Background. Bicycle-related head injuries cause >150 deaths and 45 000 nonfatal injuries among children in the United States annually. Although bicycle helmets are highly effective against head injury, only 24% of US children regularly wear one. Georgia mandated bicycle helmet use for children, effective July 1993. During that summer, 1 rural Georgia community passed an ordinance instructing police officers to impound the bicycle of any unhelmeted child. We evaluated the effect of active police enforcement of this ordinance, combined with a helmet giveaway and education program.
Methods. During April 1997, ∼580 children in kindergarten through grade 7 received free helmets, fitting instructions, and safety education. Police then began impounding bicycles of unhelmeted children. We conducted an observational study, unobtrusively observing helmet use just before helmet distribution, several times during the next 5 months, and once 2 years later.
Results. Before the program began, none of 97 observed riders wore a helmet. During the next 5 months, helmet use among 358 observed children averaged 45% (range: 30%–71%), a significant increase in all race and gender groups. In contrast, adult use did not change significantly. Police impounded 167 bicycles during the study, an average of 1 per day. Two years after program initiation, 21 of 39 child riders (54%) were observed wearing a helmet.
Conclusions. Without enforcement, the state and local laws did not prompt helmet use in this community, yet active police enforcement, coupled with helmet giveaways and education, was effective and lasting.
Bicycling is a common recreational activity and mode of transportation among children. Approximately 27.7 million American children <15 years of age ride a bicycle.1However, this activity is not without risk. In 1997, an estimated 367 700 such children sought emergency department care for a bicycle-related injury,2 of which an estimated 111 300 (30%) sustained a head, facial, or ear injury. Among the 224 children killed in 1997 from bicycle-related injuries, approximately two thirds sustained a head injury.3 Several studies indicate that bicycle helmets prevent 69% to 88% of serious head or brain injuries.4 ,5 Furthermore, as helmet use increases, hospital admissions and deaths from bicycle-related head injuries decrease.6–8 Despite such evidence supporting helmet effectiveness, a 1994 national study indicated that only approximately one half of the children who rode a bicycle reportedly owned a helmet, and only one quarter of these riders wore a helmet every time they rode.1
Bicycle helmet promotion programs have used several strategies, alone or in combination, including education, helmet giveaways, and adoption of helmet use laws at the state or local level. Passage of bicycle helmet legislation in all 50 states is a national objective inHealthy People 2000; 16 states currently have helmet use laws for children. (The 16 states currently with helmet use laws for children include: AL, CA, CT, DE, FL, GA, MD, ME, MA, NJ, NY, OR, PA, RI, TN, and WV.) However, the effectiveness of such laws has varied. In some geographic areas, enactment has coincided with an increase in reported and/or observed helmet use,9–11 whereas in others, legislation has had little effect.12 We sought to study the effectiveness of adding enforcement to these legislative efforts. We report here our evaluation of a combined giveaway, education, and enforcement program conceived and conducted by a rural community in Georgia.
The Georgia Assembly passed a state law effective July 1, 1993 mandating that all bicycle riders <16 years old wear helmets. Although the law allowed a parent to be cited for a child's noncompliance, in practice, citations have rarely been issued. Later that summer, the city council of a small, rural Georgia community passed an ordinance strengthening the state law by instructing local police to impound the bicycle of any child <13 years old seen riding without a helmet. Rather than issuing a citation or a fine, police impounded the bicycle and required a parent to retrieve it at the police station, where the safety message was reinforced to the parent and child and helmet ownership was verified or a helmet was provided. This local enforcement program allowed us the opportunity to study the effectiveness of the enforcement strategy on bicycle helmet use before and after initiation of the program.
The community has a population of ∼2400 residents (25% white and 75% black), of which 41% have an income below the federal poverty level.13 The town provided an ideal setting for this study because: 1) its geographic isolation reduced the possibility that nonresidents not exposed to the program would be observed riding; 2) its small size facilitated relatively complete observation of the entire population by only a few observers; 3) its small population allowed for the local coalition of the National SAFE KIDS Campaign and the Georgia Division of Public Health to affordably provide each rider in the target age group a bicycle helmet; 4) the police force and city officials were committed to the ongoing enforcement of this ordinance; and 5) a classroom show-of-hands in kindergarten through grade 5 before initiation of the program demonstrated a very low rate of helmet ownership.
The intervention program had 2 components, a helmet giveaway/education program (distribution) and an enforcement program. Intervention here refers to both the distribution and ongoing enforcement components. Because the local ordinance had been enforced only briefly when passed in the summer of 1993 and not subsequently, police reinstituted the program by initially issuing warnings to unhelmeted children in April 1997. The distribution program described below began in late April, after which the police began impounding the bicycle of any unhelmeted child rider. The enforcement program has continued to the present.
The distribution program consisted of a helmet giveaway with fitting instructions and 10 minutes of bike safety education in late April 1997; a parent education program; and 2 bicycle rodeos. Helmets were distributed at the only local elementary school to all 426 children in kindergarten through fifth grade (5–10 years of age). Approximately 150 local students were in grades 6 through 7 (11–12 years of age); they attended a different school with students from a different town, yet received helmets and training from the police on roadsides and in parks. To deliver the safety message to parents and to remind them of the penalty, an information pamphlet was sent home from school with each child and 2 articles were printed in the local newspaper. In addition, a bicycle rodeo was held just after the helmet distribution to reinforce safety messages and provide helmets to any target-aged child who had previously not been given one. A second rodeo was held 20 weeks after helmet distribution. Altogether, 650 helmets were distributed at the beginning of the study to the ∼576 target-aged children and to any teen requesting one. Additionally, 100 helmets were distributed during both of the 2 years since the end of the study to sustain the program.
Helmet use was determined by unobtrusive observation. This study was exempted by the Centers for Disease Control and Prevention Institutional Review Board, because we only observed public behavior, no identifying data were collected, and no interaction with the bicyclists occurred.14 Each of the same 4 trained community workers repeatedly canvassed a quadrant of town by automobile using a predetermined route according to a previously described method.15 Observers recorded helmet use, time of day, race, and gender. In addition, the rider's approximate age was estimated from the rider's size and presence or absence of secondary sexual characteristics. Each observation period lasted ∼12 hours (Friday afternoons after school until dark and daylight hours on Saturday). Observations were conducted the weekend before helmets were distributed, weekly for 3 consecutive weeks, and monthly for 4 consecutive months, totaling 7 postdistribution observation periods during the 5-month study. Although not a part of the formal study, 2 years after the initial helmet distribution most of the authors and 1 volunteer observed bicyclists 1 Thursday after school in May 1999, using the same canvassing technique to determine whether use was sustained. Because the 2-year follow-up observations were not conducted during the same 12-hour period as previous observations and are not considered part of the study, the follow-up data are reported but not included in the tables or analysis.
The unit of analysis was the observed ride rather than the observed rider. If a rider was seen more than once, observers were instructed to record subsequent observations if >30 minutes apart, because these were considered separate rides. Previous observational studies have counted rides,6 ,7 ,9 ,11 ,12 ,15 ,16 although telephone surveys have counted riders.1 ,9 ,10 These 2 methods yield identical results only if each rider is counted once or the telephone survey inquires about a single ride. However, because we thoroughly and repeatedly canvassed this small town, we assume that some unknown number of bike rides were counted 2 or more times during the observation period. We made no attempt to identify a rider or to track his/her helmet usage across observation periods because our previous experience indicated that this could not be done accurately or consistently. By counting rides instead of riders, this may incorporate some measure of exposure.
Statistical significance between predistribution and pooled postdistribution observation results for children was tested using the Pearson χ2, Mantel-Haenszel χ2, or Fisher's exact test, as appropriate. Statistical significance was established for P values <.05. Riders were classified as children if their estimated age was between 5 and 12 years old, teens if between 13 and 15 years old, and adults if 16 years of age or older. We report the observed helmet use of teens and adults but do not statistically compare them with children. Teens and adults were not considered to be valid control groups because some, but not all, teens received a free helmet, none of the teens received safety instruction, and neither the teens nor the adults were subject to bicycle impoundment because they exceeded the age limit of the local ordinance.
During the 5-month study, 777 total observations were recorded. Eighteen observations were excluded because of missing data: unknown gender (n = 7), unknown age (n = 9), and unknown race (n = 2). Eight records were excluded because the riders were estimated to be <5 years old. The remaining 751 complete observations (97 predistribution and 654 postdistribution) used in the analysis included 419 children (56%), 141 teens (19%), and 191 adults (25%).
A classroom show-of-hands before the distribution indicated that only 8% of children in kindergarten through grade 5 owned a helmet. Observations before distribution found that helmet use among 61 child bicycle riders was 0% (Table 1). After helmet distribution and initiation of enforcement, 167 bicycles were impounded. Bicycle impoundment during the study averaged 1 bicycle per day (range: 0–5) or 33 bicycles per month (range: 22–45). Observed helmet use in children increased from 0% predistribution to between 30% and 71% postdistribution (mean: 45%; P value = .001; Fig 1). By comparison, mean helmet use in adults did not change significantly, from 0% predistribution to 3% postdistribution (Table 1). Children's helmet use increased significantly in all race–gender strata for which significance could be determined (Table 2). Two years after the initiation of the intervention, 21 of 39 child riders (54%) were observed wearing helmets during 1 afternoon, compared with only 2 (15%) of 13 teens and none of the 23 adults.
In this study, the combination of a bicycle helmet giveaway, education, and active police enforcement using bicycle impoundment was highly effective in increasing helmet use among children. The observed change in helmet use was immediate, dramatic, and apparently sustained. Helmet use in this study exceeded that of another study in a similar population in which a universal helmet giveaway program was not accompanied by enforcement.15 That rural population, although lacking a state helmet law, noted an increase in helmet use among target-aged children from 0% to a maximum of 38% after universal helmet giveaways, education, and incentives; however, helmet use plummeted to 5% when the incentives program ceased. In the group of children observed here, a state bicycle helmet law with an unenforced local ordinance resulted in minimal helmet ownership and no observed helmet use. Because no other local helmet safety programs were provided during this time, our findings suggest that this combination of giveaways, education, and enforcement caused the behavioral change. Unfortunately, this study design did not allow us to determine whether enforcement alone motivated helmet use, but the results support our conclusion that the addition of an enforcement component to a program of giveaways and education is a highly effective intervention strategy.
Helmet use in children varied between 30% and 71% among the 7 postdistribution observation periods. Possible reasons for this variation include intermittent helmet use, changing perceptions of the vigor of enforcement, or the degree of peer influence by a child whose bicycle was impounded. Other observational studies have had similar variable results.15
Our results are subject to several limitations. We could not be certain that all children in grades 6 through 7 received a helmet. However, the number of helmets distributed exceeded the number of children in kindergarten through grade 7 (the target population), with the additional helmets distributed to teens. Age group misclassification was possible, although local residents conducted these observations. Age misclassification is always a concern in observational studies; however, observed helmet use is more reliable than reported use from phone or mail surveys. These results from a rural community may not be generalizable to an urban or suburban community, and such a program should be tested there. We could not study the effect of this program on teens, the age group hardest to reach, because they were not subject to the local ordinance. It is highly likely that riders were observed multiple times over the 8 observation periods, but the extent of this is unknown. In the analysis, we presumed that each observation was independent, which would lead to conservative test results. Although we hesitate to suggest program plans based on the small number of observations in each race/gender group, we provide data on race and gender that suggest that this program may be effective in each of these groups. Finally, 2 bicycle rodeos were held during the study. Each was attended by ∼40 children (∼10% of the target group). However, we do not believe that the rodeos contributed greatly to the increased helmet use. A similar study that included rodeos with the giveaway and education components was unsuccessful.15 Thus, we believe that adding the enforcement component in this study contributed to the increased helmet use.
Although bicycle impoundment may seem to be a severe penalty, in this poor community it was better accepted than a fine. The penalty was directly related to both the infraction and its solution. Thus, the police action strongly reinforced the safety messages. In fact, confiscating the bicycle actually protected these children by removing their risk of bicycle-related head injury.
Another common criticism of increased enforcement is that it requires large expenditures of otherwise limited resources. However, this program did not require financial resources or even much additional time by local police. The helmets were donated, and police officers, community volunteers, and Department of Health workers conducted the distribution component. The principal responsibility for enforcement was assigned to 1 officer each day, to be performed during routine patrol. The perception of enforcement, perhaps augmented by local conversation and media attention, might have been more important than the actual enforcement itself. If so, personnel requirements for such a program may not be prohibitive for larger communities.
Anecdotally, during the 2-year follow-up observation, we saw children participating in other activities wearing bicycle helmets not required by any law. Each of 2 children rollerblading, all 3 children jumping on a trampoline, as well as 3 children walking along a street without bicycles were observed wearing bicycle helmets. Such qualitative data suggest that wearing a bicycle helmet might have become the social norm to children in this community.
These results support the value of enforcement as 1 option in a sustained, multifaceted community-based approach among children.17 Other multifaceted programs have had a positive effect without enforcement, yet took longer to accomplish. In Seattle, Washington, helmet use increased from 6% to 40% between 1987 to 1992 without a mandatory use law.16 In Pittsburg, California, between 1994 and 1997, a sustained program of giveaways, safety education, and a state law (without a special enforcement effort) resulted in an increase from 22% to 72% helmet use in grade school children.18
Enforcement provides yet another strategy for increasing and sustaining bicycle helmet use. To our knowledge only 1 other published report involves active enforcement of bicycle helmet laws. In Victoria, Australia, legislation was passed and enforced in 1990 after a decade of comprehensive, multifaceted education, incentives, and promotion aimed at increasing helmet use in all ages. Helmet use rates there increased from 31% in 1990 (prelaw) to 75% 1 year later after initiation of tickets and fines.6
The nature and degree of enforcement make this study unique. It suggests that targeted enforcement of safety laws may be an effective means for increasing bicycle helmet use in children. The enforcement effort may not require prohibitive personnel or financial resources. In the effort to increase helmet use among children, 1 option includes local enforcement of bicycle helmet laws to supplement the traditional helmet distribution and education components of a bicycle helmet promotion program.
Helmets were provided by the Georgia Division of Public Health, the East Central Health District of Georgia, and the local SAFE KIDS campaign.
We acknowledge the support of Sally Adams of the Mayor's Office and Carl Wagster, coordinator of the local coalition of the National SAFE KIDS Campaign.
- Received June 3, 1999.
- Accepted October 20, 1999.
Reprint requests to (J.G.) Division of Unintentional Injury Prevention, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, 4770 Buford Hwy, NE, Mail Stop K63, Atlanta, GA 30341. E-mail:
- Sacks JJ,
- Kresnow M,
- Houston B,
- Russell J
- ↵US Consumer Product Safety Commission. National Electronic Injury Surveillance System Data, 1997 [machine-readable public use data tapes]. Washington, DC: US Consumer Product Safety Commission; 1999
- ↵National Center for Health Statistics. Vital Statistics Mortality Data, Underlying Cause of Death, 1997 [machine-readable public use data tapes]. Hyattsville, MD: US Department of Health and Human Service, Centers for Disease Control and Prevention, National Center for Health Statistics; 1999
- Centers for Disease Control and Prevention
- Thomas S,
- Acton C,
- Nixon J,
- Battistutta D,
- Pitt WR,
- Clark R
- Cote TR,
- Sacks JJ,
- Lambert-Huber DA,
- et al.
- ↵US Department of Commerce, Bureau of the Census. Decennial Data: 1990. Washington, DC: 1990
- ↵Human Subjects, Title 45 C. F. R. Part 46.101(2)(b)(2), 1998
- Logan P,
- Leadbetter S,
- Gibson RE,
- et al.
- Rivara FP,
- Thompson DC,
- Thompson RS,
- et al.
- ↵Foster V, Carr N, Perales D, Alberson B, Kelter A. Applying lessons learned: promoting statewide bicycle helmet use in California. In: Public Health and Managed Care, Abstracts of the 126th Annual APHA Conference; November 15–18, 1998; Washington, DC
- Copyright © 2000 American Academy of Pediatrics