Skip to main content
NCBI home page
Annual Proceedings / Association for the Advancement of Automotive Medicine logoLink to Annual Proceedings / Association for the Advancement of Automotive Medicine
. 2003;47:105–126.

Child and Adult Pedestrian Impact: The Influence of Vehicle Type on Injury Severity

Basem Y Henary 1, Jeff Crandall 1, Kavi Bhalla 1, Charles N Mock 2, Bahman S Roudsari 2
PMCID: PMC3217548  PMID: 12941221

Abstract

In the United States, the vehicle fleet is shifting from predominantly passenger cars (automobiles) to SUVs, light trucks, and vans (LTV). This study investigates how pedestrian severe injury and mortality are associated with vehicle type and pedestrian age. The Pedestrian Crash Data Study (PCDS) database for years 1994–1998 was used for a cross-sectional study design. Outcome measures were Injury Severity Score, Maximum Abbreviated Injury Score, Abbreviated Injury Scale, Pedestrian Mortality, Functional Capacity Index and Life Years Lost to Injury. Compared to children, adult pedestrians were more likely to sustain severe injury (OR = 2.81; 95% CI: 1.56–5.06) or mortality (OR = 2.91; 95% CI: 1.10–7.74) when examining all vehicle types. However, after adjusting for vehicle type and impact speed, this association was not statistically significant at p < 0.05. Compared to passenger cars, pedestrians struck by LTV were more likely to have severe injuries (OR = 1.31; 95% CI: 0.88–1.94) or mortality (OR = 1.40; 95% CI: 0.84–2.34) for all pedestrians. Adjusting for pedestrian age, this association was more obvious and significant at lower impact speeds (≤ 30 km/h); odds ratios of severe injury and mortality were 3.34 (p< 0.01) and 1.87 (p= 0.07), respectively. Adults hit by LTV had the highest risk of injury and mortality. These findings indicate that pedestrian age, vehicle engineering design and impact speed are highly contributing to risks of pedestrian injury and mortality.

Pedestrian motor vehicle crashes are associated with substantial morbidity and mortality. In the United States one pedestrian is killed by a motor vehicle every 107 minutes and one is injured every 6 minutes [NHTSA, 1999; Isenberg et al., 1998]. While pedestrian fatalities account for 13% of all traffic fatalities and 5 % of all traffic injuries in the Unites States [NHTSA, 1997; NHTSA, 1999; Stammen et al., 2001], in other parts of the world these proportions can be much higher. In Japan, 27–32% of all traffic fatalities are pedestrians [Matsui, 2002]. The percentage is nearly 30% in the United Kingdom [Stammen, 2001] and approximately 16% in Australia [ATSP Monthly Bulletin, 2001; Anderson, 2001]. In urban areas of low-income countries, such as Addis Ababa, Ethiopia, this proportion can be as high as 85% [Jacobs and Thomas, 2000].

In comparison with motor vehicle occupants, pedestrians are exposed to higher rates of being in crashes and, when involved in a collision, suffer more serious injuries. The fatality rates per mile traveled for a pedestrian are over 15 times higher than that for motor vehicle occupants [McCann and DeLille, 2000; Brass, 1998]. Langley et al. (2002) analyzed data from New Zealand to show that although pedestrians accounted for 10% of the hospitalized road users, they contributed to 18% of the total hospital cost. Similarly, the mortality rates of injured pedestrians are approximately twice that of injured vehicle occupants [Rogers et al., 1991; Hill et al., 1996; Champion et al., 1990]. Analysis of NHTSA data for the US (Fig. 1) shows that this relative pedestrian mortality rate could be even higher. For the pediatric population, the situation is even worse. Pedestrian injuries account for 61% of pediatric trauma admissions to US hospitals and 34% of pediatric critical care admissions [DiMaggio et al., 2002]. Pedestrian injuries are the first cause of death from trauma for the 5–9 years old age group [Rivara, 1990; NHTSA, 1998]. For children aged 5–14, pedestrian injuries are the second leading cause of unintentional injury-related mortality [Jarrett et al., 1998].

Fig. 1.

Fig. 1

Open in a new tab

Case Fatality Rate/1,000 Injured Person for Pedestrians only and for All Motor Vehicle Crash Victims (Data Source: Traffic Safety Facts, NHTSA)

It has been pointed out that pedestrian fatalities have steadily declined for several decades (20% since 1984, as shown in Fig. 2). This has been attributed in partial to pedestrians walking less than they did before. It was reported that there was a 37% decline in the number of trips made by children by foot between 1977 and 1995 [McCann & DeLille, 2000].

Fig. 2.

Fig. 2

Open in a new tab

Pedestrian Morbidity and Mortality: US data, 1988–2000 (Data source: Traffic Safety Facts, NHTSA)

However, future projections of injury and mortality trends should take into account expected changes in the pedestrian crash environment. The passenger vehicle fleet in the US is rapidly shifting towards an increased proportion of light trucks, sport utility vehicles and vans (LTV). These currently account for about one-third of the current vehicle registrations in the US, and almost half of all new vehicle registrations (Gabler and Hollowell, 1998). Lefler and Gabler (2001) concluded that the risk of mortality from an LTV-pedestrian impact can be two to three times greater than that from a car-pedestrian impact. Liu and Yang (2002) performed mathematical simulations of adult and child pedestrian impacts with vehicles of different geometries and have shown that the threat posed by LTV to pediatric pedestrians can be even higher than that to adult pedestrians.

In this study, real world crash data collected from the Pedestrian Crash Data Study (PCDS) is analyzed to obtain the relative risks of injury and mortality for child and adult pedestrians struck by passenger cars and LTV. Outcome measures based on Injury Severity Score, Maximum Abbreviated Injury Score, Abbreviated Injury Score, Pedestrian Mortality, Functional Capacity Index and Life Years Lost to Injury are compared and the effect of pedestrian age, vehicle type and impact speed were assessed.

PEDESTRIAN CRASH DATA STUDY BACKGROUND

In 1977, the Pedestrian Injury Causation Study (PICS) was initiated. It investigated 2,000 crashes over a thirty-month period. From 1979–1987, National Automotive Sampling System (NASS) used its Continuous Sampling System (CSS) to collect data on over 4,000 pedestrians through a representative sample of pedestrians’ involvement in all type of crashes. In 1988, when Crashworthiness Data System (CDS) was implemented, pedestrian crashes were eliminated from the data system and only general pedestrian information continued to be collected through the General Estimate System (GES). Several years later, it was felt that recent knowledge was needed to investigate the possibility that recent vehicle models were producing the same types of pedestrian injuries that occurred with older vehicle models. The Pedestrian Crash Data System (PCDS) was created as a result of the need for reliable data to establish injury criteria for pedestrians and improve understanding of real world pedestrian crashes. The PCDS offers information on pedestrian crashes that is not available through the Fatality Analysis Reporting System (FARS), the GES or the CDS. Certain case data elements were collected on scene after immediate notification of the crash through the Police Radio Log (PRL). The PCDS is a five-year compilation of vehicle-pedestrian crash data collected from 1994 to 1998 for passenger vehicles of model year 1990 and later. The PCDS data files contain 552 cases (as of 1998) with highly detailed information on pedestrian collisions. It includes pedestrian, vehicle, environmental and scene data and injury information. Data were collected from six major cities in the United States. These cities were selected based on the availability of applicable pedestrian caseload. A Pedestrian was defined as any person who is on a traffic-way or on a sidewalk or path contiguous with a traffic-way, or on private property. This includes persons who are in contact with the ground, roadway, etc. and are pushing carts, wagons, or other objects, or holding a vehicle. Case selection criteria are a forward moving, late model year vehicle (90–96) that strikes a pedestrian. Pedestrians lying or sitting before the impact were not included. The striking portion of the vehicle structure must be original equipment manufacturer. The pedestrian impact(s) are the vehicles’ only impact(s) and the first point of contact must be forward of the top of the A-pillar. All crash investigations were initiated on-scene and follow-up investigation when necessary [NASS-PCDS, 1996]. The PCDS data are not weighted as the PCDS was designed and performed to be a clinical study and not intended to be a national sample of all the US pedestrian crashes (Chidester and Isenberg, 2001).

METHODOLOGY AND DESIGN

INCLUSIONS AND EXCLUSIONS

To investigate the influence of vehicle type on pedestrian injury severity, we included all 552 pedestrian cases. To investigate the influence of age on pedestrian injury severity and fatality, only child (defined as aged 2 to 14 years) and adult pedestrians (19 to 50 years) were included (388 cases). The authors excluded subjects of other ages (15–18 and >51). Subjects aged 15–18 were excluded to rule out the possibility of overlapping physical characteristics between child and adult populations. Similarly, we excluded the older subjects group because of concerns regarding body frailty and degree heterogeneity. Due to the limited data available, aggregate age groups for children were used (age distribution for children is shown in Fig. 3). In addition, testing the risk of injuries for children at conventional age groups (2–4, 5–9 and 10–14 years), results were considerably similar. These initial results were very much comparable with findings shown by a recent study [Woods et al., 2002].

Fig. 3.

Fig. 3

Open in a new tab

Age distribution for the children group (126 subjects). (Data source: PCDS, 1994–1998)

Vehicles were categorized based on their body type to passenger cars (automobiles), sport utility vehicles (SUV), light trucks (LT) and vans (V). To meet our research question and to overcome the data shortage, we grouped sport utility vehicles, vans and light trucks into one category (LTV). Again, initial results showed that SUV, LT, and V appeared to result in higher pedestrian injury severity and more risk of mortality than passenger cars may do. These initial findings supported the authors’ decision of grouping children age 2 to 14 years in one group and grouping SUV, LT, and V into one vehicle type group (LTV). The main issue taken into consideration was availability of adequate data in each age group to allow a meaningful analysis.

PREDICTING VARIABLES

Pedestrian age (children vs. adults) and vehicle body type (passenger cars vs. LTV) were the predicting variables. The authors tested for confounders and effect modifiers and included the most influential ones in the final statistical models. We were limited by the small sample size and avoided over-adjustment and losing more data due to the missing, unknown or excluded observations in each variable. Since impact speed had a strong influence on the tested predictors and was a strong confounder and effect modifier in all the logistic regression analyses, we adjusted for impact speed and/or its interaction term in all models. We also adjusted for pedestrian age and vehicle type, when applicable.

OUTCOME VARIABLES

Various outcome measures were used to assess the level, severity and significance of sustained injuries. Each of the outcome variables (with exception of Functional Capacity Index and Life Years Lost to Injury) was categorized into a dichotomous (binary) variable based on a statistically and logically acceptable threshold value, suitable for the analysis of this study. Injury Severity Score (ISS) is an anatomical based scoring system that provides an overall score for patients with multiple injuries and ranges from 0–75 [Baker et al., 1974]. In general, subjects with ISS ≥ 16 are considered to have a poor prognosis with more than a 10% risk of death [Robertson et al., 1991; Seow et al., 1996]. Thus, we employed ISS as a binary variable with a threshold value of 16 (≤ 16 and > 16), which coincided with the mean ISS for the PCDS cases (16.11). Further, cross-tabulation of ISS and mortality outcome showed that 93.15% of those who ultimately died (68 out of 73) had an ISS > 16. Subjects with ISS >16 were considered the case group while pedestrians with ISS ≤16 were considered the control group. Abbreviated Injury Scale (AIS) is an anatomical based scoring system that is ranked on a scale of 1–6, with 1 being minor and 6 an untreatable injury. Using the AIS, the injury severity profiles for major injuries were defined by a score greater than 2 and minor injuries were defined by a score of less than or equal to 2 (AIS ≤ 2 and AIS > 2) [AAAM, 1990; Lefler and Gabler, 2003. Maximum Abbreviated Injury Score (MAIS) represents the highest AIS that a pedestrian has irrespective of the number of injuries. Pedestrians who died were assigned a score of 6, and those with no injuries were assigned a score of 0. The MAIS was categorized to a binary variable with a threshold value of 2 (MAIS ≤ 2 and MAIS > 2). Pedestrian Fatality outcome was used as a proxy for pedestrian injury severity in its dichotomous mode (fatal or non-fatal). The fatality outcomes for 20 cases were unknown but the low corresponding MAIS and ISS values indicated that those subjects were very unlikely to die, so those cases were considered survivors. Functional Capacity Index (FCI) is a multi-attribute index (10 domains) that maps anatomic descriptions of the nature and extent of injury into scores that predict the extent of functional limitations or reduced capacity at one year after injury. FCI assigns a score of “0” (no loss) to “1” (total loss) to each domain. Eventually, there is a final FCI for each body injury. Although the FCI was developed for application for adults only, we used it to assess the functional limitations and reduced capacity for both adults and children. We assigned the FCI value for each pedestrian injury. FCI was then multiplied by the person’s remaining life expectancy based on a flat life expectancy of 70 years at birth. This resulted in a measure called Life Years Lost to Injury (LLI) [MacKenzie et al. 1996; NHTSA, 2000].

DESIGN

We used a cross-sectional (prevalence) study design where the exposure and the pedestrian motor vehicle crash outcome were determined simultaneously for each subject. The outcome measures were ISS, AIS, MAIS, Fatality, FCI, and LLI. As previously described, several outcome variables were categorized into dichotomous variables at prescribed thresholds (2 for MAIS and AIS and 16 for ISS). Cases were defined as those pedestrians who died or sustained injuries relative to threshold values of AIS > 2, ISS > 16, MAIS > 2. Exposure was defined according to the pedestrian age and the vehicles body-type. Pedestrians struck by passenger cars were the non-exposed group while those struck by LTV were the exposed group. Similarly, children were hypothetically considered the non-exposed group while adults were considered the exposed group. Relative risks were estimated by the calculation of crude and adjusted odds ratios (ORs). Analyses were conducted using linear and logistic regression to estimate odds ratios, p-values and 95% confidence intervals (95% CI). A p-value of <0.05 was considered to be statistically significant. We controlled for impact speed as it proved a strong influence on the relationship between the predictors being investigated and the outcome variables. Nested Models, Pearson’s chi-square, analysis of variance (ANOVA) and analysis of covariance (ANCOVA) were used as needed in statistical analysis. Statistical analysis was conducted using Stata Statistical Software, Release 7.0 (Stata Corporation, Texas, USA) and Microsoft Excel Version 98 (Microsoft Corporation, Washington, USA).

RESULTS

Pedestrian, vehicle type and speed characteristics

Of all the cases in the NASS-PCDS, 126 (22.8%) were children and 262 (47.5%) were adults. Mean pedestrian age was higher for LTV than for passenger cars [37.9 ± (22.8) vs. 32.9 ± (21.6) year] (data distribution is shown in Fig. 4 & 5). 377 (68.3%) of all pedestrians were hit by passenger cars vs. 175 (31.7%) by LTV. Among children, 95 (75.4%) were hit by passenger cars vs. 174 (66.4%) among adults. For all vehicles, impact speed ranged between 2 and 118 km/h. The mean impact speed for all vehicles was higher for adults [29.0 ± (22.0) km/h] than for children [26.2 ± (14.5) km/h]. In general, the mean impact speed for all vehicles was higher for passenger cars than for LTV [29.3 ± (19.8) km/h and 26.3 ± (21.1) km/h respectively] (data distribution is shown in Fig. 6) (all data are shown in Table - 1).

Fig. 4.

Fig. 4

Open in a new tab

Boxplots of Pedestrian Age Distribution by Age Groups

Fig. 5.

Fig. 5

Open in a new tab

Boxplots of Pedestrian Age Distribution by Vehicle Type.

Fig. 6.

Fig. 6

Open in a new tab

Boxplots of Impact Speed Distribution by Vehicle Type.

Table 1.

Distribution of Pedestrian Characteristics, Vehicle Types and Impact Speeds by Age Groups: frequency, percentage, mean and standard error (SEM) for the mean. (PCDS, 1994–1998).

Variable Children [2–14 years] Adults [19–50 years] Total [2–93 years]
N (%) Mean ± (SEM) N (%) Mean ± (SEM) N (%) Mean ± (SEM)
Age (year) 126 (22.8) 8.9 ± (0.30) 262 (47.4) 33.3 ± (0.56) 552 (100) 34.5 ± (0.94)
Height (cm) 115 (23.9) 134.5 ± (2.15) 225 (46.8) 171.1 ± (0.65) 481 (100) 161.2 ± (0.95)
Weight (Kg) 115 (24.0) 35.8 ± (1.47) 224 (46.7) 73.8 ± (1.21) 479 (100) 63.9 ± (1.07)
Impact Speed (km/h) (by age group) 103 (22.8) 26.2 ± (1.43) 228 (50.5) 29.0 ± (1.46) 451 (100) 28.8 ± (0.94)
Vehicle type
 Passenger Car 95 (75.4) N/A 174 (66.4) N/A 377 (68.1) N/A
 LTV 31 (24.6) N/A 88 (33.6) N/A 175 (31.7) N/A
  SUV 9 (7.1) N/A 17 (6.4) N/A 36 (6.52) N/A
  Van 14 (11.1) N/A 38 (14.5) N/A 75 (13.59) N/A
  LT 8 (6.4) N/A 33 (12.6) N/A 64 (11.59) N/A
  Total 126 (100) N/A 262 (100) N/A 552 (100) N/A
Impact speed by age (km/h) group and vehicle type
 Passenger Car 79 (76.7) 27.6 ± (1.7) 149 (65.4) 29.1 ± (1.69) 311 (67.5) 29.3 ± (1.12)
 LTV 24 (23.3) 21.58 ± (2.1) 79 (34.7) 28.9 ± (2.75) 150 (32.5) 26.3 ± (1.72)
  SUV 9 (8.7) 18.3 ± (3.5) 13 (5.7) 26.3 ± (6.42) 29 (6.29) 24.3 ± (3.45)
  Van 10 (9.7) 23.5 ± (3.3) 35 (15.4) 22.6 ± (3.21) 66 (14.32) 21.8 ± (2.01)
  LT 5 (4.9) 23.6 ± (4.7) 31 (13.6) 37.2 ± (5.11) 55 (11.93) 32.9 ± (3.46)
Open in a new tab

Fig.4–6 show the 25th (Q1), 50th (median), the 75th (Q3) percentiles, Interquartile Range (IQR), upper and lower whiskers and outliers for each group. (Whiskers are lines drown to the smallest and largest observations within the calculated fences. Upper Fence = Q3 + (1.5 × IQR), Lower Fence = Q1 − (1.5 × IQR), Outliers (represented by small circles in Fig. 5) are observations values beyond the higher and lower fences. (Note: Sample size is represented by box width in Fig. 4–6).

Pedestrian injury and pedestrian age

Both descriptive and logistic regression results showed that pedestrian injury severity and mortality varied by age of the pedestrians involved in the crash. Adults had higher risk of severe injuries than children. This finding was consistent when using various injury severity outcome measures (ISS, MAIS, AIS, FCI, LLI and Mortality). As described by ISS, The mean ISS ± (SE) was much higher for adults [16.13 ± (1.30)] than for children [9.33± (1.29)]. The mortality rate for adults [11.7% (28 deaths)] was nearly 3 times higher than that for children [4.0% (5 deaths)]. Although adults had higher FCI [12.87 ± (0.99) vs. 11.23 ± (1.84) for children], results showed that children had higher Life Years Lost to Injury (LLI) [6.61± (1.07)] vs. [4.40 ± (0.34) for adults]. This finding can be explained by the fact that children had more average remaining life expectancy at the time of the crash (data are shown in Table - 2).

Table 2.

Outcome measures (ISS, MAIS, AIS, Mortality, FCI and LLI) by age groups: frequency, percentage, mean and standard error (SEM) for the mean. (PCDS, 1994–1998).

Outcome Measure Children (N =126) Adults (N =262) Total (N =552)
N (%)a Mean ± (SEM)f N (%)a Mean ± (SEM)f N (%)a Mean ± (SEM)f
ISS 9.33 ± (1.29) 16.13 ± (1.30) 16.11 ± (0.91)
0–16 110 (87.3) 4.84 ± (0.36) 186 (71.0) 4.74 ± (0.30) 402 (72.8) 4.98 ± (0.21)
>16 16 (12.6) 40.19 ± (5.41) 76 (29.0) 43.99 ± (2.24) 150 (27.2) 45.93 ± (1.67)
MAIS 1.81± (0.10) 2.26 ± (0.10) 2.27 ± (0.07)
≤ 2 100 (79.4) 1.32 ± (0.05) 164 (62.6) 1.20 ± (0.38) 355 (64.4) 1.28 ± (0.03)
> 2 26 (20.6) 3.69 ± (0.17) 98 (37.4) 4.02 ± (0.10) 196 (35.6) 4.04 ± (0.07)
AISb 779 (100) 1.35 ± (0.30) 2161 (100) 1.65 ± (0.22) 4495 (100) 1.62 ± (0.02)
≤ 2 705 (90.5) 1.12 ± (0.01) 1733 (80.2) 1.19 ± (0.01) 3663 (81.5) 1.18 ± (0.01)
> 2 74 (9.5) 3.54 ± (0.09) 428 (19.8) 3.52 ± (0.04) 832 (18.5) 3.55 ± (0.03)
Mortality
Fatal 121 (96.0) N/A 234 (89.3) N/A 478 (86.8) N/A
Not fatal 5 (4.0) N/A 28 (11.7) N/A 73 (13.2) N/A
FCIb, c 170 (10.5) 11.23 ± (1.84) 821 (50.6) 12.87 ± (0.99) 1623 (100) 13.25 ± (0.75)
LLIb, d 139 (21.7) 6.61 ± (1.07) 502 (78) 4.40 ± (0.34) 641 (100)e 4.88 ± (0.36)e
Open in a new tab
a

Numbers (N) and Percentages (%) of observations may differ from variable to other because of the missing and unknown values. Percentages are presented by column.

b

Outcome measure is for injuries and not for pedestrian cases (one pedestrian may have more than one injury).

c

FCI only for pedestrian injuries where pedestrians were reported alive a month of the crash.

d

LLI only for pedestrian injuries where pedestrians were reported alive a month after the crash assuming that life expectancy is 70 years for all pedestrians irrespective of his/her actual life expectancy. LLI=FCI × Remaining life years (70 – pedestrian age at time of crash).

e

Calculated for only the adults and the children age groups (excluding any other subjects).

f

Mean ± Standard Error for the Mean (SEM) when applicable and meaningful.

Results of Univariate logistic regression demonstrated that adults had 2 to 3 times higher risk of severe injury as that of children (Using ISS, OR = 2.81; 95% CI: 1.56 – 5.06). Similarly, adults were at higher risk of mortality than children (OR = 2.91; 95% CI: 1.10 – 7.74). Although the relationship between pedestrian age and pedestrian injury severity was not statistically significant after controlling for vehicle type and impact speed, it was still in the same direction and strength (Table - 4).

Table 4.

Relative Risk of Severe Injury or Mortality for Adults vs. Children, Crude and Adjusted (for Vehicle Type and Impact Speed). ORs, P-value and 95% CI.

Outcome Measure Univariate ORs
Adjusted ORs
OR P>|Z| 95% CI OR P>|Z| 95% CI
ISS 2.81 0.001 1.56, 5.06 1.50 0.276 0.72, 3.14
MAIS 2.30 0.001 1.40, 3.78 1.55 0.171 0.83, 2.90
AIS 2.35 0.000 1.81, 3.06 1.22 0.192 0.90, 1.65
Mortality 2.91 0.032 1.10, 7.74 1.19 0.769 0.38, 3.76
Open in a new tab

Pedestrian injury and vehicle type

Pedestrian injury severity and mortality varied by the type of vehicle involved in the crash. Pedestrians hit by LTV were at higher risk of severe injury. 31% of pedestrians hit by LTV sustained severe injuries (ISS > 16) as compared to only 25% of those hit by passenger cars. The mean ISS for pedestrians hit by LTV was higher than for those hit by passenger cars [18.13 ± (1.76) vs. 15.17 ± (1.05) respectively]. Similarly, LTV resulted in higher risk of mortality than passenger cars. Mortality rate was 16% for LTV as compared to 12% for passenger cars. FCI was higher for those pedestrians struck by LTV [13.69± (0.92)] than by passenger cars [12.4 ± (1.26)]. However, results of LLI showed no difference between the two groups (data are shown in Table - 3). Both univariate and multivariate models of logistic regression demonstrated that pedestrians hit by LTV had higher risks of sustaining severe injuries and mortality than those struck by passenger cars, this association was not statistically significant at p < 0.05 (data not shown).

Table 3.

Outcome measures (ISS, MAIS, AIS and Mortality) by vehicle type: frequency, percentages, mean and standard error (SEM) for the mean. (PCDS, 1994–1998).

Outcome Measure Passenger cars N=377 LTV N=175 Total N=552
N (%) Mean ± (SEM) N (%) Mean ± (SEM) N (%) Mean ± (SEM)
ISS 15.12 ± (1.05) 18.13 ± (1.76) 16.11 ± (0.91)
0–16 281 (74.5) 5.04 ± (0.24) 121 (69.1) 4.82 ± (0.37) 402 (72.8) 4.98 ± (0.21)
>16 96 (25.5) 44.8 ± (2.06) 54 (30.9) 47.9 ± (2.84) 150 (27.2) 45.93 ± (1.67)
MAIS 2.21 ± (0.08) 2.38 ± (0.12) 2.27 ± (0.07)
≤ 2 247 (65.7) 1.29 ± (0.03) 108 (61.7) 1.25 ± (0.05) 355 (64.4) 1.28 ± (0.03)
> 2 129 (34.3) 3.95 ± (0.09) 67 (38.3) 4.21 ± (0.13) 196 (35.6) 4.04 ± (0.07)
AIS 1.60 ± (0.19) 1.66 ± (0.03) 1.62 ± (0.02)
≤ 2 2375 (82.1) 1.18 ± (0.01) 1288 (80.4) 1.17 ± (0.1) 3663 (81.5) 1.18 ± (0.01)
> 2 519 (17.9) 3.50 ± (0.04) 313 (19.6) 3.63 ± (0.05) 832 (18.5) 3.55 ± (0.03)
Mortality
Not fatal 315(86.8) N/A 139 (82.3) N/A 454 (85.3) N/A
Fatal 45 (12.4) N/A 28 (13.7) N/A 73 (13.7) N/A
Unknown 3 (0.8) N/A 2 (1.2) N/A 5 (1.00) N/A
FCI 588 13.69 ± (0.92) 313 12.4 ± (1.26) 901 13.25 ± (0.75)
LLI 435 4.87± (0.44) 206 4.89 ± (0.62) 641 4.88 ± (0.36)
Open in a new tab

Pedestrian injury and Impact Speed

In previous studies, the influence of impact speed on injury severity was assessed as being a risk factor in impact incidents for occupants and pedestrians [Mayr et al., 2003; Ballesteros et al., 2003; Leaf and Preusser, 1999; Khattak et al., 2003]. Also, it is well known that event severity increases as a function of square of impact speed (Kinetic energy = ½ (mass × velocity 2)]. Thus, it was necessary to test for the function of impact speed in all analyses. Firstly, we used multiple linear regression models (controlling for pedestrian age and vehicle type) to assess the level and significance of this risk. Results showed that: for every 5-km/h increase in impact speed, the severity of pedestrian injury increases by average of 3.4 units on the ISS scale (β for impact speed = 0.68; p < 0.01) while the risk of mortality increases by 4% (β = 0.008; p < 0.01). Also, results of logistic regression proved that impact speed was not only a strong confounder in the relationship between vehicle type and pedestrian injury severity but an effect modifier as well. A threshold value of 30 km/h for impact speed was selected, as it was the speed point at which the slope of risk measure changed in the analysis to examine the relative risks of vehicle types. The harmful effect of LTV on pedestrian injury varied by impact speed. At low impact speed levels (≤ 30 km/h), and after controlling for pedestrian age, LTV resulted in a higher statistically significant risk of severe pedestrian injuries than did passenger cars (OR = 3.34; 95% CI: 1.35 – 8.25). No such association was seen at high impact speed levels (> 30 km/h) (Table- 5). The results of a different analysis, conducted to examine the relative risk of vehicle type on different pedestrian body regions at different impact speeds (≤ 30 km/h and > 30km/h), showed that pedestrians hit by LTV were at higher risk of severe injury (ISS > 16) for all body regions (head & face, torso, upper extremities, and lower extremities) only at low impact speed (≤ 30 km/h). This association was statistically significant when the risk for different body regions was examined after controlling for pedestrian age and confirmed the results shown above. Again, no such association was found for high impact speed (> 30 km/h) (Table - 6). Closer examination of the same previous association using other injury severity outcome measures (AIS and MAIS), revealed similar results (data not shown). These findings suggest that the vehicle design may explain the higher risk of pedestrian injury associated with LTV specifically at lower impact speeds.

Table 5.

Relative Risk of Severe Injury or Mortality for LTV vs. Passenger Cars, Crude and Adjusted (for Pedestrian Age) ORs, p-value and 95% CI.

Outcome Measure Univariate ORs
Adjusted ORs
OR P>|Z| 95% CI OR P>|Z| 95% CI
At Impact Speed Level ≤ 30 km/h
ISS 3.52 0.004 1.50, 8.27 3.34 0.009 1.35, 8.29
MAIS 2.18 0.016 1.15, 4.13 2.13 0.026 1.10, 4.16
AIS 2.32 0.000 1.50, 3.59 2.27 0.000 1.43, 3.60
Mortality 1.79 0.054 0.99, 3.25 1.87 0.072 0.95, 3.68
At Impact Speed Level > 30 km/h
ISS 1.43 0.199 0.83, 2.48 1.13 0.700 0.60, 2.13
MAIS 1.27 0.410 0.72, 2.21 0.92 0.789 0.49, 1.73
AIS 1.09 0.317 0.92, 1.30 1.02 0.863 0.84, 1.24
Mortality 5.50 0.142 5.56, 35.60 5.23 0.156 0.53, 51.41
Open in a new tab

Table 6.

Relative Risk of severe injury (ISS > 16) for LTV vs. passenger cars for different body regions after adjusting for pedestrian age.

Body region Impact speed ≤ 30 km/h Adjusted OR (95% CI) Impact speed > 30 km/h Adjusted OR (95% CI)
Head and Face 1.99 (1.22, 3.24) 1.19 (0.86, 1.66)
Torso 6.69 (2.71, 16.53) 1.29 (0.76, 2.20)
Upper ex. 5.60 (2.35, 13.34) 0.92 (0.60, 1.41)
Lower ex. 2.81 (1.66, 4.77) 1.09 (0.79, 1.49)
Open in a new tab

Assessing the combined risk of pedestrian age and vehicle type

We examined the risk of different combinations of pedestrian age groups and vehicle type groups. Thus, we established four different combinations of pedestrian age and vehicle type (children vs. passenger cars, children vs. LTV, adults vs. passenger cars, and adults vs. LTV). The highest rates of severe injury and mortality were among adults hit by LTV while the least were among children hit by passenger cars. Among adults hit by LTV, rates of severe injury (ISS > 16) and mortality were 318 and 134 per 1000 pedestrians respectively. For children hit by passenger cars, the same rates were 126 and 55 per 1000 pedestrians respectively. These findings were consistent with the results shown in the formal analyses (Table - 7). Results of logistic regression confirmed the results shown in table - 7 and demonstrated that adults hit by LTV had higher risks of severe injury and mortality than children hit by passenger cars. This association was statistically significant for both crude and adjusted odds ratios for impact speed (Table - 8).

Table 7.

Severe Injury or Mortality Rates per 1,000 Injured Pedestrians for Different Pedestrian Age Groups Struck by Different Types of Vehicles, Using Different Injury Severity Outcomes.

Outcome measure ISS MAIS AIS Mortality
Risk groups >16 >3 > 3
Children vs. Passenger cars 126.3 189.5 92.5 54.9
Children vs. LTV 129.0 258.1 102.6 N/A
Adults vs. Passenger cars 275.9 362.1 187.4 88.8
Adults vs. LTV 318.2 397.7 214.7 134.8
Open in a new tab

Table 8.

Adjusted ORs (to impact speed) of Severe Injury or Mortality Comparing Adults Struck by LTV vs. Children Struck by Passenger Cars.

ISS MAIS AIS Mortality
OR P>|Z| OR P>|Z| OR P>|Z| OR P>|Z|
Crude 1.48 0.002 1.41 0.002 1.39 0.00 1.47 0.037
Adjusted 1.38 0.044 1.37 0.026 1.10 0.14 1.07 0.00
Open in a new tab

DISCUSSION

Pedestrian injury is a multi-faceted public health problem. Various factors may influence the occurrence and severity of pedestrian injuries such as vehicle, pedestrian, driver and environmental factors. The purpose of this study is to examine particular pedestrian and vehicle factors as being possibly associated with pedestrian injuries. Assuming that different vehicle types may result in different levels of risk for pedestrians, this can be either explained by vehicle engineering design differences or because the speed on which these vehicles are driven are different. Engineering design differences include vehicle shape, size, and weight. LTV tend to be higher, bigger and stiffer than cars. Thus, any comparison between cars and LTV will generally examine these factors collectively unless being controlled for one or more of these characteristics. For our purposes in this study we compared LTV vs. passenger cars for their risk on pedestrian injury. Further work may be needed to determine which of theses characteristics is responsible for these risk variations associated with vehicle types.

In this study, the Pedestrian Crash Data Study (PCDS) was used as a trusted source of data since it had been reviewed and checked for accuracy, reliability and quality control. The present study was to examine the risk of morbidity and mortality for children (2–14 years) and adults (19–50 years) when struck by different vehicle types (passenger cars vs. LTV) before and after considering other possible contributing factors (confounders and effect modifiers). The authors excluded the elderly for reasons related to frailty and their degree of heterogeneity and pedestrians ages 15–18 years to rule out the possibility of overlapping of physical characteristics between children and adults in this transitional phase. We included 388 cases when testing age as a risk factor and included all the PCDS cases when testing striking vehicle type as a risk factor for pedestrian morbidity and mortality. Six different injury severity outcome measures were operationalized to assess level, severity and significance of sustained injuries. We noticed a potential level of consistency and agreement between results and conclusions reached by using these different estimates. The results of descriptive and statistical analyses were generally in the same strength, direction, and significance when using different outcome measures. Our risk estimates were to describe risk of aggregate age and vehicle types, though it is expected that not all the vehicles included in certain group or all the pedestrians included in certain age group would have exactly the same risk of pedestrian injury or mortality.

At level A of this analysis, we investigated the association between age (adults vs. children) and pedestrian injury severity. At level B, we investigated the effect of striking vehicle type (passenger cars vs. LTV) on pedestrian injury severity at different impact speed levels. At level C, we investigated different combinations of age groups and vehicle types for being a risk of pedestrian injury or mortality. We examined the function of impact speed in each level either being as a confounder or an interaction term. We adjusted for vehicle type when examining the function of pedestrian age, and adjusted for pedestrian age when examining the role of vehicle type. Thus, the obtained results were truly expressing the role of the predictor under investigation.

The results illustrated that there was some evidence that adults are at a higher risk of severe injury and mortality when involved in pedestrian motor vehicle crashes, though child pedestrians showed noticeable higher average Life Years Lost to Injury (LLI) than adults pedestrians, which can be explained by the higher life expectancy for children at time of crash.

Similarly, compared to passenger cars, pedestrians hit by LTV were at higher risk of severe injury and mortality. This association was evident and statistically significant at low impact speeds (≤ 30 km/h) but not at high impact speeds (> 30km/h). This finding can be explained by the fact that impact speed is a potential effect modifier in the relationship between vehicle type and severity of pedestrian injuries and that the influence of impact speed dominates over the effect of vehicle type on pedestrian injury especially at higher impact speed levels.

As previously mentioned that either vehicle engineering design or vehicle speed may be responsible for the high risk associated with LTV and because this harmful risk can only be incurred at lower impact speeds, we conclude that vehicle design is a potential factor that determines pedestrian injury severity in case of pedestrian motor vehicle impact incidents.

Analyses addressing the combined risk of vehicle type and pedestrian age demonstrated that adults hit by LTV had the highest risk of severe injury and mortality while children hit by passenger cars had the least risk. This finding supported the results obtained for each of the two formal analyses that investigate pedestrian age and vehicle type each separately.

Few published studies have attempted to address the same problem using different approaches and data sources. To our knowledge, this study is the only one to study both pedestrian age and vehicle type after testing for the role of impact speed. Lefler and Gabler (2001) analyzed the Fatality Analysis Reporting System (FARS) to examine the overall trend in pedestrian fatalities from 1991 to 2000. They found an overall decrease in pedestrian mortality rate of 18%. After broken down by vehicle type, this decrease in pedestrian fatalities was mainly in passenger cars impacts (35%), whereas it showed a 10% increase in pedestrian fatalities from LTV impacts. Lefler and Gabler (2001) concluded that pedestrians struck by an LTV are much more likely to sustain fatal injuries than pedestrians struck by passenger cars. Peng and Bongard (1999) analyzed 5,000 pedestrian patients in Los Angeles County Trauma Database for 1994–1996. They grouped patients by age: pediatrics (14 years and under), adults (15–65). In spite of both studies used the same pediatric age definition, the pediatric group represented 38.1% of the study population, which was much higher than that in our study (23%). This can be explained by the fact that the PCDS data was collected at the regular working hours where children are usually at their schools and therefore may be less represented in the PCDS database. Mortality rates in Peng and Bongard (1999) were 3.1% for children and 8% for adults compared to 4% for children and 11.7% for adults in the present study. It is obvious that mortality rates for children are very much similar between the two studies. Adult mortality rates were higher than children mortality rates in both studies but understandably different adult mortality rates because of the different adult age definitions in both studies. Mean ISS was higher for adults compared to children in both studies with respect to the different adults age definitions. Lau, Seow and Lim (1998) reviewed 369 pedestrian fatalities in Singapore from 1990 to 1994 and concluded that there was no evident relationship between type of vehicle involved and risk of fatality. They suspected that it might be the speed at which the victim is hit that is more significant than the type of vehicle. Ballesteros et al. (2003) investigated the association between pedestrian injury patterns and vehicle type using Maryland State Police, trauma registry and medical examiner data for the years 1995 to 1999, and concluded that there was an overall increase risk of pedestrian injury or mortality for SUVs and LTV compared to conventional cars.

More information is needed about the injury sources and patterns in pedestrian motor vehicle collisions. Our ultimate goal is to design less aggressive vehicles in an attempt to decrease the toll of pedestrian morbidity and mortality in pedestrian-motor vehicle collisions. This could be possible only by testing the relative pedestrian-vehicle geometry. The authors recommend further investigation on the state, national and international levels.

LIMITATIONS

Out of the 552 pedestrian cases in the PCDS, only 388 cases met our age inclusion criteria when the analysis was to investigate age as a risk factor for severe injury or mortality. In statistical analysis based on both age groups and vehicle body-types, some groups had a small number of observations. We had to group children age 2–14 years old in one group. In a review paper of pediatric pedestrian injuries at level 1 trauma center (The Children national Medical Center, Washington, D.C., USA), Woods et al., (2002) grouped children into four age groups; 0–4, 5–9, 10–14, and 15–19 years. Mean ISS and AIS were almost the same for the first 3 pediatric age groups (up to 14 years). Mean ISS was 7.5, 6.96, and 7.9 and mean AIS was 1.93, 1.97, and 1.97 respectively. This finding supported the authors’ decision to aggregate children age 14 and under in one age group.

The NASS-PCDS data were collected at the scene mostly during the regular working hours when children are expected to be at school, therefore we expect that children may be under-represented in this database. This may limit the generalization of the study findings to the general pediatric pedestrian population.

In this study, only passenger cars, sport utility vehicles, light trucks and vans were included because the PCDS database does not offer data on pedestrian impacts involving other types of vehicles such as buses, motorcycles and automobile derivatives (auto bases pickup, large limousine, three-wheel automobiles, etc.). Hence, limitation of generalization of the study results regarding these vehicle types exists.

The small number of observations may explain some of the non-statistical significance results obtained in this study but certainly explains controlling for only the most important confounders and not all.

CONCLUSIONS

This paper contributes by providing some information about the association between pedestrian age, vehicle type and impact speed and the severity of pedestrian injury and mortality. We proved that compared to passenger cars, pedestrians hit by LTV were more likely to sustain higher injury severity and mortality especially at low impact speed levels (≤ 30 km/h) but not at impact speeds (> 30 km/h). After testing the role of impact speed in this relationship, it is apparent that impact speed has a major influence on pedestrian injury severity in all pedestrian-motor vehicle crashes, and this is especially true at high impact speed levels. At higher impact speeds, the harmful effect of impact speed gradually eliminates the risk associated with different vehicle types. At higher speed levels, the impact will be tremendously devastating on the pedestrian body, irrespective of the effect of vehicle type. These results suggest that changes in vehicle fleet (i.e., more LTV) may lead to increase risk of pedestrian injury severity and mortality particularly at lower impact speed levels.

Similarly, and regardless of impact speed and vehicle type, adults were always at higher risk of severe injury or mortality compared to children in cases of pedestrian motor vehicle impact incidences. Based on the results obtained from this study, adults hit by LTV were at the highest risk of injury and mortality while children hit by passenger cars had the least risk. These conclusions were based on results of both descriptive and logistic analyses and agreed with the other conclusions from this study.

REFERENCES

  1. Anderson R, McLean J. Vehicle Design and Speed and Pedestrian Injury: Australia’s Involvement in the Harmonised Research Activities. Pedestrian Safety Expert Group, 2001. http://www.monash.edu.au/oce/roadsafety/abstracts_and_papers/110/110_revised.pdf.
  2. Association for the Advancement of Automotive Medicine (AAAM) The Abbreviated Injury Scale, 1990 revision. Des Plaines, IL: 1990. [Google Scholar]
  3. ATSB 2001 Monthly Bulletin, Road Fatalities Australia, March 2001, ATSB Canberra.
  4. Baker SP, O’Neil B, Haddon W, et al. The Injury Severity Score: A method for describing patients with multiple injuries and evaluating emergency care. J Trauma. 1974;14:187–196. [PubMed] [Google Scholar]
  5. Ballesteros MF, Dischinger PC, Langenberg P. Pedestrian Injuries and Vehicle Type in Maryland, 1995–1999. Accident Analysis and prevention. 2003;927:1–9. doi: 10.1016/s0001-4575(02)00129-x. [DOI] [PubMed] [Google Scholar]
  6. Barss P. Injury prevention: An International Prospective-Epidemiology, Surveillance and policy. New York: Oxford University Press; 1998. [Google Scholar]
  7. Champion HR, Copes WS, Sacco WJ, Lawnick MM, et al. The Major Trauma Outcome Study: Establishing Rational Norms for Trauma Care. J Trauma. 1990;30(11):1356–65. [PubMed] [Google Scholar]
  8. Chidester AB, Isenberg RA. Final Report–The Pedestrian Crash Data Study. Proceeding of thr 17th International Enhanced Vehicle Safety Vehicle Conference, Paper 248; May 2001; Amsterdam, Netherlands. [Google Scholar]
  9. DiMaggio C, Durkin M. Child pedestrian injury in an urban setting: Descriptive Epidemiology. Acad Emerg Med. 2002;9:54–62. doi: 10.1111/j.1553-2712.2002.tb01168.x. [DOI] [PubMed] [Google Scholar]
  10. Gabler HC, Hollowell WT. NHTSA’s vehicle aggressivity and compatibility research program. ESV; 1998. [Google Scholar]
  11. Hill LA, Delany LM, Duflou J. A Population –Based Study of Outcomes after Injury to Car Occupants and Pedestrians. The Journal of Trauma. 1996;40(3):351–355. doi: 10.1097/00005373-199603000-00004. [DOI] [PubMed] [Google Scholar]
  12. Isenberg RA, Chidester AB, Mavros S. Updates on the Pedestrian Crash Data Study. National Highway Traffic Safety Administration (NHTSA); Paper Number 98-S6-O-05. [Google Scholar]
  13. Jacobs G, Thomas A. Africa Road Safety Review: Final Report, US Department of Transportation, Federal Highway Administration. 2002. [Google Scholar]
  14. Jarrett KL, Saul RA. Pedestrian Injury-Analysis of the PCDS Field Collision Data. National Highway Traffic Safety Administration (NHTSA); Paper Number 98-S6-O-04.1998. [Google Scholar]
  15. Khattak AJ, Rocha MS. Are SUVs “Supremely Unsafe Vehicles?” Analysis Of Rollovers and Injuries. Transportation Research Board; 82nd Annual Meeting; Jan. 2003; Washington, D.C. [Google Scholar]
  16. Lau G, Seow E, Lim ESY. A Review of Pedestrian Fatalities in Singapore, 1990 to 1994. Ann Acad Med Singapore. 1998;27:830–7. [PubMed] [Google Scholar]
  17. Leaf WA, Preusser DF. National Highway Traffic Safety Administration, US Department of transportation Report. 1999. Literature Review on Vehicle Travel Speeds and Pedestrian Injuries. DOT HS 809 021. [Google Scholar]
  18. Lefler DE, Gabler HC. The Emerging Threat of Light Truck Impacts With Pedestrians. Proceedings of the Seventeenth International Conference on Enhanced Safety Vehicles (ESV), Paper No. 212; Amsterdam, The Netherlands. June 2001. [Google Scholar]
  19. Liu XJ, Yang JK. Improvements to vehicle front-end design for the protection of child and adult pedestrians using mathematical simulations. Int Journal of Crashworthiness. 2002 [Google Scholar]
  20. MacKenzie Ellen J, Damiano Ann, Miller Ted, et al. The Development of Functional Capacity Index. The Journal of Trauma. 1996;41(5):799–807. doi: 10.1097/00005373-199611000-00006. [DOI] [PubMed] [Google Scholar]
  21. Matsui YWA, Konosu A. Comparison of Pedestrian System safety Tests Using Impactors and Full-Scale Dummy Tests. Automotive Crash Research Side Impact, Rollover and Vehicle Aggressivity, SEA 1671; 2002. [Google Scholar]
  22. Mayr JM, Edar C, Berghold A, et al. Causes and Consequences of Pedestrian Injuries In Children. Eur J Pediatr. 2003;162:184–190. doi: 10.1007/s00431-002-1099-1. [DOI] [PubMed] [Google Scholar]
  23. McCann B, DeLille B. Mean Streets 2000: Pedestrians Safety, Health and Federal Transportation Spending. Columbia, SC: Center for Disease Control and Prevention; 2000. [Google Scholar]
  24. National Accident Sampling System, National Accident Sampling Study (NASS-PCDS) Data Collection, Coding and Editing Manual. U.S. Department of Transportation; 1996. [Google Scholar]
  25. National Highway Traffic Safety Administration. Traffic Safety facts - 1997, 1998, 1999, 2000, 2001- Pedestrians; Washington, DC: US Department of Transportation; [Google Scholar]
  26. Peng RY, Bongard FS. Pedestrian Versus Motor Vehicle Accidents: An Analysis of 5,000 patients. J Am Coll Surg. 1999;189:343–348. doi: 10.1016/s1072-7515(99)00166-0. [DOI] [PubMed] [Google Scholar]
  27. Rivara FP. Child Pedestrian Injuries in the United States: Current Status of the Problem, Potential Interventions, and Future Research needs. AJDC. 1990;144:692–6. doi: 10.1001/archpedi.1990.02150300090023. [DOI] [PubMed] [Google Scholar]
  28. Robertson C, Redmond AD. The Management of Major Trauma. New York: Oxford University Press; 1991. p. 22. [Google Scholar]
  29. Rogers CD, Pagliarello G, McLellan BA, et al. Mechanism of injury influences the pattern of injuries sustained by patients involved in vehicular trauma. Can J Surg. 1991;34:283–286. [PubMed] [Google Scholar]
  30. Seow E, Lau G. Who dies at A&E? The role of forensic pathology in the audit of mortality in the audit of mortality in an emergency medicine department. Forensic Sci Int. 1996;82:201–10. doi: 10.1016/s0379-0738(96)01997-4. [DOI] [PubMed] [Google Scholar]
  31. Stammen JA, Saul RA, Ko B. Pedestrian Impact Head Testing and PCDS Reconstructions. ESV; Jun, 2001. [Google Scholar]
  32. The Economic Impact of Motor Vehicle Crashes 2000, Appendix C. National Highway Traffic Safety Administration (NHTSA) US Department of Transportation; Washington, DC: [Google Scholar]
  33. Woods W, Sherwood C, Iversson J, Crandall J, et al. A Review of Pediatric Pedestrian Injuries At A Level 1 Trauma Center. 2002 [Google Scholar]

Articles from Annual Proceedings / Association for the Advancement of Automotive Medicine are provided here courtesy of Association for the Advancement of Automotive Medicine

RESOURCES