Shifts in Child Restraint Use According to Child Weight in the United States from 1999 to 2002

Abstract

From 1999 to 2002, 32% fewer US children between 9 and 36.4 kg (20–80 lb) were restrained inappropriately in seat belts and the most prevalent form of restraint shifted from seat belts to child restraints with harnesses. There was a significant increase in the use of combination child restraint/booster seats with harnesses by children 9.1–18.1 kg (20–40 lb). Among children weighing 14.1–18.1 kg (31–40 lb), the inappropriate use of seat belts and shield boosters decreased. Among children weighing 18.6–27.2 kg (41–60 lb), the use of belt-positioning booster seats increased while the inappropriate use of seat belts decreased. Of note, by the end of 2002, 27% of children weighing between 18.6 and 22.7 kg (41–50 lb) were restrained in child restraints with harnesses. These children were of weights typically above the manufacturer’s recommended limit. Despite progress, substantial inappropriate restraint still remains and continued investment in outreach efforts is necessary. The risk of injury for heavier children in child restraints with harnesses should be monitored.

Multiple private organizations and government agencies in the United States implemented public health and educational campaigns to promote appropriate restraint of children in motor vehicles. Appropriate restraint, as defined by the American Academy of Pediatrics (AAP, 2003) and the US National Highway Traffic Safety Administration (NHTSA, 2003), involves the use of forward-facing child restraint systems with harnesses for children over 1 year and 9.1 kg (20 lb) until the upper weight limit of the child restraint (typically 4 years old and 18.1 kg/40 lb). Children over 18.1 kg (40 lb) and under 36.3 kg (80 lb) are recommended to use belt-positioning booster seats after they outgrow their convertible child safety seats (typically 4 to 8 years old). The AAP does not recommend shield boosters for use as child restraints while the NHTSA has certified shield boosters for use only by children between 13.6 and 18.1 kg (30– 40 lb).

Recent trends in child restraint use in the United States have revealed dramatic increases in the use of child restraints with harnesses for children over the age of 4 years. (Winston et al, in review) As the weight range for these children can vary considerably, (CDC, 2003) it is important to assess how these restraint trends vary by weight of the child. Few child restraint systems manufactured in the United States are rated for use with a harness over 18.1 kg (40 lb). Thus, the purpose of this study was to examine the trends in types of child restraints used by children between 9.1 and 36.3 kg (20–80 lb).

METHODS

STUDY POPULATION AND DATA COLLECTION

Data were collected from December 1, 1998 to November 30, 2002. A description of the study methods has been published previously (Durbin et al, 2001). The project consists of a large scale, child-specific crash surveillance system: insurance claims from State Farm Insurance Co. (Bloomington, IL) function as the source of subjects, with telephone survey and on-site crash investigations serving as the primary sources of data.

Vehicles qualifying for inclusion were State Farm-insured, model year 1990 or newer, and involved in a crash with at least one child occupant <15 years of age. Qualifying crashes were limited to those that occurred in fifteen states and the District of Columbia, representing three large regions of the United States (East: NY, NJ, PA, DE, MD, VA, WV, NC, DC; Midwest: OH, MI, IN, IL; West: CA, NV, AZ). After policyholders consented to participate in the study, limited data were transferred electronically to researchers at The Children’s Hospital of Philadelphia and University of Pennsylvania. Data in this initial transfer included contact information for the insured, the ages and genders of all child occupants, and a coded variable describing the level of medical treatment received by all child occupants as reported by the policyholder (no treatment, physician’s office or emergency department only, admitted to the hospital, or death).

A stratified cluster sample was designed in order to select vehicles (the unit of sampling) for the conduct of a telephone survey with the driver. Vehicles containing children who received medical treatment following the crash were over-sampled so that the majority of injured children would be selected while maintaining the representativeness of the overall population. If a vehicle was sampled, all child occupants in that vehicle were included in the survey. Drivers of sampled vehicles were contacted by phone and, if medical treatment had been received by a passenger, screened via an abbreviated survey to verify the presence of at least one child occupant with an injury. All vehicles with at least one child who screened positive for injury and a 10% random sample of vehicles in which all child occupants who were reported to receive medical treatment but screened negative for injury were selected for a full interview; a 2.5% sample of crashes where no medical treatment was received were also selected. The full interview involved a 30-minute telephone survey with the driver of the vehicle and parent(s) of the involved children. Only adult drivers and parents were interviewed. The median length of time between the date of the crash and the completion of the interview was six days, with 95% of interviews completed within 47 days of the crash.

For the current analysis, the study sample consisted of vehicles with child occupants age 8 and younger: 8,730 vehicles with 10,195 children. About 8.7% (n=891) of the interviews on study children did not contain child weight information and 6.6% (n=670) of the study children were under the weight of 9 kglb; these children were excluded from the analyses. This resulted in a sample of 8,634 children in 6,420 crashes.

VARIABLE DEFINITIONS

Restraint status of children was determined from telephone survey responses to detailed questions about the description of the restraint and how it was used at the time of the crash. Children were classified as unrestrained or restrained, with the restraint type further classified as child safety seat (either forward- or rear-facing), combination child restraint/booster with harness, belt-positioning booster, shield booster, and seat belt. Rear-facing child safety seat included infant seats designed for rear-facing only (many of these seats have a handle to carry the seat) and rear-facing convertible seats (these seats are designed to be used either in a rear- or forward-facing orientation). Forward-facing child safety seat included forward-facing convertible seats with harness. Combination child restraint/boosters with harness are designed as a forward facing child safety seats with the harness and a belt positioning booster without the harness. Belt-positioning boosters included high-back and backless booster seats and those combination seats used without the harness. Shield boosters were seats that were installed in the vehicle with the lap portion of the adult seat belt and have a padded shield, as opposed to a harness, to keep the child in place. The restraint classification seat belt included lap only, lap/shoulder, and shoulder only seat belts.

The following are the current recommended typical weight limits for each restraint type (AAP, 2002; NHTSA, 2002). Rear-facing child safety seats include rear-facing only and rear-facing convertible seats. Rear-facing only child safety seats (infant seat) are certified for use up to 9.1–10 kg (20–22 lb). Most rear-facing convertible seats can accommodate a child weighing up to 15.9 kg (35 lb). Most forward-facing convertible seats and combination child restraint/booster seats with harness are recommended to accommodate a child weighing between 9.1 kg (20 lb) and 18.1 kg (40 lb). Shield booster seats are designed for children under 18.1 kg (40 lb). Most belt-positioning booster seats are for children weighing between 18.6 kg (41 lb) and 36.3 kg (80 lb). Seat belt use is recommended for children over 145 cm (57 in) and over approximately 36.7 kg (80 lb).

Separate verbal consent was obtained from eligible participants for the transfer of claim information from State Farm to CHOP/Penn, for the conduct of the telephone survey, and for the conduct of the crash investigation. The study protocol was reviewed and approved by the Institutional Review Boards of both The Children’s Hospital of Philadelphia and The University of Pennsylvania School of Medicine.

DATA ANALYSIS

The primary purpose of these analyses was to examine trends in restraint use by children <9 years of age in 5 different weight groups: 9.1–13.6 kg (20–30 lb), 14.1–18.1 kg (31–40 lb), 18.6–22.7 kg (41–50 lb), 23.1–27.2 kg (51–60 lb), and over 27.7 kg (61 lb +). Because sampling was based on the likelihood of an injury, subjects least likely to be injured were underrepresented in the study sample in a manner potentially associated with the type of restraint used. To account for this potential bias, and to adjust for the clustering of subjects by vehicle, analytical methods were used to account for sampling weights, sampling strata, and sampling units. The sample weights account for the differential sampling rates, defined as the inverse of the inclusion probability of the child in the sample. The weighted data with the sample weight can reflect the population. Thus the estimates can be obtained in an approximately unbiased manner. Due to the complex sample design, we used SAS-callable SUDAAN: Software for the Statistical Analysis of Correlated Data, Version 7.5 (Research Triangle Institute, Research Triangle Park, NC) for all data analyses.

Descriptive statistics, consisting of frequencies for categorical variables, were determined. In order to describe the relationship of restraint use and time, we modeled the probabilities of restraint use (predicted) as a function of time using logistic regression models for the 5 weight groups. The dependent variables examined were child safety seat use, combination child restraint/booster seat with harness use, shield booster seat use, belt –positioning booster seat use, and seat belt use. Eight six-month periods (from Dec. 1, 1998 to Nov. 30, 2002) were used as the independent variables to assess the time effects. Statistical tests were performed for the significance of the time effects. Results were plotted as probabilities of restraint use from the beginning to the end of the study period for each weight group.

RESULTS

Table 1 displays the distribution of restraint type by weight group averaged across the 4 years. The distribution of restraint type varied by weight group and very few young children were unrestrained. Therefore, the sample for the remaining analyses included only those children reported restrained.

Table 1.

Restraint Type by Weight Group in Crashes (N=8,634 crashes?)

Restraint Type/Weight	Child Safety Seat Weighted % (Unweighted N)	Combination Child Restraint/Booster with Harness Weighted % (Unweighted N)	Shield Booster Seat Weighted % (Unweighted N)	Belt Positioning Booster Seat Weighted % (Unweighted N)	Seat Belt Weighted % (Unweighted N)	Unrestrained Weighted % (Unweighted N)
9.1–13.6 kg (20–30 lb)	77.6 (1504)	14.3 (229)	3.0 (67)	1.2 (21)	3.7 (87)	0.4 (19)
14.1–18.1 kg (31–40 lb)	34.8 (716)	21.1 (370)	9.1 (197)	9.2 (171)	24.2 (656)	1.6 (57)
18.6–22.7 kg (41–50 lb)	7.5 (109)	5.8 (84)	3.1 (58)	13.7 (166)	68.1 (1450)	1.8 (86)
23.1–27.2 kg (51–60 lb)	2.4 (18)	2.0 (11)	0.9 (10)	5.9 (35)	86.9 (1029)	1.9 (48)
27.7 kg + (61 lb +)	2.4 (18)	2.2 (13)	0.8 (8)	1.8	89.9 (1226)	2.9 (81)

Note: due to rounding and unknown responses, the percentage by weight group may not equal to 100% (total number). The following parentheses shows the appropriate restraint type for each weight group: 9.1–13.6 kg/20–30 lb (forward facing child safety seat); 14.1–18.1kg/31–40 lb (forward facing child safety seat); 18.6–22.7 kg/41–50 lb (belt-positioning booster seat); 23.1–27.2 kg/51–60 lb (belt-positioning booster seat); 27.7–36.3 kg/61–80 lb (belt-positioning booster seat); 36.7 kg +/81 lb + (seat belt).

Overall (9.1–36.3 kg/20–80 lb)

Figure 1 shows the trends for the five restraint types over the study period for the overall weight group of 9.1–36.3 kg (20–80 lb). At the end of the study period, 56% of children in this weight group were restrained in a child restraint, either a child safety seat or a booster seat. From Dec. 1, 1998 to Nov. 30, 2002, the probability of child safety seat use remained stable. However, over the time period of the study, the probability of combination child restraint/booster seat with harness use increased from 0.06 to 0.18 (p≤0.001) while the probability of belt-positioning booster seat use increased from 0.03 to 0.12 (p≤0.001). Conversely, the probability of shield booster use decreased from 0.07 to 0.03 (p≤0.001) while the probability of seat belt use decreased from 0.52 to 0.33 (p≤0.001).

Fig. 1.

Probability of restraint use by type: 9.1–36.3 kg/20–80 lb⁺

9.1–13.6 kg / 20–30 lb

(Appropriate restraint: child safety seat or combination booster seat with harness) Figure 2 shows the trends in the distribution of restraint types used for the weight group of 9.1–13.6 kg (20–30 lb). Overall, at the end of the study period, 96% of children in this weight group were restrained in a child restraint (either a child safety seat or a booster seat). The low probability of either shield booster seat use or seat belt use remained stable for this weight group. However, the probability of combination child restraint/booster seat with harness use increased from 0.10 to 0.18 (p≤0.001). Concurrently, the probability of belt-positioning booster seat use increased from 0.004 to 0.02 (p=0.08). Conversely, the probability of child safety seat use decreased from 0.81 to 0.72 (p=0.08).

Fig. 2.

Probability of restraint use by type: 9.1–13.6 kg/20–30 lb⁺

⁺ Child Safety Seat (CSS); Child restraint/booster with harness (CR/BH); Belt-positioning booster (BPB); Shield booster (ShB)

* = p<0.05; ** = p<0.01; *** = p<0.001

14.1–18.1 kg / 31–40 lb

(Appropriate restraint: child safety seat or combination booster seat with harness) Figure 3 shows the trends in the distribution of restraint types used for the weight group of 14.1–18.1 kg (31–40 lb). Overall, at the end of the study period, 74% of children in this weight group were restrained in a child restraint (either a child safety seat or a booster seat). The probability of child safety seat and belt-positioning booster seat use remained stable. However, the probability of combination child restraint/booster seat with harness use increased from 0.13 to 0.29 (p≤0.001). Conversely, the probability of shield booster seat use decreased from 0.18 to 0.04 (p≤0.001) while the probability of seat belt use decreased from 0.35 to 0.16 (p≤0.001).

Fig. 3.

Probability of restraint use by type (14.1–18.1 kg/31–40 lb)⁺

18.6–22.7 kg / 41–50 lb

(Appropriate restraint: belt-positioning booster seat) Figure 4 shows the trends in the distribution of restraint types use for the weight group of 18.6–22.7 kg (41–50 lb). Overall, at the end of the study period, 30% of children in this weight group were restrained in a child restraint (either a child safety seat or a booster seat). The low probability of shield booster seat use remained stable. However, the probability of combination child restraint/booster seat with harness use increased from 0.01 to 0.15 (p≤0.001); the probability of child safety seat use increased from 0.03 to 0.13 (p= 0.003); and the probability of belt-positioning booster seat use increased from 0.05 to 0.25 (p≤0.001). Conversely, the probability of seat belt use decreased from 0.86 to 0.46 (p≤0.001).

Fig. 4.

Probability of restraint use by type (18.6–22.7 kg/41–50 lb) ⁺

23.1–27.2 kg / 51–60 lb

(Appropriate restraint: belt-positioning booster seat) Figure 5 shows the trends in the distribution of restraint types used for the weight group of 23.1–27.2 kg (51–60 lb). Overall, at the end of the study period, 11% of children in this weight group were restrained in a child restraint (either a child safety seat or a booster seat). The low probability of combination child restraint/booster seat with harness and child safety seat use remained stable. The probability of belt-positioning booster seat use increased from 0.004 to 0.21 (p≤0.001), while the probability of shield booster seat use increased from 0.000 to 0.06 (p=0.052). Conversely, the probability of seat belt use decreased significantly from 0.96 to 0.73 (p≤0.001).

Fig. 5.

Probability of restraint use by type: 23.2–27.3 kg/51–60 lb⁺

27.7 kg + (61 lb +)

(Appropriate restraint: belt-positioning booster seat) Figure 6 shows the trends in the distribution of restraint types used for children weighing more than 27.7 kg (61 lb). Overall, at the end of the study period, 7% of children in this weight group were restrained in a child restraint (either a child safety seat or a booster seat). The low probability of belt-positioning booster seat and child safety seat use remained stable over the study period. However, the probability of combination child restraint/booster seat with harness use increased from 0.004 to 0.06 (p=0.039). The probability of shield booster seat use decreased from 0.02 to 0.002 (p=0.024). The probability of seat belt use decreased from 0.93 to 0.86 (p=0.088).

Fig. 6.

Probability of restraint use by type: over 27.7 kg/61 lb⁺

DISCUSSION

Our research findings demonstrate a 33% increase (from 48% to 64%) in child restraint use (i.e., child safety seats, combination child restraint/booster seats with harness, belt-positioning booster seats, and shield booster seats) for children between 9.1 and 36.3 kg (20–80 lb) between early 1999 and the end of 2002 with a concomitant 32% decrease in seat belt use. Not only was there a dramatic shift in child restraint use but also the types of child restraints used changed as well. At the beginning of the study, the majority of children in this weight group (50%) were in seat belts while 34% were in child restraints with harness (both child safety seats and combination child restraints/booster seats with harness). By the end of the study, about half of the children in this weight group (47%) were in child restraints with harness while 35% of them used seat belts as their restraint.

A shift toward more appropriate child restraint was noted in all weight groups. The introduction of combination child restraint/booster seats with harnesses in the late 1990s was evidenced by significant increases in their use, both their appropriate use in children under 40 pounds but also beyond the typical manufacturer-recommended weight limit of 40 pounds. Belt-positioning booster seat use increased substantially. Conversely, the inappropriate use of seat belts and shield boosters decreased significantly over the time period of study.

For children weighing between 18.6 and 22.7 kg (41– 50 lb), child restraint use increased from 14% to 49%. The use of belt-positioning booster seat use in this weight group increased from 5% to 17% but so too did the use of child safety seats with harnesses (child safety seats and combination child restraint/boosters with harness) from 6% to 27%. By the end of the study period, the majority (55%) of children in this weight group using a harness type child restraint were in a combination child restraint/booster with harness.

Of note, shield booster use decreased from 15% to 3% for children weighing between 14.1 and 18.1 kg (31– 40 lb). It is important to highlight that, based on current recommendations for appropriate restraint of children, many children using shield booster seats were not, in fact, restrained appropriately. Shield booster seats have been decertified by the NHTSA for all children except those weighing between 14.1 and 18.1 kg (31–40 lb). Even in this weight group, the AAP recommends the use of a full harness child restraint as current best practice. According the AAP, shield boosters do not provide enough upper body protection and have a risk of ejection from the shield booster in a rollover crash (AAP, 2003). Our results show that over the time period of study this safety message appears to have been accepted by parents of young children.

The NHTSA began a standardized child passenger safety training and certification program in March 1998. By 2000, over 16,000 individuals had been certified as child passenger safety technicians (personal communication, Lori Miller, NHTSA, 2000). These individuals have participated in thousands of community-based child safety seat clinics and have been a source of information on appropriate restraint guidelines. Further, the NHTSA made booster seat use the focus of its Child Passenger Safety Week educational campaign in February 2000. Currently, the campaign ‘4 Steps for Kids’ conducted by NHTSA is operating to promote the appropriate use of each type of restraint. State Farm Insurance Co. has conducted a nationwide education campaign of its 35 million policyholders via a mass mailing of information on booster seats. Ford Motor Co., through its BoostAmerica campaign, has also increased the public’s awareness of the need for booster seats through distribution of 1 million booster seats across the country. Programs such as ‘Safe Kids Buckle Up’ sponsored by General Motors and ‘Fit For A Kid’ sponsored by DaimlerChrysler have also played an important role in educating the public about the importance of using age/size appropriate child restraints. Our findings of increased appropriate restraint use suggest the success of these joint efforts by federal, nonprofit, corporate, and public health organizations in promoting appropriate restraint for children.

The use of harnesses by children weighing more than 18.6 kg (41 lb) increased dramatically over the study period, a weight beyond the limit recommended by most manufacturers. Two engineering modifications may address this issue: tether straps and novel designs. First, the addition of a tether strap may allow the combination child restraint/booster with harness to be utilized by heavier children. (Weber, 2000) After September 1999, all child safety seat models with harnesses sold in the US are required to be equipped with top tether straps (64 FR 36657). In addition, NHTSA has considered the modification of US regulations to certify the use of tethered child safety seat with harness and combination child restraint/booster with harness for children weighing over 18.1 kg (40 lb). (Weber, 1997) Second, the promotion of appropriate restraint for children appears to have had an impact on the development of novel child restraint designs. Several of the new child safety seat models are designed to be used by children up to 27.2 kg (60 lb) (NHTSA, 2002; AAP, 2002). The safety implications of heavier children in child safety seat with harness or combination child restraint/booster with harness are not currently known and should be a focus of future work.

Our data identified areas of inappropriate use of belt-positioning booster seats and rear-facing infant seats. There was a marginally significant increasing trend for belt-positioning booster seat use for children weighing between 9.1 and 13.6 kg (20–30 lb). Additional analysis showed that a number of children weighing more than 9.1 kg (20 lb) were still restrained in rear facing infant-only seats, none of which are certified for use past 10.0 kg (22 lb). Because a child’s weight and restraint fit may have an effect on the risk and pattern of injury in a collision, future studies should be directed at monitoring injury risk for children who are using restraints outside of manufacturers’ weight recommendations.

Many parents appeared to delay graduation of children to seat belts until older ages. Educational efforts aimed at promoting booster seat use by children over 18.1 kg (40 lb) appeared to “raise the bar” for all children, reducing premature graduation out of child safety seat and into seat belts. However, by the end of 2002, few children over 27.7 kg (61 lb) were restrained in belt-positioning booster seats. This points to the need for continued and improved intervention efforts that promote the use of belt-positioning booster for larger children who are not of sufficient size for appropriate restrain in a seat belt.

LIMITATIONS

This research was conducted on crashes involving State Farm Insurance Co. policyholders only. Although State Farm is the largest insurer of automobiles in the United States, with over 38 million vehicles covered; therefore, its policyholders are likely representative of the insured public in this country, but the results might not be generalizable to children in uninsured vehicles. Further, the results might not apply to crashes involving vehicles older than Model Year 1990.

The self-report nature of child’s weight by the adult driver (mostly parents) via telephone survey might have an effect on the validity of child’s weight. However, in a study examining the validity of parents’ knowledge of their child’s weight, it was found that parent-reported weight was highly correlated with weight as reported on the medical record (r=0.98, p<0.0005; Spearman’s rho=0.96, p<0.0005) (personal communication, Zsolway, 2003). Also, the emphasis on size appropriate child restraint use might have biased the parent’s report of child’s weight due to the effect of social desirability. Thus, future research is necessary to ensure the accuracy of reported child’s weight by parents.

CONCLUSIONS

These results reflect very current trends in child restraint usage by weight groups for children under 80 pounds in crashes. While considerable achievements have been realized over a short period of time, substantial inappropriate restraint still remains. Also, new parents may not have heard this important information. Continued investment is necessary in educational and outreach efforts promoting appropriate child restraint use. Of note, there is an increasing trend of children over 18.6 kg (41 lb) restrained in child restraints with harnesses. The risk of injury for heavier children in such restraints needs to be monitored and further studied.