Journal of Surgical Research 167, 1418 (2011)

doi:10.1016/j.jss.2010.10.007

ASSOCIATION FOR ACADEMIC SURGERY

Does the Pattern of Injury in Elderly Pedestrian Trauma Mirror
That of The Younger Pedestrian?1
Suryanarayana M. Siram, M.D.,2 Victor Sonaike, M.D., Oluwaseyi B. Bolorunduro, M.D., MPH.,
Wendy R. Greene, M.D., Sonja Z. Gerald, M.D., David C. Chang, Ph.D., Edward E. Cornwell III, M.D.,
and Tolulope A. Oyetunji, M.D., M.P.H
Department of Surgery, Howard University College of Medicine, Washington D.C
Submitted for publication January 8, 2010

six to eight times more likely to die (OR 6.24 and 8.27,
P < 0.001).
Conclusion. Elderly patients have higher rates of
fractures and intracranial injuries with an extremely
worse mortality after pedestrian trauma. 2011 Elsevier

Background. Walking is the primary mode of transportation for people aged 65 y and over; hence pedestrian injuries are a substantial source of morbidity
and mortality among elderly patients in the United
States. This study is aimed at evaluating the pattern
of injury in the elderly pedestrians and how it differs
from younger patients.
Methods. Retrospective analysis of the National
Trauma Data Bank (20022006) was performed, with inclusion criteria defined as pedestrian injuries based on
ICD-9 codes, excluding age < 15 y. The following age
categories in years were created: 1524 (reference
group), 2534, 3544, 4554, 5564, 6574, 7584, and
8589. The injury prevalence was compared, and multivariate regression for mortality was conducted adjusting for demographic and injury characteristics.
Results. A total of 79,307 patients were analyzed. Superficial injuries were the most common at 29.1%, with
lower extremity fractures and intracranial injuries following at 25.1% and 21.4% respectively. The very elderly (7584 and 8589) had significantly higher rates
of fractures of the pelvis(16.2% and 16.8% versus 8.1%
in the youngest group), upper (19.3% and 18.4% versus
9.8%), lower extremities (31.1% and 31.9% versus
22.5%) and intracranial injuries (25.5% and 28.7% versus 22.4%), but sustained lower rates of hepatic (2.3%
and 1.7% versus 3.0%) injuries, with no difference
seen in pancreatic, splenic, and genitourinary injuries.
On multivariate analysis, very elderly patients were

Inc. All rights reserved.

Key Words: elderly pedestrian; trauma; lower extremity injuries.

INTRODUCTION

Pedestrian injuries remain a substantial source of

morbidity and mortality in the United States, especially in large urban areas [1]. The management of
these injured pedestrians is often complex and involving multiple systems [2]. Anatomic changes with aging
make it possible that older patients sustain different injuries than their younger counterparts, even with the
same mechanism [3].
Improvements in the standard of living and life expectancy have increased the elderly population in developed countries. Consequently, the number of older
trauma patients has increased as well. Presently, individuals aged 65 y or older represent approximately 12%
of the population in the United States, and expenditures for their trauma care account for 25% of the total
[4, 5]. Approximately 28% of all trauma related deaths
occur in the people who are 65 y of age or older, and
injury is the fifth most common cause of death in this
age group [69]. It has been stated that the
percentage of travel time spent walking is highest for
people over 65 y [7]. In the year 2030, the
65 y age
group is expected to rise to 21% of the population [10].
Given the preponderance of recent literature associating insurance status with outcomes in trauma

1
Presented at the 5th Annual Academic Surgical Congress
meeting, February 35, 2010 in San Antonio, Texas.
2
To whom correspondence and reprint requests should be addressed at Department of Surgery, Howard University Hospital,
2041 Georgia Ave., N.W., Washington, D.C. 20060. E-mail: ssiram@
howard.edu.

0022-4804/$36.00
2011 Elsevier Inc. All rights reserved.

14

Downloaded from ClinicalKey.com at Universidad Ces September 18, 2016.

For personal use only. No other uses without permission. Copyright 2016. Elsevier Inc. All rights reserved.

15

SIRAM ET AL.: ELDERLY PEDESTRIAN TRAUMA

patients, age 65 (the age at which Americans achieve

universal insurability) possesses more than just anatomic interest as a point of analysis [11, 12].
The purpose of this study is to evaluate the pattern of
injury in the elderly pedestrians and explore how it differs from the non-elderly. This study also hopes to determine the appropriate age cutoff at which elderly
injuries defer by outcomes.
METHODS
This was a retrospective analysis of the National Trauma Data
Bank (NTDB ver. 7.1) between 2002 and 2006. The NTDB is managed
by the American College of Surgeons and is the largest repository of
data on trauma inpatients ever assembled with information on over
1.8 million trauma admissions. Data is voluntarily reported from
over 700 trauma centers across the United States and its territories.
Detailed information on the NTDB data collection procedures have
been published elsewhere [9, 13, 14]. This study was reviewed by
the Johns Hopkins University School of Medicine Institutional
Review Board and approved for exempt status.
For purposes of statistical analysis, age was assigned to one of eight
groups: 1524, 2534, 3544, 4554, 5564, 6574, 7584, and 8589.
The 1524 y category was the reference group. We excluded children
< 15 y. The age categories differ from a similar study [3]. The choice to
bin the ages into 10-y categories will help to delineate subtle differences that may otherwise have been missed with larger age categorizations. The extremely elderly group was the only 5-y group due to the
intrinsic limitation of the dataset, which does not collect specific age
data in individuals 90 y and above.
Patients with pedestrian injuries were included in the study, and
we excluded all burn patients. The outcome variables were the occurrence of specific types of injury defined by International Classification
of Diseases, ver. 9 (ICD-9) codes. Demographic and injury severity
characteristics such as age, race, injury severity score (ISS), and injury intent were compared within age groups. Crude mortality was
also calculated for each group. The race and gender distributions of
these groups of patients were also compared. The prevalence of the
different injury types was compared using c2 and t-tests for bivariate
analysis as appropriate.
Multivariate regression analysis for mortality was conducted comparing the elderly patients to the reference age group controlling for
age, gender, race, ISS, shock, insurance status, head injury, extremity
injury, and the teaching status of the hospital. Each of these variables
has been associated with outcomes in large database studies in the
trauma literature. Statistical significance was defined as P < 0.05.
All analysis was performed with STATA ver. 10 (Stata Corp., College
Station, TX).

RESULTS

Table 1 shows the demographic distribution of the

study population, with 79,307 patients meeting inclusion criteria. The mean age was 34.2 y, with a median
ISS of 9. The population was about two-thirds male
and three-quarters were insured. Minorities were
disproportionately represented, accounting for 52% of
patients.
Males accounted for the majority of patients in each
age group, but less substantially so among seniors
65 y and older (Table 2). Patients > 64 y were less likely
to be Black, or uninsured. Patients > 55 y were more

TABLE 1
Study Population Demographic and Injury Severity
Characteristic
N
Gender
Male
Female
Age [mean (SD)]*
1524
2534
3544
4554
5564
6574
7584
8589
Race
White
African American
Hispanic
Asian or Pacific Islander
American Indian/Alaska
Other
Insurance (n 66295)
Private insurance
Government insured
Self pay/none
Other
Injury severity score [median (IQR)]**
<9

9 and <15

15 and <25

25
Shock
Length of stay [median (IQR)]**
Death
*

79,307

51,904
27,235
34.2 (22.1)
12,578
9313
11,611
10,619
6391
4018
3497
760

65.6
34.4

21.4
15.8
19.8
18.1
10.9
6.8
6.0
1.3

36,055
18,149
13,222
2096
526
4435

48.4
24.4
17.8
2.8
0.7
6.0

18,931
14,469
15,610
17,285
9 (417)
31,030
21,581
10,612
11,434
5806
3 (17)
5,508

28.6
21.8
23.6
26.1

41.6
28.9
14.2
15.3
7.7

7.0

SD standard deviation.
IQR interquartile range.

**

likely to experience shock on admission, longer length

of stay, and mortality than patients < 55 y. However,
mortality was significantly higher in the very elderly
groups (7584 y and 8589 y, see Table 2).
The different injury types seen in this patient population are shown in Table 3 stratified by body regions. The
different injury types were defined by specific ICD-9 codes. Over three-quarters of the patients had one form of
external injuries or another, which typically were not
anatomically defined by body regions. Specifically, superficial injuries were the most common at about
29.1%, followed by lower extremity fractures and intracranial injuries 25.1% and 21.4%, respectively. Pelvic
fractures accounted for about 9.2% of the injury types,
with splenic injuries accounting for about 2.5%.
Table 4 describes the prevalence of each type of injury
in the different age groups. Compared with patients
aged 1524 y, the very elderly (7584 years and 85
89 y) had significantly higher rates of fractures of the
pelvis (16.2% and 16.8% versus 8.1% in the youngest
group, P < 0.01), upper (19.3% and 18.4% versus

Downloaded from ClinicalKey.com at Universidad Ces September 18, 2016.

For personal use only. No other uses without permission. Copyright 2016. Elsevier Inc. All rights reserved.

16

JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 1, MAY 1, 2011

TABLE 2
Demographic and Injury Characteristics by Age Categories

Gender (%)
Male
Race (%)
White
African American
Hispanic
Asian or Pacific Islander
American Indian/Alaska
Other
Insurance (%)
Private insurance
Government insured
Self pay/none
Other
Injury severity Score (%)
<9
914
1524

25
Median (IQR)*
Shock
Death
LOS [Median (IQR)]*

1524 (%)

2534 (%)

3544 (%)

4554 (%)

5564 (%)

6574 (%)

7584 (%)

8589 (%)

61.5

69.4

69.6

68.8

63.7

55.7

51.7

47.9

49.5
23.8
18.0
2.1
0.8
5.9

45.2
22.5
23.2
2.3
0.9
5.8

47.3
27.6
17.2
2.01
1.12
4.9

51.5
27.3
13.2
2.3
0.7
5.0

56.2
20.3
13.9
3.9
0.5
5.3

60.9
13.3
13.1
6.5
0.4
5.9

68.9
8.6
11.2
6.5
0.3
4.6

73.9
6.6
7.9
7.4
0.0
4.2

31.9
15.5
26.8
25.8

25.4
12.8
35.0
26.8

25.5
15.1
32.4
27.1

27.6
16.9
28.6
26.9

31.3
18.6
23.1
27.0

22.2
44.4
9.5
24.0

20.5
45.1
9.2
25.3

24.8
41.9
7.5
25.8

39.7
29.2
14.9
16.3
9 (518)
7.9
6.5
4 (19)

36.4
30.5
15.8
17.3
10 (519)
9.3
8.4
4 (211)

32.6
30.2
18.3
18.9
10 (521)
10.1
10.3
5 (211)

30.7
30.2
18.0
21.1
10 (422)
9.6
12.3
5 (210)

28.7
31.8
16.4
23.1
12 (622)
10.3
17.0
5 (210)

28.0
31.5
17.5
23.1
10 (622)
10.5
19.2
5 (211)

46.2
27.7
12.7
13.5
9 (416)
5.2
4.5
3 (16)

43.3
27.8
13.8
15.1
9 (417)
6.6
5.5
3 (18)

All P values <0.001.

IQR inter-quartile range.

TABLE 3
Injury Pattern by Body Regions
Injury type by region
External injuries (Total %)
Superficial injury
Contusion of head/neck/
trunk/extremity
Open wound of lower extremity
Open wound of upper extremity
Open wound of head/neck/trunk
Extremities (total %)
Fracture of upper extremity
Fracture of lower extremity
Sprain of lower extremity
Head and neck (total %)
Intracranial injury
Fracture of skull
Chest (total %)
Heart and lung injury
Pneumothorax/hemothorax
Fracture of ribs, sternum,
larynx, trachea,
Abdomen, pelvis (total %)
Hepatic injury
Splenic injury
Genitourinary injury
Retroperitoneal, peritoneal and
extra-hepatic biliary injury
Gastrointestinal tract injury
Pancreatic injury
Pelvic fracture

ICD9 Codes

Percentage

910919
920924

73.9
29.1
17.5

890894
880884
870879

861
860
807

4.3
3.9
19.1
39.0
11.6
25.1
2.3
31.9
21.4
10.5
20.1
6.0
5.2
8.9

864
865
866867
868

19.7
3.0
2.5
2.3
1.6

863
863
808

0.9
0.2
9.2

810819
820828
843845
850854
800802

9.8%), lower extremities (31.1% and 31.9% versus 22.5%). and intracranial injuries (25.5% and
28.7% versus 22.4%), but sustained lower rates of hepatic (2.3% and 1.7% versus 3.0%, P < 0.001) injuries,
with no difference in the rates of pancreatic, splenic
and genitourinary injuries (see Table 4).
On multivariate regression controlling for demographics and injury severity characteristics, older patients significantly had increasing odds of mortality
after pedestrian trauma (Fig. 1). The 2534 y age
group was not statistically different from the reference group (OR 1.08, CI 0.901.30) in contrast to
the other age categories with significantly worse
mortality with increasing age compared to the reference group (Fig. 1). Statistically, overlapping confidence interval in the younger four age groups
comparison (2564 y) suggests the likelihood of no
difference amongst the group. However, the three
age categories > 65 y showed a significant difference
compared with the younger age groups (non-overlapping confidence interval), and when compared within
groups (Fig. 1). Elderly patients 6574, 7584, and
8589 years were approximately four, six, and eight
times more likely to die after pedestrian injury compared with those 1524 y, and the difference in mortality seen within subsets of patients > 65 y was also
statistically significant.

Downloaded from ClinicalKey.com at Universidad Ces September 18, 2016.

For personal use only. No other uses without permission. Copyright 2016. Elsevier Inc. All rights reserved.

17

SIRAM ET AL.: ELDERLY PEDESTRIAN TRAUMA

TABLE 4
Comparison of Injuries by Body Regions Among the Different Age Categories
1524

2534

3544

4554

5564

6574

7584

8589

31.3
17.2

26.2
16.6

24.5
16.2

23.9
15.8

22.5
17.6

24.6
20.7

24.0
22.5

22.8
22.1

<0.001
<0.001

4.4
4.4
19.1

4.7
3.9
19.3

4.7
3.9
19.9

3.9
3.9
19.5

3.9
3.9
20.8

4.2
5.0
21.6

5.1
7.5
24.1

5.9
9.9
25.7

<0.001
<0.001
<0.001

9.8
22.5
3.3

10.9
23.0
3.0

11.5
26.4
2.7

12.9
27.4
2.4

15.1
29.1
2.3

16.2
29.0
2.3

19.3
31.1
2.0

18.4
31.9
2.2

<0.001
<0.001
<0.001

22.4
10.2

19.6
9.9

20.0
9.6

20.2
10.2

23.0
11.2

25.3
11.7

25.5
12.4

28.7
11.6

<0.001
<0.001

5.8
4.2
4.6

5.4
4.6
7.1

5.6
5.9
10.8

5.7
6.1
13.0

5.7
6.9
14.8

5.5
7.1
16.0

6.6
7.1
17.9

6.3
5.1
16.6

0.315
<0.001
<0.001

3.0
2.7
2.3
1.4

3.2
2.3
2.8
1.4

3.0
2.4
2.8
1.7

3.0
2.4
2.4
1.8

2.6
2.6
2.4
2.0

2.4
2.2
2.4
2.0

2.3
2.1
2.9
2.0

1.7
1.3
2.4
2.5

0.01
0.083
0.161
0.004

0.8
0.2
8.1

0.9
0.2
8.5

1.2
0.2
9.2

1.1
0.2
9.8

1.2
0.4
11.9

1.3
0.3
15.3

1.2
0.2
16.2

1.1
0.4
16.8

0.027
0.257
<0.001

DISCUSSION

This study clearly demonstrates a difference in injury

pattern among elderly patients > 65 y compared with
younger patients. The present study describes the different patterns of pedestrian injury by age stratification with comparative analysis directed towards
patients aged
65 y with those aged 1524. While other
age subsets have been utilized in other studies as typified by the paper by Demetriades et al. [3], this study
further highlight differences that otherwise can easily
be missed among patients > 65 y of age.
From our study, patients > 65 y sustained more fractures of the skull, pelvis, upper limbs, lower limbs, and
intracranial injuries than the reference age group.
However, they suffered fewer sprains and hepatic injuries. There was no difference in the rates of pancreatic, splenic, and genitourinary injuries. One might
summarize these findings with the observation that
bony injuries of all types are prevalent in the elderly,
and they are also more prone to intra-cranial injuries.
These findings in themselves are not surprising if one
considers the physiologic and anatomic changes that occur with increasing age. The higher fracture rates
observed are probably be due to the effect of osteoporosis, muscle atrophy, and decreased subcutaneous tissue

in elderly patients. The increased prevalence of intracranial injuries, and pneumothorax/hemothorax could
be associated with the higher rates of skull and rib fractures, respectively. The prevalence of skull fracture
would appear to overcome any theoretic benefit, one

12.00

Odds of Death

External injuries
Superficial injury
Contusion of head/neck/trunk/
extremities
Open wound of lower extremity
Open wound of upper extremity
Open wound of head/neck/trunk
Extremities
Fracture of upper extremity
Fracture of lower extremity
Sprain of lower extremity
Head and neck
Intracranial injury
Fracture of skull
Chest
Heart and lung injury
Pneumothorax/hemothorax
Fracture of ribs, sternum, larynx,
and trachea
Abdomen and pelvis
Hepatic injury
Splenic injury
Genitourinary injury
Retroperitoneal, peritoneal and
extra-hepatic biliary injury
Gastrointestinal tract injury
Pancreatic injury
Pelvic fracture

Adjusted Odds of mortality after

Pedestrian injury (Reference 15-24
year category)

10.00
8.00
6.00
4.00
2.00
0.00

25-34 35-44 45-54 55-64 65-74 75-84 >=85

yrs
yrs
yrs
yrs
yrs
yrs
yrs

Upper CI

1.30

1.48

2.07

2.79

4.46

7.60

11.36

Lower CI

0.90

1.05

1.48

1.95

3.02

5.13

6.02

Odds Ratio 1.08

1.25

1.75

2.33

3.67

6.24

8.27

Age Categories
FIG. 1. Adjusted odds of mortality by age categories after pedestrian trauma.

Downloaded from ClinicalKey.com at Universidad Ces September 18, 2016.

For personal use only. No other uses without permission. Copyright 2016. Elsevier Inc. All rights reserved.

18

JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 1, MAY 1, 2011

might otherwise assume occurs from the fact that brain

atrophy accommodates small intracranial bleeds without significant rise in intracranial pressure. Similar
finding is corroborated by Demetriades et al., which
showed that the elderly were more likely to sustain
fractures after pedestrian trauma [3]. Weakened ligaments at joints also explain the decreased prevalence
of sprains with increased fractures instead. By body regions, elderly pedestrians suffer significantly higher
head injuries, extremity injuries and in the abdomen
and pelvic region, mostly pelvic fractures.
An important finding in this study is the difference in
mortality observed in subsets of elderly patients as presently defined, age
65 y. While the elderly age cutoff
has typically been accepted to be 65 y and older, mostly
based on payer status, this study clearly demonstrates
that based on statistics, a difference exists by outcomes
from age 65 y. A sharp rise in mortality with a nonoverlapping confidence interval in the age group 6574
clearly shows that elderly outcome is worse compared
with the relatively younger group. Among pedestrianrelated trauma patients, this study shows that patients
1564 y have similar mortality outcomes regardless of
age stratification or injury and demographic characteristics. However, patient > 65 years, though different from
younger age groups, also differ compared with one another. In essence, while 65 y age cutoff correlate with the
present payer status system of Medicare as definition of
elderly citizens, it is critically important to appreciate
the difference that exists even among elderly patients.
The coming decades will be the first time in the history
of the United States when a large proportion of the population will be in the Medicare defined elderly age group. It
may be a grievous error to assume a universal outcome after injury based on age
65 y. This study clearly demonstrates worse outcome with increasing age, and it is
important that health care providers consciously factor
this into the care delivered to elderly patients, as we progressively care for more elderly patients in the forthcoming decades
To our knowledge, this is the largest study of injured
elderly pedestrians ever published. A study by Schwab
et al. suggested that there has been an overall decrease
in pedestrian deaths per capita in recent years, but this
decline has been significantly less for older pedestrians
[6, 15, 16]. If the prevalent trends persist, the numbers
of older pedestrians killed and injured can only be
expected to grow given the aging of the population.
With a retrospective study using a large dataset, limitations do exist in its utilization. Variables like for comorbidities present in the elderly, collision speed and
type of the vehicles involved in the accident were not accounted for. This could have further enhanced the independent effect of age on the outcome. However, we
believe the effect of co-morbidities would have been

strongly correlated with age and may not both fit in the
model.
In conclusion, this study clearly demonstrates that elderly injuries after pedestrian injury differ by injury
patterns and are associated with a significantly worse
mortality. We also showed that age 65 y still remains
an appropriate cutoff for the definition of the elderly
as it statistically represents the age when the mortality
becomes significantly different from the younger subsets. Finally, we outlined the importance of appreciating the subtle differences that exist even amongst the
elderly cohort, with older citizens experiencing a doubling in odds of mortality with increasing decades of
age. Therefore, management of elderly patients in the
emergency department after pedestrian trauma needs
to be individualized, and this can serve as a guide in policy creation for those in charge of elderly care as a whole.
REFERENCES
1. Yee WY, Cameron PA, Bailey MJ. Road traffic injuries in the elderly. Emerg Med J 2006;23:42.
2. Starnes M, Hadjizacharia P, Chan L, et al. Automobile versus
Pedestrian Injuries: Does gender matter? J Emerg Med 2008;
23:507.
3. Demetriades D, Murray J, Martin M, et al. Pedestrians injured
by automobiles: Relationship of age to injury type and severity.
J Am Coll Surgeons 2004;199:382.
4. Akkose Aydin S, Bulut M, Fedakar R, et al. Trauma in the elderly
patients in Bursa. Ulus Travma Acil Cerrahi Derg 2006;12:230.
5. Ma OJ, DeBehnke DJ. Geriatric trauma. In: Tintinalli J, Kelen GD,
Stapcznski JS, eds. Emergency medicine: A comprehensive study
guide. 5th ed. New York: McGraw-Hill, 1999. p. 1623.
6. Schwab CW, Kauder DR. Trauma in the geriatric patient. Arch
Surg 1992;127:701.
7. Keall MD. Pedestrian exposure to risk of road accident in New
Zealand. Accid Anal Prev 1995;27:729.
8. Ferrera PC, Bartfield JM, DAndrea CC. Geriatric trauma: Outcomes of elderly patients discharged from the ED. Am J Emerg
Med 17;629.
9. Arthurs ZM, Starnes BW, Sohn VY, et al. Functional and survival
outcomes in traumatic blunt thoracic aortic injuries: An analysis of
the National Trauma Data Bank. J Vasc Surg 2009;49:988.
10. Stutts J, Waller P, Martell C. Older driver population and crush
involvement trends 19741986. Proceeding of the 33rd Annual
Meeting of the Association for the Advancement of Automotive
Medicine. Baltimore, MD, October 24, 1989.
11. Greene WR, Oyetunji TA, Bowers U, et al. Insurance status is
a potent predictor of outcomes in both blunt and penetrating
trauma. Am J Surg 2010;199:554.
12. Haider AH, Chang DC, Efron DT, et al. Race and insurance status as risk factors for trauma mortality. Arch Surg 2008;143:945.
13. Moore L, Hanley JA, Lavoie A, et al. Evaluating the validity of
multiple imputations for missing physiological data in the National Trauma Data Bank. J Emerg Trauma Shock 2009;2:73.
14. Committee on Trauma, American College of Surgeons. NTDB
ver. 7.0. Chicago, IL. 2008.
15. Projections of the Population of the United States, by Age, Sex,
and Race: 1983 to 2080. Washington, DC: US Bureau of the Census, 1990. US Census Statistics, ser P-25, no. 952.
16. Spencer G. Projections of the Population of the United States by
Age, Sex and Race: 1988 to 2080. Washington, DC: US Bureau of
the Census, 1989. Current Population Reports, ser P-25, no. 1018.

Downloaded from ClinicalKey.com at Universidad Ces September 18, 2016.

For personal use only. No other uses without permission. Copyright 2016. Elsevier Inc. All rights reserved.