Challenging Dogma

...Using social sciences to improve the practice of public health

Sunday, April 22, 2007

Riding At Risk: How NHTSA Initiatives To Increase Use of Child Restraint Systems Have Failed by Ignoring Sociocultural Factors- Marni Francer

It is well established that child restraint systems such as infant safety seats and booster seats reduce morbidity and mortality in children in the event of an automobile accident. Approximately 7,500 lives have been saved by the proper use of child restraints during the past 20 years (1). However, according to the National Highway Traffic Safety Administration (NHTSA), an average of five children aged birth to 14 years are killed and 640 injured in motor vehicle crashes every day. Motor vehicle crashes are the number one killer of children ages 4 to 14 in America (1). The reason? Too often it is the improper use or non-use of child safety seats. The proportion of children who are properly restrained in a car is appallingly low despite numerous public health efforts to improve usage. So, as a pediatrician and a concerned citizen I agonize: Why are these children dying preventable deaths? Why have public health efforts to improve child restraint usage failed our children?

The largest organizer of initiatives to improve child restraint system use is the NHTSA through its “Child Passenger Safety Program” (CPS) (1). As a part of this initiative, the New York State Police implemented a child passenger safety program centered on four activities: 1) the development of a comprehensive educational program available to parents and caregivers, 2) the promotion of the program through various media outlets, 3) the creation of permanent fitting stations where parents can obtain education on child passenger safety and assistance with the installation of their child restraint devices, and 4) the encouragement of "in-service" child passenger safety training for all New York State Police uniformed officers (2). It is based on the health belief model that increased education and the availability of assistance with car seat installation will lead to the intention and subsequent adoption of a healthy behavior.

However, the Child Passenger Safety Program in New York is destined to fail because it does not take sociocultural and financial factors that influence car seat use into account. The program does not specifically target those groups known to be at increased risk of failure to use child restraints and does not address three important barriers to use. First, it fails to address the caregiver’s sense of self-efficacy to enforce child restraint use in the face of a child’s resistance. Second, it ignores important cultural influences that may work against use. Finally, it neglects significant barriers among those of lower SES such as the high cost of car seats and low levels of literacy that impede self-education about child restraint devices.

Numerous studies have consistently demonstrated risk factors for failure to use child restraint devices. Caretakers who have limited education, are of non-white race (3-6), have low family income (3,4,7), are not married (2, 7 ), are male (5) , and are older (generally >40 years) (3, 4) are less likely to use child restraint devices. The New York State Police initiative states that it aims to address its message to all inhabitants of New York state without mention of specific interventions targeted to high risk groups such men, those with lower SES, or racial and ethnic minorities (2) .

The first means by which the NY State Police initiative is limited is its failure to consider the caregiver’s sense of self-efficacy to enforce the use of a restraint device despite adamant protests from the child. As a pediatrician, I observe daily that if a parent does not believe they can successfully get their child to comply with positive behaviors, any advice I give about “what you should do” will never be acted upon no matter how much the parent wants a change in the child’s behavior. The first step to changing a parent’s behavior, and thus a child’s behavior, is to endow the parent with the skills and accompanying confidence to take on and conquer the change they seek. My observations are supported by the Social- Cognitive Theory which states that behavior change hinges on the belief that one has the capabilities to execute the action required to manage prospective situations (8). Self-efficacy plays a central role in the Social-Cognitive Theory of Motivation because people self-regulate the level and distribution of effort they will expend in accordance with the effect they are experiencing from their actions. In other words, people are more likely to take on a task in which they believe they will succeed (8). It follows that caregivers who feel confident that they can successfully triumph over a child’s resistance will be more likely to engage in child restraint use, whereas those who feel that they do not have the self-efficacy to control their child’s behavior to get them into and to remain in a car seat will not even attempt to use a car seat for fear of failure.

In a study that used focus groups, Simpson et al found that one key difference between parents who used child restraints and those who did not seemed to be what they termed “negotiability”(9). Parents who used child restraints for their children commented that they insist every time that their children use the restraint device every time they are in the vehicle, with no exception. Thus use was non-negotiable. One such participant commented “Those kinds of strategies (for convincing a child to comply) aren’t used when it comes to booster seats. Maybe because they know they can’t win the argument.” These users demonstrate a high level of self-efficacy related to enforcing use.

However those who did not use restraint devices for their child commented that they had trouble getting their child into the restraint device and felt that use of enforcing use of a restraint device would be a struggle that was too much for them to handle. One such participant commented “I think it would be a constant struggle with my daughter,and after giving up on its use, “I can’t picture going back to that battle.” Thus use was negotiable for non-users and these parents demonstrated low self-efficacy related to enforcing child restraint use.

Simpson’s findings of the correlation between self-efficacy, parenting style, and child behavior are corroborated by Sanders and Wooley, who studied the relationship between maternal self-efficacy, dysfunctional discipline practices and child conduct problems (11). Task-specific self-efficacy (which includes self-efficacy around specific parenting tasks, as would be relevant to the use of child restraints) was assessed in mothers of 2- to 8-year-old children with conduct problems and mothers with children who did not have conduct problems. Mothers of children with behavior problems reported significantly lower self-efficacy than mothers of control children for all but one of the parenting tasks assessed. In the sample as a whole self-efficacy measures were significant predictors of appropriate maternal discipline style after controlling for other parent, child and risk factors. This study supports the assertion that low maternal self-efficacy is associated with children who display more behavioral problems and that mothers with low self-efficacy use less effective parenting styles which hinder the successful completion of parenting tasks. Thus, parents with low self-efficacy have a double disadvantage in that their children are more difficult to control and they are less effective at controlling them. It follows that parents with low self-efficacy will have children who will more often and more adamantly resist staying in child restraints and that these parents will be less effective at using appropriate parenting skills to force them to use the restraint.

An important factor that affects self-efficacy is prior experience (8). Positive past experiences (successes) will lead to increased self-efficacy while past failures will lower self-efficacy. Therefore caregivers who have tried to adopt any positive behavior on behalf of their child, anything from regular tooth brushing to not hitting others, and perceived their efforts as unsuccessful will have low self-efficacy about influencing their child’s behavior. Therefore, they will not be motivated to try to use child restraint systems, anticipating the same failure with car seat use that they experienced in the past with other healthy interventions. To support this view, Webb, Sanson-Fischer, and Bowman demonstrated that non-users of child restraints were less likely to engage in other preventative health behaviors on their child’s behalf than users of child restraints (7). Thus it is likely that car seat users have high self-efficacy about car seat use that was reinforced by other successes with preventative actions, whereas non-users do not. Education about the importance of child restraint use and the availability of car seat installation, the cornerstones of the NY initiative, will not improve the self-efficacy of a caregiver to motivate them to feel that they will be able to successfully implement car seat use. In my experience, it will only widen the gap between what they know they should do and what they feel they can truly accomplish.

Evidence supports the efficacy of parenting interventions for reducing problematic child behavior. Gardner, Burton, and Klimes tested the effectiveness of the Webster-Stratton Incredible Years video based parenting program for reducing conduct problems in children aged 2-9 years (11). They found post-treatment improvements in both parent-reported and observed identified child problem behavior, increases in the use of positive parenting techniques, and most importantly, improved parent-reported confidence. They concluded that change in observed positive parenting appeared to mediate change in child problem behavior. It follows that if parents were taught positive parenting skills and their self-efficacy improved, they would be better able to facilitate child compliance with restraint use. However, few if any programs exist for increasing child restraint use that included an intervention to improve parental self-efficacy or positive parenting skills .

The second major limitation of the New York State Police imitative is the failure to consider cultural factors that influence child restraint use. Numerous studies have demonstrated that non-whites are significantly less likely to use child restraints than whites (3-6). To illustrate an example, I will focus on unique cultural factors in Latino populations that are barriers to child restraint use. It has been demonstrated that Latino children in Washington State were one third less likely to use a booster seat than non-Latino children. In response, Lee et al conducted a study that examined Latino attitudes and beliefs about booster seats and barriers to booster seat use, using two focus groups with Spanish-speaking parents (12). Lee’s findings echo the barriers to change I have noticed during my 8 years as a provider in the Latino community. Lee identified characteristics of the Latino community such as the view of the man as the household decision-maker, the presence of larger families with many children, and the lack of use of safety belts in their native country as deterrents to child restraint use.

In Latino communities the man is generally considered the decision maker for the household. Numerous studies have demonstrated the men as compared to women as less likely to use safety belts themselves (5 ), less likely to use child restraint systems for child passengers (5), and are less likely to be educated about the importance of restraint systems (12). Thus the decision-makers in Latino communities are less likely understand the importance of or use child restraint systems. Furthermore, females in the Lee study felt that they would not have enough influence to persuade the male householder to use child restraint systems (12). In terms of Social-Cognitive Theory, these Latino women had low self-efficacy surrounding their ability to negotiate child restraint use with their male partners and enforce its use in the face of resistance from male householders. The irony of the NY State Police intervention is that woman are the ones most often in the position of receiving educational messages, since they are often delivered through channels that preferentially engage women—visits to the child’s doctor, interactions with child care providers, commercials during female-centered TV shows. However, in the Latino communities those who receive the intervention are the least likely to be able to affect change given their low self-efficacy in the face of resistance from men.

The next barrier to child restraint use in Latinos is the tendency for Latino families to be large, either through large immediate families or the closeness of extended families. This results in many people trying to fit into one car at the same time. Statistically, Latino families tend to be poorer than Caucasians and less likely to have multiple cars per family. Thus large families must all fit into one car. Anecdotally, as a practicing pediatrician, I have been told by Latino parents that they have not adopted child restraint use because it is easier to fit more people in a car if someone holds the child on their lap rather than take up the extra room with an infant/booster seat. This is echoed by the participants in Lee’s study (12).

Another barrier to use among Latinos, especially more recent immigrants, is the lack of use of child restraints in their native countries (12). For these people, having a child ride unrestrained is the cultural norm. The Social Norms Theory supports that if one perceives that other members of their social network do not engage in a behavior, that person will be less likely to engage in that behavior (13). Thus to affect a change in behavior, one would have to change the perception of what the social norm is for that person’s social network. Unfortunately, the NY State Police educational or media campaigns did not include any component that reframed the use of child restraints as the social norm among any group, let alone in cultural minorities. Furthermore, because child restraints were not used in their native countries, such Latinos have a low perception of risk of injury from lack of child restraint use. Numerous studies have correlated perception of risk of injury to likelihood of child restraint use, with those who perceive low risk of injury less likely to use restraints (6,7,9,12 ).

A further limitation of the New York State Police intervention is the lack of consideration of barriers among those with lower socioeconomic status, both directly related to financial barriers as well other barriers among this population such as lower literacy level. The current guidelines suggest the need for numerous types of child restraint devices over the course of childhood until they are safe to use adult type seat belts (14). It is useful to examine the cost of all of the recommended child restraint devices for one child as they progress from infancy to young adulthood (14). An infant seat is recommended from birth to age 1 year (or > 20 lbs) which ranges from $49.99-$219.99 at Babies R Us (15). From age 1 -4 years (or 20 lbs-40 lbs) a front-facing seat is recommended, costing from $59.99- $199.99 (15). Then from age 4 – 8 years, a booster seat is recommended, costing from $29.99- $269.99 (15). Thus, choosing the least expensive models for each device, the minimum cost of proper restraint from infancy to age 8 is approximately $140. However, this does not reflect an accurate measurement of cost because many other factors increase the actual price. For example, individual models will not fit into every car, and older cars often necessitate the use of more expensive models which have features that adapt to the seat belts found in older cars (14-15). Furthermore, if the child travels in another car, an additional base is needed, which costs another $49.99 – $79.99 (15). Furthermore, NHTSA recommends that a child restraint device be discarded and replaced if the vehicle is in a moderate or severe crash. However, the NHTSA definition of moderate- severe crash is quite loose, such that many crashes would fall under this category (16). Additionally, when there is more than one child who falls into a certain category of restraint device (i.e. two or more children between the age of 4 and 8 years), the parents must purchase an additional device for each additional child. This becomes extremely costly for families with many children or twins/triplets etc.

Additionally, older cars (model years before 1996) often do not have the necessary features to accommodate child restraint use (17). Unfortunately, those of lower SES are more likely to have these older cars. Many have automatic (passive) front safety belts either attached to the door of the vehicle or moving on a track above the door and some have separate lap belts that do not lock tightly around child safety seats. Special parts may be ordered from the vehicle manufacturer to solve some of these problems, but often a charge is incurred for both the parts and the labor for the modification (17). Also, dealerships are generally unfamiliar with ordering these infrequently requested parts (17). Thus even if an older car can be modified to accommodate a child restraint device, additional money and burden are needed to do so.

A final hurdle to use among those with lower socioeconomic status is the presence of low literacy levels which hinder self-education about the importance of and correct use of child restraint devices. Vaca et al demonstrated that lower knowledge about car seat safety was associated with lower income, fewer years of education, and less fluency in English, all of which contribute to lower SES (18).

Populations with lower SES are more likely to have lower literacy levels, and as such, they are less likely to be able to gain knowledge about the importance and proper use of restraints from flyers and written media. Much of the NHTSA campaign centers on giving parents written educational materials or placing flyers or ads in public places. Frequently these pediatric educational materials are written for a reading level far too advanced for most people of low SES. Lingbeil, Speece, and Schubiner found that the majority of pediatric education pamphlets had readabilities of grade nine or above which is far too high for the average low SES reader (19). Furthermore, an evaluation of the readability of car seat safety instruction manuals revealed that they too are written at a reading level that exceeds the reading skills of most American consumers, let alone those with lower literacy levels (20). Even as a person with a post-graduate education, I was unable to clearly understand how to install a car seat using the manual. Therefore, even if the New York State Police initiative was able to convince parents of the need for restraint devices, those of lower SES with accompanying low literacy would not be able to use educational materials or instruction manuals to install and use the seats.

The NHTSA Child Passenger Safety Program in New York State has the opportunity to increase the usage of child restraint devices of considerably more children than it is currently affecting. The lives of hundreds of children each year are on the line. By linking strategies to increase caretaker self-efficacy in the face of child resistance with efforts to increase child restraint use, more caretakers would feel competent to enforce the use of car seats. Addressing cultural factors to reduce barriers in racial and ethnic minorities would also strengthen the campaign. This could be accomplished by targeting men as household decision makers, suggesting reasonable strategies for restraint usage in large families, and reframing the use of child restraint devices as a social norm among their cultural group. Finally, addressing issues of relevance to those of lower SES such as providing free car seats, ensuring free and easy access to refitting older cars to allow restraint use, and adjusting the content of promotional materials to ensure they are at a reading level appropriate to those with lower education would make the messages and behavior modification recommendations more realistic and accessible. If these modifications can be made, the NHTSA campaign still has a chance to make a significant impact on child restraint device use and eliminate car crashes as the greatest killer of children aged 4 to 14 in the U.S.


1. National Highway Safety Administration. Child Passenger Safety Program. National Highway Safety Administration:

2. National Highway Safety Administration. New York State Child Passenger Safety Program. National Highway Safety Administration:

3. Winston F, Chen I, Smith R, Elliot MR. Parent driver characteristics associated with sub-optimal restraint of child passengers. Safety Injury Prevention 2006. 7(4):373-80.

4. Wagenaar AC, Molnar LJ, Margolis LH. Characteristic of child safety seat users. Accid Anal Prev 1988; 20(4):311-22.

5. Voas RB, Fisher DA, Tippetts AS. Children in fatal crashes: driver blood alcohol concentration and demographics of child passengers and their drivers.Addiction 2002;97(11):1439-48.

6. Margolis LH, Wagenaar AC, Molnar LJ.Use and misuse of automobile restraint devices. Am J Dis Child 1992;146(3):361-6.

7. Webb GR, Sanson-Fisher RW, Bowman JA. Psychosocial factors related to parental restraint of preschool children in motor vehicles. Faculty of Medicine, University of Newcastle, Australia 2002;

8. Wikipedia. Social Cognitive Theory. Wikipedia:

9. Simpson EM, Moll EK, Kassam-Adams N, Miller GJ, Winston FK. Barriers to booster seat use and strategies to increase their use. Pediatrics 2002;110(4):729-36.

10. Sanders, M R. Woolley, M L. The relationship between maternal self-efficacy and parenting practices: implications for parent training. Child Care, Health & Development 2005;3:65-73.
11. Gardner F, Burton, J, Klimes I.Randomised controlled trial of a parenting intervention in the voluntary sector for reducing child conduct problems: outcomes and mechanisms of change. Journal of Child Psychology & Psychiatry & Allied Disciplines. 2006; 47(11):1123-32.

12. Lee JW, Fitzgerald K, Ebel BE. Lessons for increasing awareness and use of booster seats in a Latino community. Inj Prev 2003; 9(3):268-9.

13. Social Norms Theory. Higher Ed Center. Social Norms Theory:

14. NHTSA. Child Passenger Saftey. NHTSA: http: // / people/injury/ childps/ AreYouUsing/pages/GeneralCSSUse.html

15. Babies R US. Car Seats. Babies R US:

16.National Highway Traffic Safety Administration. Car Seat Reuse . National Highway Transportation Safety Administration:


17. National Highway Traffic Safety Administration. Car Seat Addon. National Highway Transportation Safety Administration: injury/childps/Add-On04/index.htm

18. Vaca et. al. Child safety seat knowledge among parents utilizing emergency services in a level I trauma center in southern California. Pediatrics 2002; 110(5): e61.

19. Lingbeil C, Speece M W, Schubiner H. Readability of pediatric patient education materials:Current perspectives on an old problem.Clinical Pediatric 1995;34(2):96-102.

20. Wegner MV,Girasek DC.How readable are child safety seat installation instructions? Pediatrics 2003;111(3):588-91.

Labels: , ,


Post a Comment

Subscribe to Post Comments [Atom]

<< Home