Significance of Adult Pedestrian Torso Injury

Abstract

The current paper uses data from two trauma registries to evaluate the significance of adult pedestrian torso injury relative to head and lower extremity injuries and to determine the relative importance of injuries to individual torso organs/structures. Analyses are conducted with and without adjusting for striking vehicle body type (car versus LTV). Although the incidence of torso injury is approximately 50% higher in pedestrians struck by LTVs than in those struck by cars, torso injury appears to be as an important contributor to the overall cost of pedestrian morbidity as is lower extremity injury. The most frequently injured torso organs/structures include the rib cage, lung & pleura, and liver. The results indicate a need for an increased focus on the prevention of torso injury in the design of pedestrian safety countermeasures.

Full text

49th ANNUAL PROCEEDINGS
ASSOCIATION FOR THE ADVANCEMENT OF AUTOMOTIVE MEDICINE
September 12-14, 2005
SIGNIFICANCE OF ADULT PEDESTRIAN TORSO INJURY
B. Johan Ivarsson
Basem Henary
Jeff R. Crandall
Center for Applied Biomechanics
University of Virginia
Charlottesville, Virginia, USA
Douglas Longhitano
Honda R&D Americas, Inc.
Raymond, Ohio, USA
ABSTRACT
The current paper uses data from two trauma registries to
evaluate the significance of adult pedestrian torso injury relative to
head and lower extremity injuries and to determine the relative
importance of injuries to individual torso organs/structures. Analyses
are conducted with and without adjusting for striking vehicle body
type (car versus LTV). Although the incidence of torso injury is
approximately 50% higher in pedestrians struck by LTVs than in
those struck by cars, torso injury appears to be as an important
contributor to the overall cost of pedestrian morbidity as is lower
extremity injury. The most frequently injured torso organs/structures
include the rib cage, lung & pleura, and liver. The results indicate a
need for an increased focus on the prevention of torso injury in the
design of pedestrian safety countermeasures.
Pedestrian-motor vehicle trauma is a common source of morbidity
and mortality in all motorized societies throughout the world. More
than a third of the approximately 11.2 million people that are killed
or injured in road traffic crashes every year are pedestrians [Crandall
et al., 2002]. Considering fatalities only, approximately 760,000 or
65% are pedestrians [World Bank, 2001]. In the US alone,
approximately 70,000 or 2.4% of the 2,889,000 who were injured
and 4,749 or 11.1% of the 42,643 who were killed in road traffic
crashes during 2003 were pedestrians [National Highway Traffic
Safety Administration (NHTSA), 2005].

264
While automotive safety research traditionally has focused on
developing knowledge and systems for protecting vehicle occupants,
the frequency and severity of injuries resulting from pedestrian
crashes has led to an increased focus on pedestrian safety in recent
years. Based on the same safety design principles that have resulted
in substantial benefits for occupants, several automotive
countermeasure concepts for minimizing the frequency and severity
of injuries to the pedestrian lower extremities [Aldman et al., 1985;
Harris and Grew, 1985; Ishikawa et al., 1994; Nagatomi et al., 1996;
Detweiler and Miller, 2001] and head [Okamoto et al., 1994;
Fredriksson et al., 2001] have been proposed. A further step towards
reducing the frequency and severity of injuries to pedestrian
casualties is the pedestrian test protocol included in the European
New Car Assessment Programme (EuroNCAP). As part of a program
to provide consumers and manufacturers with information on the
safety performance of new cars sold in Europe, EuroNCAP evaluates
the vehicle aggressiveness towards the pedestrian lower extremities
and head by measuring the impact response of mechanical leg and
head forms propelled into the vehicle front and hood structures.
The majority of countermeasure concepts for pedestrian
safety as well as the EuroNCAP pedestrian test protocol focus solely
on the head and lower extremities. This focus most likely stems from
the findings of epidemiological studies indicating that the head and
lower extremities are the most frequently injured body regions in
pedestrian victims [Chidester and Isenberg, 2001; Mizuno, 2003;
Ballesteros et al., 2004]. While head and lower extremity injuries
both are major contributors to pedestrian morbidity, the results from
several epidemiological studies on pedestrian casualties indicate that
a substantial proportion of the non-trivial injuries are to the torso
region. Ashton et al. (1979) reviewed 2,066 pedestrian crashes that
occurred in Birmingham, United Kingdom in 1976. For the 15-59-
year-old victims struck by the front of cars and light goods vehicles,
chest injuries were sustained by 9.4% of the AIS 2+ injured non-
fatalities and by 58.6% of the fatalities. The occurrence of abdominal
injury for the same population was 4.9% and 27.6% among the AIS
2+ injured non-fatalities and fatalities, respectively. Civil (1986)
evaluated the 45 pedestrian victims administered to the resuscitation
room at Auckland Hospital in New Zealand during a six-month-
period in 1983 and reported that 27% and 22% had chest and
abdominal injuries, respectively. Hill et al. (1996) reviewed the 101
severe adult pedestrian cases (Injury Severity Score (ISS) > 15) that
occurred in Sydney, Australia during the period from July, 1991
through June, 1994 and reported that 60% of the victims sustained
chest injury, 24% abdominal injury, 11% AIS 3+ injuries to the liver
and/or spleen, 4% ruptured thoracic aorta, and 23% fractured
thoracolumbar spine. Brundage et al. (1998) reviewed the 220 cases

265
of adult pedestrian fatality that occurred in metropolitan Seattle/King
County during the period of 1990 through 1995 and reported that
13% had thoracic aorta lacerations. Harruff et al. (1998) analyzed the
records of 217 of the 220 fatalities reviewed by Brundage et al.
(1998) and reported that 73% suffered at least one chest injury and
56% at least one abdominal injury. In their analysis of 521 Pedestrian
Crash Data Study (PCDS) cases, Chidester and Isenberg (2001)
found that injuries to the torso region and cervical spine made up
23% of the AIS 2+ injuries. The corresponding numbers for head and
lower extremity injuries were 30% and 32%, respectively.
Results from recent epidemiological studies on pedestrian-
motor vehicle trauma suggest that there is a difference in the injury
pattern and outcome of pedestrians struck by regular passenger cars
as compared to those struck by a vehicle of LTV (Light Trucks and
Vans) body type which includes pick-up truck, sport utility vehicle
(SUV), and mini-van [Lane et al., 1994; Henary et al., 2003;
Ballesteros et al., 2004; Lefler and Gabler, 2004; Longhitano et al.,
2005a, b; NHTSA, 2005]. While all of those studies suggest that
LTVs present a greater risk of serious injury and fatality to
pedestrian casualties than passenger cars, some specifically indicate
that pedestrians struck by LTVs are of higher risk of sustaining
thoracic and/or abdominal injuries [Lane et al., 1994; Ballesteros et
al., 2004; Lefler and Gabler, 2004; Longhitano et al., 2005a,b]. The
elevated risk of pedestrian torso injury associated with LTVs in
combination with the fact that the vehicle fleets in the US and in
other countries are shifting towards a higher proportion of LTVs
suggest that increased focus should be put on pedestrian safety
countermeasures for prevention of torso injury. Using data from two
trauma registries of pedestrian casualties, the current paper aims to
evaluate the significance of adult (19-60 years of age) pedestrian
torso injury relative to head and lower extremity injury and to
determine the importance of injuries to individual torso organs/
structures. Analyses will be conducted both with and without
adjusting for striking vehicle body type (car versus LTV).
METHODS
DATA SOURCES - The PCDS trauma registry is a
compilation of detailed information on a total of 552 pedestrian
crashes that occurred during the period from 1994 through 1998 in
six metropolitan areas in the US [Chidester and Isenberg, 2001].
Among the variables specified is the body type of the striking
vehicle. The types and severities of the injuries sustained by the
pedestrian casualties are coded according to the 1990 revision of the
Abbreviated Injury Scale (AIS-90). A “pedestrian” is defined as any
person located in a traffic-way, on a sidewalk or path contiguous

266
with a traffic-way, or on private property. The striking vehicle had to
be forward moving and of model year 1990-1996. An additional
requirement on the striking vehicle was that the striking portion of
the vehicle structure had to be original equipment. Crashes in which
a person was lying or sitting while struck were not included. The
pedestrian impact had to be the only impact and the first point of
contact had to be forward of the top of the A-pillar. The PCDS data
are not weighted since the study was designed to be clinical rather
than providing a national sample of all US pedestrian crashes.
The Fairfax INOVA pedestrian trauma registry includes the
ICD-9-CM (International Classification of Diseases, Revision 9,
Clinical Modification) coded injuries sustained by pedestrian
casualties who were treated at the Fairfax INOVA hospital, a level
one trauma center in northern Virginia, during the period 1991-2002.
A pedestrian victim was defined in agreement with E-code 814.7,
which specifies “pedestrian dragged, hit, or run over by motor
vehicle”. In contrast to the PCDS, the Fairfax INOVA registry does
not include enough information on the striking vehicle to
differentiate between cars and LTVs.
FILTERING OF THE PCDS DATA - All cases involving
pedestrian victims of age 0-18 years and 61 years and older were
eliminated. Of the remaining 306 19-60-year-old victims, 208 were
struck by cars and sustained 1,538 injuries while the remaining 98
were struck by LTVs and sustained 1,011 injuries. Of those 2,549
injuries, 1,651 were of severity AIS 1 while two injuries were of
unknown severity (AIS 9). Those 1,653 injuries were eliminated
leaving 554 AIS 2+ injuries sustained by 106 victims struck by cars
and 342 AIS 2+ injuries sustained by 54 casualties struck by LTVs.
FILTERING AND CONVERSION OF THE FAIRFAX
INOVA DATA - All cases involving pedestrian casualties of age 0-
18 years and 61 years and older were eliminated. Of the remaining
979 19-60-year-old casualties, 103 were eliminated for fulfilling one
or more of the following exclusion criteria: struck by a vehicle other
than a passenger car or LTV, striking portion of the vehicle other
than the front, pedestrian sitting or lying while struck, pedestrian
pinned/crushed between two vehicles or between a vehicle and
another object, pedestrian struck multiple times, and pedestrian
grabbed and dragged by striking vehicle. The remaining 876
pedestrian casualties had a total of 3,978 ICD-9-CM coded injuries.
Using the software ICDMAP-90 [Johns Hopkins University,
Baltimore, MD], the set of ICD-9-CM codes assigned to each victim
was mapped to the corresponding set of AIS-90 codes. ICDMAP-90
offers a high and low severity mapping option. If an ICD-9-CM
descriptor contains two or more specific injuries of different severity
combined using the connector “or” or “and”, high severity mapping
will result in that ICDMAP-90 maps to AIS-90 using the highest

267
severity option, whereas the low severity option will result in that the
lowest severity option is used. For instance, the ICD-9-CM
description “Cerebral Laceration and Contusion” (code: 800.65)
refers to the presence of a cerebral laceration and/or a cerebral
contusion. Using low severity mapping, ICDMAP-90 returns the
AIS-90 code 140602 (Cerebrum contusion – Not Further Specified
(NFS)) of severity AIS 3, whereas high severity mapping returns the
AIS-90 code 140688 (Cerebrum laceration) of severity AIS 4. Using
the high severity mapping option, the mapping of the 3,978 ICD-9-
CM coded injuries resulted in a total of 4,339 AIS-90 coded injuries
of which 4,246 were of known severity (AIS 1+) and 93 of unknown
severity (AIS 9).
Forty-four of the injuries of unknown severity had invalid
AIS-90 codes and were therefore eliminated. The remaining 49
injuries of unknown severity had all been assigned AIS-90 code
115099 (Closed head injury – NFS). Those 49 casualties were traced
in the original trauma registry and it was found that all of them
originally had been assigned one of the ICD-9-CM codes described
as “Brain Injury - Not Elsewhere Classified (NEC)”. All the AIS-90
codes for the 49 casualties that had been assigned code 115099 were
then carefully reviewed. For the 24 cases in which the casualties had
been assigned at least one valid AIS 2+ head injury code in addition
to the 115099 code, the 115099 code was eliminated. For the
remaining 25 cases, the 115099 code was reassigned a severity of
AIS 2. Of the final 4,271 AIS 1+ injuries, 1,512 were of severity AIS
1 and therefore eliminated leaving a total of 2,759 AIS 2+ injuries
sustained by 693 casualties.
PROCESSING OF THE DATA – The filtered data from the
two trauma registries were broken down by severity and body region
into groups of torso, head & face, lower extremity, and other injuries.
A torso injury was defined as an injury to the thorax, abdomen and
pelvic content, thoracic spine, or lumbar spine while the lower
extremity group included injuries distal to the pelvis, i.e. injuries to
the thigh, knee, leg, and foot & ankle regions. The incidence and
frequency per victim of torso, head & face, and lower extremity
injury were then determined at various severity levels. In addition,
the costs of the AIS 2+ torso and head & face injuries relative to the
cost of the AIS 2+ lower extremity injuries were determined for both
trauma registries using the unit costs for comprehensive HARM from
crash injuries published by NHTSA (1996). Finally, the incidence of
injury to specific torso organs and structures were determined at
various severity levels. The abovementioned analyses were
conducted without adjusting for striking vehicle body type (non-
vehicle specific analysis) using the data from both trauma registries
as well as after stratifying for vehicle type (car vs. LTV) using the
data from PCDS only (vehicle specific analysis).

268
RESULTS
NON-VEHICLE SPECIFIC ANALYSIS – Figure 1
compares the incidence of at least one torso injury to the
corresponding incidence of head & face and lower extremity injury at
three different severity levels. As shown, head & face injury has the
highest incidence at all three severity levels except at the AIS 2+
level in the PCDS registry where the incidence of lower extremity
injury is slightly higher. However, the relative importance of torso
injury appears to increase with increasing injury severity level being
approximately equal to the importance of lower extremity injury at
the AIS 3+ level. At the AIS 4+ level, the incidence of torso injury is
second only to the incidence of head & face injury. Figure 2
compares the average numbers of torso, head & face, and lower
extremity injuries per pedestrian victim at various severity levels.
Again, the importance of torso injury is outnumbered by both head &
face and lower extremity injuries at the AIS 2+ level but second only
to the importance of head & face injury at the AIS 3+ and AIS 4+
levels. Figure 3 shows the total costs of the AIS 2+ torso and head &
face injuries relative to the total cost of the AIS 2+ lower extremity
injuries as determined using the using the unit costs for
comprehensive HARM from crash injuries [NHTSA, 1996]. While
head & face injury gives rise to a substantially higher cost than torso
and lower extremity injuries, it is interesting to note that the cost of
AIS 2+ torso injury exceeds that of AIS 2+ lower extremity injury by
as much as 92% in the PCDS registry. The same is, however, not true
for the Fairfax INOVA registry in which the cost of AIS 2+ torso
injury is 23% lower than the cost of AIS 2+ lower extremity injury.
0
5
10
15
20
25
30
35
40
45
50
PC-AIS 2+ FI-AIS 2+ PC-AIS 3+ FI-AIS 3+ PC-AIS 4+ FI-AIS 4+
Trauma registry and injury severity
Tor so
Head & face
Lower extremities
Fraction of victims with ≥ 1 injury (%)
Figure 1 – Fraction of the 19-60-year-old pedestrian victims with at
least one injury by body region, injury severity, and trauma registry
(PC = PCDS, FI = Fairfax INOVA).

269
0
0.2
0.4
0.6
0.8
1
1.2
PC-AIS 2+ FI-AIS 2+ PC-AIS 3+ FI-AIS 3+ PC-AIS 4+ FI-AIS 4+
Trauma registry and injury severity
Tor so
Head & face
Lower extremities
Average number of injuries per victim
Figure 2 – Average number of injuries per 19-60-year-old pedestrian
victim by body region, injury severity, and trauma registry (PC =
PCDS, FI = Fairfax INOVA).
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
PCDS Fairfax INOVA
Trauma r eg i s tr y
Relative cost
Tor so
Head & face
Lower extremities
Figure 3 – Total costs of the AIS 2+ torso and head & face injuries
relative to the total cost of AIS 2+ lower extremity injury for the 19-
60-year-old pedestrian victims in the two trauma registries.
Figure 4 shows the fractions of the pedestrian victims who
sustained at least one AIS 2+ injury to specific torso organs/
structures. The most commonly injured organ/structure at this level is
the rib cage with an incidence of 10.5% and 8% in the PCDS and
Fairfax INOVA registry, respectively. Other organs/structures of
high priority in both registries are the liver and lung & pleura. Figure
5 shows the fractions of the victims who sustained at least one AIS
3+ injury to specific torso organs/structures. Just like at the AIS 2+
level, the most commonly injured torso organ/structure at the AIS 3+

270
level is the rib cage with an incidence of 9.2% and 5.1% in the PCDS
and Fairfax INOVA registry, respectively. However, in contrast to
the AIS 2+ incidence distribution, the incidence of liver injury is
surpassed by that of lung & pleura injury and approximately equal to
that of spleen injury at the AIS 3+ level. Figure 6 shows the fractions
of the pedestrian victims who sustained at least one AIS 4+ injury to
specific torso organs/structures. Again, the rib cage is the most
commonly injured torso organ/structure although the incidence is low
in both the PCDS (3.9%) and Fairfax INOVA (2.1%) registries. The
organs of any importance at the AIS 4+ level are the lung & pleura
and, to a lesser extent, the spleen, thoracic aorta, and liver.
0
1
2
3
4
5
6
7
8
9
10
11
Rib cage Liver Lung &
pleura
Spleen Thoracic
spine
Kidney Thoracic
cavity
Lumbar
spine
Torso organ/structure
PCDS
Fairfax INOV
A
% of victims with ≥ 1 AIS 2+ injury
Figure 4 – Fractions of the victims with ≥ 1 AIS 2+ injury by the
most commonly injured torso organs/structures and trauma registry.
0
1
2
3
4
5
6
7
8
9
10
Ri b cage Lung &
pleura
Spleen Liver Thoracic
cavity
Tho rac i c
aorta
Torso organ/structure
% of victims with ≥ 1 AIS 3+ injury
PCDS
Fairfax INOVA
Figure 5 – Fractions of the victims with ≥ 1 AIS 3+ injury by the
most commonly injured torso organs/structures and trauma registry.

271
0
1
2
3
4
Rib cage Lung &
pleura
Spl een Thoracic
aorta
Liver
Torso organ/structure
% of victims with ≥ 1 AIS 4+ injury
PCDS
Fairfax INOVA
Figure 6 – Fractions of the victims with ≥ 1 AIS 4+ injury by the
most commonly injured torso organs/structures and trauma registry.
VEHICLE SPECIFIC ANALYSIS – Figure 7 shows the
fraction of the 306 included PCDS cases with at least one injury by
body region, striking vehicle body type, and injury severity. While
striking vehicle body type appears to have a relatively minor
influence on the incidence of head & face and lower extremity
injuries, the incidence of torso injury at the AIS 2+, 3+, and 4+ levels
is 51%, 71%, and 98%, respectively, higher in the LTV cases than in
the car cases. It should, however, be emphasized that vehicle body
type appears to affect the incidence of severe head injury (42%
higher risk associated with LTVs at the AIS 4+ level). Figure 8
shows the average number of injuries per pedestrian victim for the
306 included PCDS cases by body region, striking vehicle body type,
and injury severity. Again, there appears to be a relatively low
influence of striking vehicle body type on head & face and lower
extremity injuries while the average number of AIS 2+, 3+, and 4+
torso injuries per victim are 176%, 177%, and 214%, respectively,
higher in the LTV than in the car cases. In fact, the data in Figure 8
suggest that adult pedestrians struck by LTVs tend to sustain a higher
number of torso injuries than lower extremity injuries regardless of
severity level. Figure 9 shows the average costs of AIS 2+ torso,
head & face, and lower extremity injuries per victim for the 306
included PCDS cases normalized by the average cost of AIS 2+
lower extremity injuries per LTV victim. The costs are determined
using the unit costs for comprehensive HARM from crash injuries
[NHTSA, 1996]. As shown, the cost of AIS 2+ torso injury exceeds
that of AIS 2+ lower extremity injury by as much as 369% in the
LTV cases, whereas injuries to those two body regions appear to
contribute equally to the total morbidity cost of pedestrians struck by
cars. Furthermore, it is interesting to note that the average costs of

272
AIS 2+ head & face and lower extremity injuries per LTV victim are
17% higher and 35% lower, respectively, than the corresponding
costs per car victim, whereas the cost of AIS 2+ torso injury per LTV
victim exceeds the corresponding cost per car victim with as much as
205%. Figure 10 shows the fractions of the 306 included PCDS cases
with at least one AIS 2+ injury by striking vehicle body type and the
most commonly AIS 2+ injured torso organs/structures. In agreement
with the AIS 2+ torso data shown in Figure 7, Figure 10
demonstrates a higher incidence for the LTV victims for all of the
torso organs/ structures. The fact that some of the organs/structures
demonstrate a relatively large difference in incidence between LTV
and car victims while others show fairly similar incidence for the two
vehicle types suggests that the influence of vehicle body type on
torso injury could be structure/organ dependent. However, due to the
low sample size, it is impossible to decide whether this trend is valid.
DISCUSSION
The current study has evaluated the significance of adult
pedestrian torso injury relative to head & face and lower extremity
injuries and determined the importance of injuries to individual torso
organs/structures. Analyses have been conducted with and without
adjusting for vehicle body type (car versus LTV). Children of age 18
and under were excluded due to confounding factors such as size and
biomechanical characteristics associated with growth and
development. Similarly, adults over 60 were excluded to avoid
potential effects of age related issues such as degradation in bone
mineral density.
0
5
10
15
20
25
30
35
40
Car-AIS 2+ LTV-AIS 2+ Car-AIS 3+ LTV-AIS 3+ Car-AIS 4+ LTV-AIS 4+
Stri king vehi cle body type and injury severi ty
Tor so
Head & face
Lower extremities
Fraction of victims with ≥ 1 injury (%)
Figure 7 - Fractions of the 19-60-year-old pedestrian victims in the
PCDS registry with at least one injury by body region, striking
vehicle body type, and injury severity.

273
0
0.2
0.4
0.6
0.8
1
1.2
Car-AIS 2+ LTV-AIS 2+ Car-AIS 3+ LTV-AIS 3+ Car-AIS 4+ LTV-AIS 4+
Stri king vehicle body type and injury severity
Tor so
Head & face
Lower extremities
Average number of injuries per victim
Figure 8 – Average number of injuries per 19-60-year-old pedestrian
victim in the PCDS registry by body region, striking vehicle body
type, and injury severity.
0
1
2
3
4
5
6
7
8
Car LTV
Stri king vehi cl e body type
Relative cost
Tor so
Head & face
Lower extremities
Figure 9 - Average costs of AIS 2+ torso, head & face, and lower
extremity injuries per pedestrian victim for the 306 included PCDS
cases normalized by the average cost of AIS 2+ lower extremity
injury per pedestrian LTV victim.
In agreement with the results from previous investigators, the
results from the non-vehicle specific analyses of both the PCDS and
Fairfax INOVA trauma registries indicate that torso injury is a
substantial contributor to the overall burden of pedestrian morbidity.
However, due to different inclusion and exclusion criteria and
different measures for quantifying the importance of pedestrian torso
injury, it is not possible to directly compare the results from the
current study to those of previous investigators. Nevertheless, it is

274
interesting to note that the incidence of AIS 2+ torso injury in the
current study is approximately 19% (17.3% in the PCDS and 20.3%
in the Fairfax INOVA) and thus falls in between the incidence of
chest and abdominal injury for AIS 2+ injured non-fatalities reported
by Ashton et al. (1976) and the incidence of chest and abdominal
injury in the 45 resuscitation room administered pedestrian victims
reported on by Civil (1986). Not surprisingly, the incidence of AIS
2+ torso injury in the current study is substantially lower than the
incidence of chest and abdominal injury in fatally injured pedestrian
victims reported on by Ashton et al. (1976) and Harruff et al. (1998).
When comparing the results from the two registries in Figures
1 and 2, there is a trend towards higher incidence and average
number of AIS 2+ injuries per victim in the Fairfax INOVA registry
than in the PCDS registry. In contrast, the trend is the opposite at the
AIS 3+ and AIS 4+ levels. These trends should partly be due to the
different sources for the two trauma registries. Since the criteria for
inclusion of a pedestrian crash in the PCDS did not include any
limitations on the outcome of the pedestrian, PCDS includes a wide
range of outcomes from those who were struck at low speed and
suffered mainly bruises to those who died at the scene. In contrast,
the Fairfax INOVA registry includes only those pedestrian victims
who were severely enough injured to be taken to the level one trauma
center but not worse off than that they made it there alive.
Consequently, the PCDS registry should include higher proportions
of victims with relatively minor injuries and victims with critical and
fatal injuries. A direct indication of that PCDS includes a higher
proportion of victims with fatal injuries is the higher incidence of
thoracic aorta injury among the PCDS victims (Figures 5 and 6).
Additional factors that influenced the incidence, average number of
injuries per victim, and relative cost distributions as determined from
the Fairfax INOVA data include the choice of mapping severity for
the conversion from ICD-9-CM to AIS-90 and, for the head & face
distributions only, the filtering method used for the 49 head injuries
that ICDMAP-90 had assigned AIS-90 code 115099. The use of the
ICDMAP-90 low severity mapping option would most likely have
resulted in a slight shift of the Fairfax INOVA severity distributions
towards lower injury severities and as such, a slightly increased
difference between the PCDS and Fairfax INOVA distributions in
Figures 1-6. Considering the large number of AIS 2+ head injuries,
the effect of the filtering method used for the 49 115099 codes had a
relatively minor influence on the Fairfax INOVA head & face data in
Figures 1-3. If all of the 115099 codes had been eliminated, the
incidence of AIS 2+ head face & injury according to the Fairfax
INOVA registry shown in Figure 1 would have been 44.4% instead
of 47.3% while the average number of AIS 2+ head & face injuries
per victim shown in Figure 2 would decrease from 1.03 to 1.00.

275
0
2
4
6
8
10
12
14
16
18
Rib cage Lung &
ple ura
Liver Spleen Thoracic
spine
Kidney Mesentery Thoracic
aorta
Torso organ/structure
% of victims with ≥ 1 AIS 2+ injury
Car
LTV
Figure 10 – Fraction of 19-60-year old pedestrian victims with at
least one AIS 2+ injury by striking vehicle body type and the most
commonly AIS 2+ injured torso structures/organs.
The results from the vehicle specific analysis support the
findings of previous investigators indicating that LTVs pose a higher
risk of pedestrian torso injury than cars [Lane et al., 1994;
Ballesteros et al., 2004; Lefler and Gabler, 2004; Longhitano et al.,
2005a,b]. A comparison of the torso data shown in Figures 7 and 8
clearly indicates that striking vehicle body type has a larger influence
on the average number of torso injuries per victim than on the
incidence of torso injury. As such, pedestrian LTV victims who
sustain torso injury are also likely to have a higher number of torso
injuries than car victims with torso injury.
While the results shown in Figures 7-10 clearly illustrate the
substantial contribution of torso injury to the overall morbidity of
pedestrians struck by LTVs, it is important to emphasize that the
results from the current study indicate that torso injury is a problem
in pedestrians struck by cars as well. In fact, the relative cost
distribution shown in Figure 9 suggests that the current focus on
pedestrian safety countermeasures for head and lower extremity
protection may not even be optimal for minimizing the overall
morbidity of pedestrians struck by cars. An increased focus on
countermeasures for the prevention of pedestrian torso injury should
therefore reduce the overall morbidity among pedestrians struck by
LTVs as well as regular passenger cars.
CONCLUSIONS
Torso injury is second only to head injury as the most
important contributor to the morbidity among adult pedestrian
victims struck by LTVs. In case of pedestrians struck by cars, torso
injury and lower extremity injury appear to be approximately equal

276
contributors. Regardless of striking vehicle type, torso injury appears
to be as an important contributor to the overall cost of pedestrian
morbidity as is lower extremity injury. The most frequently injured
torso organs/structures include the rib cage, lung & pleura, and liver.
REFERENCES
Aldman, B, Kajzer, J, Bunketorp, O, et al. An experimental study of
a modified compliant bumper. Proceedings, 10
th
International
Technical Conference on the Experimental Safety Vehicles,
July 1-4, 1985
, pp 1035-1040.
Ashton, SJ, Bimson, S, Driscoll, C. Patterns of injury in pedestrian
accidents. 23
rd
Annual Proceedings, American Association
for Automotive Medicine, October 3-6, 1979
, pp 185-202.
Ballesteros, MF, Dischinger, PC, Langenberg, P. Pedestrian injuries
and vehicle type in Maryland, 1995-1999. Accident Analysis
& Prevention
36: 73-81; 2004.
Brundage, SI., Harruff, R, Jurkovich, GJ, et al. The epidemiology of
thoracic aorta injuries in pedestrians. J Trauma
45: 1010-
1014; 1998.
Chidester, AB, Isenberg, RA. Final report – the pedestrian crash data
study. Proceedings, 17
th
International Technical Conference
on the Enhanced Safety of Vehicles, June 4-7, 2001
.
Civil, ID. Patterns of injury in motor vehicle trauma. New Zealand
Medical Journal
, November 26, 905-906; 1986.
Crandall, JR, Bhalla KS, Madeley NJ. Designing road vehicles for
pedestrian protection. BMJ
324: 1145-1148; May 11, 2002.
Detweiler, DT, Miller, RA. Development of a sport utility front
bumper system for pedestrian safety and 5 mph impact
performance. Proceedings, 17
th
International Technical
Conference on the Enhanced Safety of Vehicles, June 4-7,
2001
, Paper 01-S6-W-145.
European New Car Assessment Programme (EuroNCAP). Pedestrian
testing protocol, version 4.0, 2003, www.euroncap.com
.
Fredriksson, R, Håland, Y, Yang, J. Evaluation of a new pedestrian
head injury protection system with a sensor in the bumper and
lifting of the bonnet’s rear part. Proceedings, 17
th
International Technical Conference on the Enhanced Safety
of Vehicles, June 4-7, 2001
, Paper 131.
Harris, J, Grew, ND. The influence of car design on pedestrian
protection. Proceedings, 10
th
International Technical
Conference on the Experimental Safety Vehicles, July 1-4,
1985
, pp 1009-1022.
Harruff, RC, Avery, A, Alter-Pandya, AS. Analysis of circumstances
and injuries in 217 pedestrian traffic fatalities. Accident
Analysis & Prevention
30: 11-20; 1998.

277
Hill, DA, Delaney, LM, Duflou, J. A population-based study of
outcome after injury to car occupants and to pedestrians. J
Trauma
40: 351-355; 1996.
Ishikawa, H, Kajzer, J, Ono, K, et al. Simulation of car impact to
pedestrian lower extremity: influence of different car-front
shapes and dummy parameters on test results. Accident
Analysis & Prevention
26: 231-242; 1994.
Lane, PI, McClafferty, KJ, Nowak, ES. Pedestrians in real world
collisions. J Trauma
36: 231-236; 1994.
Lefler, DE, Gabler, HC. The fatality and injury risk of light truck
impacts with pedestrians in the United States. Accident
Analysis & Prevention
36: 295-304; 2004.
Longhitano, D, Bhalla, K, Henary, B, et al. Influence of vehicle body
type on pedestrian injury distribution. SAE World Congress,
April 11-14, 2005a,
SAE Technical Paper 2005-01-1876.
Longhitano, D., Ivarsson, J, Henary, B, et al. Torso injury trends for
pedestrians struck by cars and LTVs. Proceedings, 19
th
International Technical Conference on the Enhanced Safety
of Vehicles, June 6-9, 2005b
, Paper 05-0411.
Mizuno, Y. (2003) Summary of IHRA Pedestrian safety WG
activities ( 2003 ) – proposed test methods to evaluate
pedestrian protection afforded by passenger cars.
Proceedings, 18
th
International Technical Conference on the
Enhanced Safety of Vehicles, May 19-22, 2003
, Paper 580.
Nagatomi, K, Akiyama, A, Kobayashi, T. Bumper structure for
pedestrian protection. Proceedings, 15
th
International
Technical Conference on the Enhanced Safety of Vehicles,
May 13-17, 1996
, Paper 96-S4-O-02.
National Accident Sampling System - Pedestrian Crash Data Study
(NASS-PCDS). Data Collection, Coding and Editing Manual.
US Department of Transportation, Washington, D.C., 1996.
National Highway Traffic Safety Administration. The
Economic Costs of Motor Vehicle Crashes, 1994.
Department
of Transportation HS 808 425, Washington, D.C., 1996.
National Highway Traffic Safety Administration. Traffic
Safety Facts 2003: A compilation of motor vehicle crash data
from the Fatality Analysis Reporting System and the General
Estimates System. 2005, http://www-
nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSFAnn/TSF2003.pdf
Okamoto, Y, Akiyama, A, Nagatomi, K, et al. Concept of hood
design for possible reduction in pedestrian head injury.
Proceedings, 14
th
International Technical Conference on the
Enhanced Safety of Vehicles, May 23-26, 1994
, pp 1035-
1040, Paper 94-S7-W-14.
World Bank. Road Safety. 2001,
http://www.worldbank.org/transport/roads/safety.htm
.