greenbergsethfactorsassociatedrearseating2004.md 39 KB
category: literaturenote citekey: greenbergsethfactorsassociatedrearseating2004 title: "Factors associated with rear seating of children in motor vehicles: a study in two low-income, predominantly Hispanic communities" authors: "Greenberg-Seth, Jennifer; Hemenway, David; Gallagher, Susan S.; Lissy, Karen S.; Ross, Julie B." year: 2004 date: 2004-00-00 2004 doi: 10.1016/s0001-4575(03)00070-8 publication: "Accident Analysis & Prevention" url: "https://doi.org/10.1016/s0001-4575(03)00070-8" zotero_key: N6TZMP7X zotero_storage: 46N62PGR collections: transpordi_planeerimine folder: 001_artiklid firstAuthor: "Greenberg-Seth, Jennifer"
status: converted
Accident Analysis and Prevention 36 (2004) 621–626
Factors associated with rear seating of children in motor vehicles: a study in two low-income, predominantly Hispanic communities
Jennifer Greenberg-Seth a, David Hemenway a,∗, Susan S. Gallagher b, Karen S. Lissy a, Julie B. Ross b
a Injury Control Research Center, Harvard School of Public Health, 677 Huntington Avenue, 3rd Floor, Boston, MA 02115, USA b Education Development Center Inc., 55 Chapel Street, Newton, MA 02458, USA
Received 13 January 2003; received in revised form 21 March 2003; accepted 31 March 2003
Abstract
This study examined child seating patterns in two predominantly low-income, Hispanic communities in Massachusetts. The purpose was to determine the factors associated with child rear seating in the community as a whole and for a subset of Hispanic motorists. Five hundred and five vehicles carrying child passengers and no adult other than the driver were observed in parking lots of fast food restaurants and grocery stores. Four hundred and thirty-two vehicle drivers agreed to be interviewed. A child was defined as a passenger younger than age 12 as determined by appearance and height (head below the vehicle headrest when seated). Variables under study included driver gender, age, ethnicity, and educational attainment; driver shoulder belt use; driver perception of passenger-side airbag presence; and the number and ages of children in the car. Overall, 51% of vehicles were observed with all children seated in the rear. In a bivariate analysis, child rear seating was strongly associated with female drivers (P = 0.01), younger drivers (P = 0.02), driver shoulder belt use (P < 0.00), perceived presence of a passenger-side airbag (P < 0.00), all children in the vehicle ≤6 years old (P < 0.00), and fewer than three children in the vehicle (P < 0.00). Non-Hispanic ethnicity was weakly associated with child rear seating (P = 0.09). In the multivariate analysis, only the association between child rear seating and all children in the vehicle ≤6 years old remained strong (RR 2.4, 95% CI 1.7–3.3). Factors associated with rear seating were similar between Hispanic and non-Hispanic motorists. Controlling for other factors, the presence of a passenger-side airbag appeared to be more strongly associated with rear seating among non-Hispanics compared to Hispanics (RR 1.5 versus 1.1). Fewer Hispanic motorists drove vehicles with all children in the rear compared with the non-Hispanic motorists (48% versus 56%); this difference appears to be associated with the finding that fewer Hispanic drivers wore their seat belt. For both Hispanic and non-Hispanic motorists, efforts to increase child rear seating should focus on older children, male drivers, and on increasing general motor vehicle safety behavior (e.g. seat belt use). Messages should be culturally appropriate and should emphasize driver seat belt use in conjunction with rear seating and the importance of rear seating regardless of the presence of a passenger-side airbag. © 2003 Elsevier Ltd. All rights reserved.
Keywords: Child passenger safety; Rear seating; Hispanics; Motor vehicle; Safety belt use; Airbags
1. Introduction
In the United States, motor vehicle injuries to child occupants under age 12 account for more than 1600 fatalities and 200,000 emergency room visits annually (Centers for Disease Control and Prevention, 2001). Hispanic children are at increased risk for both fatal and non-fatal injuries as motor vehicle occupants compared to non-Hispanic white children (Centers for Disease Control and Prevention, 2001). Baker et al. (1998) found that the death rate per billion vehicle-miles traveled was 60% higher for Hispanic children aged 5–12 compared to non-Hispanic white children.
Seating children in the back seat of motor vehicles has been shown to reduce the risk of injury or fatality in a crash (Williams and Zador, 1977; Huelke and Lawson, 1978; Evans and Frick, 1988; Berg et al., 2000; Glass et al., 2000; Durbin et al., 2001; Winston et al., 2001). Braver et al. (1998) found that the fatality risk was 36% lower for children aged 12 and under seated in back compared to those seated in front (95% CI for RR 1.49–1.64). This risk reduction was considerably larger for children riding unrestrained in vehicles with passenger-side airbags (95% CI for RR 1.63–2.78). The increasing number of vehicles with passenger-side airbags in recent years has raised the risk of injury and fatality to children who ride in the front seat and has focused attention on the role of child seating position in injury prevention (Graham et al., 1998).
∗ Corresponding author. Tel.: +1-617-432-4493; fax: +1-617-432-4494. E-mail address: hemenway@hsph.harvard.edu (D. Hemenway).
Despite the safety benefit of rear seating, studies have indicated that approximately one-third of US children travel in the front seat of motor vehicles (Segui-Gomez et al., 1998; Glass and Graham, 1999; Ferguson et al., 2000; Glass et al., 2000). This proportion varied across states. The percentage of children in the front seat increased to 56% when there were no adult or teenage passengers in the vehicle (Glass and Graham, 1999).
To date, few states have passed legislation mandating rear seating for children. In order to effectively promote rear seating for children, it is important to understand what factors are associated with this behavior. A few studies have investigated factors associated with rear seating using observational and crash fatality data. These studies showed that predictors of rear seating included the presence of adult or teenage passengers in the vehicle, female driver, driver seat belt use, younger age of children in the car, and passenger-side airbag presence (Glass and Graham, 1999; Ferguson et al., 2000; Wittenberg et al., 2001). No study has examined these factors for Hispanic populations or in samples where there were no adult or teenage passengers in the vehicle. The present study looked at child seating patterns in two predominantly low-income, Hispanic communities in Massachusetts. The purpose was to determine what factors are associated with child rear seating in the overall community and for a subset of Hispanic motorists.
2. Methods
2.1. Sample
Interviews were conducted with a convenience sample of motorists transporting child passengers in two Massachusetts cities. The cities were selected to ensure a sizable percentage of low income and Hispanic motorists. The Hispanic population in the two cities is comprised predominantly of Puerto Ricans and Dominicans. According to 2000 US Census data, Holyoke, MA has a population of approximately 40,000. Forty-one percent of the population is Hispanic and the median household income is US\$ 30,000 per year. The population of Lawrence, MA is approximately 72,000. Fifty-nine percent of the population is Hispanic and the median household income is US\$ 28,000 (United States Census Bureau, 2000).
Drivers of non-commercial passenger vehicles (e.g. sedan, wagon, sport utility vehicle, minivan, pickup truck) with Massachusetts license plates, child passengers, and no adult passengers were eligible for the study. Vehicles with any adult or teenage passengers were excluded based on previous research, which indicated that children almost always sit in the rear in this situation (Wittenberg et al., 2001; Segui-Gomez et al., 1998). A child was defined as a passenger younger than age 12 as determined by appearance and height (head below the vehicle headrest when seated). Passengers aged 12 and older were considered adults. At the start of the interview, participants were asked the ages of all passengers and if a passenger aged 12 or older was present, the interview was terminated and the observation was excluded from the sample.
2.2. Data collection
Several popular fast food restaurants and grocery stores in each city were asked to participate as data collection sites. Sites were selected in both higher and lower income areas of each city. Two sites in each city gave permission for interviews to be conducted in their parking lots.
Nine paid data collectors attended a standard 3 h training before data collection began. Data were collected in the summer of 2000, on Friday afternoons, Saturdays, and Sundays during daylight hours and in good weather. Two data collectors were stationed at each site, however only one collected data on each vehicle.
Data collectors observed motor vehicles as they entered the parking lot. When a vehicle met the inclusion criteria, a data collector approached the driver after he or she had parked and exited the vehicle to request participation in a brief interview about car safety. Drivers agreeing to participate received a US\$ 1 gift certificate to a local business establishment. The interview questions were translated so that interviews could be conducted in either English or Spanish. At least one interviewer per site was proficient in Spanish. Traffic flow was sufficiently light that all eligible vehicles were observed and approached for interview.
The study protocol was approved by the Institutional Review Board of the Harvard School of Public Health and the Human Subjects Committee of Education Development Center Inc. Oral consent was obtained from participants in the driver interviews.
2.3. Instrument
The data collection instrument consisted of observational elements and a 14-item driver interview. As vehicles entered the parking lot, data collectors recorded child seating patterns, driver gender, and driver shoulder belt use. Child restraint use was not recorded because appropriate use could not be determined by observation.
Drivers were asked questions about their vehicle, including whether they thought it had a passenger-side airbag. Driver perception of passenger-side airbag presence has been shown to correlate well with actual airbag presence (Wittenberg et al., 1999). Participants were asked their age, race/ethnicity, and educational achievement, as well as the ages of all children traveling in their vehicle. Each interview lasted approximately 3 min.
2.4. Analysis
We chose to use the vehicle as the unit of analysis to study the outcome of all children in the car seated in the rear. After analyzing the association with rear seating at different levels and finding no significant differences in the relationship with rear seating at different levels, some independent variables were dichotomized. Driver race/ethnicity was categorized as Hispanic or Non-Hispanic; driver age as ≤30 years old or >30 years old; driver education as college degree or no college degree; age of children in the vehicle was categorized as all children in the vehicle ≤6 years old or at least one child >6 years old. The cut points for the variables were determined by an analysis of the data. The cut points for the age of children in the car was also based on previous studies which found significant differences in child rear seating using this cut-off value (Wittenberg et al., 2001; Ferguson et al., 2000; Wittenberg et al., 1999). The total number of children in the vehicle was categorized as one child, two children, or three or more children.
The statistical significance of the observed association between independent variables and rear seating was determined using a χ2 test. A χ2 test was also used to determine the statistical significance of the associations among independent variables. Multivariate analysis between the dependent variable (all children seated in the rear) and associated independent variables was performed using a generalized linear model with a Poisson distribution. Confounding in the model was assessed by including various combinations of correlated variables in the model and evaluating the impact on other parameter estimates. The same process was used to evaluate the data for subsets of Hispanic motorists and Non-Hispanic motorists. All analyses were conducted using SAS software version 8.02 (SAS Institute, 1999–2001).
3. Results
3.1. Overall sample
Five hundred and five motor vehicles met the inclusion criteria for the study. Of those, 432 vehicle drivers agreed to participate for a response rate of 85.5%. In 51% of those vehicles, all children were observed seated in the rear. Child seating patterns were similar among refusals compared to those who participated in the interview. In both groups, 51% of vehicles were observed with all children seated in the rear. Forty-four percent of refusers were female compared to 42% of participants; 44% of refusers were observed wearing shoulder belts compared with 52% of participants.
Child seating patterns in the two cities were also similar. Responses from the two cities were combined.
Sixty-two percent of respondents were Hispanic and 52% were female. Thirty-five percent were ≤30 years old and 25% had a college degree. Forty-seven percent of drivers were observed wearing a shoulder belt; 33% reported that their vehicle had a passenger-side airbag; and 38% were driving vehicles with all children who were 6 years old. In 56% of the vehicles, only one child was riding in the vehicle with the driver.
In the χ2 analysis, positive associations were found between rear seating and female driver (P = 0.01), driver ≤30 years old (P = 0.02), driver shoulder belt use (P < 0.00), driver reported presence of a passenger-side airbag (P < 0.00), and all children in the vehicle ≤6 years old (P < 0.00). Hispanic ethnicity was negatively associated with rear seating although the relationship was not strong (P = 0.09). Vehicles carrying three or more children were less likely to have all children seated in the rear compared with vehicles carrying one child or two children (P < 0.00). The educational achievement of the driver was not significantly associated with child rear seating (P = 0.20) (Table 1).
In the multivariate analysis, the factor most strongly associated with child rear seating was having only young children (≤6 years old) in the vehicle (RR 2.4, 1.7–3.3). Driver reported passenger-side airbag (RR 1.3, 1.0–1.8) and driver shoulder belt use (RR 1.3, 1.0–1.8) were also associated with child rear seating (Table 1).
Female gender (RR 1.1, 0.8–1.5) and Hispanic ethnicity (RR 1.0, 0.7–1.4) were not associated with child rear seating
Table 1 Factors associated with child rear seating for the overall sample
| Factor | N | All children (%) seated in rear (P-value) |
Risk ratio (95% CI) |
|||
|---|---|---|---|---|---|---|
| Overall | 432 | 51 | ||||
| Driver gender | ||||||
| Female | 224 | 57 (0.01) | 1.1 (0.8–1.5) | |||
| Male | 179 | 45 | 1.0 | |||
| Missing | 29 | |||||
| Driver ethnicity | ||||||
| Hispanic | 267 | 48 (0.09) | 1.0 (0.7–1.4) | |||
| Non-Hispanic | 165 | 56 | 1.0 | |||
| Driver age | ||||||
| ≤30 years | 150 | 59 (0.02) | 1.1 (0.8–1.5) | |||
| >30 years | 277 | 47 | 1.0 | |||
| Missing | 5 | |||||
| Driver education | ||||||
| College degree | 106 | 57 (0.20) | 1.0 (0.7–1.4) | |||
| 324 | 49 | 1.0 | 324 | 49 | 1.0 | |
| Missing | 2 | |||||
| Driver shoulder belt use | ||||||
| Yes | 202 | 59 (<0.00) | 1.3 (1.0–1.8) | |||
| No | 280 | 41 | 1.0 | |||
| Missing | 44 | |||||
| Passenger airbag | ||||||
| Yes | 142 | 61 (<0.00) | 1.3 (1.0–1.8) | |||
| No | 280 | 46 | 1.0 | |||
| Missing | 10 | |||||
| All children ≤6 years old | ||||||
| Yes | 163 | 79 (<0.00) | 2.4 (1.7–3.3) | |||
| No | 264 | 34 | 1.0 | |||
| Total children in vehicle | ||||||
| Three or more | 67 | 30 (<0.00) | 0.9 (0.5–1.7) | |||
| Two | 125 | 58 | 1.4 (1.0–1.9) | |||
| One | 240 | 53 | 1.0 |
in the multivariate analysis, nor were driver age and educational achievement (Table 1).
Vehicles carrying two children were more likely to have both children seated in the rear compared to vehicles carrying only one child (RR 1.4, 1.0–1.9). There was no association between vehicles carrying three or more children and rear seating compared to those carrying only one child (RR 0.9, 0.5–1.7).
3.2. Hispanic and non-Hispanic subsets
Among the 267 Hispanic motorists in the sample, 54% were female, 41% were ≤30 years old, 16% had a college degree, 42% wore a shoulder belt, and 28% reported that their vehicle had a passenger-side airbag. Thirty-five percent of Hispanic motorists drove vehicles with all children ≤6 years old and 53% were driving with only one child in the vehicle.
Compared to the 165 non-Hispanics in the sample, Hispanic drivers were more likely to be ≤30 years old (41% versus 24%, P < 0.00) and less likely to have a college education (16% versus 39%, P < 0.00). Hispanics were less likely to wear a shoulder belt (42% versus 68%, P < 0.00) and less likely to drive a vehicle with a passenger-side airbag (28% versus 42%, P = 0.00). Hispanics were less likely to have only young children in the vehicle (35% ver-
Table 2 Multivariate associations between independent variables and child rear seating among Hispanic motorists and Non-Hispanic motorists
| Factor | | Hispanics
(N = 267) | | Non-Hispanics
(N = 165) | |
|------------------------------------------------------------------------------|----------|------------------------|----------|----------------------------|--|
| | N
(%) | Risk ratio
(95% CI) | N
(%) | Risk ratio
(95% CI) | |
| Driver gender | | | | | |
| Female | 54 | 1.2 (0.8–1.8) | 57 | 1.2 (0.7–1.8) | |
| Male | 46 | 1.0 | 43 | 1.0 | |
| Driver age | | | | | |
| ≤30 years | 41 | 1.1 (0.7–1.8) | 24 | 0.9 (0.5–1.5) | |
| >30 years | 59 | 1.0 | 76 | 1.0 | |
| Driver education | | | | | |
| College degree | 16 | 0.08 (0.4–1.4) | 39 | 1.1 (0.7–1.7) | |
|
sus 44%, P = 0.07) and more likely to drive cars with three or more child passengers (19% versus 10%, P = 0.03) (Table 2).
In the multivariate analysis, the factors associated with child rear seating were similar for Hispanic and non-Hispanic drivers. The presence of a passenger-side airbag seemed to be more strongly associated with rear seating among non-Hispanics (Table 2).
4. Discussion
Child rear seating is a simple, yet effective way of reducing the risk of injury and fatality to children in a crash. Studies have found that children are more likely to be seated in the rear if there is an adult or teenage passenger in the vehicle (Segui-Gomez et al., 1998; Wittenberg et al., 1999, 2001). This study is the first population-based survey of child seating patterns that looked specifically at vehicles with only child passengers to determine which factors may influence the decision of where to seat children.
It is also the first to examine factors associated with rear seating in a subset of Hispanic motorists. Other studies have shown that Hispanic children are at increased risk for injury and fatality in motor vehicle crashes (Baker et al., 1998; Centers for Disease Control and Prevention, 2001), however, no one has examined the risk factors for this population. Hispanics in our study were less likely to have all children seated in the rear compared to the sample as a whole (48–56%). Hispanic drivers were also less likely to wear shoulder belts; including the shoulder belt variable in the multivariate analysis eliminated the significance of the ethnicity variable.
Overall, a rear seat was available 91% of the time and results were similar if we excluded the 9% of vehicles where a rear seat was not available. The factor most strongly associated with child rear seating for both Hispanic motorists and non-Hispanic motorists was having only young children in the vehicle. In vehicles carrying only one child, this relationship remained strong; 78% of 0–6-year old were seated in the rear compared to 25% of 7–12-year old. Previous research has also documented that children aged 7–12 are less likely than younger children to be seated in the rear seat (Ferguson et al., 2000). Similar patterns have also been found in restraint use for children, with older children less likely to be restrained (Haaga, 1986; Russell et al., 1994; Agran et al., 1998).
The perceived presence of a passenger-side airbag was strongly associated with child rear seating overall. A similar effect of a passenger-side airbag was also found in studies by Wittenberg et al. (1999) and Ferguson et al. (2000). This finding may be due to national efforts to educate parents about the dangers of airbags to children. The perceived presence of a passenger-side airbag was more strongly associated with rear seating among non-Hispanic motorists compared to Hispanic motorists. This may suggest the messages about the danger of airbags to children have not been effective in reaching this Hispanic population. In our study, the presence of a passenger-side airbag was also associated with newer vehicles (RR 7.2, 6.5–7.9), which may be a proxy for socio-economic status. Newer vehicles have been previously associated with child rear seating (Wittenberg et al., 2001) as well as with child restraint use (Agran et al., 1998).
We found that driver shoulder belt use was associated with rear seating. Edwards and Sullivan (1997) reported that restrained drivers were more likely to place children in the rear seat. It has also been widely documented that driver seat belt use is significantly correlated with child restraint use (Russell et al., 1994; Edwards and Sullivan, 1997; Agran et al., 1998; Ferguson et al., 2000; Eby et al., 2001).
In our sample, females were more likely to seat their children in the rear. The association was weakened in the multivariate model, in part due to a positive association between female gender and driver shoulder belt use (52% of females were belted compared to 44% of males). Previous studies have found that female drivers are more likely to use seatbelts (Insurance Institute for Highway Safety, 2001) and to use child safety seats (Agran et al., 1998; Eby and Kostyniuk, 1999).
Previous studies have found that child safety seat use is positively associated with younger drivers (Agran et al., 1998; Eby and Kostyniuk, 1999), and with adult educational attainment (Russell et al., 1994). We found an association between younger drivers and child rear seating in the bivariate analysis, but there was no association in the multivariate analysis. Younger drivers were more likely to carry younger children compared to older drivers (51% versus 33%), and the age of children in the vehicle was a stronger determinant of child rear seating than the age of the driver. We did not find an association between driver educational attainment and child seating patterns.
Other studies have found that child restraint use decreases with increased vehicle occupants (Agran et al., 1998; Ramsey et al., 2000). We did not observe a linear association between number of children in the vehicle and rear seating. With three or more children in the vehicle, drivers were less likely to seat all children in the rear. However, when two children were in a vehicle, they were slightly more likely to be rear seated compared to when there was only one child in a vehicle.
4.1. Limitations
Our study is limited by several factors. The study was conducted in two cities in Massachusetts with large proportions of low-income and Puerto Rican and Dominican Hispanic residents. Consequently, results may not be generalizable to other communities. Additionally, interviews may not be representative since they were conducted with a convenience sample of drivers (on weekends at fast food restaurants and grocery stores). No identifying information about vehicles or drivers was recorded so a family could have been observed and interviewed on more than one occasion. Our study has limited power because of the small sample size, particularly in the Hispanic and non-Hispanic subsets.
While some variables were directly observed (e.g. child seating, driver shoulder belt use) others relied on self-report (e.g. driver educational attainment, passenger airbag, child age) and results are subject to the limitations of such self-report information. We did not measure inter-rater reliability among the data collectors, however they received consistent training and a practice data collection session was held in which data collectors were observed and feedback was given.
4.2. Implications
Drivers appeared to be more likely to seat younger children (age ≤6) in the back seat of motor vehicles compared to older children. Additionally, driver gender, driver shoulder belt use, and the perceived presence of a passenger-side airbag were associated with child rear seating. Factors associated with rear seating were similar for Hispanics and non-Hispanics.
Efforts to increase child rear seating should focus on older children. While Hispanics were only slightly less likely to seat their children in the rear, it is important to direct education efforts toward this group because Hispanic children are more likely to be injured or killed in car crashes than non-Hispanic whites. Male drivers should also be specifically targeted because they are less likely to seat their children in the rear and are also less likely to wear seat belts themselves.
Messages to promote rear seating should be culturally appropriate, should emphasize driver seatbelt use in conjunction with rear seating, and should address the importance of rear seating regardless of the presence of a passenger-side airbag.
Additionally, as the fleet of passenger vehicles ages, more vehicles with passenger-side airbags will be entering the used sales market. Several major car companies have introduced child passenger safety campaigns that include information about the dangers of seating a child in front of an airbag and promote rear seating (e.g. Daimler-Chrysler's "The Back is Where it's At" and "Fit for a Kid" campaigns and Ford's "Boost America" campaign). We speculate that people who buy vehicles through independent used car dealers or through private transactions may not be as exposed to these messages. Efforts should be made to target this population.
Acknowledgements
This research was supported by Grant # H28/CCH117598- 01 from the Centers for Disease Control and Prevention. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the CDC. The authors would like to thank John D. Graham who co-conceptualized and initiated the project; George Gray, the Harvard Center for Risk Analysis, and the Harvard Injury Control Research Center for their ongoing support; Matthew Miller for his help with statistical methods; Jennifer Koenig, Susan Lew, Eve Wittenberg, and Roberta Glass for their work in the early stages of the project; and the managers of local business establishments who allowed us to use their parking lots as sites for the driver interviews.
References
- Agran, P.F., Anderson, C.L., Winn, D.G., 1998. Factors associated with restraint use of children in fatal crashes. Pediatrics 102 (3), e39.
- Baker, S.P., Braver, E.R., Chen, L.H., Pantula, J.F., Massie, D., 1998. Motor vehicle occupant deaths among Hispanic and black children and teenagers. Arch. Pediatr. Adolesc. Med. 152 (12), 1209–1212.
- Berg, M.D., Cook, L., Corneli, H.M., Vernon, D.D., Dean, M.J., 2000. Effect of seating position and restraint use on injuries to children in motor vehicle crashes. Pediatrics 105 (4), 831–835.
- Braver, E.R., Whitfield, R., Ferguson, S.A., 1998. Seating positions and children's risk of dying in motor vehicles crashes. Inj. Prevent. 4 (3), 181–187.
- Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS) [online], 2001 [accessed 2002 Nov 24]. National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. Available from: URL: http://www.cdc.gov/ncipc/wisqars.
- Durbin, D.R., Elliot, M., Arbogast, K.B., Anderko, R.L., Winston, F.K., 2001. The effect of seating position on injury for children in side impact collisions. Annu. Proc. Assoc. Automot. Med. 45, 61–72.
- Eby, D.W., Kostyniuk, L.P., 1999. A statewide analysis of child safety seat use and misuse in Michigan. Accid. Anal. Prevent. 31 (5), 555–566.
- Eby, D.W., Kostyniuk, L.P., Vivoda, J.M., 2001. Restraint use patterns for older child passengers in Michigan. Accid. Anal. Prevent. 33 (2), 235–242.
- Edwards, J., Sullivan, K., 1997. Where are all the children seated and when are they restrained? Report Number 971550. Society of Automotive Engineers Inc., Warrendale, PA.
Evans, L., Frick, M.C., 1988. Seating position in cars and fatality risk. Am. J. Public Health 78, 1456–1458.
Ferguson, S.A., Wells, J.K., Williams, A.F., 2000. Child seating position and restraint use in three states. Inj. Prevent. 6 (1), 24–28.
Glass, R., Graham, J.D., 1999. Kids at risk: where children sit in passenger vehicles. J. Safety Res. 30 (1), 17–24.
Glass, R.J., Segui-Gomez, M., Graham, J.D., 2000. Decisions about seating location, airbag exposure, and restraint use. Risk Anal. 20 (4), 521–527.
Graham, J.D., Goldie, S.J., Segui-Gomez, M., Thompson, K.M., Nelson, T., Glass, R., Simpson, A., Woerner, L.G., 1998. Reducing risks to children in vehicles with passenger airbags. Pediatrics 102 (2), e3.
Haaga, J., 1986. Children's seatbelt useage: evidence from the National Health Interview Survey. Am. J. Public Health 76 (12), 1425– 1427.
Huelke, D., Lawson, T.E., 1978. The rear seat automobile passenger in frontal crashes. In: Proceedings of the 22nd Conference of the American Association for Automotive Medicine. Morton Grove, IL.
Insurance Institute for Highway Safety. Fatality facts [online], 2001. Available from URL: http://www.highwaysafety.org/safety facts/ safety.htm.
Ramsey, A., Simpson, E., Rivara, F.P., 2000. Booster seat use and reasons for nonuse. Pediatrics 106 (2), E20.
Russell, J., Kresnow, M., Brackbill, R., 1994. The effect of adult belt laws and other factors on restraint use for children under age 11. Accid. Anal. Prevent. 26 (3), 287–295.
SAS Institute, Inc., 1999–2001. SAS version 8.02. The SAS Institute, Cary, NC.
Segui-Gomez, M., Glass, R., Graham, J.D., 1998. Where children sit in motor vehicles: a comparison of selected European and American cities. Inj. Prevent. 4, 98–102.
United States Census Bureau. Census 2000. American Factfinder database [Online]. 2000 [cited 2002 June 10]. Available from: URL: http://factfinder.census.gov/servlet/BasicFactsServlet.
Williams, A., Zador, P., 1977. Injuries to children in automobiles in relation to seating location and restraint use. Accid. Anal. Prevent. 9, 69–76.
Winston, F.K., Durbin, D.R., Kallan, M.J., Elliott, M.R., 2001. Rear seating and risk of injury to child occupants by vehicle type. Annu. Proc. Assoc. Automot. Med. 45, 51–60.
Wittenberg, E., Goldie, S.J., Graham, J.D., 2001. Predictors of hazardous child seating behavior in fatal motor vehicle crashes: 1990–1998. Pediatrics 108 (2), 438–442.
Wittenberg, E., Nelson, T., Graham, J.D., 1999. The effect of passenger airbags on child seating behavior in motor vehicles. Pediatrics 104 (6), 1247–1250.