no lower limit can be placed on alcohol impairment of driving-related skills." The authors further stated that "the weight of existing empirical evidence is sufficient to scientifically justify the setting of legal BAC limits at 0.05% or lower," and that further studies could better define BAC levels at which different behavioral areas are impaired.

An analysis of the data of the second national roadside breath testing survey and the NHTSA Fatal Accident Reporting System (FARS) 12 found that the relative fatality risk for drivers in single-vehicle crashes was slightly elevated for drivers with BACs of 0.02% to 0.04% (1.4 times greater risk), and sharply increased for drivers with BACs between 0.05% to 0.09% BACs (11.1 times greater risk), with 48 times greater risk between 0.10% and 0.15%, and 385 times greater risk at BACs higher than 0.15%. These risks were considerably higher for those aged 16 to 20 compared to older populations.

Analyses of the scientific literature focusing on the epidemiology of drinking drivers and pilots, as well as on alcohol consumption 13,14 also indicated impairment at levels below 0.05%. They highlighted the need to consider a range of factors that affect the relationship of alcohol consumption and motor vehicle safety including: BAC level; age; gender; severity of alcohol problem; drinking patterns; demographic characteristics; driving-related attitudes; driving behaviors; involvement in deviant and criminal behavior; expectations about the effects of alcohol; and personality characteristics. In another more recent summary of the most relevant literature, 1984 to 1995, the National Institute on Alcohol Abuse and Alcoholism noted:

The many skills involved in driving are not all impaired at the same BACs. For example, a driver’s ability to divide attention between two or more sources of visual information can be impaired by BACs of 0.02 percent or lower. However, it is not until BACs of 0.05 percent or more are reached that impairment occurs consistently in eye movements, glare resistance, visual perception, reaction time, certain types of steering tasks, information processing, and other aspects of psychomotor performance.15

Relevant research from other countries 16-20

Since the 1970s, despite different measuring tools and standards, there has been a world-wide decline in alcohol-related traffic fatalities. Efforts to lower legal BAC limits in other countries continue. At present, BAC limits in western Europe and other industrialized nations range from 0.02% to 0.08%: 0.02% in Sweden; 0.04% in Portugal; 0.05% in Australia, Belgium, Finland, Greece, Iceland, Japan, the Netherlands and Norway; 0.07% in France; and 0.08% in Austria, Canada, Denmark, Germany, Great Britain, Ireland, Italy, Luxembourg, Mexico, and Spain.

It is difficult to compare alcohol-related crash rates in the United States to those in other countries. Most countries do not routinely test the BAC of drivers involved in accidents. Many have different attitudes towards alcohol use, intoxication, and enforcement of DUI laws. Moreover, with the exceptions of Australia and Canada, there is greater reliance on use of public transportation, bicycles, and walking. Studies in Australia that examined lower BAC limits (0.05%) combined with random breath test (RBT) enforcement methods showed substantial drops in alcohol-related driving fatalities. In January 1991, the Australian Capital Territory and 3 Australian states reduced their legal BAC from 0.08% to 0.05%. Retrospective analysis of roadside RBT data from 1990 through the first 6 months of 1991, following passage of the 0.05% limit, showed a 39% decline in the incidence of drivers with a BAC above 0.15% for the first half of 1991 compared to the same period in 1990. In addition, the percent of drivers with a BAC above 0.02% decreased 61% during the same period. Of 14 government initiatives to reduce Australian driving-related fatalities only the 0.05% limit and RBTs had statistically significant effects on the reduction of total crashes. The researchers felt that further lowering of the BAC limit would be only marginally effective. They proposed additional measures, such as targeting young and novice drivers, increasing public education, and employing greater use of RBTs, to further reduce the nation's alcohol-related morbidity and mortality rates.19

The Australian state of Victoria has combined a 0.05% limit (with "zero tolerance" for new drivers), RBTs, graduated licensing, and numerous other aggressive measures to reduce DUI traffic fatalities. This combination of efforts has maintained the reductions described above and resulted in one of the lowest fatality rates among western nations.20

The Scandinavian countries (Norway, Finland, and Sweden) have lower BAC limits than the United States and have achieved considerably lower average percentages of drivers who have been found to be drinking and driving at night (8.4% in the United States vs 1% in Scandinavia). These countries also have strong norms against drunken driving along with extensive enforcement via RBTs, strict alcohol control policies, and severe sanctions for intoxicated drivers.

Scandinavians who continue to drink and drive have been identified as problem drinkers with repeatedly high BAC levels. This group constitutes a high proportion of drivers arrested, convicted, and/or involved in fatal crashes. Unlike the rest of the Scandinavian population, they do not appear to be deterred by legal BAC limits or DUI laws. Furthermore, repeat offenders (convicted at least once) frequently have higher BACs than first-time offenders. Management of the repeat offender problem in Sweden and other European countries has increasingly focused on the need for alcohol treatment services to reduce recidivism.

Sweden's experience in lowering the BAC limit to 0.02%, combined with RBTs, is difficult to compare with "zero tolerance" efforts in the United States. Implementation of "zero tolerance" (i.e., <0.02%) in Sweden resulted in only slight declines in the numbers of drinking drivers at 0.02% to 0.05% levels. However, before Sweden adopted "zero tolerance" it had a comparatively low level of DUI arrest levels and strong cultural norms that viewed DUI offenses negatively.

RBTs (i.e., tests conducted without cause) are legal in Sweden and provide greater latitude for enforcement. However, RBTs are not legal in the United States because they do not meet the constitutional requirement that there be probable cause to selectively stop an individual to find evidence of the crime of driving while intoxicated. The use of RBTs in other nations has also occurred within areas of centralized enforcement structures. The multiplicity of law enforcement agencies and jurisdictions in the United States would make funding, coordination, education of law enforcement agencies, consistency (e.g., of techniques, publicity, fines, etc) difficult. Thus a consistently credible deterrent effect would be much more difficult to achieve in this country.21 An alternative method used in the United States is the setting up of a sobriety checkpoint. In this method, law enforcement officers stop drivers at specified locations to conduct roadside tests of inebriation. Since all drivers, or a predetermined sample, are stopped, arbitrary selection is prevented. The primary purpose is to create a "sentinel effect" to discourage DUIs, as well as to catch violators. Frequent, publicized checkpoints can have a deterrent effect. This method was upheld by the US Supreme Court in 1990 (although some states interpret their legislation as prohibiting the practice). Furthermore, there is evidence that the combined effects of multiple strategies available in the United States (i.e., "zero tolerance" and minimum drinking age for youth, checkpoints, comprehensive and coordinated community interventions, etc.), if consistently used, can have substantial impact in reducing alcohol-related traffic injury and deaths without the use of RBTs.22-24

European countries, unlike the United States, generally have not drafted legislation specifically covering drivers under the age of 21. Minimum legal drinking ages, if they exist, are not enforced. In the United States, maintenance and enforcement of minimum drinking ages, combined with the adoption and enforcement of "zero tolerance" laws for underage drivers, may lead to overall reductions in alcohol-related driver fatalities that are comparable to the combined use of lower adult BACs, enforcement via RBTs, and strict DUI measures in other nations.

While RBTs of the general population in the United States would violate constitutional prohibitions against random search and seizure, federal laws permit specific groups of drivers to be stopped and randomly tested without cause. The Omnibus Transportation Employee Testing Act of 1991, based on the rationale that commercial drivers may be held to more stringent standards, requires random and post-accident testing for alcohol without need for reasonable suspicion.

In all countries, enforcement of lower BAC limits has been difficult. Problem drinkers ignore current laws; enforcement measures across jurisdictions are inconsistent; intoxication and impairment, even at higher BACs, are not always recognized by law enforcement officials; and services to screen and treat DUI offenders are often not available:

Progress in reducing the incidence of drinking and driving in the United States

From 1982 through 1994, the percentage of crashes and deaths (including drivers, occupants, and pedestrians) involving BACs of 0.10% or greater declined almost every year. During these years, the proportion of alcohol-involved drivers in fatal crashes, by age group, declined in all age categories. The rates declined, for drivers aged 15 to 17, by 56% (from 31.5% to 13.9%); for ages 18 to 20 by 44% (from 48.2% to 27.1%); and for ages 25 and older by 30% (from 34.2% to 23.9%). During that time period, the estimated number of alcohol-related traffic fatalities also dropped from 1,556 to 752 (ages 15 to 17), from 3,824 to 1,590 (ages 18 to 20), from 4,593 to 2,351 (ages 21 to 24), and from 14,093 to 11,225 ( ages 25 and older).3 In approximately the same time period (1985 to 1995), intoxication rates decreased for drivers of all age groups involved in fatal crashes; the youngest and oldest drivers experienced the largest decreases. Among drivers aged 16 to 20, intoxication rates fell 47% (from 23.9% in 1985 to 12.7% in 1995). Even so, the alcohol-related crash rate among young adults remains high. In 1994, alcohol-related deaths accounted for 29% of crash-related deaths among persons aged 15 to 17; 44% for ages 18 to 20; and 55.6% for ages 21 to 24. High rates of binge-drinking (5 or more drinks on a single occasion), and thus high BACs within this age group, are believed to play a significant role in these fatalities.3,23-28 In 1994, 1.4 million drivers were arrested for DUI (alcohol or narcotics) at an arrest rate of 1 for every 127 licensed drivers in the United States.5 In the 1993 national Behavioral Risk Factor Surveillance System (BRFSS),29 2.5% of the adults surveyed self-reported a total estimated 123 million episodes of alcohol-impaired driving in 1993, which corresponds to 655 episodes per 1000 adults. The frequencies among adults were highest for men aged 21 to 34 (1,739 episodes per 1,000 adults) and men ages 18 to 20 (1,623 episodes per 1,000 adults), despite laws in all states prohibiting sales to persons under age 21. Current estimates of the likelihood that an alcohol-impaired driver will be arrested for DUI range from 1 in 250 to 1 in 2,000.29

Taken as a whole, these data indicate that despite considerable progress, the DUI problem remains a large one. While efforts targeting drivers under the age of 21 have been effective, there remain considerable problems among minors aged 18 to 20 and among young adults. Several other indicators also raise ongoing concern:

  1. Legislative efforts to reduce the BAC limit from 0.10% to 0.08% have been defeated in a number of states, despite positive indications that 0.08% BAC legislation is effective (see below). Yet, in 1995 32% of all traffic fatalities occurred in crashes in which at least one driver or non-occupant had a BAC of 0.10% or greater. More than two thirds of the 13,545 individuals who died were intoxicated; the other one third of deaths were passengers, unintoxicated drivers, or non-occupants (pedestrians, pedal-cyclists, etc).5
  2. In 1995, alcohol-related traffic deaths increased for the first time since 1986 (from 16,589 in 1994 to 17,274 in 1995). Although it is too early to identify this as a trend, this represents 41% of the total traffic fatalities in the United States for the year (an average of one alcohol-connected fatality every 30 minutes).3,5
  3. The decline of driving-related fatalities in the youngest age group (15 to 17-year-olds) began only in 1990, and the proportion of alcohol-involved drivers aged 18 to 20 was higher than the proportion of alcohol-involved drivers aged 25 and older in all years from 1982 to 1994.3 However, recent data suggest that alcohol use in older age groups continues to pose a significant public health challenge.5,29 The highest intoxication rates in fatal crashes in 1995 were recorded for drivers aged 21 to 24 years (27.8%) followed by ages 25 to 34 (26.8%) and 35 to 44 (22.8%). These three groups also have had the smallest percent reductions in alcohol-involved fatal crashes since 1985.5 It is notable that the successful passage of "zero tolerance" laws will only affect drivers under 21 years of age and not those in the high-risk category of drivers aged 21 to 24 and who would be covered by legislation for 0.08% or 0.05% BAC for adults.

The impact of lower BACs for adults in the United States19,25-28

Lower adult BAC limits (e.g., 0.08% for the general population and 0.04% for repeat offenders and commercial drivers) have led to a decrease in fatal crashes. A new BAC of 0.08% for adults (California-1990, Maine-1988), with a 0.04% limit for those with a previous DUI convictions in Maine, was associated with a 12% decrease in fatal crashes and a 19% decrease in fatal nighttime crashes. Comparison of the first 5 states having lowered BAC of 0.08% with 5 states retaining 0.10% BAC limits showed a 4% decline in fatal single vehicle crashes in the former states.26 The investigators note that while the reduction from 0.10% to 0.08% did not result in as large a proportional decline in fatal crashes as "zero tolerance" laws for drivers under 21, the 0.08% laws for adults have an impact on a much larger driver population and thus can potentially save a larger number of lives.

This country’s experience with "zero tolerance" laws for minors suggests the potential benefits of greater reductions in BAC limits for adult drivers. Analysis of the experience of the first 4 states that lowered their legal BAC limit for minors to 0.02% or less (Maine, New Mexico, North Carolina, Wisconsin) found a 34% decline in teenage fatal nighttime crashes, and a 7% decline in adult fatal nighttime crashes, compared to states that had not passed "zero tolerance" laws. The analysis of this data was incomplete and other factors, such as increased enforcement levels, may also have played a role.19 Data from the first 12 states to lower BAC limits for minors were more conclusive. These states experienced a 17% decline in fatal single-vehicle nighttime crashes in states with 0.02% BAC, and a 22% decline in states with 0.00% limits (vs a 4% and 2% increase in comparison with states that did not raise their limits). Those states with "zero tolerance" laws had a 20% greater decline, whereas those that lowered their rates to 0.04% or 0.06% had no significant change. The researchers looking at this data hypothesized that the more favorable results among the 0.00% BAC states were attributable to the clear public health message that 0.00% sends; i.e., no drinking at all is permissible.26,27

With the exception of some "hangover" effects, no one is impaired at 0.00% BAC. (Figures 1, 2, 3.) As was discussed above, both laboratory and road studies indicate that all individuals exhibit major impairments at 0.08% while many may be impaired at 0.04% or less. Given this and the greater risks presented by youthful drinking and driving, the United States has arrived at a national consensus favoring passage of "zero tolerance" laws for youth. However, the costs and practical difficulties of detection and enforcement are significantly increased when the BAC limit is reduced to 0.05% or lower. At these levels, police cannot see visible signs of impairment. Unless drivers are stopped for other reasons and found to have either alcohol on the breath or an open container of alcohol in the vehicle, there is no "probable cause" to compel individuals to submit to sobriety or BAC testing. Although testing equipment can be calibrated to measure levels from 0.00% to 0.08%, recalibrations and the use of more sensitive breath-alcohol analyzers may increase costs. While many states intermittently use sobriety checkpoints as effective deterrents, extensive use would require substantial increases in fiscal and human resources including more enforcement personnel and hours and expanded public awareness, information, and education efforts.

Educating the public: Public and law enforcement attitudes

The public has a general understanding of the impact of alcohol on driving and its relationship to BAC limits. How these apply to specific situations and individuals is less clear to the public. A 1991 nationally representative telephone survey of 4,000 persons, aged 16 and older, conducted by the NHTSA,19 identified significant gaps in the American public's understanding of alcohol-related driving risk. Of the 2,100 respondents who indicated they drink alcohol and drive, 51% reported drinking at a location to which they had driven in the past year, and 75% of these indicated that they had planned in advance to drink at those places. Only 7% of drinker-drivers gauged their driving capacity by the amount they drank, and only 2% mentioned the BAC limit as the specific reason for the limit they set (although 87% said they felt they were at or within the legal limit).

One could assume from such results that BAC was not an important factor in the public mind. However, responses to questions regarding BAC indicate at least a general "environmental effect" of BAC policies; i.e., BAC policies create a framework within which individual decisions are made. For example, 61% of drinker/driver respondents said there should be a legal limit to the number of drinks before driving (31% said there should not be a limit), 84% had heard of BAC, and 87% knew there was a state limit (although almost half could not correctly identify what it was). The respondents also displayed a general understanding of the relationship between alcohol levels and driving impairment. Of those who drank and drove during the last year, 41% said that they set a limit on the number of drinks they would consume and 87% felt their threshold for safe driving was at a BAC level lower than the legal limit. Only one third thought they would be safe drivers at BACs above 0.04%.19 Another national public opinion survey conducted by NHTSA in 1991 found high levels of support for 0.08%, 0.04%, and "zero tolerance" levels for adults. Fifty-three percent of respondents (aged 16 to 64) strongly agreed that "people should not be allowed to drive if they had been drinking any alcohol at all." An additional 24% somewhat agreed, while only 10% strongly disagreed.19

These surveys demonstrated that the general population and drinker/drivers believe drinking and driving is unacceptable. BAC is a known and accepted measure of what is safe, and the public thinks that alcohol use should not be combined with driving. At the same time, few individuals identify themselves as being part of the drinking and driving problem, and most consider "having a drink or two" and driving socially permissible and safe. Most also think of themselves as staying within the legal BAC limits, although they do not know how to apply specific BAC levels to their own situations.19

Reviewing both NHTSA studies and a similar one in California,19 NHTSA concluded that:

A majority of the public perceives alcohol concentrations below current per se limits as being dangerous. Most agree that people should not be allowed to drive if they have been drinking any alcohol at all. If lower limits were the law, they would assign penalties similar to what they think the penalties should be for higher levels.

These studies combined with those on the effects on DUI levels of changing BAC levels also suggest that, rather than personally calculating and setting their own BAC level in response to the precise level set by law, people change their drinking-driving behaviors when they perceive that the BAC limit (whatever it has been) has gone down and will be strictly enforced. Thus BAC limits provide a guideline and background against which individuals formulate their own drinking and driving decisions and behaviors.

At two workshops conducted by NHTSA in 1991, and attended by leaders from law enforcement, courts, motor vehicle departments, treatment services, and other concerned individuals, the group consensus was that enforcement of BAC levels below 0.08% would be difficult without increased training and some new legislation. The workshop members endorsed a 2-tiered enforcement system, with one set of administrative or civil (noncriminal) penalties covering all BACs up to 0.079%, and continuing application of existing penalties for individuals with 0.08% or 0.10% BACs and above. Although it was recognized that this system might be less practical, workshop participants feared public and legislative rejection of a single lower BAC limit. Furthermore, they anticipated that court and enforcement resources might be diverted to less serious cases (although administrative decisions rather than court decisions for lower level BACs could offset this). They agreed that no one should drink and drive and that efforts to enact and enforce a lower BAC could have positive effects due to widespread news media coverage and increased awareness of the effects of low levels of alcohol on drivers.

In its 1992 Report to Congress19 NHTSA reviewed the scientific literature, existing legislation, and data in the United States and other countries on BAC limits and their effects. This report documents a continued consensus in the literature and among practitioners on the following important issues:

There is no safe alcohol level for drivers. Alcohol is a major causal factor in traffic crashes and is involved in almost one half of all fatal crashes. Crash risks (especially severe crashes) consistently increase above the 0.05% BAC and escalate even more rapidly above 0.08%. Thus the more alcohol in a driver’s bloodstream, the greater the crash risk.

Performance of driving-related tasks decreases as alcohol consumption increases. As BAC increases, more functions are degraded and more people show the effects. Even so, less than 10% of individuals who drink and drive specifically consider the capacity to drive as a gauge of the amount they drink.

The current role of BAC limits in mediating drivers' behavior is complex because individuals determine their drinking limits by considering a variety of factors. A state's specific BAC limit may only be a minor consideration in specific situations. Of those who drive to a location to drink and set a limit before drinking, only 2% mention the BAC limit as the reason for the limit they set

Nevertheless, drivers do set limits when BAC limits are known to exist and where enforcement is publicized. There is no practical tool for consumers to measure their BAC, and because many factors influence an individual's BAC (age, sex, weight, fatigue level, food, medications), individuals cannot estimate their BAC level and its effects. However, many drivers do use less precise measures, including perceived physical impairment such as vision, speech, or motor skills or feeling certain ways such as numb, tipsy, ill, off balance, etc, to gauge their level of intoxication.

Lower BACs (i.e., under 0.10%) and publication of new enforcement efforts are likely to reduce fatalities, but only if new methods and procedures are adopted to effectively implement them.

Discussion

Scientific data regarding the effects on driving at all levels of alcohol consumption are not matched by the public’s perceptions of their own safety risks when they drink and drive or by policy maker responses. The public health and medical costs resulting from alcohol-impaired driving underscore the need for intensified efforts to bring legislation and public awareness in line with current scientific information.

The scientific literature on alcohol and driving shows some impairment at all BACs. There is no threshold, or lower level, at which impairment starts and below which no impairment is found. The highest levels of fatal crashes occur among drivers under age 20, and among adult drivers with a BAC of 0.08% and above.

However, studies have demonstrated driver impairment at blood-alcohol levels much lower than the 0.10% and 0.08% legal limits.28 As BAC increases, so do the degree of impairment on a given task, the number of tasks and functions that are impaired, and the risk of crashes.

All states and the District of Columbia continue to have BAC limits between 0.10% and 0.08% for adult drivers. Despite efforts by some states to lower the BAC below 0.08%, none have adopted the AMA's recommendation for a 0.05% limit. National consensus resulted in standards that prohibit commercial drivers from having a BAC greater than 0.04%, but this limit does not currently extend to expectations for other adult drivers. Attempts at further BAC reductions for adults have met resistance from the beverage alcohol industry and many groups that benefit from the sale of alcohol.30,31 This resistance is strengthened by lack of full public understanding of driver impairment at even lower (0.04% and 0.05%) BACs.

To lower the legal BAC in states would require substantial changes in state legislation and public perceptions. First, if states were to adopt BAC limits of 0.05% or lower for adults and 0.00% to 0.02% for underage drivers (as the AMA recommends), they would have to revise the statutory definition of driving while intoxicated or adopt a two-tiered policy. As AMA Policies H-30.962, 30.963, and 30.967 indicate, if one is to be identified as legally driving under the influence of alcohol "regardless of outward signs of intoxication," a new definition of DUI must be drafted and new definitions of presumptive and per se BAC would be needed. Second, law enforcement officers would need new criteria for probable cause to stop drivers, assess sobriety, make arrests, administer breath-alcohol analyzer tests, and press charges. Third, those who prosecute DUI charges would require new standards of evidence to seek convictions. Fourth, additional legislation would be needed to establish penalties for conviction of both first-time and repeat offenders, and to determine new regulations for administrative license suspension and revocation. Finally, preventive measures, such as interlock devices that block intoxicated drivers from operating a vehicle or pharmaceutical agents that could counteract intoxication and its effects, are unavailable at this time. Thus reliance would have to be placed on the development of practical ways for consumers to measure their BAC and for educational and news media activities to be developed that persuade them of the hazards of drinking and driving even at low BACs.

The scientific literature indicates several additional considerations regarding the adoption of a 0.02% BAC for adults:

  • There have been no successful efforts in the United States to legislate a BAC limit below 0.08% for drivers aged 21 or older, and efforts in many states to achieve 0.08% levels have stalled.
  • The predominant public health message in the United States regarding drinking and driving is "If you drink, don’ t drive. If you drive, don’t drink." Mothers Against Drunk Driving promotes the message " Impairment Begins With the First Drink." These messages indicate that there is no safe level of drinking when driving, and the NHTSA survey indicates public support for such a concept. However, the extensive use of the automobile for work, leisure, and general transportation, combined with the legal use and availability of alcohol and individual variations in metabolizing alcohol, make adherence to a "zero tolerance" law for all drivers in the United States very difficult. Given the high percentage of drivers in this and other countries who drive with some level of alcohol, a goal of "zero tolerance" for all drivers is unrealistic.
  • Opponents of lower BAC limits use negative associations with Prohibition to thwart 0.08% BAC laws. They argue that such laws would mean that people who drive could not use alcohol for social purposes, since most individuals who drink socially use their cars to do so (e.g., to go to and from bars, parties, sports events, etc.). A "zero tolerance" policy for Amer ican adults, if adhered to, would result in decreased alcohol sales and consumption and hence would be strongly opposed by the industries that profit from alcohol sales.

Discussions among highway safety experts suggest there is no consensus as to whether "zero tolerance" would significantly reduce motor vehicle-related injuries and deaths. Some experts express concern that lowering the BAC limit to zero, if not accompanied by measures that allow for effective enforcement (e.g., through RBTs), might undermine public respect for all drink-drive laws, without achieving gains commensurate with the additional enforcement costs and public inconvenience. Without random testing or use of other legally permissible indicators, the disparity between detection and infractions would be even greater.

On the other hand, some highway safety experts cite the benefits of using a legal BAC of 0.00% or 0.02% as a goal. Any extensive efforts to lower the limit would receive considerable news media and public attention. Difficulties in obtaining grounds for reasonable suspicion or probable cause might make officials hesitant to prosecute individuals or enforce such laws. Yet such laws could also make successful prosecution of drivers with a higher BAC limit (e.g., 0.10% or 0.08%) easier and more acceptable. Furthermore, the establishment of "zero tolerance" for all drivers, as with enforcement of "zero tolerance" laws for youth, might simplify enforcement by eliminating many of the complexities that result from having the different BAC limits now in place.

The BAC limit is only one factor in what must be a comprehensive effort to reduce alcohol-impaired driving. Changes in legislation governing legal BAC limits must provide for funding and policies that include improved and highly visible enforcement; strategies to identify impaired drivers; public education; alternative means of transportation; screening, referral, and treatment services for repeat offenders; and, ultimately, measures to reduce the abuse of alcohol in our society.

Conclusions

The research and field literature indicate that there is some driving-related impairment for many people at any level of BAC although substantial and consistent impairment begins at 0.04% to 0.05% BAC. As BAC levels increase, so does individual impairment and the likelihood of impairment across the entire population. The risk of fatal crashes greatly increases at 0.04% to 0.05% BAC with greatly increased risk at 0.10% BAC.12,32,33

The effects of alcohol are mediated by a number of additional factors such as driver experience, age, gender, tolerance to alcohol and the complexities of any given driving situation. So too, effective DUI management is possible only through the implementation of an array of policy, media, education, and enforcement strategies. However, based on the data from field and experimental studies in this and other countries, it is clear that enforced policies that reduce the existing legal BAC limits in the direction of 0.00% BAC have a positive effect on both youth and adult populations. The population that appears to be least affected by these policies are those drivers who have alcohol problems (problem drinkers and alcoholics). This population is disproportionately found among repeat DUI offenders and among those with higher BAC levels in reported DUI incidents. Increased screening, referral, and treatment for all DUI offenders, accompanied by stronger enforcement policies and lower BAC limits for repeat offenders, are likely to reduce the DUI burden from this population.

There is broad based public understanding of the concept of BAC and its relationship to driving behaviors. Surveys show that most members of the public believe that alcohol use combined with driving is risky and undesirable. Nevertheless, the public and policy makers must be further educated about the negative effects that even low BAC levels have on driving before reduced blood alcohol limits will be widely accepted. In addition, adults (aged 20 to 34), especially males 21 to 24 years of age, must be the focus of new research, policy development, and existing prevention strategies.

In summary, broad scale, multi-faceted environmental strategies to reduce driving under the influence of alcohol have been very effective. BAC limits are an important component of these strategies from both law enforcement and public education perspectives. Technological development may give rise to further improvements (e.g., by making it physically impossible for intoxicated drivers to use their vehicles) but enforced lower BAC limits for adults and youth warrant increased support and attention.

Recommendations

The following statements, recommended by the Council on Scientific Affairs, were adopted by the AMA House of Delegates as AMA policy and directives at the 1997 AMA Annual Meeting:

  1. The AMA acknowledges that all alcohol consumption, even at low levels, has a negative impact on driver skills, perceptions, abilities, and performance and poses significant health and safety risks. The AMA will be involved in efforts to educate physicians, the public, and policy makers about this issue and urges national, state, and local medical associations and societies, together with public health, transportation safety, insurance industry, and alcohol beverage industry professionals to renew and strengthen their commitment to preventing alcohol-impaired driving.
  2. The AMA encourages physicians to participate in educating the public about the hazards of chemically impaired driving.
  3. The AMA urges public education messages that now use the phrase "drunk driving," or make reference to the amount one might drink without fear of arrest, be replaced with messages that indicate that "all alcohol use, even at low levels, impairs driving performance and poses significant health and safety risks. "
  4. The AMA reaffirms that alcohol is a drug of addiction (Policy H-30.958) and that alcoholism is a disease (Policy H-95.983) that is a disabling and handicapping condition (Policy H-30.995) and treatable. The AMA also reaffirms and endorses alcoholism's classification under both the psychiatric and medical sections of the International Classification of Diseases (Policy H-30.997).
  5. Thee AMA urges all states to pass legislation mandating all drivers convicted of first and multiple DUI offenses be screened for alcoholism and provided with referral and treatment when indicated.
  6. The AMA further recommends the following measures be taken to reduce repeat DUI offenses: (a) Aggressive measures be applied to first-time DUI offenders (e.g., license suspension and administrative license revocation). (b) Stronger penalties be leveled against repeat offenders, including second-time offenders. (c) Such legal sanctions must be linked, for all offenders, to substance abuse assessment and treatment services, to prevent future deaths in alcohol-related crashes and multiple DUI offenses. (d) The AMA calls upon the states to coordinate law enforcement, court system, and motor vehicle departments to implement forceful and swift penalties for second-time DUI convictions to send the message that those who drink and drive might receive a second chance but not a third.
  7. The AMA encourages the National Highway Traffic Safety Administration to investigate the feasibility of technologies that would prevent an automobile from being started or driven by an individual with an excessive blood alcohol level.

Reports by topic

 References

  1. American Medical Association, Committee on Medicolegal Problems. Alcohol and the Impaired Driver. A Manual on the Medicolegal Aspects of Chemical Tests for Intoxication. Chicago: American Medical Association; 1968.
  2. Centers for Disease Control and Prevention. Alcohol-related traffic fatalities among youth and young adults--United States, 1982-1989. MMWR. 1991; 40:178-9,185-187.
  3. Centers for Disease Control and Prevention. Reduction in alcohol-related traffic fatalities--United States, 1990-1992. MMWR. 1993;42:905-909.
  4. Centers for Disease Control and Prevention. Update: alcohol-related traffic crashes and fatalities among youth and young adults--United States, 1982-1994. MMWR. 1995;44:869-874.
  5. National Highway Traffic Safety Administration (NHTSA). Alcohol. Traffic Safety Facts. 1995.
  6. Rice DP. The economic cost of alcohol abuse and alcohol dependence: 1990. Alcohol Health Res World. 1993;17:10-11.
  7. Blincoe LJB. The Economic Cost of Motor Vehicle Crashes, 1994. NHTSA Technical Report. Washington, DC: National Highway Traffic Safety Administration; 1994.
  8. Miller T; National Public Service Research, Institute, Landover, MD. "Beer Bottle Blues: U.S. Heavy Drinking Costs and Prevention Savings." Paper presented at Alcohol Policy X, May 4-8, 1996, Toronto, Canada.
  9. Mothers Against Drunk Driving (MADD). Rating the States. A Report Card on the Nation’s Attention to the Problem of Alcohol- and Other Drug-Impaired Driving. Irving, TX: MADD National Office; 1996.
  10. Voas RB, Lacey JH. Issues in the enforcement of impaired driving laws in the United States. In: Surgeon General's Workshop on Drunk Driving. Washington, DC: U.S. Department of Health and Health and Services; Dec 14-16, 1988:136-156.
  11. U.S. Department of Transportation (USDOT). Effects of Low Doses of Alcohol on Driving-Related Skills: A Review of the Evidence (DOT HS 807 280) Washington, DC: USDOT. National Highway Traffic Safety Administration; July 1988.
  12. Zador PL. Alcohol-related relative risk of fatal driver injuries in relation to driver and age and sex. J Stud Alcohol. 1991;52:302-310.
  13. Perrine MWB. Who are the drinking drivers? Alcohol Health Res World. 1990;14:26-35.
  14. US Department of Health and Human Services. Eighth Special Report to the U.S. Congress on Alcohol and Health. Washington, DC:USDHHS. National Institute on Alcohol Abuse and Alcoholism; 1993.
  15. National Institute on Alcohol Abuse and Alcoholism. Alcohol Alert, No.25- Alcohol-Related Impairment. PH 351; July 1994.
  16. Office of the Surgeon General. Surgeon General’s Workshop on Drunk Driving - Background Papers. Rockville, MD: USDHHS, Public Health Service; 1988.
  17. Wilson RJ. Drinking and driving: in search of solutions to an international problem. Alcohol Health Res World. 1993;17:212-220.
  18. European Transport Safety Council. Reducing Traffic Injuries Resulting From Alcohol Impairment. Brussels: European Transport Safety Council; January 1995:1-29.
  19. United States Department of Transportation (USDOT). Driving Under the Influence: A Report to Congress on Alcohol Limits. National Highway Traffic Safety Administration; October 1992; D1-3.
  20. VicNet, Ministry for Roads and Ports. Safety First. Victoria’s Road Safety Strategy 1995-2000. Victoria, Australia: Department of Infrastructure; 1995.
  21. Homel RJ. Random breath testing the Australian way: a model for the United States? Alcohol Health Res World. 1990;14:70-75.
  22. Blomberg RD. Lower BAC Limits for Youth: Evaluation of the Maryland .02 Law. (Final Report, DOT HS 807 860) Washington, DC: US DOT. National Highway Traffic Safety Administration; March 1992.
  23. Centers for Disease Control and Prevention. Risky driving behavior among teenagers--Gwinnet County, Georgia, 1993. MMWR. 1994:43:405-409.
  24. Hingson R, McGovern T, Howland J, Heeren T, Winter M, Zakocs R. Reducing alcohol impaired driving in Massachusetts: the saving lives program. Am J Public Health. 1996; 86:791-797.
  25. Hingson R. Prevention of alcohol-impaired driving. Alcohol Health Res World. 1993;17:28-34.
  26. Hingson R, Berson J, Dowley K. Review of research on interventions to reduce college student drinking and related health and social problems. Report to the Robert Wood Johnson Foundation. January 1995.
  27. Hingson R, Heeren T, Winter M. Lower legal blood alcohol limits in young drivers. Public Health Rep. 1994; 109:738-744.
  28. National Highway Traffic Safety Administration, US Department of Transportation. Effects of Low Doses of Alcohol on Driving-related Skills: A Review of the Evidence. DOT HS 897 280; Literature Review; July 1988.
  29. Liu S, Siegel PZ, Brewer RD, et al. Prevalence of alcohol-impaired driving. Results from a national self-reported survey of health behaviors. JAMA. 1997;277:122-125.
  30. Berman R, General Counsel, American Beverage Institute. MADD should focus on real drunks. Chicago Tribune. February 4, 1997.
  31. Stewart K. .08 Blood Alcohol Content Laws: Facts, Myths and Fictions. Rockville, MD; Center for Substance Abuse Prevention; 1993.
  32. American Medical Association, Council on Scientific Affairs. Alcohol and the driver. JAMA. 1986;255:522-527.
  33. Perrine MW, Peck RC, Fell JC. Epidemiologic perspective on drunk driving. In Surgeon General’s Workshop on Drunk Driving. Washington, DC: U.S. Department of Health and Social Services; Dec 14-16, 1988:35-76.

Figure 1. BAC, legal limits and driver impairment

Legal limits for blood alcohol content (BAC)

Drivers covered

Driving aspects impaired by alcohol and collision risks*

0.00-0.05%

("zero tolerance")

U.S.: Under age 21 in 35 states, District of Columbia

Other: all drivers inSweden (0.02%)

Visual functions (voluntary eye movement, rapid tracking of moving target)

Ability to divide attention between two or more sources of visual information (at 0.02%)

Divided attention (tendency to focus more on steering, less on safety)

Steering (possibly significant at 0.035%)

Driver performance (e.g., errors in planning, performance, procedures and vigilance)

Increased risk of collision compared to non-drinkers (greater for 16- to 20- year old drivers than older drivers)

1.4 times increased risk of single-vehicle fatal crash (higher for men than women)

0.04%

U.S.: Commercial drivers in all states

Other: all drivers in Portugal

Attentional field (narrowed)

Simple reaction/response time to unexpected, emergency, even simple situations

Problem solving

0.05%

U.S.: AMA policy (all adult drivers; not enacted in any states)

Other: all drivers in Australia, Japan, 6 west European nations

Consistent effects on eye movements, glare resistance, visual perception, reaction.

Concentrated attention

Information processing

Task tracking (especially multiple)

Skilled motor performance and coordination

Standing steadiness

Twice the normal risk of collision (may be higher for women than men)

0.08%

U.S.: Ages 21 and older in 13 states

Other: all drivers in Canada and 8 west European nations [France is 0.07%]

Information processing

Perception

Concentrated attention and vigilance (over a long time)

Accuracy (steering, braking, speed control, lane tracking, gear changing)

Judgment (speed, distance)

0.10%

U.S.: Ages 21 and older in 37 states

Consistent impairment of all the above

Short-term memory

Six times the normal risk of collision

0.10-0.14% BAC: 48 times higher risk for single-vehicle fatal crash

0.15% BAC or greater: 380 times higher risk for single-vehicle fatal crash

National Highway Traffic Safety Administration, US Department of Transportation. Effects of Low Doses of Alcohol on Driving-related Skills: A review of the evidence. DO T HS 897 280; Literature Review; July 1988.

Romelsjo A. Alcohol consumption and unintentional injury, suicide, violence, work performance, and inter-generational effects. In: Holder H, Edwards G, eds. Alcohol and Public Policy: Evidence and Issues. New York: Oxford University Press; 1995:114-142.

Resolution 423 (I-95)

Resolution 423, introduced at the 1995 Interim Meeting by the Hawaii Delegation and referred to the Board of Trustees, resolved:

That the American Medical Association campaign against driving under the influence of alcohol and other drugs with the enthusiasm shown for the campaign against violence; and

That the AMA gather injury data from countries that have blood alcohol levels for drivers of 0.02 percent or less; and

That the AMA propose national legislation that makes it per se illegal to drive at any age with a blood alcohol level of greater than 0.02 percent; and

That the AMA encourage physicians across America to educate the public on the hazards of chemically impaired driving and set personal examples in this regard.

Reports by topic

Last updated: Sep 24, 2007
Content provided by: CSAPH