Article Text
PDF
Abstract
Objective To assess the impact of a booster seat law in Wisconsin on booster seat use in relation to race, ethnicity and socioeconomic status.
Methods A longitudinal study in Milwaukee County, Wisconsin, involving repeated direct observational assessments of booster seat use rates by child passengers aged 4–7 years over five time periods, before and after legislation mandating booster seat use.
Results Overall, booster seat use increased from 24% to 43%, whereas proper restraint use increased pre to post-legislation from 21% to 28%. Proper use increased after legislation in white, but not in black or Latino children. White individuals had a proper booster use increase from 48% to 68% over the time period of the study. Black children's proper use dropped from 18% to 7% over the study period and Latino children's proper use rates were stable at 10%. Driver-reported household income had a significant impact on overall use, but not on proper use.
Conclusions Racial/ethnic minority groups and those of lower socioeconomic status have significantly lower use and proper use of booster seats. Legislation may increase the total use of booster seats but not necessarily the correct use of the restraint, particularly in racial/ethnic minorities.
- Child
- legislation
- passenger
- restraint
- socioeconomic status
Statistics from Altmetric.com
Motor vehicle crash injuries are a leading cause of death for children over the age of 3 years in the USA.1 Belt-positioning booster (BPB) seats are well documented to reduce injuries in children 4–7 years of age.2 Socioeconomic and racial disparities exist for both the use of booster seats and for motor vehicle crash-related morbidity and mortality of children.3–5
Children involved in a motor vehicle crash in a state with a booster seat law are 39% more likely to be restrained in a BPB seat than children living in states without booster seat laws.6 A study of BPB legislation in Tennessee found less impact of the legislation on black than on white children; compared with pre-law BPB use, post-law use increased by a statistically significant 49% in white individuals but only by a non-significant 31% in black individuals.7 Proper use of booster seats is important to document when assessing improvement in use rates. Although proper restraint use reduces injuries for children,8 improper use decreases the effectiveness of the device, resulting in a higher risk of injury.9–11 This study evaluated the impact of strengthened child passenger safety legislation, without a comprehensive community campaign, on different racial groups in an urban population with a particular focus on proper BPB use. We hypothesised BPB legislation would result in minimal improvement in the proper use of BPB seats in children in a metropolitan centre and proper use would vary by racial and socioeconomic groups.
Methods
Survey sites
We conducted a prospective, non-randomised pre-post direct observational study of BPB seat use by means of convenience sampling of vehicles transporting children aged 4–7 years using the methodology employed by the National Highway Traffic Safety Administration (NHTSA) in its observational studies.12 To expand demographic diversity, observations were at multiple urban daycare and community centres and at the primary children's hospital. The locations were chosen from institutions with more than 10 children 4–7 years of age served and a safe area to conduct interviews in the parking lot. Bilingual observers were utilised at centres with primarily Spanish-speaking clients. Cars were approached by a minimum of two observers simultaneously after the children had been placed in the vehicle and as the vehicle was leaving the parking lot or when the vehicle was initially entering the parking lot. Drivers were asked if there was a child in the car aged 4–7 years. If yes, they were asked to participate in the confidential research study survey that took less than 1 min to complete. All participants provided verbal consent; the protocol was approved by the Children's Hospital of Wisconsin Institutional Review Board.
Data collection
Wisconsin enacted a booster seat law in June 2006, which applies to children 4–7 years of age and mandated back seat placement if one was available. The law included a 6-month grace period from 1 June to 31 December 2006 during which no fines were imposed. On 1 January 2007, fines were imposed for infractions. Observations were conducted during the following periods: (1) ‘time period 1, pre-law,’ June 2005, a year before the implementation of the law; (2) ‘time period 2, immediate pre-law,’ May 2006, immediately before law implementation; (3) ‘time period 3, post-law,’ September 2006, post-law but before fine implementation; (4) ‘time period 4, post-law’ October 2007, 16 months post-law implementation; and (5) ‘time period 5, post-law’ October 2008, 28 months post-law implementation. The observations conducted in 2005, 2007 and 2008 included an assessment of misuse of the BPB seats to distinguish improper use from proper use. The 2006 observations only assessed any use and did not include an assessment of misuse. A team of three certified child passenger safety technicians from the firm Program Professionals performed all observations except for those in June 2005, which were performed by 51 trained community volunteers. The training and methods for the year 2005 observation have been described previously.5
Survey tool
A survey tool modified from the NHTSA survey instrument was used. Information obtained included a classification of the vehicle type, the residential ZIP code of the driver, the driver's age, gender, race (categorised as white, black, Latino or other/unknown), and observed use of restraint by the driver. Each passenger was identified by their seating position, age, gender, observed race and restraint type. During time periods 1, 4 and 5, observers recorded the misuse of restraints to assess the proper use of BPB seats.
Variable definitions
Two outcomes were examined, any use of BPB seats and proper use of BPB. Any use of a BPB seat for 4–7 year olds, consistent with the methodology of NHTSA and other experts in the field, included high-back and low-back boosters as well as combination seats used as boosters. Analysis of data from 2005, 2007 and 2008 divided any restraint use into proper or improper to assess any shift from no use of a BPB to having and using a BPB, but not using the restraint in a manner that would maximise child passenger restraint protection. Proper use required the correct fastening and positioning of the lap and shoulder belt in front of the body. Improper use included using a BPB with the shoulder strap placed behind the back or under the arm and the use of a lap belt only. Children restrained in forward or rear-facing car seats were excluded from analysis because the brief interaction with the driver did not allow time to gather sufficient details to determine whether the restraint was appropriate for the child.
Neighbourhood and income definitions
Residential ZIP code median income was used to examine possible ‘neighbourhood’ or living environment effects. ZIP codes were divided into those with a 1999 median income less than US$40 000 or greater than or equal to US$40 000 or unknown, based on the 2000 federal census. Self-reported household income categories were divided arbitrarily into less than US$30 000, US$30 000–39 999, US$40 000–49 999, US$50 000 or more and unknown, as used in other published studies.5 13
Statistical analysis
χ2 Tests were used to compare outcomes among unordered categorical variables. The Cochran–Armitage trend test was used to perform bivariate examination of a trend in restraint use over pertinent time periods (all five for any use; time periods 1, 4 and 5 for proper use).14 Multivariate logit regression was used to test for the simultaneous, independent effects of race, household income and neighbourhood on the outcomes of any use and proper BPB use, during time periods 4 and 5. Time periods 4 and 5 were the only observation time periods when respondent-reported household income was available, in addition to the median income of the ZIP code of residence (which was collected in all five study periods). The interaction of neighbourhood and household income on BPB use was examined, but was not significant and was removed from the models. Data for periods 4 and 5 were combined for the multivariate logit regression because results were not significantly different between the two periods. To account for child passengers being nested within a single vehicle and vehicles nested within observation sites (clustering), we used a three-level random-intercept logit regression model.15 The final models, containing only the three covariates race, neighbourhood and reported income, were used to produce in-sample estimates of booster use for subgroups of race, neighbourhood and reported income.
Results
A total of 1775 children aged 4–7 years was observed in 1540 vehicles over the five time periods (table 1). The range of observed children was 263 (time period 1) to 470 (time period 4). The racial/ethnic makeup of the observed sample was 33% black, 24% Latino, and 37% white. Fewer Latino and more white children were observed during time periods 2 and 3, the 2006 immediate pre and post-law observations. As expected, demographic diversity varied between the children's hospital and the community sites (table 2).
Children 4–7 years old observed for BPB seat use by time period and race
Demographics of children observed at hospital and daycare/community centres combined over five time periods
Overall, children 4–7 years of age showed some improvement in the use of BPB seats following legislation. From time periods 1 to 5, the use of BPB seats improved from 24% to 43%, with a peak of 51% in time period 3, immediately post-legislation (Cochran–Armitage trend test, p=0.01; figure 1). Most of the improvement occurred in white children who had observed BPB use rates of 56% in time period 1 and 83% during time period 5 (Cochran–Armitage trend test, p<0.00001). Black individuals showed no change in use of a booster seat despite the implementation of legislation, with use rates of 19% in time period 1 and 21% in time period 5. Latino individuals showed a modest, non-significant, increase in use rates, with a change from 14% in time period 1 to 23% in time period 5 (Cochran–Armitage trend test, p=0.3).
Any use of belt-positioning booster seats over time by racial/ethnic grouping.
Proper BPB use increased slightly following the legislation, with proper use increasing from 21% in time period 1 to 28% in time period 5 (Cochran–Armitage trend test, p=0.05; figure 2 and table 3). The percentage of children improperly restrained increased (from 4% to 15%) as the percentage of unrestrained children decreased from (76% to 57%) from time periods 1 to 5. There were racial and ethnic differences identified in proper BPB use. White children had proper use rates of BPB seats at 48% in time period 1, which rose to 68% in time period 5 (Cochran–Armitage trend test; p=0.02). Black individuals had a proper use rate of 18% in time period 1 which dropped to 7% in time period 5 (Cochran–Armitage trend test; p=0.02). Latino individuals had proper use of BPB seats at a rate of 10% in both time period 1 and time period 5.
Rate of proper belt-positioning booster seat use over time among all children 4–7 years by race.
Percentage of observed proper and improper use of BPB seats over time
Among children using BPB seats, it was not just proper BPB use that showed a slight increase, but improper use increased from time period 1 to time period 5 (16% and 35%, respectively; Cochran–Armitage trend test; p=0.006). There were also racial and ethnic disparities in the rate of improper BPB use among those using a booster seat. When time periods 1, 4 and 5 were combined, white individuals had an 18% rate of improper BPB use, whereas black and Latino individuals had 41% and 42% rates of improper use, respectively (χ2 tests; white vs black and white vs Latino; both p<0.0005).
In multivariate models controlling for race, neighbourhood income and household income, different effects were seen for the outcomes of any BPB use and proper use. Household income and race had significant independent effects on estimated booster use among all 4–7 year olds, while controlling for neighbourhood (figure 3). Children riding with drivers reporting an annual income of less than US$50 000 were 90–94% less likely to be using a booster seat than those reporting an income of US$50 000 or more. Within the subset of 4–7 year olds using a booster seat, race, median neighbourhood income and household income did not reach significance in predicting proper use, although race and neighbourhood income showed differences at the p≤0.15 level (figure 4). Children living in lower median income ZIP code areas (<US$40 000) had lower proper usage of booster seats than those in the higher median income areas (OR 0.29; 95% CI 0.05 to 1.58). Black children were 85% less likely to be properly restrained (OR 0.15; 95% CI 0.01 to 1.54) and Latino children were 58% less likely to be properly restrained (OR 0.42; 95% CI 0.07 to 2.58).
Percentage of any belt-positioning booster use (proper or improper) by neighbourhood income, household income, and race during time periods 4 and 5. Note: Data for all 4–7-year-old children from time periods 4 and 5 only; multivariate logit regression model controlled for race, reported household income, neighborhood income derived from year 2000 federal census median ZIP code income, and accounted for nesting of children within vehicles and vehicle within site.
Percentage of proper use among children 4–7 years using a belt-positioning booster (BPB) seat, by neighbourhood income, household income, and race. Note: Data for 4–7-year-old children using a BPB from time periods 4 and 5 only; multivariate logit regression model controlled for race, reported household income, neighbourhood income derived from year 2000 federal census median ZIP code income, and accounted for nesting of children within vehicles and vehicle within site. The estimated percentages are higher than those for figure 3 because this analysis was restricted to the subset of children using a booster seat.
Discussion
Implementation of Wisconsin legislation mandating BPB use was associated with a significant increase in any use of BPB for the general population in metro Milwaukee, but limited improvement in proper use. The legislation was associated with increased racial and ethnic disparities in total and proper BPB use. The proportion of children properly restrained varied by race/ethnicity and neighbourhood, with black and Latino children less likely to be properly restrained.
Other studies have shown safety belt misuse among children using booster seats.16–18 The work of O'Neil et al18 identified a high frequency of safety belt misuse, almost 65%, for children in BPB. Although O'Neil et al18 identified younger drivers, younger children and children who weighed less then 40 lb to be at increased risk of improper booster seat use, safety belt misuse for children driven by those who were non-white and with lower income and education levels was also identified. This study is the first to show that legislation intended to increase BPB use is associated with an increase in racial/ethnic disparities in proper BPB use within metro Milwaukee. Causation should not be assumed in this non-controlled study and further investigation is warranted.
The NHTSA reported an increase in the proportion of children aged 4–8 years using BPB in Wisconsin following the legislation.12 Our research suggests NHTSA's measurement of any BPB use, including improper use, may overestimate the benefits of legislation, particularly among minorities. Racial differences in rates of children unrestrained after legislation have been reported previously.4 7 19 Our study suggests that racial differences in proper BPB use were not just present after legislation but widened compared with pre-legislation disparities. Improvement in the rate of the overall use of booster seats may provide false security that legislation alone is improving the safety of child passengers when proper use does not increase to the same degree in all subgroups and may, in poor and minority children, actually decrease following legislation.
We found residence in a ZIP code area with a median household income less than US$40 000 was associated with a child being 71% less likely to be properly restrained in their BPB. While not reaching significance, this neighbourhood effect, which was controlled for the impact of race and household income, suggests the environmental or cultural aspect of the neighbourhood may be an important determinant of proper BPB use. Social norms, a concept that refers to implicit standards of behaviour inferred by individuals from the behaviour of others in their social milieu,20 might be the mechanism accounting for the neighbourhood influence. Social norms have been useful in understanding behaviours in other injury control fields.21 It is not well understood why certain populations modify their health behaviour at varying rates (or not at all) despite legislation. It has been well established that BPB rates can be significantly improved in the context of a comprehensive community campaign, steeped in behavioural theory and providing the public health message through social marketing.22 Using qualitative data, it has been shown that while legislation is important, other factors such as safety may serve as the primary motivator to use a BPB.23 Wisconsin's booster law did not directly include funding for educational programming. Culturally appropriate, community-based education that is sensitive to social norms as an adjunct to legislation needs to be implemented and evaluated within Wisconsin. Ultimately, a comprehensive approach of engaging behavioural science theories and methods at the community as well the individual level may be needed to increase the proper use of BPB seats in marginalised groups.
In addition, legislation will not work unless it is effectively enforced. City of Milwaukee police issued 151 child safety seat restraint violation citations in 2006, 153 in 2007 and 225 in 2008 for 4–7 year-old children compared with 434, 484 and 525 citations, respectively, for infants to age 3 years (personal communication, City of Milwaukee Police Department). There was a minimal increase in police enforcement of this new legislation at the time of the study. This may reflect a need to educate law enforcement officers about how best to enforce this legislation.
A limitation of this research, similar to those in other child passenger observational studies, is that convenience samples, while allowing the study to focus on areas with high representation of booster seat use, may not be representative of a wider geographical area. This study was not controlled or randomised, with relatively small sample sizes in some subgroups, so there is the possibility of historical bias and unmeasured confounders. There may have been a general increase in BPB use due to factors other than the passage of the law. Due to the collection of data across multiple sites over several years a specific cohort of children was not followed and the racial distribution varied. Observation of children after they were placed in the vehicle and restrained, or before being released from their restraint, allowed for accurate assessment of the ‘true’ use of child restraints, but did require brevity in the number of questions asked so as not to disrupt traffic flow at the sites. This limited questioning about other demographic variables such as education level and exposure to educational information about BPB, which may have been helpful in secondary analyses. Although race was asked and self-reported frequently, at times it was observed due to the brief interaction and the need to maintain traffic flow. Therefore, observed race was used in some instances and may be less accurate then self-reported race. The 2006 observations did not collect misuse data, preventing the inclusion of those data in the examination of proper use. The legislated exemption of the use of a BPB if over 80 lb could not be accounted for as weight was unknown. Annual household income was assessed in time periods 4 and 5 but not in time periods 1–3 so reported income could only be examined in periods 4 and 5. The city where observations took place has a 75% dissimilarity index indicating significant segregation in living situations.
In conclusion, Wisconsin's strengthened child passenger safety legislation did increase the total use of BPB, but gains were not equally shared by minority families living in poorer neighbourhoods. Black and Latino children did not show significant benefit from legislation mandating BPB use and disparities in use widened. Expanding research on interventional public health laws is critical as more states use injury prevention legislation to improve BPB use rates.24 Analysis of appropriate child passenger restraint use, with a focus on proper use, is vital to determine if all children are benefitting from the passage of injury prevention legislation. Culturally appropriate behavioural interventions and education, such as using peer leaders within communities to support the message of child passenger safety, need to be developed and evaluated in an effort to improve proper BPB seat use if all children are to benefit from child passenger restraint regulations.
What is already known on the subject
BPB seats save lives when used properly.
Legislation is an important tool for injury prevention.
What this study adds
Legislation mandating BPB seats may result in limited improvement in proper use.
Black and Latino children did not show significant benefit from legislation mandating BPB use and disparities in use widened.
Acknowledgments
The authors are grateful to Program Professionals and the families of Milwaukee that participated in this research.
References
Footnotes
Funding This work was funded in part by the Medical College of Wisconsin Injury Research Center (CDC grants R49/CE001175 and R49/CCR519614) and in part by the Healthier Wisconsin Partnership Program, a component of the Advancing a Healthier Wisconsin Endowment at the Medical College of Wisconsin.
Competing interests None.
Patient consent Obtained.
Ethics approval Ethics committee approval was obtained from the Children's Hospital of Wisconsin Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.