Skip over navigation

Journal Issue: Unintentional Injuries in Childhood Volume 10 Number 1 Spring/Summer 2000

Community-Based Injury Prevention Interventions
Terry P. Klassen J. Morag MacKay David Moher Annie Walker Alison L. Jones

The Effectiveness of Community-Based Approaches

Overall, the impact of community-based interventions on child and adolescent safety practices and injuries is mixed. Relatively strong evidence suggests that such interventions can increase bicycle helmet and motor vehicle restraint use among children. However, the success of these interventions at improving child pedestrian safety, influencing adolescent alcohol use and vehicle safety, and reducing rates of a broad array of childhood injuries within communities is less evident.

Elements of Successful Community-Based Approaches

Successful community-based programs share a number of common elements. The use of multiple strategies consistent with an underlying theory of behavior change is critical to success (see Box 1).34–36 Interventions are more effective when they are integrated into the community and when approaches are tailored to address unique community characteristics such as ethnicity or socioeconomic status. Effective programs actively involve community stakeholders in the program-development process or hold public consultations to determine a program's process and goals. As a program is spread across a community, its impact may be increased by peer pressure and modeling by other community members.

In addition to successful programmatic elements, the design of community-based trials and the evaluation methodology used also influence program effectiveness. The most informative and reliable results come from RCTs (see the article by DiGuiseppi and Roberts in this journal issue). When randomization of groups to intervention and control conditions is not possible, comparison communities should be carefully selected that are similar to the intervention community in characteristics—such as socioeconomic status, age, race/ethnicity, and baseline use of safety behaviors—that may be associated with the effectiveness of the program.

The outcomes that were measured also influence the interpretation of a program's effectiveness; therefore, injury rates should be used whenever possible. However, measuring safety behaviors may be a suitable alternative when evidence shows that adoption of the behavior (that is, bicycle helmet use) is strongly associated with a decrease in injury rates (that is, head injury). Less rigorous outcomes, such as changes in knowledge or attitudes, should be avoided since the correlation between knowledge, attitudes, and behaviors can be weak.37 Finally, regardless of the outcomes selected, assessment of program effectiveness requires careful and thorough statistical analyses to control for spurious effects due to repeated measures over time or bias.

Summarized below is the impact of community-based interventions on child and adolescent safety practices and injury outcomes based on this systematic review (see Tables 1 to 5). Results from specific studies are discussed if they substantially contribute to the knowledge of the effectiveness of community-based interventions in general or if they illustrate critical aspects of a particular program that influenced the program's success or failure.

Programs Targeting Bicycle Helmet Use

Bicycle injuries are a leading cause of injury death and disability among children.37 Bicycle-related head injuries result in approximately one-third of emergency department visits, two-thirds of hospitalizations, and three-fourths of deaths related to bicycling (see the article by Grossman in this journal issue).38 Several studies have shown that helmet use decreases the risk of head injury by 70% to 88%.38–43

Eleven community-based programs aimed at increasing the use of bicycle helmets by children and adolescents were identified in this review (see Table 1).2–13 Eight studies, representing seven different programs, reported a significant increase in helmet use,2,4–7,9–11 and one study reported a significant decrease in serious bicycle-related injuries12 associated with the intervention in at least some subgroups examined. Two studies found no difference in helmet use following the intervention,3,8 and one study reported no difference in self-reported risk-taking behaviors.13 A common element of successful programs was the use of multiple strategies targeted at different audiences to address three pivotal barriers to helmet use: lack of awareness about the risks of bicycling and the effectiveness of helmets, the cost of helmets, and the perception of negative peer pressure regarding helmet use. Use of multiple strategies allows a program to tackle more than one barrier to behavior change and to target more than one sub-population within the community. The result is a broad reception of the main campaign message, which is reinforced because it is sent and received in several ways.

An RCT conducted in Ontario, Canada, illustrates how multiple strategies can be used in a community-based program aimed at increasing bicycle helmet use among children and how the effects of each strategy can be examined in combination and separately.2 Three schools similar in socioeconomic status and catchment area were randomly assigned to one of the following conditions: no intervention, an educational program aimed at increasing awareness of the importance of bicycle helmet use, or the educational awareness program plus a subsidy that allowed helmets to be purchased at a discounted price.

Baseline observations revealed that none of the children in any of the schools wore bicycle helmets. One month after the program, 22% of the children in the education-plus-subsidy school were observed wearing helmets, but none of the children in the other groups had adopted this safety practice. These findings indicate that an educational intervention supplemented by a subsidy to decrease financial barriers was successful at increasing helmet use in this population.

Although educational strategies coupled with economic incentives are often successful at increasing bicycle helmet use, their effectiveness may vary by the intensity and duration of the intervention. That is, the presence of multiple strategies is a necessary, but not always sufficient, condition to ensure a successful program. For example, Towner and colleagues developed an elementary school–based intervention with multiple components.3 Discount coupons toward the purchase of bicycle helmets also were provided. However, although helmet ownership increased in the study group following the intervention (13% preintervention versus 27% postintervention), ownership increased similarly in the control group (19% preintervention versus 28% postintervention). Thus, the increase cannot be attributed to the intervention. In addition, the program failed to increase helmet use in the targeted schools, perhaps because it was short in duration and intensity.

Legislation is another strategy that has been paired with educational campaigns in community-based interventions. In Howard County, Maryland, legislation was passed that requires children under 16 years of age to wear helmets when riding bicycles on county roads and paths. An educational campaign also was adopted that included both school and community components. The combined effect of legislation and education in Howard County was compared with the effect of education alone in adjacent Montgomery County in two separate evaluations.4,5 A third community, Baltimore County, served as a control community, where no formal educational or legislative efforts were in place.

One evaluation found that approximately 10 months after the legislation went into effect, there was a significant increase in observed helmet use in the target population (4% preintervention versus 47% postintervention). 4 Although observed helmet use also increased following the educational intervention in Montgomery County (8% preintervention versus 19% postintervention), this increase was not statistically significant. In Baltimore County, use actually decreased somewhat during the same time period (19% preintervention versus 4% postintervention). In the second evaluation, fourth-, seventh-, and ninth-grade students from randomly selected schools in the three counties were asked about helmet use before and one year after the Howard County law want into effect.5 Although self-reported helmet use increased in all three counties, the increase was greatest in Howard County. One year following the legislative mandate, schoolchildren in Howard County were 2.3 times more likely to report helmet use on the most recent ride than children in the other two counties. Thus, the combined effect of legislation and education increased helmet use more than education alone.

The Seattle Bike Helmet campaign is another example of a comprehensive community-based trial that used multiple strategies to increase helmet use.6 This campaign had three specific objectives: to increase parental awareness of the need to wear a helmet when bicycling, to change peer pressure to make helmets “cool,” and to reduce financial barriers to purchasing a helmet. Financial barriers were reduced by distributing 100,000 helmet discount coupons to families through physician offices, schools, youth groups, and community events. In addition, 1,300 helmets were sold at cost through the Parent-Teacher Association, and another 1,300 were donated to low-income children.

The Seattle campaign was successful at increasing observed helmet use among children ages 5 to 15, from a baseline rate of 5.5% to nearly 16%, 16 months following the intervention. During the same time period, observed helmet use among 5- to 15-year-olds remained stable in the control community of Portland, Oregon. Since the 16-month campaign follow-up in Seattle, the observed rate of helmet use among children has increased even more, reaching 60% in 1998. This increase in helmet use was accompanied by a two-thirds reduction in bicycle-related head injuries in the target population.44

Although these studies indicate that the use of multiple strategies within a community-based intervention increased helmet use, other studies have found that the impact of comprehensive community-based programs varies substantially with the socioeconomic status of the target population.45 In particular, some community-based interventions aimed at increasing bicycle helmet use among children have been effective in high-income, but not low-income, communities.7–9 This is especially discouraging since children in low-income communities are often at higher risk for unintentional injuries.45 The studies discussed below illustrate the importance of targeting economic barriers in a community as part of a community-based approach.

A series of evaluations were conducted of “Be Bike Smart,” a program that promotes bicycle helmet use among children ages 5 to 14 in low- and high-income schools in a suburb of Toronto, Canada.7,8 This weeklong multiple-strategy program included classroom teaching, peer presentations, and celebrity appearances to encourage helmet use among children. Parents were targeted via mailings and were invited to school activities. In conjunction with this educational program, bicycle helmets were sold at a 20% discount off the regular price.7 In a subsequent “Be Bike Smart” campaign conducted only in low-income schools, the discount was raised to 75% to further minimize financial barriers.8

Following both programs, children across groups—regardless of whether or not they received the intervention—demonstrated an increase in observed helmet use. However, the intervention had a statistically significant effect on helmet use only in one subgroup: High-income children in the program that combined educational activities with a 20% helmet discount were more likely to wear a helmet following the intervention than were children in the control group (36% versus 15%, respectively). The lack of an intervention effect in low-income communities was attributed to perceived dangers or impracticalities of bicycling in high-density areas, differing belief systems, and financial and language barriers. Thus, results from the “Be Bike Smart” evaluations indicate that additional reinforcement or broader strategies may be necessary to overcome barriers to helmet use in low-income communities.

In summary, results from the studies included in this systematic review indicate that community-based efforts can effectively increase bicycle helmet use among children. To maximize the likelihood of success, efforts should be targeted broadly to increase parents' awareness of the importance of helmet use, overcome resistance to helmet use by educating children, subsidize helmet costs, and pass regional legislation enforcing bicycle helmet use.

Even when these approaches are adopted, helmet use among children may not be increased in all circumstances, and different interventions may be required for different populations. For example, low-income communities may be more difficult to influence with community-based bicycle injury prevention programs. Regardless of community socioeconomic status, however, a common thread through several of these studies was the influence of peer pressure and modeling by adults. Children were more likely to wear a bicycle helmet if their friends also wore a helmet or if they were with an adult who also wore a helmet. This effect is consistent with social learning theory, which recognizes modeling and peer group behaviors as both stimuli and active reinforcers of behavior change.35

Despite the positive impact of numerous community-based interventions on children's bicycle helmet use, the design of these evaluations and the limited outcomes explored suggest that caution is warranted when interpreting these results. Importantly, only two RCTs of community-based helmet promotion programs were included in this review. In addition, nearly all studies focused on helmet ownership or use; only one looked at actual injuries.12 Future research should investigate the importance of bicycle paths and general bicycle safety measures.

Programs Targeting Motor Vehicle Restraint Use

In 1996, motor vehicle occupant injuries resulted in the deaths of 4,970 children and adolescents in the United States. Child passenger restraints for infants and young children, and adult shoulder-lap belts used correctly by children older than eight years of age, substantially reduce the probability of serious injury or fatality (see the article by Grossman in this journal issue).46 Although child safety restraint legislation has been passed in all 50 states, restraints are used for only about 85% of infants and 60% of toddlers.47 Furthermore, a large proportion of children are incorrectly restrained in car seats, and car seats are often installed incorrectly in vehicles.48

This review identified five community-based studies aimed at increasing motor vehicle restraint use among children 14–17 or adolescents 18 (see Table 2). (Interventions focused on increasing infant car seat use are primarily delivered in clinical settings and are discussed in the article by DiGuiseppi and Roberts in this journal issue). Findings from these studies indicate that multiple-strategy programs that educate preschoolers, and programs that rely on parent education in concert with coercive techniques, can increase child motor vehicle restraint use. However, interventions may benefit more from the “carrot” than the “stick” approach to behavior change. Targeting children's behavior directly is particularly important, because child motor vehicle restraint use is often inhibited when a child resists being restricted in a child safety seat and a weary parent succumbs to this resistance. Some of the most successful community-based interventions aimed at increasing child safety restraint use focus primarily on increasing compliance among children, with the theory that children may be trained to serve as monitors of the family's motor vehicle restraint behaviors.

The “Bucklebear” preschool-based curriculum, designed to increase child motor vehicle restraint use among preschool-age children in California, focused primarily on children to effect behavior change.14 Six preschools that adopted this curriculum were compared with seven control preschools before they implemented the program and again three weeks after implementation. The program was successful, as evidenced by a significant increase in safety seat use among children in the intervention group (22% preintervention versus 44% postintervention) with no change among controls (22% preintervention versus 24% postintervention).

The Children's Traffic Safety Program in Tennessee is another school-based program aimed at increasing children's motor vehicle restraint use by targeting children and their parents with a public education campaign.15 The program was based on the premise that if young children (grades K–12) are taught to adopt safety behaviors, they can then influence motor vehicle restraint use within their family. A community education campaign that included television, radio, and newspaper features and public service announcements ran parallel to the 10-week school-based curriculum. Control schools within the same district did not receive the intervention. In addition to evaluating the effect of the program on motor vehicle restraint use, researchers also examined the variation in impact by socioeconomic status and the quality of program implementation at each school.

Program effectiveness was influenced by both the quality of program implementation and socioeconomic status. Observed safety restraint use increased significantly only in low-income schools with good program implementation (13% preintervention versus 25% postintervention), measured 5 to 14 days after the intervention. In other intervention and control schools, the difference in motor vehicle restraint use before and after the intervention was not significant.

Findings from this trial highlight the critical importance of both program implementation and the target population to a program's success. If program coordinators/teachers are not adequately trained or do not present the program in the manner in which it was designed, the impact may be diminished. Results from this trial mirror those observed in bicycle helmet promotion trials, reported earlier, in terms of differential impact across socioeconomic status. However, whereas community-based programs to increase bicycle helmet use were generally more effective in high-income communities, this trial had a greater impact in low-income communities. This variation in results may be attributed to the much lower baseline rates of restraint use in low-income than high-income schools in this study, which allowed more room for improvement. Despite the fact that low-income children demonstrated a significant increase in motor vehicle restraint use, their postintervention use (16% to 25%) still was lower than the baseline use among high-income children (49% to 68%).

A community-based trial in Australia compared the effect of coercive tactics aimed at parents with an educational approach aimed at children.16 The coercive intervention threatened parents with random police checks and fines, and it reminded them of legislation mandating child safety restraint use. The educational intervention relied on kits used by educators to teach preschool children the importance of wearing automobile safety restraints. The rationale for teaching children was that this information would “filter up” to the parents and further increase the use of child restraints.

Observations before and two weeks after the intervention revealed that only the educational approach was effective at increasing automobile safety restraint use among preschoolers (61% preintervention versus 75% postintervention). Use did not increase in either the control or the coercive intervention groups. It was postulated that the coercive intervention failed to impact safety restraint use because it did not deal with the fact that children often resist being restrained. In addition, the mere threat of police involvement may not have been a sufficient deterrent, and a more visible police presence may be necessary.

A positive approach to changing the safety behavior of parents may have a greater impact on child safety restraint use than coercive tactics. In an RCT aimed at increasing child safety restraint use, Stuy and colleagues sought to alter the behavior of parents by requiring them to sign a policy statement in which they agreed to comply with state laws and day-care center recommendations regarding child safety restraint use.17 Based on the tenets of social learning theory, Stuy and colleagues hypothesized that this program “buy-in” would help foster a sense of membership in a social group and that parents would alter their behavior as a result of modeling or reinforcement by other members or because of their own active involvement in the group. This strategy was combined with a comprehensive educational safety program in the school.

Child safety restraint use increased significantly following the intervention in both the intervention (54% preintervention versus 75% postintervention) and the control (20% preintervention versus 30% postintervention) groups. Although child care centers were randomly assigned to the intervention or control conditions, some differences between the groups are worth noting. Specifically, the control centers had fewer white families, more single-parent families, and more families with an annual income of less than $15,000. It is unlikely that these differences substantially influenced the findings, however, and this “positive” approach was successful at increasing child safety restraint use. It also may be a less costly alternative to the coercive approach described above, since it does not require visible enforcement.

Only one trial in this review, The Oregon Head and Spinal Cord Injury Prevention Program, was aimed at increasing shoulder and lap belt use among adolescents through an educational campaign.18 This program relied on a single strategy—providing emotionally charged information at a high school assembly to change knowledge, attitudes, and behaviors regarding seat belt use.

The effectiveness of the program was measured by observed seat belt use and questionnaire measures of knowledge and attitudes taken two weeks before and two weeks after the intervention. The intervention was not associated with changes in seat belt use or attitudes toward use. Only knowledge about the importance of seat belt use increased following the intervention. This may be of little consequence, however, given the weak relationship between changes in knowledge and changes in injury behavior.37 Another study of this curriculum implemented in Washington State also found no consistent change in knowledge, attitudes, or seat belt use associated with the intervention.49 Consistent with other community-based programs, these findings suggest that single, one-time interventions to change behavior are not successful.

In summary, several community-based interventions delivered in day-care or school settings have increased motor vehicle restraint use among young children. The use of multiple strategies to promote behavior change and a focus on increasing children's acceptance of motor vehicle safety seats appear critical to a program's success. Additional well-designed RCTs of community-based interventions that use these strategies would be useful to corroborate the findings of the numerous nonrandomized studies reported in this article. Rigorous community-based studies focusing on motor vehicle restraint use among adolescents also are needed.

Pedestrian Safety Interventions

Pedestrian injuries among children and youths accounted for nearly 19% of all pedestrian deaths in 1996 (see the article by Grossman in this journal issue). Preschool-age children are at particular risk, as they lack the ability to judge the safety of street crossings and may be inadequately supervised by their parents.

Four community-based studies aimed at reducing child pedestrian injuries met the inclusion criteria for this systematic review. These studies all targeted children between the ages of three and six in school20–22 or home19 settings (see Table 3). Two of these were RCTs that measured children's traffic safety behavior in simulated environments; they reported marginal21 or no improvement associated with the interventions.22 The two nonrandomized studies of child pedestrian interventions focused on parents as supervisors and facilitators of children's behavior change.19,20 Both studies found that with appropriate training, parents or other adults can positively impact children's traffic safety behavior. However, the one study that reported numeric data found that even after the intervention nearly three-quarters of children still ran ahead of their parents near traffic, and participating parents did not differ from control parents in the amount or quality of their supervision.19

In summary, results from this review indicate that the benefit of community-based education aimed at improving traffic safety behavior among young children is limited. The benefit of such interventions delivered in a simulated setting is marginal at best, and there is no evidence that children will behave the same way in real-life settings. Although some evidence suggests that community-based interventions involving parents or other instructors as supervisors and facilitators can improve children's traffic safety behaviors, this improvement is modest, and even after training, young children remain at substantial risk for pedestrian injuries.

Developmentally, preschool-age children are not prepared to learn and react appropriately to traffic.50 Therefore, physically separating young children from traffic may be a more effective approach. Uncontrolled evaluations of environmental approaches to reducing pedestrian injuries—such as by lowering speed limits, using speed bumps and signs, or narrowing roads—have been conducted, primarily in Europe, and are promising.51–53 Future investigations using rigorous methodological designs are necessary to quantify the benefits or shortcomings of environmental approaches, because they are gaining popularity over educational interventions.

General Safety Campaigns

This article has discussed community-based studies, each focused on a specific type of injury. However, a major strength of the community-based approach is the opportunity to target more than one type of injury within a community. Four nonrandomized community-based trials were found that used general safety campaigns to target multiple child-injury problems (see Table 4).23–26 Three of these used changes in injury rates as a measure of program effectiveness, though only one trial reported statistically significant decreases in injury rates associated with the intervention.23 General safety campaigns differ from other community-based interventions in that they tend to be long term and adapted over time to meet community needs. The success of these interventions largely relies on input from community members and community ownership of program activities.

The Safe Kids/Healthy Neighborhoods Coalition illustrates how a successful community-based intervention grows and changes within a community.23 This program, implemented in the Harlem neighborhood of New York City, was aimed at reducing a variety of childhood injuries resulting from outdoor activities. The initiative was started because parents and educators in central Harlem requested a program in playground safety from health professionals. Surveys of playgrounds revealed that they were being used by drug dealers and were in poor repair, and that children using playgrounds had little adult supervision. The Harlem Hospital Injury Prevention Program, in collaboration with a variety of community groups and city agencies, started the Safe Kids/Healthy Neighborhoods Coalition with the following goals: (1) to renovate central Harlem playgrounds; (2) to involve children and adolescents in safe, supervised activities that teach useful skills; (3) to provide injury and violence prevention education; and (4) to provide safety equipment (for example, bicycle helmets) at reasonable cost. During the first three years of the program, 26 organizations participated in the coalition, and different programs and activities were developed.

The effectiveness of this community-based coalition was evaluated by examining changes in injury rates in the targeted age group (5- to 16-year-olds) for injuries targeted by the campaign (for example, all injuries related to vehicles, outdoor falls, assaults, and guns) over nine years (1983 to 1991). These rates were compared with changes in the rates of nontargeted injuries (that is, poisoning, burns, and so on) and changes in injury rates in a comparison community (the suburb of Washington Heights) during the same time period.

Targeted injuries were reduced by an estimated 44% during the intervention period, with no significant decrease for non-targeted injuries; this decrease was noted mainly in the targeted age group. Unexpectedly, a 30% decline in severe injuries among school-age children also was observed in Washington Heights. However, the decline in Washington Heights occurred in both targeted (motor vehicle injuries only) and nontargeted categories. Whether this reduction in motor vehicle injuries observed in the comparison community occurred independently of the intervention, or whether the effect of the Safe Kids/Healthy Neighborhoods Coalition “spilled over” into Washington Heights, could not be ascertained. Notwithstanding this concurrent change in the control community, the authors concluded that the specific decrease in targeted injuries within the targeted age group in central Harlem demonstrated a positive effect of the intervention.

A second general injury prevention campaign, the Lidkoping Accident Prevention Programme (LAPP), was associated with a modest, though not significant, rate decrease in injuries leading to hospital admission.24 Nonetheless, the comprehensiveness of the strategies used and the collaborative approach adopted illustrate critical components of promising general safety campaigns.

LAPP was developed by a community health unit in an effort to plan and coordinate health and safety in Lidkoping, Sweden. The program included five elements: (1) surveillance of inpatient and outpatient injuries; (2) information, such as telephone hotlines and bicycle safety campaigns; (3) training of coaches, child care staff, and parents in first aid or other injury prevention areas; (4) supervision, including municipal safety; and (5) environmental improvements, such as improving gym floors to decrease slipping injuries. Key to this campaign was the involvement of representatives from existing community organizations, municipal administrations, welfare services, and the Red Cross, as well as community professionals such as engineers, nurses, teachers, and police.

The impact of LAPP was evaluated by measuring decreases in rates of injuries leading to hospital admissions over nine years (1983 to 1991). During this time, there was an average annual decrease in hospital admissions in the intervention group of 2.4% for boys and 2.1% for girls, without a similar decrease in neighboring comparison municipalities. These differences were not significant, and many methodological limitations were associated with the intervention, such as changes in recording procedures at the hospitals and inconsistent implementation during the early years of the program.

In summary, findings from general community-based safety campaigns are mixed at best. Only the Safe Kids/Healthy Neighborhoods Coalition has reported statistically significant decreases in targeted injuries, and even in this study it is questionable whether the decrease in injury rates can be wholly attributed to the intervention. Nonetheless, general safety campaigns that adopt a broad array of strategies and garner support from numerous community constituencies are promising.

Future general safety campaigns should rely on these strategies and employ rigorous study designs to evaluate program impact. RCTs may not be possible since the success of these broad-based programs likely depends on an impetus for change that comes from community constituencies, as with the Safe Kids/Healthy Neighborhoods Coalition. Thus, randomly assigning a community to receive the intervention, whether or not the community expresses a desire for change, may diminish program success. Nevertheless, careful selection of a comparison community that is similar to the intervention community with respect to characteristics associated with the injury outcomes of interest, but distant enough in proximity to avoid “spill over” effects, is critical. Evaluation of general safety campaigns also requires carefully standardized data coding that is narrow enough to capture the different components of the program, yet broad enough to ensure that data collection is manageable. Finally, more sophisticated data analyses are required so that possible confounds, such as variations in time, are statistically controlled.

Programs Targeting Adolescent Alcohol Use and Vehicle Safety

Motor vehicle collisions are a major cause of death and disability in youths, with alcohol playing an important role in many instances (see the article by Grossman in this journal issue).54 Although interventions are needed to alter unsafe adolescent driving behaviors, risk-oriented behaviors common among teenagers make this a difficult challenge. Peer pressure also is a powerful motivator among adolescents, and individual behavior change often is predicated on behavior change in one's peer group.

This review identified three community-based programs aimed at decreasing alcohol misuse and reckless driving among teenagers (see Table 5).27–29 All three programs were delivered as part of a high school educational curriculum; only one was an RCT.28 While all programs reported that the intervention increased knowledge of the risks of consuming alcohol and operating a motor vehicle, only one trial found a change in behavior (self-reported increased seat belt use when riding as a passenger) associated with the intervention.29 Because the link between knowledge, attitudes, and subsequent behaviors is tenuous at best,37 interventions that affect the first two domains, but do not influence behavior change, should be viewed with caution.

Newman and colleagues developed an innovative program, “Resisting Pressures to Drink and Drive,” designed to teach ninth-grade students the physiological effects of alcohol, myths about alcohol use, and skills to resist peer pressure to drink.27 Students from nine schools were assigned to either the “Resisting Pressures to Drink and Drive” intervention group or to a control group that received the traditional alcohol education program. Increases in students' knowledge associated with the intervention were observed after the first year of the program. The intervention was not successful, however, at changing behaviors (that is, reducing the rate of riding with a driver who has been drinking), changing teenagers' perceived ability to resist peer pressure, or reducing alcohol consumption. In fact, both alcohol consumption and the number of occurrences of riding with a driver who had been drinking increased over time in both groups.

A second school-based curriculum was aimed at preventing alcohol use among 10th grade students in nine Michigan high schools by preparing them to cope effectively with peer pressures to misuse alcohol. 28 Classes within each school were randomly assigned to either the intervention or control groups, with the intervention delivered by trained teachers during five sessions. A significant increase in knowledge about alcohol, its physiological effects, and resisting pressure to drink was observed in the intervention group following the intervention and again in 12th grade, with no change among controls. However, despite this sustained gain in knowledge, self-reported alcohol misuse and driving a motor vehicle after drinking still increased over time among students in both groups.

A similar program aimed at improving vehicle safety included a weeklong module on injury control and crash safety information in a high school physics course.29 The intervention covered forms of energy, injury prevention, car safety features, types of vehicular collisions, seat belts, and g forces, as well as hands-on activities. At baseline, the groups were similar in terms of knowledge, self-reported seat belt use, speeding, and driving under the influence of alcohol. Two years following the intervention, however, knowledge, reported seat belt use when riding as a passenger, and intention of always wearing a seat belt were significantly higher in the intervention group. Despite the positive behavioral change reflected in seat belt use—which increased from 70% to 80% in the intervention group, but only 67% to 70% among controls—this study suffered from notable limitations. These included differential dropout of risk takers (that is, teenagers most inclined to drive fast, drive after drinking one to two alcoholic beverages, and not wear a seat belt) between follow-up assessments, and scores on drinking and driving attitudinal questions that left little room for improvement.

In summary, alcohol use among adolescents is a difficult behavior to modify, and community-based programs aimed at reducing the likelihood that adolescents will drive or ride with a driver under the influence of alcohol have been unsuccessful. In fact, results from these studies show that adolescent alcohol consumption actually increases with age, and increased knowledge regarding alcohol misuse negatively correlates with subsequent alcohol-related behavior.

The failure of these studies to influence adolescent behavior may reflect interventions that were poorly designed for this age group. Because modeling peer group behaviors is so important for teenagers, future community-based strategies should appeal to the ability of peer groups to facilitate changes in normative behavior. Differences within teenage peer groups also should be examined to determine whether certain groups are more amenable to change following different strategic interventions. It may be that different strategies need to be implemented for different peer groups to maximize program success. Teenagers who are involved in sports, for example, may respond most favorably to interventions that focus on the potential detrimental effects of alcohol on athletic performance. Alternatively, educating young women about the high caloric content of alcoholic beverages may have the greatest impact on their drinking behaviors. Because teenagers are in a unique developmental stage, in which they are testing boundaries, interventions also may be more successful if they provide additional rewards and incentives for adopting well-defined vehicle safety behaviors. Finally, future interventions should explore more innovative and less pedagogical approaches to promote adolescent safety behaviors. Given the increased use of computer technology by youths, for example, Internet-based interventions should be tested as a potential venue for influencing adolescent safety behaviors.

Even if the effectiveness of community-based strategies can be improved by more carefully designed and executed interventions, programs that rely on educational strategies to alter adolescent behaviors are likely to be more successful when combined with regulatory or legislative approaches such as graduated licensure (see the article by Schieber, Gilchrist, and Sleet in this journal issue). A recent systematic review of graduated licensure found that this approach is likely to reduce motor vehicle crashes and crash-related injuries involving teenagers, although few evaluations of graduated licensing systems have been conducted.55