Child Passenger Safety Practices in the U.S

osuoo

Child Passenger Safety Practices in the U.S.Disparities in Light of Updated Recommendations

Michelle L. Macy, MD, MS, Gary L. Freed, MD, MPH

Background: Children are best protected in motor vehicle collisions when properly using theappropriate restraint and sitting in a rear row. Racial and ethnic disparities have been reported ininjury statistics and use of any restraint; however, predictors of safety seat use, being unrestrained,and sitting in the front seat have not been explored previously.

Purpose: To determine factors associated with child passenger safety practices by race/ethnicity ina national sample of child passengers aged �13 years.

Methods: Secondary analysis conducted in 2011 of the 2007, 2008, and 2009 National Survey of theUse of Booster Seats in which child passenger restraint use was observed directly. Age-stratifıed,survey-weighted chi-square and logistic regression analyses were conducted.

Results: Restraint use was observed for 21,476 children aged �13 years. A decline in child safetyseat use and increase in being unrestrained were observed with increasing child age. In multivariateanalyses, race/ethnicity, unrestrained drivers, and sitting in the front seat were associated with lowerodds of child safety seat use among children aged�8 years.Older child agewas associatedwith sittingin the front seat and being unrestrained. The presence of multiple child passengers was associatedwith lower odds of sitting in the front but higher odds of being unrestrained.

Conclusions: Few children use the recommended child passenger restraints. Understanding thereasons for the suboptimal child passenger restraint practices identifıed in this study is essential forthe development of effective programs to reduce or eliminate preventable motor vehicle collision–related injuries.(Am J Prev Med 2012;xx(x):xxx) © 2012 American Journal of Preventive Medicine

UNDER EMBARGO UNTIL AUGUST 7, 2012, 12:01 AM ET

s4Agabfpmosh

t

Introduction

Motor vehicle collisions (MVCs) remain a lead-ing cause of death among children after thefırst year of life and the leading cause of death

among children aged �3 years in the U.S.1 In addition,more than 140,000 children aged �13 years are seen inhospital emergency departments each year for nonfatalinjuries sustained as occupants in MVCs.2 Increased riskf injury and death as children age may be attributable touboptimal passenger safety practices, including beingnrestrained, sitting in the front seat, and premature usef adult seat belts, which increase during the fırst decadef life.3

From the Department of Emergency Medicine (Macy), Child Health Eval-uation and Research (CHEAR) Unit (Macy, Freed), Division of GeneralPediatrics, University of Michigan, Ann Arbor, Michigan

Address correspondence to: Michelle L. Macy, MD, MS, University ofMichigan, Division of General Pediatrics, 300 North Ingalls 6E08, AnnArbor MI 48109-5456. E-mail: [email protected].

0749-3797/$36.00
(http://dx.doi.org/10.1016/j.amepre.2012.05.023
© 2012 American Journal of Preventive Medicine • Published by Elsev

Parents have long been advised to keep their infantsrear-facing until age 1 year and 20 pounds4 and manytates legally require booster-seat use for those aged–7 years.5 According to the 2011 update of the Americancademy of Pediatrics (AAP) Guidelines for Child Passen-er Safety,6,7 children should (1) remain rear-facing untilt least age 2 years or the highest weight or height allowedy the child safety seat manufacturer; (2) use a forward-acing car seat with a fıve-point harness for as long asossible up to the highest weight or height allowed by theanufacturer; (3) use a booster seat from the time theyutgrow their car seat until they fıt properly in an adulteat belt, expected at a height of 5=7�, the 50th percentileeight for a child aged 11 years8; and (4) always ride in the

back seat until age 13 years.Although use of any restraint has clear safety benefıts

in anMVC, studies9–15 also have demonstrated that childpassengers are best protected from MVC-related injur-ies when properly using age-appropriate restraintsand seated in a vehicle rear row. Prior reports16–19 fromhe National Highway Traffıc Safety Administration
NHTSA) National Survey on the Use of Booster Seats
ier Inc. Am J Prev Med 2012;xx(x):xxx 1

mailto:[email protected]

http://dx.doi.org/10.1016/j.amepre.2012.05.023

sc

a4tsuc

2 Macy and Freed / Am J Prev Med 2012;xx(x):xxx

(NSUBS) demonstrate persistent racial/ethnic differ-ences in the use of ANY restraint. However, racial/ethnicdisparities in relation to the recommended child passen-ger safety practices across early childhood have not beenexplored previously. The primary objective of the presentstudywas to analyze 3 years ofNSUBSdata to evaluate forracial/ethnic disparities in the types of child safety seats inuse across childhood. A secondary objective was to iden-tify child, driver, and vehicle characteristics associatedwith child passengers being unrestrained and sitting inthe front seat.

MethodsStudy Design

This secondary analysis was conducted in 2011 using 3 years ofcombined data (2007, 2008, and 2009) from the NHTSA NSUBS.The NSUBS is a recurrent direct observational study of child pas-senger restraint use across the U.S. Details of data collection andsurveyweights have been described previously.20,21 This analysis ofde-identifıed data was deemed exempt from review by the Univer-sity of Michigan IRB.

Study Setting and Population

Data from NSUBS were collected during summer months from aprobability sample of drivers with child passengers who appearedto be aged �13 years arriving to community sites including gastations, fast-food chains, recreation centers, and child careenters.

Measurements

Trained data collectors directly observed and recorded child re-straint type (rear-facing car seat, forward-facing car seat, boosterseat, or seat belt) and seat row (front vs rear); child and drivergender; driver restraint use; and vehicle type (passenger car, van/sport utility vehicle [SUV], or pickup truck). Drivers’ report oftheir own age, child age, child race, and Hispanic ethnicity wereobtained through a brief interview. Driver report of child weightand height also was obtained; however, there was no independentverifıcation of these data. Because the authors believe driver reportof child age to be more reliable than size estimates, only child agewas included in the analyses.In the data set, driver and child age were categoric variables. Driver

agewas categorizedas16–24years, 25–69years, and�70 years.Childge was categorized within the following groups: �1 year, 1–3 years,–5 years, 6–7 years, 8–10 years, and 11–12 years. The �1-year andhe 1–3-year age groups were combined because of the small sampleize of children aged�1 year who were not in a car seat and the 2011pdated Guidelines for Child Passenger Safety, which recommendhildren remain rear-facing beyond age 1 year.6 Unique vehicle iden-tifıers were used to determine the number of child passengers, aged�13 years, in the vehicle.Racial/ethnic groups were categorized as non-Hispanic white,

non-Hispanic black, Hispanic, or other race for analyses because ofthe small number of children in the other racial groups. Childrenwere excluded from analysis if data fıelds for age and/or racecontained missing values (missing age, n�102; missing age and
race, n�5106).
Outcome Measures

According to the 2002 Guidelines for Child Passenger Safety inplace at the time of data collection, all children were designated tobe in one of three specifıc child passenger restraint categories:(1) Age-Appropriate, (2) Premature Transition, (3) Unrestrained.The category Age-Appropriate was assigned to children aged0–3 years in a rear-facing or forward-facing car seat; children aged4–7 years in a car seat or booster seat, and children aged 8–12 years ina safety seat or seat belt. The Premature Transition category in-cluded children aged 0–3 years in a booster seat or seat belt andchildren aged 4–7 years in a seat belt alone. Unrestrained wasassigned if the child was using no restraint. Sitting in the front seatwas compared with sitting in a rear seat as a separate child passen-ger safety behavior.

Data Analysis

Analyses were conducted in 2011with Stata, version 10.1, using thesvy commands and survey weights provided by NHTSA to permitinferences about national patterns. Children aged 0–10 years werethe focus of the Age-Appropriate Restraint analysis because mostchildren in this age range would benefıt from use of a safety seat.Analyses were stratifıed by age group. Child passenger restrainttype and sitting in the front seat were compared across racial/ethnic group using chi-square tests. Proportions and 95%CIs werecalculated.Bivariate logistic regression analyses stratifıed by age group and

limited to children using any restraint were conducted in order todetermine child, driver, and vehicle characteristics associated withAge-Appropriate Use consistent with the 2002 Guidelines. Bivari-ate logistic regression analyses also were conducted to assess forchild, driver, and vehicle characteristics associated with being un-restrained and sitting in the front seat. Multivariate logistic regres-sion analyses were conducted using variables that were associatedin the bivariate analyses. The interaction between child race and anunrestrained driver was not signifıcant (results not shown). Toassess the gap between current safety practices and 2011Guidelinesfor Child Passenger Safety, a subanalysis was conducted to deter-mine the proportions of restrained children aged 1 years usingrear-facing car seats and the proportions of restrained childrenaged 8–10 years using child safety seats.

ResultsSample CharacteristicsPassenger restraint use was observed for 21,476 childrenaged �13 years with available data on age and race/ethnicity, representing a weighted sample size of 123,082.The sample consisted of 59% white children, 11% blackchildren, 21% Hispanic children, and 9% children ofother race. Child and driver characteristics across racial/ethnic group are presented in Table 1.

Restraint UseA decline in child safety seat use and an increase in beingunrestrained were observed with increasing child age.
Racial/ethnic disparities in age-appropriate restraint use
www.ajpmonline.org

Macy and Freed / Am J Prev Med 2012;xx(x):xxx 3

Table 1. Sample characteristics by child race and ethnicity, weighted row % unless otherwise noted

White n�12,047 Black n�2,713 Hispanic n�4,801 Other n�1,915

Unweightedn

Weightedn�72,294

(59.0%)

Weightedn�12,954

(10.6%)

Weightedn�25,976

(21.2%)

Weightedn�11,358

(9.3%)

Child age (years)

0–3 6,164 61.7 10.6 18.2 9.5

4–5 3,935 58.7 11.7 20.2 9.3

6–7 4,050 57.1 10.6 22.4 9.8

8–10 5,127 55.4 9.6 26.1 8.9

11–12 2,200 59.9 8.8 23.9 7.4

Child gender

Male 11,043 59.7 10.1 21.4 8.8

Female 10,433 58.2 11.0 21.0 9.7

Restraint type**

Rear-facing car seat 1,068 69.1 6.5 16.3 8.0

Forward-facing car seat 4,771 64.4 8.6 17.3 9.7

Booster seat 4,041 69.0 7.6 14.8 8.5

Seat belt 9,009 54.5 12.0 23.9 9.6

Unrestrained 2,587 34.0 18.9 38.0 9.1

Seat position*

Front row 2,922 53.5 10.9 27.2 8.3

Rear row 18,554 59.7 10.5 20.4 9.4

Driver age (years)

16–24 1,203 49.1 9.4 29.6 11.8

25–69 19,979 59.6 10.6 20.6 9.2

�70 years 294 51.6 14.5 29.8 4.2

Driver gender

Male 7,639 58.5 9.8 21.3 10.3

Female 13,837 59.1 10.9 21.1 8.7

Driver belt use**

Restrained 18,841 61.0 9.6 20.2 9.2

Unrestrained 2,635 45.4 17.2 28.0 9.4

Vehicle type**

Van/sport utility vehicle 10,967 63.7 7.8 19.9 8.5

Passenger car 9,057 53.9 14.6 21.3 10.2

Pickup truck 1,452 57.6 4.2 29.9 8.2

Child passengers

One 1,795 57.8 12.2 19.2 10.9

Two 3,333 62.1 9.9 20.2 7.9

(continued on next page)

Month 2012

pdspHwa

o6ttb

cC

7Hd

rscamaatowieoAalAtgwaol(rC


were present in age-stratifıed chi-square analyses (Figure 1,top and lower left, p�0.001; Figure 1, lower right,�0.005). Within each age group, minority childrenemonstrated lower proportions of age-appropriate re-traint use compared with white children. There wereersistent differences in the proportions of black andispanic childrenwhowere unrestrained comparedwithhites, ranging from a tenfold difference among infantsnd toddlers to a twofold difference for older age groups.Among children aged 0–3 years (Figure 1, top left), the

proportions of rear-facing car seat usewere lowest amongminority groups, but even among whites just 17% wererear-facing. It is important to note that 84% of the chil-dren in rear-facing car seats were infants aged �1 year.Premature Transitions to booster seats or seat belts weresimilar across racial/ethnic groups.Age-Appropriate Restraint Use was lower among mi-

nority children aged 4–7 years comparedwith white chil-dren (Figure 1, top right, lower left). Higher proportionsofminority children aged 4 and 5 years were PrematurelyTransitioned to seat belts compared with white children:white, 16% (95% CI�13%, 19%); black, 35% (95%CI�25%, 46%); Hispanic, 26% (95% CI�21%, 31%);ther, 27% (95% CI�20%, 36%). Among those aged–7 years, the gap in Age-Appropriate Restraint Use be-ween minority and white children decreased largelyhrough a disproportionate increase in white childreneing Prematurely Transitioned to seat belts.A seat belt was themost common restraint used among

hildren aged 8–10 years, observed for 80% (95%

Table 1. Sample characteristics by child race and ethnici

White n�12,04

Unweightedn

Weightedn�72,294

(59.0%)

Three 4,212 59.0

Four or more 12,136 58.2

Area

Suburban 15,350 59.9

Urban 4,185 49.7

Rural 1,941 65.7

Region**

Midwest 8,987 81.1

South 6,312 49.4

West 4,317 33.2

Northeast 1,860 72.7

*p�0.002, **p�0.001

I�77%, 83%) of white children; 68% (95% CI�57%, s

6%) of black children; 71% (95% CI�64%, 77%) ofispanic children; and 76% (95% CI�74%, 78%) of chil-ren in other racial groups (Figure 1, lower right).Table 2 presents results of the age-stratifıed logistic

egression analyses of predictors of Age-Appropriate Re-traint Use. Among children aged 0–3 years, black raceompared with white and sitting in the front seat weressociated with lower odds of car seat use in bivariate andultivariate analyses. Unrestrained drivers were associ-ted with lower odds of car seat use in the bivariatenalysis, but this variable was not signifıcant in the mul-ivariate analysis. Among children aged 4–7 years, oddsf using a child safety seat, either a car seat or booster seat,ere lower for minority groups and racial/ethnic dispar-ties persisted inmultivariate analysis. Unrestrained driv-rs and driving in a car comparedwith a van or SUVwerether predictors of lower odds of child safety seat use.mong those aged 8–10 years, sitting in the front seatnd riding with a younger driver were associated withower odds of using a child safety seat. The odds of Age-ppropriate Restraint Use were signifıcantly higher inhe Northeast compared with the Midwest for each ageroup. Child gender, driver gender, and year of studyere not associatedwithAge-AppropriateRestraintUse forny age group. Taking into consideration the expanded rec-mmendations for child safety seat use in the 2011 guide-ines published after data collection for this study, just 3%95% CI�2%, 4%) of children aged 1–3 years using anyestraint were sitting in rear-facing car seats and 10% (95%I�6%, 11%) of children aged 8–10 years who were re-

eighted row % unless otherwise noted (continued)

Black n�2,713 Hispanic n�4,801 Other n�1,915

Weightedn�12,954

(10.6%)

Weightedn�25,976

(21.2%)

Weightedn�11,358

(9.3%)

9.7 21.9 9.4

10.8 21.6 9.4

10.4 20.0 9.7

15.8 26.0 8.5

4.0 24.2 6.1

8.0 4.9 6.0

16.5 22.1 12.0

2.3 56.3 8.2

10.3 7.3 9.6

ty, w

7

trained were sitting in a car seat or booster seat.

www.ajpmonline.org

�rvmco


Front SeatFewer than 5% of children aged �4 years were front seatpassengers, followed by a steady increase in sitting in thefront seat with increasing child age (Table 3). Child age5 years, Hispanic ethnicity, and unrestrained driversemained predictors of sitting in the Front Seat in multi-ariate regression analysis (Table 4). The presence ofore thanone child passenger and living in theNortheastompared with the Midwest were associated with lowerdds of sitting in the front seat.

UnrestrainedChildren with an unrestrained driver had a 23 timeshigher adjusted odds of being unrestrained themselves.Child age �3 years, minority race/ethnicity, and havingfour or more child passengers compared with one childpassenger also were associated with child passengers be-

0102030405060708090

100

Rear-facing Forward-facing Booster/Belt Unrestrained

Restraint type

Per

cent

0 10 20 30 40 50 60 70 80 90

100

Car seat/booster Seat belt Unrestrained

Perc

ent

Restraint type

White

p<0.001

p<0.001

Age-appropriate Premature transition

Unrestrained


Unrestrained

Booster/belt

Figure 1. Child passenger restraint use category by childNote: Bars indicate the percentage of children usingGuidelines for Child passenger Safety state that childreyears. Because of the categoric nature of the age variachildren aged �2 years separately; therefore rear-faappropriate for children aged 0–3 years.

ing unrestrained (Table 5).

Month 2012

DiscussionThemost important fınding from the current study is thatalthough age and racial disparities exist, overall low pro-portions of U.S. child passengers are using the age-appropriate restraint andmany are placed at risk by sittingin the front seat. This research reveals three specifıc op-portunities to increase the number of children who arerestrained according to the 2011 Guidelines for ChildPassenger Safety.6,7 First, few children remain rear-facing after age 1 year. Second, after age 7 years, lessthan 2% use a booster seat. Third, many children aged�6 years sit in the front seat, one in seven aged 6–7 years,one quarter of those aged 8–10 years, and more thanone third of those aged 11–12 years. Clinical encoun-ters should serve as an opportunity to convey childpassenger safety information to parents and childrenin relation to the child’s size because more can be doneto better protect children in the unexpected event of

0 10 20 30 40 50 60 70 80 90

100

Car seat Booster Seat belt Unrestrained

Per

cent

Restraint type

Hispanic Other

p<0.001


Unrestrained

0 10 20 30 40 50 60 70 80 90

100

Car seat/booster Seat belt Unrestrained

Perc

ent

Restraint type

Age-appropriate Unrestrained

p=0.005

/ethnicity across ageh type of restraint or being unrestrained. The 2011hould remain rear-facing from birth until at least age 2in the NSUBS data set, it was not possible to analyzeand forward-facing car seats were considered age-

Black

raceeacn sblecing

an MVC.

ptn

*


These fındings demonstrate that not all children havebeen reached equally by community-based public educa-tion campaigns and the passage of child safety seat laws in48 states.22–24 Although education campaigns and lawshave led to both an increase in restraint use and a reduc-

Table 2. Predictors of age-appropriate restraint use amon

Rear-facing or forward-facing restraintuse, age 0–3 years

OR (95% CI) AOR (95% CI) OR

Child race/ethnicity

White 1.0 (ref) 1.0 (ref) 1.0

Black 0.44* (0.30, 0.65) 0.48* (0.34, 0.69) 0.33*

Hispanic 0.82 (0.59, 1.13) 0.81 (0.59, 1.10) 0.40*

Other 0.77 (0.46, 1.31) 0.76 (0.46, 1.25) 0.60*

Driver age (years)

16–24 0.82 (0.59, 1.56) — 0.55*

25–69 1.0 (ref) — 1.0

�70 1.87 (0.59, 5.91) — 1.17

Driver belt use

Belted 1.0 (ref) — 1.0

Unrestrained 0.66* (0.47, 0.91) 0.72 (0.51, 1.00) 0.23*

Vehicle type

Van/sport utilityvehicle

1.0 (ref) — 1.0

Car 1.17 (0.97, 1.41) — 0.64*

Pickup truck 0.86 (0.53, 1.39) — 0.62*

Seat row

Rear 1.0 (ref) 1.0 (ref) 1.0

Front 0.29* (0.17, 0.50) 0.31* (0.20, 0.48) 0.11*

Child passengers

One 1.0 (ref) 1.0 (ref) 1.0

Two 1.02 (0.51, 2.04) 0.97 (0.50, 1.90) 1.03

Three 1.64* (1.02, 2.63) 1.49 (0.92, 2.39) 0.78

Four or more 1.15 (0.71, 1.87) 1.07 (0.65, 1.76) 0.75

Location

Suburban 1.0 (ref) — 1.0

Urban 1.10 (0.67, 1.80) — 0.65

Rural 1.38 (0.88, 2.15) — 1.33

Region

Midwest 1.0 (ref) — 1.0

South 0.97 (0.70, 1.33) 1.07 (0.80, 1.45) 0.78

West 1.37* (0.97, 1.94) 1.42 (1.00, 2.02) 1.01

Northeast 2.12* (1.45, 3.11) 2.14* (1.44, 3.18) 2.41*

p�0.05

tion in child injuries as a result of MVCs,25–27 the adop- b

tion of booster seat laws also may result in increaseddisparities in proper booster seat use.28,29 Tailored childassenger safety programs are needed to address the mo-ivations of parents from various cultural and socioeco-omic backgrounds, and for drivers who do not use seat

ildren using any restraint

at or booster seat use,age 4–7 years

Car seat or booster seat use,age 8–10 years

CI) AOR (95% CI) OR (95% CI) AOR (95% CI)

1.0 (ref) 1.0 (ref) —

, 0.48) 0.37* (0.24, 0.57) 1.03 (0.44, 2.42) —

, 0.56) 0.42* (0.31, 0.58) 0.87 (0.47, 1.60) —

, 0.87) 0.61* (0.41, 0.89) 1.10 (0.46, 2.66) —

, 0.90) 0.73 (0.47, 1.13) 0.43* (0.20, 0.93) 0.44* (0.19, 0.98)

1.0 (ref) 1.0 (ref) 1.0 (ref)

, 1.80) 1.27 (0.72, 2.25) 0.21* (0.05, 0.82) 0.24* (0.07, 0.80)

1.0 (ref) 1.0 (ref) —

, 0.33) 0.24* (0.16, 0.38) 1.58 (0.73, 3.45) —

1.0 (ref) 1.0 (ref) —

, 0.74) 0.77* (0.65, 0.90) 0.82 (0.60, 1.10) —

, 0.85) 1.07 (0.68, 1.66) 0.53 (0.24, 1.16) —

1.0 (ref) 1.0 (ref) 1.0 (ref)

, 0.15) 0.13* (0.09, 0.18) 0.16* (0.08, 0.32) 0.16* (0.08, 0.32)

— 1.0 (ref) —

, 1.52) — 0.84 (0.44, 1.60) —

, 1.29) — 0.70 (0.36, 1.38) —

, 1.15) — 0.82 (0.46, 1.46) —

— 1.0 (ref) 1.0 (ref)

, 1.16) — 0.80 (0.48, 1.33) 1.06 (1.02, 2.67)

, 2.21) — 2.00* (1.30, 3.10) 1.65 (1.02, 2.67)

1.0 (ref) 1.0 (ref) 1.0 (ref)

, 1.03) 0.95 (0.74, 1.23) 0.92 (0.48, 1.76) 1.03 (0.54, 1.95)

, 1.59) 1.34 (0.94, 1.91) 1.57 (0.75, 3.28) 1.65 (0.81, 3.36)

, 3.07) 2.58* (2.12, 3.13) 3.16* (1.91, 5.24) 3.10* (1.76, 5.46)

g ch

Car se

(95%

(ref)

(0.22

(0.29

(0.41

(0.34

(ref)

(0.76

(ref)

(0.16

(ref)

(0.55

(0.46

(ref)

(0.08

(ref)

(0.70

(0.47

(0.49

(ref)

(0.36

(0.80

(ref)

(0.59

(0.64

(1.90

elts themselves. Clinic visits can be used to address pa-

www.ajpmonline.org

pwcretdc

aa

dgacpfgglc

s

fp�b


rentalmisinformation related to child safety seats, such asconcerns that booster seats can cause injury.30 Furtherdevelopment and dissemination of culturally specifıcprograms that have demonstrated success in promotingrestraint use among minority children are necessary.23,31

The current study also brings new insights to the pre-viously reported and striking disparities between minor-ity and white child passengers who are unrestrained andpresents information regarding children sitting in thefront seat. Older children had higher odds of both beingunrestrained and sitting in the front seat. Having four ormore child passengers was associated with children beingunrestrained. However, similar to other research,32 theresence ofmore than one child passenger was associatedith decreased odds of sitting in the front seat. The asso-iation between child age and being unrestrained mayelate to a child’s ability to escape from a seat belt moreasily than they can from a car seat. Another possibility ishat older children refuse to wear seat belts because ofiscomfort associated with incorrect fıt that could beorrected with the use of a booster seat.33

These fındings alsomay relate to the limited number ofseat belt positions available in a vehicle back seat, whichcreates a barrier to restraint use for parents who carpoolmultiple children.34 In addition, vehicle crowding hasbeen identifıed as a barrier to restraint use in focus groupsconducted with parents from minority racial and ethnicgroups.30,35–37 Whenever possible, drivers should seeklternatives to transportingmore children than they havevailable seat belts.There was a strong association between unrestrainedrivers and unrestrained child passengers. Child passen-ers also had higher odds of sitting in the front seat withn unrestrained driver. One novel approach to increasinghild passenger restraint use may be to mandate childassenger safety education for parents who are ticketedor driving unrestrained. Because parents look to laws foruidance about how to restrain their child passen-ers,38,39 itmust be noted thatmost child passenger safetyaws in the U.S. provide only a minimum standard for

Table 3. Sitting in the front seat by child age and race/et

Age (years) White Bla

0–3 1.8 (1.3, 2.5) 3.5 (1.8

4–5* 5.1 (3.1, 8.2) 8.9 (6.1

6–7* 11.4 (8.8, 14.7) 14.2 (10.

8–10 23.7 (20.9, 26.8) 21.8 (15.

11–12 38.8 (35.0, 42.8) 41.3 (30.

*p�0.03

hild safety seat and seat belt use.

Month 2012

The use of any restraint is superior to being unre-trained13,40 but children who are transitioned prema-turely to seat belts are more likely to experience incorrectseat belt fıt, exposing them to greater risk of injury in anMVC.41-43 Because the NSUBS does not assess seat beltıt or verify driver report of child height, it was notossible to estimate the proportion of children aged7 years who have been transitioned prematurely to seatelts and would still benefıt from the use of a booster seat.33

Although the safety benefıts of using a rear-facing car seatfor a child aged 18 months may be greater than the safetybenefıts of using a booster seat for a child aged 10 years, thecurrent results indicate that additional efforts to promotebest-practice child passenger safety recommendations areneeded across the fırst decade of life.

LimitationsThe fırst limitation of the current study relates to thefact that no information related to SES of the partici-pants is collected for the NSUBS. Although culturalnorms contribute to child passenger safety practices,30

the observed racial/ethnic differences in the currentstudy may be a marker for disparities related to lowSES, low educational attainment, or limited Englishprofıciency. LowSESmaydecrease the ability of a family toown child safety seats. Low educational attainment and lim-ited English profıciency may decrease access to the recom-mendations for child passenger safety provided in printma-terials andmass media campaigns.Second, there is no verifıcation of child age. This could

result in drivers stating that a child is older or youngerthan his or her true age. The authors hypothesize thatdrivers would bemore likely to report that a child is olderthan his or her true age if they have been transitionedprematurely, which would result in an underestimationof premature transitions.Third, the recommended restraint for a child’s agemay

not match the recommended restraint for their size.Given the possible range of child heights and weights fora given age and the wide range of manufacturer specifı-

ity, weighted % (95% CI)

Hispanic Other

) 3.4 (1.7, 6.9) 3.6 (1.5, 8.3)

7) 9.8 (6.7, 14.1) 3.4 (1.3, 8.6)

.3) 17.2 (13.8, 21.2) 12.3 (7.7, 19.1)

.1) 25.7 (21.2, 30.9) 19.4 (13.7, 26.8)

.5) 34.8 (29.2, 40.7) 48.2 (33.7, 63.1)

hnic

ck

, 6.8

, 12.

3, 19

9, 29

9, 52

cations for maximum height and weight allowable for

mTt

f


child safety seats on the market, it is not possible toestimate the direction of this bias in the current study.

Table 4. Predictors of child passengers sitting in theront seat

OR (95% CI) AOR (95% CI)

Child age (years)

0–5 1.0 (ref) 1.0 (ref)

6–7 3.76* (2.91, 4.62) 3.81* (3.02, 4.79)

8–10 7.55* (5.66, 10.08) 7.89* (6.16, 10.11)

11–12 15.41* (11.09, 21.42) 17.64* (13.27, 23.62)

Child gender

Male 1.0 (ref) 1.0 (ref)

Female 0.78* (0.70, 0.87) 0.77* (0.67, 0.89)


White 1.0 (ref) 1.0 (ref)

Black 1.15 (0.86, 1.56) 1.14 (0.83, 1.58)

Hispanic 1.48* (1.21, 1.82) 1.21* (1.04, 1.41)

Other 0.99 (0.73, 1.33) 0.91 (0.69, 1.20)

Driver age (years)

16–24 1.0 (ref) 1.0 (ref)

25–69 1.06 (0.78, 1.45) 1.06 (0.80, 1.44)

�70 1.66* (1.06, 2.61) 1.08 (0.67, 1.74)

Driver gender

Male 1.0 (ref) 1.0 (ref)

Female 0.85* (0.74, 0.98) 1.19* (1.06, 1.37)

Driver belt use

Belted 1.0 (ref) 1.0 (ref)

Unbelted 1.77* (1.47, 2.13) 1.64* (1.36, 1.99)

Vehicle type


1.0 (ref) 1.0 (ref)

Car 1.57* (1.38, 1.79) 1.69* (1.44, 1.93)

Pickup truck 5.66* (4.00, 8.01) 5.92* (4.17, 8.32)

Child passengers

One 1.0 (ref) 1.0 (ref)

Two 0.64* (0.46, 0.90) 0.61* (0.43, 0.85)

Three 0.68* (0.56, 0.82) 0.64* (0.52, 0.80)

Four or more 0.56* (0.47, 0.67) 0.53* (0.44, 0.63)

Region

Midwest 1.0 (ref) 1.0 (ref)

South 1.02 (0.80, 1.30) 1.23 (0.98, 1.57)

West 1.16 (0.87, 1.54) 0.87 (0.73, 1.06)

Northeast 0.47* (0.39, 0.57) 0.65* (0.45, 0.94)

*p�0.05

Fourth, these results are only generalizable to the com- r

unity settings in which the surveys were conducted.he use of passenger restraints by children who wereraveling to survey locations may not be representative of

Table 5. Predictors of child passengers beingunrestrained

OR (95% CI) AOR (95% CI)

Child age (years)

0–3 1.0 (ref) 1.0 (ref)

4–5 2.65* (1.77, 4.00) 3.22* (2.13, 4.87)

6–7 3.14* (2.01, 4.90) 4.04* (2.70, 6.07)

8–10 3.55* (2.32, 5.43) 4.46* (3.00, 6.61)

11–12 3.24* (2.03, 5.18) 4.68* (3.16, 6.91)


White 1.0 (ref) 1.0 (ref)

Black 3.60* (2.51, 5.16) 2.85* (1.85, 4.39)

Hispanic 3.60* (2.42, 5.34) 3.74* (2.42, 5.77)

Other 1.79* (1.10, 2.92) 1.57* (1.05, 2.35)

Driver age (years)

16–24 1.36 (0.87, 2.11) 0.93 (0.76, 1.13)

25–69 1.0 (ref) 1.0 (ref)

�70 2.06* (1.15, 3.70) 1.87 (0.77, 4.53)

Driver belt use

Belted 1.0 (ref) 1.0 (ref)

Unbelted 21.45* (15.34, 30.01) 22.86* (15.14, 34.54)

Vehicle type


1.0 (ref) 1.0 (ref)

Car 1.74* (1.46, 2.07) 1.39* (1.16, 1.66)

Pickup truck 2.2* (1.43, 3.37) 1.51 (0.96, 2.38)

Child passengers

One 1.0 (ref) 1.0 (ref)

Two 0.92 (0.63, 1.36) 0.88 (0.55, 1.40)

Three 1.20 (0.82, 1.75) 1.16 (0.71, 1.89)

Four or more 1.66* (1.20, 2.30) 1.59* (1.05, 2.39)

Seat row

Rear 1.0 (ref) 1.0 (ref)

Front 1.95* (1.46, 2.61) 0.98 (0.85, 1.67)

Region

Midwest 1.0 (ref) 1.0 (ref)

South 1.16 (0.76, 1.77) 1.19 (0.85, 1.67)

West 1.00 (0.59, 1.71) 0.63 (0.40, 1.00)

Northeast 0.58* (0.48, 0.72) 0.50* (0.39, 0.64)

*p�0.05

estraint use for other types of trips.

www.ajpmonline.org

fpra

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3


ConclusionThere are substantial opportunities to improve childpassenger safety practices in the U.S. The new AAPrecommendations for child passenger safety provideguidance to clinicians counseling parents on childsafety seat use. Child age, race/ethnicity, and driverrestraint use are associated with less-safe child passen-ger behaviors. Understanding the reasons why these dif-erences exist is essential for the development of effectiverograms to increase recommended restraint use andeduce or eliminate preventableMVC-associated injuriesnd deaths.

The authors thank Timothy M. Pickrell for his assistance withobtaining the data fıles from the Mathematical Analysis Divi-sion, National Center for Statistics and Analysis, NationalHighway Traffıc Safety Administration, and SubrahmanyamPilli and Acham Gebremariam for statistical support from theUniversity of Michigan.No fınancial disclosures were reported by the authors of this

paper.

References1. CDC. Ten leading causes of death and injury charts: Causes of death by

age group, and causes of njury death: highlighting unintentional injury,2007. www.cdc.gov/injury/wisqars/LeadingCauses.html.

2. CDC.Injury prevention & control: data & statistics.Web-based InjuryStatistics Query and Reporting System (WISQARS TM), 2009. www.cdc.gov/injury/wisqars/index.html.

3. Greenspan AI, Dellinger AM, Chen J. Restraint use and seating posi-tion among children less than 13 years of age: is it still a problem? JSafety Res 2010;41(2):183–5.

4. Committee on Injury and Poison Prevention; American Academy ofPediatrics. Selecting and using themost appropriate car safety seats forgrowing children: guidelines for counseling parents. Pediatrics2002;109(3):550–3.

5. Insurance Institute for Highway Safety. Child restraint laws. www.iihs.org/laws/ChildRestraint.aspx.

6. Durbin DR. Child passenger safety—technical report. Pediatrics2011;127(4):e1050–e1066.

7. Durbin DR. Child passenger safety—policy statement. Pediatrics2011;127(4):788–93.

8. CDC, National Center for Health Statistics. CDC growth charts: U.S.2000;www.cdc.gov/growthcharts/.

9. Carlsson G, Norin H, Ysander L. Rearward-facing child seats—thesafest car restraint for children? Accid Anal Prev 1991;23(2/3):176–82.

0. Henary B, Sherwood CP, Crandall JR, et al. Car safety seats forchildren: rear facing for best protection. Inj Prev 2007;13(6):398–402.

1. Elliott MR, Kallan MJ, Durbin DR, Winston FK. Effectiveness ofchild safety seats vs seat belts in reducing risk for death in children inpassenger vehicle crashes.ArchPediatrAdolescMed2006;160(6):617–21.

2. Javouhey E, Guerin AC, Gadegbeku B, Chiron M, Floret D. Are re-strained children under 15 years of age in cars as effectively protected asadults? Arch Dis Child 2006;91(4):304–8.

3. Lee KC, Shults RA, Greenspan AI, Haileyesus T, Dellinger AM. Child
passenger restraint use and emergency department–reported inju-
Month 2012

ries: a special study using the National Electronic InjurySurveillance System—All Injury Program, 2004. J Safety Res2008;39(1):25–31.

4. ArbogastKB, Jermakian JS, KallanMJ,DurbinDR. Effectiveness of beltpositioning booster seats: an updated assessment. Pediatrics 2009;124(5):1281–6.

5. Arbogast KB, Kallan MJ, Durbin DR. Front versus rear seat injury riskfor child passengers: evaluation of newer model year vehicles. TraffıcInj Prev 2009;10(3):297–301.

6. Glassbrenner D, Ye T. Child restraint use in 2006—demographicresults. Washington DC: National Center for Statistics and Analysis,National Highway Traffıc Safety Administration, 2007.

7. Glassbrenner D, Ye T. Child restraint use in 2007—demographic re-sults. Washington DC: National Center for Statistics and Analysis,National Highway Traffıc Safety Administration, 2008.

8. Pickrell TM, Ye T. Child restraint use in 2008—demographic results.Washington DC: National Center for Statistics and Analysis, NationalHighway Traffıc Safety Administration, 2009.

9. Pickrell TM, Ye T. The 2009 national survey of the use of booster seats.Washington DC: National Center for Statistics and Analysis, NationalHighway Traffıc Safety Administration, 2010.

0. Glassbrenner D. The 2006 national survey of the use of boosterseats—methodology report. Washington DC: National Center forStatistics and Analysis, National Highway Traffıc Safety Adminis-tration, 2009.

1. Pickrell TM, Ye T. Child Restraint Use in 2008—use of correct re-straint types.Washington DC: National Center for Statistics and Anal-ysis, National Highway Traffıc Safety Administration, 2009.

2. Ebel BE, Koepsell TD, Bennett EE, Rivara FP. Use of child booster seatsin motor vehicles following a community campaign: a controlled trial.JAMA 2003;289(7):879–84.

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

4. Decina L, Lococo K, AshburnW, Hall W, Rose J. Identifying strategiesto improve the effectiveness of booster seat laws. Washington DC:National Highway Traffıc Safety Administration, 2008.

5. Winston FK, Kallan MJ, Elliott MR, Xie D, Durbin DR. Effect ofbooster seat laws on appropriate restraint use by children 4 to 7years old involved in crashes. Arch Pediatr Adolesc Med 2007;161(3):270–5.

6. Sun K, Bauer MJ, Hardman S. Effects of upgraded child restraint lawdesigned to increase booster seat use in New York. Pediatrics 2010;126(3):484–9.

7. Pressley JC, Trieu L, Barlow B, Kendig T. Motor vehicle occupantinjury and related hospital expenditures in children aged 3 years to 8years covered versus uncovered by booster seat legislation. J Trauma2009;67(1S):S20–S29.

8. Brixey S, Ravindran K, Guse CE. Legislating child restraint usage—itseffect on self-reported child restraint use rates in a central city. J SafetyRes 2010;41(1):47–52.

9. Brixey S, Corden TE, Guse CE, Layde P. Booster seat legislation: does itwork for all children? Injury Prev 2011;17(4):233–7.

0. Johnston BD, Bennett E, Quan L, Gonzalez-Walker D, Crispin B,Ebel B. Factors influencing booster seat use in a multiethnic com-munity: lessons for program implementation. Health Promot Pract2009;10(3):411–8.

1. Falcone R, Brentley A, Ricketts C, Allen S, Garcia V. Development,implementation and evaluation of a unique African-American faith-based approach to increase automobile restraint use. J Natl Med Assoc2006;98(8):1335–41.

2. Durbin DR, Chen I, Elliott M, Winston FK. Factors associated withfront row seating of children in motor vehicle crashes. Epidemiology
2004;15(3):345–9.
http://www.cdc.gov/injury/wisqars/LeadingCauses.html

http://www.cdc.gov/injury/wisqars/index.html

http://www.cdc.gov/injury/wisqars/index.html

http://www.iihs.org/laws/ChildRestraint.aspx

http://www.iihs.org/laws/ChildRestraint.aspx

http://www.cdc.gov/growthcharts/


33. Reed MP, Ebert SM, Sherwood CP, Klinich KD, Manary MA. Evalua-tion of the static belt fıt provided by belt-positioning booster seats.Accid Anal Prev 2009;41(3):598–607.

34. Macy ML, Clark SJ, Freed GL, et al. Carpooling and booster seats: anational survey of parents. Pediatrics 2012;129(2):290–8.

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

36. Agran PF, Anderson CL, Winn DG. Violators of a child passengersafety law. Pediatrics 2004;114(1):109–15.

37. Winston FK, Erkoboni D, Xie D. Identifying interventions that pro-mote belt-positioning booster seat use for parents with low educationalattainment. J Trauma 2007;63(3S):S29–S38.

38. Bingham CR, Eby DW, Hockanson HM, Greenspan AI. Factors influ-encing the use of booster seats: a state-wide survey of parents. Accid
Anal Prev 2006;38(5):1028–37.
39. Winston FK, Chen IG, Smith R, Elliott MR. Parent driver characteris-tics associated with sub-optimal restraint of child passengers. TraffıcInj Prev 2006;7(4):373–80.

40. Agran PF, Castillo DN,Winn DG. Comparison of motor vehicle occu-pant injuries in restrained and unrestrained 4- to 14-year-olds. AccidAnal Prev 1992;24(4):349–55.

41. Anderson PA, Rivara FP, Maier RV, Drake C. The epi-demiology of seatbelt-associated injuries. J Trauma 1991;31(1):60–7.

42. Arbogast KB, Kent RW, Menon RA, Ghati Y, Durbin DR, RouhanaSW. Mechanisms of abdominal organ injury in seat belt-restrainedchildren. J Trauma 2007;62(6):1473–80.

43. Winston FK, Xie D, Durbin DR, Elliott MR. Are child passengersbringing up the rear? Evidence for differential improvements in injuryrisk between drivers and their child passengers. Annu Proc Assoc Adv
Automot Med 2007;51:113–27.
www.ajpmonline.org

Documents

Child Passenger Safety Practices in the U.S