Factors Influencing Pediatric Injury in Side Impact Collisions

Abstract

Side impacts collisions pose a great risk to children in crashes but information about the injury mechanisms is limited. The heights and weights of children vary widely and as a result, the injury patterns may vary across the pediatric age range. This study involves a case series of children in side impact collisions who were identified through Partners for Child Passenger Safety, a large child-focused crash surveillance system. The aim of the current study was to use in-depth crash investigations to identify injury mechanisms to children in side impact collisions.

93 children in 55 side impact crashes were studied. 23% (n=22) of the children received an AIS ≥2 (clinically significant) injury. In these 22 children, head (39%), extremity (22%), and abdominal injuries (17%) were the most common significant injuries. The cases revealed that serious injuries occur even in minor crashes. Cases that illustrate body region-specific injury mechanisms are discussed.

Side impact crashes as compared with other impact directions have been shown to have higher rates of death and serious injury for adult occupants. Analysis of the Fatality Analysis Reporting System (FARS) shows that side impacts represent 27% of all occupant fatalities. [www.nhtsa.dot.gov 2000] In a study of police reported crashes, side impacts represented 19% of the crashes but 32% of the fatalities. [Vander-Lugt 1999] Dischinger et al. showed in a study of trauma patients, drivers in near-side impacts had twice the mortality rate than those in frontal impacts. [Dischinger, Cushing et al. 1993] For non-fatal injuries, occupants in side crashes sustained an Injury Severity Score (ISS) of 25 compared with an ISS of 20 for crashes of other impact directions. In particular, occupants in side crashes experienced more significant chest and intra-abdominal injuries. [McLellan 1996]

This increased risk has also been demonstrated for children. Agran and colleagues examined injury patterns in children aged 4 to 9 years restrained in motor vehicle crashes through a hospital based monitoring system. They found that children age 5 to 9 in near side impacts sustained injuries with a maximum Abbreviated Injury Scale (MAIS) score of 2+ at a far greater rate than other impact directions (41% versus 15% for frontal impacts and 3% for rear impacts). [\Agran, Winn et al. 1989] Langweider and colleagues stated that “side impacts are the most substantial risk for children” as evidenced by the fact that 35% of restrained children in their study with AIS 3+ injuries were in a side impact. [Langweider 1989; Langweider 1994;Langewieder, Hell et al. 1996] They placed the importance of side impacts for children as equal to that of frontal impacts with respect to serious injury risk. Recent FARS analysis for child occupants calculated the fatality rate for children in side impacts at 30% compared to 17% for frontal impacts. [Braver, Whitfield et al. 1998] This increased potential for harm points to the importance of side impacts as a focus of study for the safety community. [Vander-Lugt 1999]

Biomechanically, children encompass a wide range of variability in terms of height, weight, and other factors that may put them at risk for injury in side impact collisions. Although we know that children are at increased risk for injury and death in side impacts, little information is available about the specific injury patterns. This information is needed to inform the current discussions about side impact safety for these youngest passengers.

Agran et al. demonstrated that all the serious injuries (AIS>2) in side impacts were to children seated on the side of impact. [Agran, Winn et al. 1989] The most frequent serious injuries were to the head and face; extremity and torso injuries were not common. This work was performed however, during the 1980’s. Since then there have been substantial vehicle safety modifications such as the implementation of FMVSS 214. [Vander-Lugt 1999] Furthermore, most of the children in Agran’s study were in lap belts since that was the primary form of restraint in the rear seat during that time period. It is important to assess if similar injury patterns exist with changes to current vehicles, restraint systems, and restraint behavior. In addition, other pediatric age groups need to be investigated to explore age-specific injury mechanisms.

Previous work by Langwieder et al. described the characteristics and injury mechanisms of a series of children in serious side impact collisions, focusing on crashes where child passengers received MAIS scores ≥ 2. While this study provided an important first look at injury mechanisms of children in side impacts, the study analysis was limited to cases where restraint misuse, ejection, and catastrophic intrusion were not present. Only children with MAIS ≥ 2 were included in analyses. Most importantly, the potential roles of misuse of child restraint systems (CRS) and of inappropriate restraint were not investigated. The authors stated that the differences in characteristics of children and of restraint use indicated that additional investigation is needed to fully understand injury mechanisms. [Langewieder, Hell et al. 1996]

Current large, population-based crash surveillance systems in the United States, such as NASS, that are needed to perform these analysis, however, are not designed exclusively for children and consequently, relatively few children are included. [Edwards and Sullivan 1997] Results specific to pediatric injury mechanisms are difficult to extract from this system.

In contrast, the Partners for Child Passenger Safety (PCPS) surveillance system, used in this study, focuses exclusively on children who are injured in motor vehicle crashes as well as those who are not. The PCPS system allows for the comparison of a wider range of child injury severity in minor, moderate, and severe motor vehicle crashes. CRS misuse and inappropriate restraint are carefully documented using interviews and evidence in the vehicle such as child contact points and belt marks. This in-depth investigation of children in crashes allows for a more detailed look at children in side impact crashes and provides data to evaluate the role of CRS misuse and inappropriate restraint in these crashes. The aim of the current study was to use in-depth crash investigations to identify injury mechanisms to children in side impact collisions.

METHODS

Cases were identified through PCPS, a multi-year collaboration between The Children’s Hospital of Philadelphia, the University of Pennsylvania, and State Farm Insurance Companies. The methods of this research project have been explained in detail previously. [Arbogast, Durbin et al., 1999; Winston, Durbin et al., 1999; Winston, Durbin et al., 2000] Briefly, claims qualifying for inclusion were those reporting a crash including at least one child occupant < 15 years of age riding in a model year 1990 or newer insured vehicle. Only children riding as occupants (non-drivers) in the insured vehicle were eligible for inclusion.

Qualifying claims were also limited to crashes that occurred in 15 states (DE, MD, NC, NJ, NY, PA, VA, WV, IL, IN, MI, OH, AZ, CA, NV) and the District of Columbia. On a daily basis, data from consenting claims were transferred electronically from all field offices to State Farm corporate headquarters in Bloomington, IL. Data included in this initial transfer were contact information for the driver/insured, the vehicle identification number, the ages and genders of all child occupants, and coded variables describing the severity of vehicle damage and the medical treatment received by all child occupants. In addition, text fields were available for the claim representatives to provide a brief description of the damage to the vehicle and the injuries sustained by the occupants. The data were then forwarded via email to researchers at The Children’s Hospital of Philadelphia/University of Pennsylvania (CHOP/Penn) on a daily basis, 6 days per week (no transmissions on Sundays).

Once received in Philadelphia, the data were reviewed at CHOP to identify potential cases of children in side impact crashes for in-depth crash investigation. Cases were screened via telephone to confirm the status and details of the case. Contact information from selected cases was then forwarded to a crash investigation firm and a full-scale on-site crash investigation was conducted using custom child-specific data collection forms.

Crash investigation teams were dispatched to the scenes to measure and document the crash environment, damage to the vehicles involved, and occupant contact points according to a standardized protocol. The on-scene investigations were supplemented by information from witnesses, crash victims, physicians, hospital records, police reports, and emergency medical service personnel. From this information, reports were generated that include estimates of the vehicle dynamics and occupant kinematics during the crashes and detailed descriptions of the injuries sustained in the crashes by body region, type of injury, and severity of injury.

Crashes with a principal direction of force (PDOF) of 45–135° and 225–325° were considered side impact crashes (see Figure 1). For side impact collisions, the estimation of crash severity by a calculation of delta V has many limitations [Frampton, Brown et al., 1998, Foret-Bruno, Hartemann, et al., 1980, Strother, Warner, et al., 1990], and therefore intrusion was chosen as a suitable proxy. In order to compare intrusion over a variety of vehicle sizes, the width of the vehicle was defined into 8 zones. [Perry , 1994] The first 7 were equi-distant across the impact half of the occupant compartment and the eighth zone represented the other half of the vehicle. The maximum intrusion was coded according to the highest zone to which the measured value corresponds. For this study, no or minor intrusion was defined as zones 0–2, moderate intrusion as zones 3–5, and severe intrusion as zones 6–8. Descriptive statistics were calculated and illustrative cases were selected to highlight interesting injury mechanisms.

FIGURE 1.

Top view of exemplar vehicle demonstrating PDOF criteria for side impact

RESULTS

Ninety-three children in 55 side impact crashes were included in the case series. The average delta V was 19.6 kph (range = 3–46 kph). 69% (n=38) of crashes resulted in no or minor intrusion (defined above) into the occupant compartment. 27% (n=15) resulted in moderate intrusion and 4% (n=2) resulted in severe intrusion.

The mean age of the child occupants in this series was 7.5 years. The age distribution for the child occupants is depicted in Figure 2. 72% (n=67) of the children were sitting in the rear seat of the vehicle at the time of the crash. 48% (n=45) of the children were seated on the side of the impact (near-side). Those children seated in the center of the rear seats were considered far-side for either a right or left side impact. Seating position by age group is contained in Figure 3.

FIGURE 2.

Age distribution

FIGURE 3.

Seating row by age group

The distribution of restraint type for each age group is presented in Figure 4. 55% of the children were appropriately restrained according to current US best practice guidelines. Best practice according to the National Highway Traffic Safety Administration (NHTSA) and the American Academy of Pediatrics (AAP) is as follows [Winston and Durbin , 1999]: infants should ride in a rear-facing child safety seat in the rear vehicle seat until they reach one year AND 9 kg (20 lbs). Children over 9 kg (20 lbs) AND at least one year old should ride in the rear vehicle seat in a child safety seat that faces the front of the vehicle until they weigh 18 kg (40 lbs). Children over 18 kg (40 lbs) should ride in a belt-positioning booster seat and use the vehicle’s lap and shoulder belts. Proper adult seat belt fit is not achieved until they have reached approximately 145 cm (57 inches) and 36 kg (80 lbs) (about 8 years of age).

FIGURE 4.

Restraint types by age group

62% (n=58) of children in the study received an injury. 22 of these children sustained 48 AIS>2 injuries. 42% of significant injuries occurred in crashes with no or minor intrusion. Among the 22 children with MAIS ≥ 2 injuries, the median MAIS of inappropriately restrained children was 4 while the median MAIS for appropriately restrained children was 2.

CHILDREN UNDER FOUR YEARS OLD

25% (8/32) of the children aged 4 and younger in the study were significantly injured (AIS>2). The MAIS scores ranged from 2 to 6. Table 1 shows the child, crash, and injury details for the significantly injured children in this age group.

TABLE 1.

Significantly injured 0 to 4 year-olds (n=8)

Case #	Age (yrs)	Ht (cm)	Wt (kg)	Restraint	Intrusion+	Vehicle year	Vehicle type	Near- or far-side	Injury (AIS)	Injury source
1	2	89	13	CRS	Moderate	96	4 door sedan	Far	-Closed basilar skull fx (3) -R frontal lobe contusion (3) -R orbital rim fx (2) -Unspecified LOC (2) -L distal tibia fx (3) -L fibula fx (2)	-RF seat back -RF seat back -RF seat back -RF seat back -Unknown -Unknown
2	4	122	20	L/S belt*	Moderate	91	4 door sedan	Near	-Splenic rupture (5)	-Arm rest on door
3	3	117	20	L/S (shoulder belt behind back)*	Moderate	93	Minivan	Near	-Spinal cord injury (4) -Concussion + brief LOC (2)	-Head impact -Right door panel
4	1.3	81	11	CRS	None/minor	91	4 door sedan	Far	-Concussion/no LOC (2)	-CRS
5	4	122	19	Low back booster	None/minor	97	4 door sedan	Far	-R orbit fx (2)	-Adjacent CRS
6	3	91	17	Unrestrained*	Moderate	96	4 door sedan	Near	-Ruptured spleen (3)	-Instrument panel
7	3	Unk	Unk	CRS	Severe	97	Minivan	Near	-Atlanto-occipital spine fx (6)	-Right door panel
8	4	110	21	L/S belt*	Moderate	93	2 door sedan	Near	-Pulmonary contusion (3) -Liver laceration-Gr II/III (2) -Closed head injury + LOC >1 hr (2)	-Door panel -Shoulder belt -Right door or B pillar

CRS = Child restraint system

L/S belt = Lap-shoulder belt

⁺To calculate intrusion, the width of the vehicle was defined into 8 zones. The first 7 were equi-distant across the impact half of the occupant compartment and the eighth zone represented the other half of the vehicle. The maximum intrusion was coded according to the highest zone to which the measured value corresponds. For this study, no or minor intrusion was defined as zones 0–2, moderate intrusion as zones 3–5, and severe intrusion as zones 6–8.

^* Inappropriate restraint for child’s age and weight based on guidelines from NHTSA and AAP [Winston and Durbin , 1999]

58% of the AIS >2 injuries to this age group were to the head, face, or cervical spine. 59% of the children received significant injuries in crashes with intrusion. Two children aged 4 and younger received significant injuries in crashes with no or minor intrusion. Both of these children were seated away from the impact, appropriately restrained based on age and weight and received AIS 2 head injuries. One was in a low back belt-positioning booster, moved away from the shoulder belt upon impact and struck an adjacent child restraint. She suffered a fracture to her right orbit. The other child was in a CRS and experienced a concussion – possibly from increased head excursion due to loose harness straps or from impact with the side wings of the CRS.

FIVE TO NINE YEAR OLDS

19% (6/31) of the children aged 5 to 9 years sustained significant injuries. The MAIS scores ranged from 2 to 5. Table 2 shows the crash, restraint, and injury details for the significantly injured children in this age group.

TABLE 2.

Significantly injured 5 to 9 year-olds (n=6)

Case#	Age (yrs)	Ht (cm)	Wt (kg)	Restraint	Intrusion+	Vehicle year	Vehicle type	Near- or Far-side	Injury (AIS)	Injury source
9	8	119	37	L/S belt	Moderate	96	2 door sedan	Near	-Concussion (2)	-Grille of intruding vehicle
10	5	104	23	CRS	None/minor	90	Minivan	Far	-Left ulna fx (2) -Left femur fx (2) -Right lower leg fx (2)	-Roof -Seat back -Seat back
11	8	119	28	L/S belt*	Moderate	95	4 door sedan	Near	-Closed head injury + LOC ~10 min (2)	-Intruding window frame
12	5	113	19	L/S belt*	None/minor	94	4 door sedan	Near	-Right frontal/parietal subdural hematoma (4) -Right temporal bone fx (3) -Left parietal bone fx (2)	-Interior door -Interior door -Interior door
13	8	140	24	Shoulder belt only*	Moderate	90	4 door sedan	Near	-Cranial trauma + LOC >96 hrs (5) -Splenic laceration-Gr III/IV (5) -Right superior/inferior pubic ramus fx (2) -Complex facial laceration (2)	-Hood of intruding vehicle -Shoulder belt -Intruding door -Hood of intruding vehicle
14	7	124	38	L/S belt	None/minor	94	Minivan	Far	-Perforated jejunum (3) -2 mesenteric tears (3) -Omental tear (2)	-Seat belt -Seat belt -Seat belt

CRS = Child restraint system

L/S belt = Lap-shoulder belt

⁺To calculate intrusion, the width of the vehicle was defined into 8 zones. The first 7 were equi-distant across the impact half of the occupant compartment and the eighth zone represented the other half of the vehicle. The maximum intrusion was coded according to the highest zone to which the measured value corresponds. For this study, no or minor intrusion was defined as zones 0–2, moderate intrusion as zones 3–5, and severe intrusion as zones 6–8.

^* Inappropriate restraint for child’s age and weight based on guidelines from NHTSA and AAP [Winston and Durbin , 1999]

Head injuries were the most common AIS>2 injury (44%). Three of the 4 significant head injuries in this age group occurred in near-side crashes and were in crashes with intrusion.

TEN TO FIFTEEN YEAR OLDS

27% (8/30) of children age 10 to 15 years had clinically significant injuries. MAIS scores ranged from 2 to 4. Table 3 shows the crash, restraint, and injury details for the significantly injured children in this age group.

TABLE 3.

Significantly injured 10 to 15 year-olds (n=8)

Case #	Age (yrs)	Ht (cm)	Wt (kg)	Restraint	Intrusion+	Vehicle year	Vehicle type	Near- or Far-side	Injury (AIS)	Injury source
15	10	142	41	Unrestrained*	None/minor	90	Minivan	Near	-Non-displaced C2 fx (2) -Closed head injury + LOC (3) -Right frontal bone fx (2) -Major laceration to face (2) -Major laceration to left arm (2)	-Ground (ejected) -Ground -Ground -Glass -Left window
16	14	Unk	Unk	L/S belt	Moderate	93	4 door sedan	Near	-Pelvic fx (2) -Ruptured bladder (2) -Urethral disruption (2) -Head trauma + LOC (3)	-Intruding door -Intruding door -Intruding door -Intruding tree
17	10	127	34	L/S belt*	None/minor	91	Minivan	Near	-Left frontal lobe hemorrhage with epidural hematoma (4)	-Right window
18	14	157	54	L/S belt	None/minor	97	4 door sedan	Near	-Concussion with amnesia (2)	-B-pillar
19	15	Unk	Unk	L/S belt	Severe	97	4 door sedan	Near	-Pelvic fx (2) -Head trauma + LOC (2)	-Right door -Right door
20	13	183	100	L/S belt	Moderate	98	4 door sedan	Near	-Right clavicle fx (2)	-Intruding B-pillar
21	13	150	45	L/S belt	None	90	4 door sedan	Near	-Closed head injury w/unspecified intercranial hemorrhage (2)	-Side window
22	11	Unk	Unk	L/S belt	Moderate	96	2 door sedan	Near	-Femur fx (2)	-Right vehicle interior

L/S belt = Lap-shoulder belt

⁺To calculate intrusion, the width of the vehicle was defined into 8 zones. The first 7 were equi-distant across the impact half of the occupant compartment and the eighth zone represented the other half of the vehicle. The maximum intrusion was coded according to the highest zone to which the measured value corresponds. For this study, no or minor intrusion was defined as zones 0–2, moderate intrusion as zones 3–5, and severe intrusion as zones 6–8.

^* Inappropriate restraint for child’s age and weight based on guidelines from NHTSA and AAP [Winston and Durbin , 1999]

Although head injuries were still the most common significant injury (44%), extremity injuries occurred more often than in any other age group (36%). All 8 children with significant injuries were involved in near-side collisions and 6 of the children were appropriately restrained. 50% of the injuries were in crashes with no or minor intrusion.

BODY REGION

The most common body regions injured for all children in the study were the face (29%), extremities (29%), and head (18%). For the AIS>2 injuries, 39% were to the head, 22% were to the extremities, and 19% were to the abdomen or thorax. Figure 5 shows the number of children by age group who sustained a significant injury to each body region.

FIGURE 5.

Number of children who sustained a significant injury to the head, abdomen and thorax, and extremities

50% of the significant head injuries were sustained in crashes with no or minor intrusion. 50% of significant head, face, and spine injuries, and 29% of abdominal and thoracic injuries were sustained by children who were appropriately restrained for their age and weight. In contrast, 75% of significant extremity injuries were sustained by appropriately restrained children. Exemplar cases are presented below to illustrate common mechanisms of head, abdomen, and lower extremity injuries.

Case #1 – Head injury

A 5-year-old was sitting in the right front passenger seat of a 4-door sedan (case vehicle, Figure 6) traveling through an intersection. The front bumper of a minivan struck the right front side of the sedan. The sedan rotated counter-clockwise and side-slapped the minivan. It then proceeded forward and struck the left rear and right side of a second minivan. The delta v for the sedan’s first impact was 14 kph, a minor crash. No intrusion into the occupant compartment occurred in this crash.

FIGURE 6.

Case vehicle for exemplar case #1

The 5-year-old (114 cm, 19 kg) was using the available lap/shoulder belt. The child suffered a right frontal/parietal subdural hematoma (AIS 4), a right temporal bone fracture (AIS 3), and a right parietal bone fracture (AIS 2). These injuries were attributed to contact with the right interior of the door. The child also sustained an abrasion to the torso on the left side (AIS 1), which was attributed to the seat belt webbing. The child was hospitalized for three days, two of which were in the Intensive Care Unit. Both the driver and the right rear seated occupant sustained only minor (AIS 1) injuries. In this case, the acceleration of the child towards the right front door may have been exacerbated by the deploying front passenger air bag.

Case #2 – Abdominal injury

A 4-year-old was seated in the left rear seating position of a four-door sedan (case vehicle, Figure 7). The sedan was attempting to make a right turn from a stop and pulled into the path of an on-coming four-door sedan. The second vehicle struck the case vehicle on the left side. The delta v for the crash was 18 kph and there was moderate intrusion into the occupant compartment on the left side.

FIGURE 7.

Case vehicle for exemplar case #2

The 4-year-old (122 cm, 20 kg) was using the lap/shoulder belt in the left rear seat of the vehicle. The child sustained a Grade III splenic laceration (AIS 5) most likely due to contact with the armrest on the interior door. The child also sustained a contusion to the left cheek (AIS 1) from the shoulder belt webbing, and a contusion to the right side of the abdomen (AIS 1).

Case #3 – Lower extremity injury

An 8-year-old was sitting in the right front seating position of a small sedan (case vehicle, Figure 8) that was approaching an intersection. As the driver attempted to stop at the intersection, the tires locked and the vehicle initiated a counter clockwise spin that resulted in entry into the opposite travel lane. The front of a large sport utility vehicle impacted the right side of the sedan. The delta v for the initial crash was 46 kph and there was severe intrusion into the right side of the occupant compartment.

FIGURE 8.

Case vehicle for exemplar case #3

The 8-year-old occupant (140 cm, 25 kg) was using the automatic shoulder belt in the right front seat without the manual lap belt. Upon impact, the child moved toward the right intruding door panel, which resulted in numerous injuries. The child sustained a fractured right rib (AIS 1), lacerations to the chin and posterior scalp (AIS 1), contusions to both lower legs, superior and inferior pubic rami fractures (AIS 2), a complex forehead laceration (AIS 2), and a Grade III-IV splenic laceration (AIS 5). The child also suffered moderate to severe cranial trauma and was unconscious for greater than 96 hours (AIS 5). The child was hospitalized for two weeks and was subsequently transferred to a specialized trauma recovery institution.

DISCUSSION

This study of children in side impact collisions demonstrates that children receive clinically significant injuries even in low severity crashes (based on intrusion). Appropriate restraint appeared to be an important factor in determining injury severity. While there was high restraint usage in the study population, many children were inappropriately restrained according to current U.S. best practice guidelines. Children who were inappropriately restrained sustained more serious injuries than children who were using the recommended restraint system for their age and weight.

In particular, the use of adult seat belts by young children appears to put them at risk in side impact collisions. According to recommendations from the NHTSA and AAP, the vehicle seat belt will likely not fit children properly until they have reached approximately 145 cm (57 inches) and 36 kg (80 lbs) (about 8 years of age). [Winston and Durbin , 1999] Using an improperly fitting adult seat belt may lead to increased head and upper torso excursion.

Previous work has documented that the effect of restraints for adult occupants in near-side lateral crashes is minimal. [Rouhana 1985; Farmer, Braver et al. 1997] In this impact configuration use of a restraint protects against ejection and injuries due to secondary contacts. There is a documented benefit to restraint use in near-side crashes that have an oblique PDOF (primary impact outside the occupant compartment). A study by Otte and colleagues showed that 64% of the near side belted occupants in oblique lateral collisions had no injury while only 30% of the near side belted occupants in perpendicular impacts had no injury. [Otte 1984] Jones documented a shift towards lower injury severity (from MAIS 4 to MAIS 3) in belted versus unbelted near side occupants. (Jones 1982)

In contrast, belt use on the non-struck side has been shown to be much more effective in both studies of real world crashes as well as laboratory simulations with anthropometric dummies. [Otte 1984; Rouhana 1985; Frampton, Brown et al. 1998] Seat belts reduced the incidence of MAIS 2+ injuries from 32% to 16% for adult occupants seated away from the side of impact. [Thomas 1999] Horsch documented that although the shoulder escapes from the shoulder belt in far side impacts of 60–90° PDOF, the kinetic energy of the upper body is significantly reduced. [Horsch 1980] Mackay illustrated that although there is an overall benefit of seat belt use for far-side occupants, these passengers often slip out of the belts, strike their head, and due to the rotation and movement of their upper torso, concentrate the lap belt loads on their abdomen. [Mackay 1993]

More case studies would need to be performed to formally assess the relationship between belt use and protection in side impacts for children. The advantage of our dataset is that it can evaluate an intermediate level of protection, the inappropriately restrained child. The current cases show a trend toward improved protection (a reduction of median MAIS from 4 to 2 among those with significant injuries) for children that are appropriately restrained compared to those that are inappropriately restrained. The benefit of restraint for adults in far side impacts may be diminished for children too small for adult belts. These children may slip out of the shoulder belt earlier and the reduction in kinetic energy of the torso seen in adults may not be realized. Furthermore, hypothesized but not yet confirmed differences in injury tolerance between adults and children may result in more abdominal injuries with the same amount of force applied. The cases in this study support this hypothesis as evidenced by the high incidence of head and abdominal injuries in inappropriately restrained children. Pretensioners may be a solution but current studies indicate that the reduction in head excursion is at the expense of higher neck loads. [Stolinski 1999]

As in previous studies of side impact, the majority of the significant injuries were to the head (39%). There were 4 children 4 years of age or less who sustained AIS≥2 head injuries in the study. Two of four of these children were seated away from the point of impact (far side), were appropriately restrained, and struck their heads on the interior of the vehicle. These children were sitting in child safety seats, which if properly used, should reduce head excursion and prevent the child from striking either his own seat or the back of the seat in front of him as these children did. It is possible however, that the harnesses of the child safety seats were not tight which would allow for increased head excursion upon impact. Looseness of the harness has been documented by previous studies as an extremely common misuse. [http://www.safekids.org/buckleup/study.html , 1999;Arbogast, Durbin et al., 2000] In addition to looseness of the harness, looseness of the CRS attachment can be another important factor that may lead to increased head excursion.

Children ages 5 through 15 who sustained head injuries as their most significant injury (n=9) were all involved in near-side crashes. With one exception (an ejection of an unrestrained child), all of these children were restrained and were injured by contact with interior side of the vehicle. Two children sustained AIS 4 head injuries from contact with the side window or interior door surface in a crash with little intrusion. Overall, 50% of the serious head injuries occurred in crashes with no or minor intrusion. Because of their smaller stature and lower sitting height, children have the potential to sustain serious head injuries from contact with the interior side of the vehicle in even minor crashes. The US Federal Motor Vehicle Safety Standard that covers occupant protection and results in padding of interior surfaces (FMVSS 201) does not apply to the interior door structure below the belt line where many children may have head contact. [United States Department of Transportation, National Highway Traffic Safety Administration]

The majority of the seriously injured children ages 5 to 9 years were seated on the side of impact (4/6). These results are remarkably similar to cases investigated by Agran and colleagues that showed all of the seriously injured children aged 4 to 9 years were in near side crashes. [Agran, Winn et al. 1989] These results are striking in that Agran’s work examined crashes in 1980’s model year cars while this study looked at 1990’s model year cars. The two children in our study in this age group who did not fit the scenario described by Agran et al. and were injured on the non-struck side were both children with special health care needs. One was a child with musculo-skeletal deficiencies that prevented him from being restrained in a typical CRS. He was seated in a special needs CRS that came loose upon impact. The other child had a Meckel’s diverticulum (a defect of the gastro-intenstinal wall) that may have put him at risk in high force situations such as motor vehicle crashes. These were the only two children with special needs. More research needs to be performed to fully understand how these children with special health needs respond to the forces of a crash.

While significant abdominal injury in young children using seat belts has been well documented [Gotschall, Better et al., 1998], exemplar case #3 illustrates another potential mechanism for abdominal injuries in young belted children. The young child in this case was inappropriately restrained in a lap and shoulder belt (appropriate restraint would be a belt positioning booster seat). We hypothesize that the position of the armrest on the door relative to the location of the child’s spleen was an important factor in injury causation. Previous work has documented that in a side impact, discontinuities in the loading surface (both indentations as well as protuberances) and changes in interior compliance can cause significant injury. [Rouhana 1989; Lau 1991] Due to differences in sitting height, the armrest may have contacted an adult passenger much lower on the abdomen, most likely interacting with the bones of the pelvis, rather than the internal abdominal organs. Most of the serious injuries to the abdomen were sustained by children who were inappropriately restrained for their age.

A total of 6 children in the study sustained AIS ≥ 2 lower extremity injuries. Three of these children were ages 10 to 15 years and were appropriately restrained in lap and shoulder belts. All 3 were on the side of impact and suffered pelvic or femur fractures due to contact with the intruding door structure. These cases as well as exemplar case #2 illustrate the importance of contact with the interior vehicle surface as it moves inward due to intrusion in side impact crashes. Researchers have shown that the most important parameter in assessing injury risk in side impacts is velocity with which the occupant strikes the interior. [Strother 1990] When load is applied to the door, the force from the striking vehicle is often not adequately transferred through the door to the center of gravity of the struck vehicle and is subsequently transferred to the occupant. The time point along the velocity time profile of the door when contact between the occupant and the door occurs determines the magnitude of that force. [Warner 1990]

Of note, extremity injuries, and in particular pelvic injuries, did not occur in children less than 8 years. One possible reason for this result is that the younger and smaller children still in adult seat belts have more space between them and the intruding door and as a result, the contact between the child and the door may have occurred farther along the velocity time history of the door. It has been shown that delaying the occupant contact with the door significantly reduces the energy transferred to the near side passenger. [Lau 1991] However, an initial standoff between the occupant and door may be detrimental. [Rouhana 1989] A second hypothesis is that inherent biomechanical differences with age provide protection to the young school age children in this impact configuration while allowing for injuries to occur in young teenagers. More research must be conducted to clarify these relationships. It is difficult to assess the exact position of the occupant at the time of impact using crash investigation methodology. A mathematical simulation parametrically measuring the effect of a change in position of the child with respect to the door may answer some of these questions.

The most ideal countermeasure for side impacts would reduce the energy transferred to the occupant upon impact. Decreasing the peak velocity of the door by increasing door structure may be a possible solution, though one which may be difficult to obtain in design. Increased padding may be another alternative, however there is conflicting evidence of its benefit for reducing thoracic injuries. [Lau 1991] Research has shown that increased padding decreases head injury measures such as peak head acceleration and HIC. [Barbat 1995] More padding where children most likely have head impact may reduce the incidence of these injuries. Further understanding of how children sit with respect to the door structure and subsequent education to encourage safer positioning is necessary.

SUMMARY

This study highlighted mechanisms of head, abdominal, and lower extremity injuries in children involved in side impact crashes. These cases demonstrate the importance of contact with the intruding door in lower extremity injuries and abdominal injuries. In addition, the occurrence of head injuries in minor crashes (as indicated by little or no intrusion) suggests that the interior of vehicles should be designed to be more child occupant friendly. Specifically, efforts should be made to control the velocity and magnitude of the crush so that the force transferred from interior contacts is minimized.

Side impacts are an extremely complex issue in occupant protection. The results of this study have provided guidance for further research and offered insight into relevant factors for child occupant protection in side impacts. Prior to this study, the role of appropriate versus inappropriate restraint for children in side impact crashes was unclear. Most research efforts had focused on the importance of properly fitting restraints in frontal crashes and the benefit of restraint compared to no restraint in side crashes. Although the results of this study are not conclusive, they suggest that appropriate restraint for the child’s age and size provides protection in side impacts. Specifically, in these cases, when an injury occurs to an appropriately restrained child, it is of lesser severity than when the child in not restrained according to best practice. The specifics of restraint performance in near and far side crashes for children as well as investigation into whether restraint use is a proxy for another determinant factor deserves additional attention. Future work will compare national population-based data to our study population of children in side impact crashes and further delineate age-specific and body region specific injury mechanisms.

(Presenter: Kristy Arbogast)

Uwe Meissner: You mentioned in the beginning that you had a 55% appropriately restrained population. How did you determine that because when you go to the checkpoints where the installation of the child seat is checked, there is a failure rate of 95–98% so I cannot see that you have a total 55% success rate.

K. Arbogast: I want to distinguish between appropriate restraint and proper restraint. What I was defining was that the children were in the restraint system that was appropriate for their age and weight. I was not commenting on whether they were using that properly. We have that information but have not analyzed that yet so I agree with you that proper use is a different ballgame.

Sue Baker: I want to thank you for what I think is one of the most significant papers here and say that it was a beautiful presentation with very thoughtful interpretations. It’s nice to see someone really thinking about what those data mean and how they are going to protect our children in the future.

K. Arbogast: Thank you.

Unidentified questioner: What in the child restraint system lowered the AIS from 4 to 2? You mentioned that the AIS for unrestrained or inappropriately restrained occupants was an average of 4, and then you said for properly restrained children, it was an average of 2, or median of 2. Was there anything specific in the child restraint system that lowered that because typically in a frontal system, you would expect the restraint to work more effectively, but in side impacts, on an adult, the seat belt wouldn’t necessarily offer an improvement. So what in the child restraint system offered the improvement?

K. Arbogast: I agree with your comment. In fact, we were perplexed when we saw that there was a difference between appropriate and inappropriate restraint. Most of the kids who had the AIS 2 injuries were not in child restraints but really because of how it fell out, we were looking at kids who were appropriately restrained in seat belts versus inappropriately restrained in seat belts. One guess is that because of where the seat belt rests on their lower abdomen, they’re held more in place than possibly an adult might be, but we haven’t really figured out why that result came out. Another idea we had was that restraint may be a proxy for another determinate factor that we haven’t really teased out yet.

Same questioner: So you haven’t gone back to the cases to look at specifically why a child didn’t get a higher injury value when an unrestrained child did.

K. Arbogast: On first look, nothing stood out.

Narayan Yoganandan: Maybe I missed this. Was it all AIS 2+ injuries you looked at?

K. Arbogast: Yes. The 93 kids were all injuries, and 23% of them had AIS 2+.

N. Yoganandan: Did you have a chance to look at the AIS 3+ group?

K. Arbogast: I do know that of those kids with AIS 2+, it was a 60/40 split between 3 and greater and just those with AIS 2. And mainly those with just AIS 2, these were extremity fractures and those kids with concussion and brief loss of consciousness. But we haven’t looked at those three discrete groups yet.

N. Yoganandan: So your spectrum may change of shortening the abdominal bars and increasing the lower extremity bars.

K. Arbogast: Yes, particularly the extremities because a lot of our AIS 2 were extremities.

N. Yoganandan: Another comment I have is you may want to include the type of fracture when you group these from 0 to 40 lbs. The mechanism of injury could be entirely different because of the ossification process which has not taken place particularly in the 5 or 6 year old or younger, but for the brain by the time you are 5 years old, you almost have 80–90% of your brain developed. We have to be careful when we mix this pediatric population and use only one AIS rating to look at the mechanisms.

K. Arbogast: That’s very true. It’s a very moving target as to whether a one year old is the same as a two year old, or even whether two one year olds are the same developmentally which makes it a challenge to amass enough cases to be able to cut it small enough.