Factors Associated with Rural Run-Off-Road and Urban Run-off-road Crashes: A Study in the United States

Abstract

A Run-Off-Road (ROR) crash occurs when a vehicle leaves the travel lane resulting in a collision. ROR crashes have become a major cause of serious injuries and fatalities in the United States. Data from Kansas Crash and Analysis Reporting System database during the period 2007 to 2011 were used in this study to examine ROR crashes. Identification of various characteristics related to environment, roadway, driver, and vehicle as well as factors contributing to rural ROR and urban ROR crashes is important because potential countermeasures can be developed to improve roadside safety. It was found that avoidance/evasive actions; driver being ill, falling asleep or fatigued; or animal at the road are more common on rural roadways than urban roadways, leading to ROR crashes.

Full text

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
FACTORS ASSOCIATED WITH RURAL RUN-OFF-ROAD AND
URBAN RUN-OFF-ROAD CRASHES: A STUDY IN THE UNITED STATES
Abstract: A Run-Off-Road (ROR) crash occurs when a vehicle leaves the travel lane resulting in a
collision. ROR crashes have become a major cause of serious injuries and fatalities in the United
States. Data from Kansas Crash and Analysis Reporting System database during the period 2007 to
2011 were used in this study to examine ROR crashes. Identification of various characteristics related
to environment, roadway, driver, and vehicle as well as factors contributing to rural ROR and urban
ROR crashes is important because potential countermeasures can be developed to improve roadside
safety. It was found that avoidance/evasive actions; driver being ill, falling asleep or fatigued; or
animal at the road are more common on rural roadways than urban roadways, leading to ROR crashes.
Key Words: Run-off-Road crashes, crash data analysis, lane departure crashes, rural/ urban nature of
ROR crashes
1. INTRODUCTION
Each year ROR crashes cause serious injuries
and fatalities throughout the world, where ROR
crashes is a major safety concern in the United
States of America (U.S.A.) as well. Data from
Fatality Analysis Reporting System (FARS)
illustrated that ROR crashes cause around 33%
of fatalities in the U.S.A. in 2009 (FARS 2012).
The statistics about single vehicle ROR crashes
is alarming; FARS reported about 53% of the
total traffic fatalities in the U.S.A. are due to
single-vehicle ROR crashes where vehicles leave
the travel lanes resulting in collisions with fixed
objects or overturning (FARS 2012). ROR
crashes usually involve running off the road onto
the right or left shoulder and hitting a fixed
object or a parked vehicle. ROR crashes also
involve crossing into an opposite lane and
colliding with an oncoming vehicle. Those
crashes resulting in between two moving
vehicles may be potentially more severe. An
estimated societal cost of $110 billion has been
imposed each year due to roadside crashes
(McGinnis et al. 2001). Dedicated efforts are
necessary to be taken in order to reduce the
severity and frequency of ROR crashes. The
efforts that have been taken until now are mostly
related to removing or relocating roadside
hazards for designing of forgiving roadsides, or
designing better safety features to mitigate the
severity of these hazards in case these hazards
cannot be removed. (Mak 2010). The societal
costs associated with roadside crashes must be
recognized before developing cost-effective
strategies to improve roadside safety. All
elements of roadside vehicle-driver system need
to be taken into consideration while addressing
the roadside safety problem (McGinnis et al.
2001). The roadside safety characteristics in
rural roadways are different than urban
roadways. Therefore, recognition of various
Niranga Amarasingha
Senior Lecturer
Department of Civil Engineering
Sri Lanka Institute of Information
Technology
New Kandy Road, Malabe, Sri Lanka
Fax: +94 11 241 3901
E-mail: niranga.a@sliit.lk
Sunanda Dissanayake
Professor
Department of Civil Engineering
Kansas State University
2118, Fiedler Hall, Manhattan,
KS 66506, U.S.A.
Fax: +1 785 532 7717
E-mail: sunanda@ksu.edu
39

Niranga Amarasingha, Sunanda Dissanayake
driver, roadway, vehicle, and environmental
characteristics of ROR crashes in rural areas
from those in urban areas is necessary, which
will help to develop appropriate countermeasures
to reduce ROR crashes.
2. LITERATURE REVIEW
Numerous studies have been conducted in recent
years on roadside safety crashes. Hallmark et al.
(2009) investigated different characteristics
associated with large truck ROR crashes using
the data from Large Truck Crash Causation Study
database. While investigating roadway
characteristics, it was found that large truck ROR
crashes are more common in urban freeways
with more than four lanes. For lane departure
crashes, the most common condition was
identified as no flow restrictions. ROR crashes
were found to occur more in dry condition, while
rain was more common for multi-vehicle ROR
crashes in comparison to other ROR crashes.
Single vehicle ROR found to be more common
at nights without streetlights than the multi-
vehicle ROR crashes. Among different driver
related factors, it was identified that fatigue was
the topmost reason for ROR crashes to occur.
Distraction by either an internal or an external
event prior to the crash was also responsible for a
large number of ROR crashes. Aggression was
more likely to be the reason of single-vehicle
ROR crashes. When alcohol and drug
involvement for drivers were compared, the
research found out that large truck drivers were
more likely to be using illegal drugs (10% to
11%) for ROR crashes than alcohol (around 1%).
Liu and Subramanian (2009) examined
environment, roadway, vehicle, occupants, and
drivers’ performance-related factors for fatal
single-vehicle ROR crashes. While investigating
environment-related factors, it was found that
fatal ROR crashes were more likely to take place
in curved roads, rural roads, roadways with
posted speed limit more than 60 mph, and
roadways with fewer lanes. Also, ROR crashes
were more common in adverse weather
conditions and during night time. Among
different driver related factors, it was found that
ROR crashes were more likely to occur when
drivers were traveling with passengers. Male
drivers, younger drivers, and alcohol impaired
drivers were more likely to be involved in ROR
crashes as well. When considering different
vehicle-related factors, ROR crashes were more
common for passenger cars and speeding
vehicles.
A study done by McLaughlin et al. (2009)
investigated the effects of levels of precipitation,
lighting, and roadway surface conditions in
occurrence of ROR crashes. Precipitation (fog,
mist, and rain) increased the occurrence of ROR
crashes 2.5 times more than clear conditions and
snow or ice increased the likelihood by seven
times than that of dry conditions. While
investigating driver related factors, it was found
that multiple factors were responsible for an
event but distraction/inattention (40% of ROR
events) was the most common contributing
factor. Changes in roadway boundaries such as
the start of a median, narrowing of the lane from
the right, loss of a lane, or atypical roadway
geometry were considered as a contributing
factor in 22% of the events.
The role of ambient light in fatal single-vehicle
ROR crashes was studied by Sullivan and
Flannagan (2002). There were three scenarios:
pedestrian crashes at intersections, pedestrian
crashes on dark, straight, high-speed roads and
single-vehicle ROR crashes that were studied to
see the role of light. The research identified that
single-vehicle ROR crashes on dark, curved
roads had little sensitivity to light compared to
first two scenarios and suggested factors other
than the light level that play a dominant role in
fatal single-vehicle ROR crashes. In a research to
evaluate frequency and injury outcomes of ROR
crashes, Benavente et al. (2006) found that ROR
head-on crashes to be more severe than single-
40

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
vehicle ROR crashes. The study also identified
that ROR head-on crashes had higher associated
costs and longer length of stay than single
vehicle ROR crashes. Montella et al. (2010) in
an in-depth investigation of ROR crashes in a
motorway of Italy found that ROR crashes were
more severe for motorcycles than any other
vehicle types. The same study also showed that
crashes against roadside safety barriers tend to be
less severe than crashes against ditches, walls,
fore-slopes, and back-slopes.
Spainhour and Mishra (2008) in a study done in
Florida examined human, roadway, vehicle, and
environmental factors associated with ROR
crashes. The analysis of the data revealed the fact
that approximately 36 percent of the vehicles
crossed the entire roadway and departed on the
opposite side from the initial roadway departure.
Among different contributory factors, alcohol
was the major one, followed by speed,
inattention and fatigue/sleep. It was also found
that overcorrection had a strong positive
association with the presence of rumble strips,
inclement weather, rural locations, incapacitated
drivers, and running off the road to the left or
straight and a strong negative association with
male drivers, speeding, paved or curbed
shoulders, wet or slippery roads, and larger
vehicles. Fewer than 20 percent of fatal ROR
crashes occurred where rumble strips were
present; drivers were more than fifty percent
were more likely to overcorrect than when they
were not present.
In another research conducted by Dissanayake
(2003), a sequential binary logistic regression
modeling was carried to identify the most
important factors related to ROR crashes and to
estimate the severity of young driver ROR
crashes. The study investigated that use of
alcohol or drugs, ejection due to the crash,
gender, impact point of the vehicle, restraint
device usage, urban/rural nature, grade/curve
existence of the crash location, lighting
condition, and speed were the most important
factors affecting the severity of young drivers in
single vehicle ROR crashes.
3. DATA AND METHODOLOGY
Crash data from 2007 to 2011 were obtained
from the Kansas Department of Transportation
(KDOT) for analysis in this study. This data set,
Kansas Crash and Analysis Reporting System
(KCARS) database, is comprised of all police-
reported crashes that occurred in Kansas, U.S.A.
Police officers fill an accident report form with
all details including contributory causes and send
it to KDOT within ten days of the investigation
for any crash which occurs on a public roadway
and which results in death or injury to any person
or total property damage of $1,000 or more
(KDOT 2013). More details of the recording
each of the variables can be found from the
KDOT Accident Reporting Manual (KDOT
2013).
KCARS is Microsoft Access based database
which consists of several tables describing each
crash. The definition for ROR crashes in this
study was the crashes where the vehicles leaving
the roadway encroach upon the median,
shoulders, or beyond and either overturns,
collides with fixed objects or leads to head-on
crashes with other vehicles; sideswipe with
opposing vehicles; or crashes where the first
harmful events occur off the roadway or median-
off roadway in case of divided highway sections.
The tables in the KCARS database were
combined and queries were made to filter all
ROR crashes in order to compare rural and urban
ROR crashes. From the data, it has been found
that ROR crashes were approximately 18% of
total crashes for combined crash data from 2007
to 2011. For the same time period in Kansas
injury ROR crashes were found to be
approximately 24% of the total injury crashes
and fatal ROR crashes were 54% to that of total
fatal crashes. Data were used to investigate ROR
crashes, and calculating their frequencies and
41

Niranga Amarasingha, Sunanda Dissanayake
percentages. The next step of the research was to
develop separate crash severity models for
investigate the further differences.
4. RESULTS AND DISCUSSION
During 2007 to 2011, there were 43,335 ROR
crashes on rural Kansas roadways that accounted
for 48.8% of total ROR crashes occurring in the
state. Figure 1 shows number of ROR crashes
occurred on rural Kansas roadways and urban
Kansas roadways by year.
Figure 1. Rural and Urban ROR crashes by Year
Though the number of ROR crashes in rural
Kansas decreased in 2011 in comparison to prior
four years, they continued to comprise of a
uniform percentage (approximately 8.5%) of
total crashes. Percentages in each sub-category
were calculated by taking the total number of
ROR crashes as the base value. Data such as
numbers of crashes and percentages for each
characteristic and contributory cause were
presented in tabular format in the next sections.
The variables were organized under driver,
environmental, road, vehicle, and crash-related
characteristics, as well as contributory causes.
Information such as “unknown” and/or “other”
for some of variables was not presented in the
tables. Hence, the sum of the percentage for a
particular variable is slightly less than 100.
The frequencies and percentages of ROR crashes
for environmental-related characteristics are
given in Table 1. About 37.2% of ROR crashes
on rural roads occurred under the dark conditions
and approximately 39.1% of ROR crashes on
urban roads occurred in such conditions. The
percentage of ROR crashes represented on rural
roads (16.2%) during 00:00 am- 6:00 am was
slightly lower than that of crashes represented
urban roads (18.0%). During weekends, the
percentage of ROR crashes on rural roads
(32.1%) was slightly higher than that of crashes
on urban roads (30.0%).
42

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
Table 1. ROR Crash Frequencies and Percentages by Urban/Rural Nature:
Environmental Related Factors
Environmental-Related
Characteristics
Number of ROR Crashes
Urban
Rural
Total
Number
%
Number
%
Number
%
Light Conditions
Daylight
27,178
59.9
26,927
62.1
54,167
61.0
Dark
17,733
39.1
16,138
37.2
33,922
38.2
Weather Conditions
Normal conditions
32,490
71.6
30,753
71.0
63,330
71.3
Adverse conditions
12,477
27.5
12,288
28.4
24,791
27.9
Time of Crash
6.00-12.00-Morning
11,251
24.8
11,621
26.8
22,904
25.8
12.00-18.00-Afternoon
14,571
32.1
13,605
31.4
28,209
31.8
18.00-24.00-Evening
11,364
25.1
11,094
25.6
22,485
25.3
00.00-6.00-Night
8,175
18.0
7,015
16.2
15,211
17.1
Day of Week
Week days
31,698
69.9
29,382
67.8
61,162
68.9
Week end
13,594
30.0
13,919
32.1
27,544
31.0
Total Number of Crashes
45,361
100.0
43,335
100.0
88,809
100.0
The frequencies and percentages of ROR crashes
for driver-related characteristics are given in
Table 2. Higher percentage of male drivers was
involved ROR crashes on rural roads compared
to ROR crashes on urban rods. Also, the
percentages of middle age and older drivers in
ROR crashes were slightly higher when they
were traveling on rural roads than traveling on
urban roads. The percentage of restricted
licenses holders’ representation in rural ROR
crashes (34.1%) was slightly higher than in
urban ROR crashes (27.6%). The percentage of
unrestrained drivers in rural ROR (11.9%) was
higher than the drivers in urban ROR crashes
(7.5%).
43

Niranga Amarasingha, Sunanda Dissanayake
Table 2. ROR Crash Frequencies and Percentages by Urban/Rural Nature: Driver Related Factors
Driver-Related Characteristics
Number of ROR Crashes
Urban
Rural
Total
Number
%
Number
%
Number
%
Gender
Female
15,568
34.3
14,953
34.5
30,558
34.4
Male
26,565
58.6
27,405
63.2
54,045
60.9
Age
Young (<24 years)
17,814
39.3
16,104
37.2
33,937
38.2
Middle Age (25-64 years)
25,284
55.7
24,561
56.7
49,931
56.2
Old (>64 years)
2,263
5.0
2,670
6.2
4,941
5.6
License Compliance
Valid licensed
36,047
79.5
38,827
89.6
74,973
84.4
Not valid licensed
4,881
10.8
3,039
7.0
7,933
8.9
Restriction Compliance
No restrictions on driver license
25,496
56.2
25,242
58.2
50,795
57.2
Restricted license
12,529
27.6
14,795
34.1
27,367
30.8
Safety Equipment used
Safety belt used
33,759
74.4
33,352
77.0
67,190
75.7
Safety belt not used
3,408
7.5
5,159
11.9
8,592
9.7
Airbag
Airbag deployed
3,535
7.8
2,658
6.1
6,218
7.0
Airbag not deployed
38,991
86.0
35,951
83.0
75,016
84.5
Alcohol/drug related
Alcohol/drug related
5,560
12.3
4,238
9.8
9,835
11.1
No alcohol or drug
39,801
87.7
39,097
90.2
78,974
88.9
Total Number of Crashes
45,361
100.0
43,335
100.0
88,809
100.0
44

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
It is important to investigate the frequencies and
percentages of ROR crashes for road-related
characteristics by rural and urban nature. As
shown in Table 3, higher percentage of ROR
crashes were reported on black tops in rural areas
compared to urban areas. Also, the percentages
of ROR crashes were slightly higher when they
were recorded on gravel/brick and other road
surfaces in rural areas than urban areas. The
percentages of ROR crashes on dry surfaces and
debris surfaces in rural areas were slightly higher
than that of urban areas. The percentage of rural
ROR crashes (53.9%) on roadways with posted
speed limit between 35 mph and 60 mph was
higher than that of urban ROR crashes (39.0%).
When the roadway posted speed limit was higher
than 60 mph, the percentage of rural ROR
crashes (34.1%) was nearly twice that of urban
ROR crashes (17.7%).
Table 3. ROR Crash Frequencies and Percentages by Urban/Rural Nature: Road Related Factors
Road-Related Characteristics
Number of ROR Crashes
Urban
Rural
Total
Number
%
Number
%
Number
%
Road Surface Type
Concrete
16,796
37.0
5,490
12.7
22,321
25.1
Black top
27,024
59.6
26,414
61.0
53,516
60.3
Gravel/brick or other
1,370
3.0
11,347
26.2
12,717
14.3
Road Surface Condition
Dry
28,224
62.2
27,788
64.1
56,097
63.2
Wet
7,609
16.8
4,974
11.5
12,589
14.2
Debris
9,093
20.0
10,330
23.8
19,445
21.9
Road Surface Character
Straight and level
27,923
61.6
25,250
58.3
53,228
59.9
Straight not level
8,355
18.4
10,374
23.9
18,761
21.1
Curved
8,601
19.0
7,356
17.0
15,983
18.0
Posted Speed Limit
Less than 35 mph
19,646
43.3
5,195
12.0
24,848
28.0
35-60 mph
17,705
39.0
23,365
53.9
41,115
46.3
More than 60 mph
8,010
17.7
14,775
34.1
22,846
25.7
Total Number of Crashes
45,361
100.0
43,335
100.0
88,809
100.0
45

Niranga Amarasingha, Sunanda Dissanayake
The frequencies and percentages of ROR crashes
for vehicle-related characteristics are shown in
Table 4. Automobiles had the largest share of
both for rural and urban ROR crashes although
the percentage of urban ROR crashes involving
automobiles was slightly higher than the
percentage of rural ROR crashes involving
automobiles. Pickup trucks and camper-rv have a
slightly higher percentage for rural ROR crashes
than that for urban ROR crashes. The percentage
of rural truck ROR crashes were approximately
twice that of urban crashes. When examining the
percentages of ROR crashes by vehicle age, there
distribution of rural ROR by vehicle age was
close to the distribution of urban ROR crashes by
vehicle age. The percentage of sole drivers
involving rural ROR crashes was slightly lower
than urban ROR crashes.
As shown in Table 5, the percentage of rural
ROR fatalities was higher than urban ROR
fatalities. Also, the percentages of disabled
injuries, injuries, or possible injuries for ROR
crashes occurred in rural areas were higher than
those of urban areas. Higher percentage of ROR
crashes involving drivers ejected or trapped in
rural areas than those in urban areas. The
percentage of vehicle destroyed at the time of
ROR crashes occurred in rural areas slightly
higher than twice that of crashes occurred in
urban areas. The percentages of ROR crashes
occurred driving on straight-following roadways
in rural areas were higher than that of urban
areas. The percentage of ROR crashes occurred
in rural areas when attempting to turn or lane
change was lower than that of crashes in urban
areas, may be due to low complexity.
Table 4. ROR Crash Frequencies and Percentages by Urban/Rural Nature: Vehicle Related Factors
Vehicle-Related Characteristics
Number of ROR Crashes
Urban
Rural
Total
Number
%
Number
%
Number
%
Vehicle Type
Automobile
24,461
53.9
18,062
41.7
42,572
47.9
Van
2,267
5.0
2,018
4.7
4,289
4.8
Pickup-truck, camper-rv
7,200
15.9
11,593
26.8
18,809
21.2
Sport utility vehicle
7,018
15.5
6,819
15.7
13,856
15.6
Truck
1,561
3.4
3,195
7.4
4,772
5.4
Vehicle Age
Year 4 or newer
12,005
26.5
11,617
26.8
23,651
26.6
5-9 years
17,627
38.9
16,852
38.9
34,518
38.9
10-14 years
12,566
27.7
12,048
27.8
24,646
27.8
Year 15 or older
5,967
13.2
5,490
12.7
11,477
12.9
Occupants
Only driver
33,666
74.2
30,971
71.5
64,716
72.9
Driver and passengers
11,194
24.7
12,079
27.9
23,306
26.2
Total Number of Crashes
45,361
100
43,335
100
88,809
100
46

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
Table 5. ROR Crash Frequencies and Percentages by Urban/Rural Nature: Crash Related
Factors
Crash-Related Characteristics
Number of ROR Crashes
Urban
Rural
Total
Number
%
Number
%
Number
%
Crash Severity
Fatal crashes
376
0.8
1,081
2.5
1,485
1.7
Injury crashes
13,180
29.1
15,651
36.1
28,878
32.5
Non-injury crashes
31,805
70.1
26,603
61.4
58,446
65.8
Driver Injury Severity
Fatal injury
245
0.5
716
1.7
977
1.1
Disabled injury
1,177
2.6
1,965
4.5
3,154
3.6
Injury
4,980
11.0
7,147
16.5
12,150
13.7
Possible injury
4,248
9.4
4,780
11.0
9,041
10.2
Not injured
30,066
66.3
26,570
61.3
56,684
63.8
Ejection
Ejected
1,086
2.4
1,586
3.7
2,693
3.0
Not ejected
40,200
88.6
37,857
87.4
78,138
88.0
Trapped
594
1.3
1,515
3.5
2,120
2.4
Vehicle Damage
Not damage
695
1.5
677
1.6
1,373
1.5
Minor damage
6,276
13.8
4,886
11.3
11,169
12.6
Functional
10,477
23.1
8,311
19.2
18,801
21.2
Disabling
21,326
47.0
18,399
42.5
39,767
44.8
Destroyed
4,681
10.3
10,177
23.5
14,907
16.8
Vehicle Maneuver Before Un-stabilized Situation
Straight-following
27,103
59.7
31,762
73.3
58,940
66.4
Turn or changing lanes
9,055
20.0
3,723
8.6
12,788
14.4
Avoiding maneuver
2,509
5.5
1,334
3.1
3,845
4.3
Stopped, parking or backing
4,056
8.9
4,848
11.2
8,924
10.0
Accident Class
Collision with vehicle
12,966
28.6
6,420
14.8
19,425
21.9
Collision with object
29,629
65.3
26,854
62.0
56,541
63.7
Collision with animal
70
0.2
438
1.0
508
0.6
Collision with pedestrian
58
0.1
18
0.0
76
0.1
Non-collision & overturned
2,622
5.8
9,590
22.1
12,228
13.8
Manner of Collision
Head on
8,001
17.6
2,934
6.8
10,962
12.3
Rear end
248
0.5
150
0.3
398
0.4
47

Niranga Amarasingha, Sunanda Dissanayake
Table 5. ROR Crash Frequencies and Percentages by Urban/Rural Nature: Crash Related
Factors
Crash-Related Characteristics
Number of ROR Crashes
Urban
Rural
Total
Number
%
Number
%
Number
%
Angle side impact
428
0.9
247
0.6
679
0.8
Sideswipe
4,891
10.8
3,227
7.4
8,128
9.2
Backed into
55
0.1
55
0.1
110
0.1
Total Number of Crashes
45,361
100
43,335
100
88,809
100
Also, when the vehicle maneuver was stopping,
parking or backing, higher percentage of rural
ROR crashes were reported than urban ROR
crashes. When examining the percentages of
ROR crashes by accident class, approximately
22.1% of rural ROR crashes were non-collision
or overturn crashes which were higher than that
of urban ROR crashes 5.8%. A higher percentage
of urban ROR crashes were related to collision
with a vehicle than that of rural ROR crashes.
About 17.6% of urban ROR crashes was head-on
crashes, that was higher than the percentage of
rural head-on ROR crashes (6.8%).
Contributory causes for rural and urban ROR
crashes were also investigated using Kansas
crash data. Since a single cause is rarely to blame
for a traffic crash, many factors have been
considered. Driver-related contributory causes
involve the condition of or actions taken by the
driver of the vehicle, while environmental, road
or vehicle-related contributory causes involve the
environmental, road or vehicle conditions at the
time of crash. The frequencies and percentages
of ROR crashes for contributory causes are
shown in Table 6. The contributory causes were
reported according to the opinion of the accident
investigating officer. Driving fast was the top-
ranked driver contributory cause in rural ROR
crashes (22.2%), followed by inattention
(18.9%), and avoidance/evasive or slow (8.3%),
and improper actions taking by the driver (8.3%).
Driving fast and inattention were also the most
frequent contributory causes in urban ROR
crashes. The percentages of urban ROR crashes
due to driving fast, improper action, aggressive
driving, disregarding traffic signs/signals, turning
or lane changing, and failure to yield right of
way were higher than that of rural ROR crashes.
The percentage of total urban ROR crashes
occurred due to driver contributory causes was
higher than that of rural ROR crashes.
The most frequent environmental-related
contributory causes was animal at the road and
its percentage for rural ROR crashes (11.7%)
was slightly higher than that of urban ROR
crashes (9.2%). The percentage of total rural
ROR crashes (16.1%) occurred due to
environmental contributory causes was higher
than that of urban ROR crashes (10.6%). On the
other hand, the percentage of total rural ROR
crashes (2.4%) occurred due to vehicle
contributory causes was slightly lower than that
of urban ROR crashes (2.6%). The percentage of
rural ROR crashes (14.5%) due to road related
factors was slightly lower than that of urban
ROR crashes (16.1%).
48

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
Table 6: ROR Crash Frequencies and Percentages by Urban/Rural Nature: Contributory Causes
Contributory Causes
Number of ROR Crashes
Urban Rural Total
Number
%
Number
%
Number
%
Driver Action Related
Driving fast
11,800
26.0
9,613
22.2
21,426
24.1
Avoidance/ evasive or slow
3,098
6.8
3,581
8.3
6,694
7.5
Improper action
3,080
6.8
2,017
4.7
5,116
5.8
Aggressive driving
3,543
7.8
1,509
3.5
5,057
5.7
Disregarded traffic signs/signals
3,056
6.7
1,457
3.4
4,529
5.1
Turning or lane changing
2,731
6.0
784
1.8
3,522
4.0
Failure to yield right of way
3,304
7.3
708
1.6
4,014
4.5
Driver Condition Related
Alcohol impaired
5,736
12.6
3,755
8.7
9,515
10.7
Ill, falling asleep or fatigued
2,413
5.3
2,981
6.9
5,407
6.1
Driver Distractions Related
Inattention
9,891
21.8
8,201
18.9
18,118
20.4
In vehicle distraction
1,556
3.4
1,599
3.7
3,160
3.6
Total Crashes Occurred Due to Driver
Factors
35,645
78.6
26,322
60.7
62,046
69.9
Environmental Related
Animal
4,194
9.2
5,062
11.7
9,267
10.4
Weather related
335
0.7
1,744
4.0
2,081
2.3
Vision obstruction
293
0.6
253
0.6
550
0.6
Total Crashes Occurred Due to
Environmental Factors
4,793
10.6
6,991
16.1
11,801
13.3
Total Crashes Occurred Due to Vehicle
Factors 1,177 2.6 1,055 2.4 2,234 2.5
Total Crashes Occurred Due to Road
Factors
7,285 16.1 6,283 14.5 13,576 15.3
Total Number of Crashes
45,361
100
43,335
100
88,809
100
49

Niranga Amarasingha, Sunanda Dissanayake
5. SUMMARY AND CONCLUSIONS
This study compared the characteristics and
contributory causes of rural ROR crashes and
urban ROR crashes using Police reported crash
data. The variables which have greater
association with rural ROR crashes compared
with urban ROR crashes were identified. Higher
percentage of rural ROR crashes were reported
during driving on day light condition, adverse
weather condition, morning (6:00 am to 12:00
am) hours, or weekends. Also, male drivers or
middle age drivers were more likely to be
involved in rural ROR crashes compared to
urban ROR crashes. When considered on
roadway characteristics, higher percentage of
rural ROR crashes were reported on gravel/brick
road surface, dry road surfaces, debris road
surfaces, driving on straight not level road
surfaces, or posted speed more than 35 mph
compared to urban ROR crashes. Pick-up trucks,
camper-rv, trucks were more likely to involve
rural ROR crashes compared to urban ROR
crashes. When considering vehicle maneuver,
higher percentage of rural ROR crashes were
reported for straight-following or non-collision
and overturn than urban ROR. When considering
contributory causes for ROR crashes,
avoidance/evasive maneuvers or being slow,
driver ill, falling asleep, or fatigued, or animal on
the road were more common for rural roadways
than urban roadways. These results will help to
develop appropriate countermeasures in
preventing ROR crashes.
REFERENCES
Benavente, M., H. A. Rothenberg, and M. A. Knodler Jr., 2006. Evaluation of Frequency and Injury
Outcomes of Lane Departure Crashes, in 2006 ITE Annual Meeting and Exhibit Compendium of
Technical Papers.
Dissanayake, S., 2003. Young Drivers and Runoff Road Crashes. Proceedings of the 2003
MidContinent Transportation Research Symposium, Iowa.
Fatality Accident Report System (FARS), 2012. National Highway Traffic Safety Administration, U.S.
Department of Transportation. http://wwwfars.nhtsa.dot.gov/Main/index.aspx. Accessed June 14,
2012.
Hallmark, S. L., Y.Y. Hsu, T. H. Maze, T. J. McDonald, and E. J. Fitzsimmons, 2009. Investigating
Factors Contributing to Large Truck Lane Departure Crashes Using the Federal Motor Carrier
Safety Administration’s Large Truck Crash Causation Study (LTCCS) Database. Center for
Transportation Research and Education, Iowa State University.
Kansas Department of Transportation, 2013. Kansas Motor Vehicle Accident Report Coding
Manual, http://www.ksdot.org/burtransplan/prodinfo/lawinfo/
2012_Motor_Vehicle_Coding_Manual_v1_web.pdf, Accessed March 3, 2013.
Liu, C., and R. Subramanian, 2009. Factors Related to Fatal Single-vehicle Run-off-road Crashes,
Report No. DOT-HS-811-232, U.S. Department of Transportation.
Mak, K. K., 2010. Identification of Vehicular Impact Conditions Associated with Serious Run-off
50

Journal of Society for Transportation and Traffic Studies (JSTS) Vol.5 No.4
road Crashes, Report No. 665, Transportation Research Board of the National Academies.
McGinnis, R. G., M. J. Davis, and E. A. Hathaway, 2001. Longitudinal Analysis of Fatal Run-off-
road Crashes, 1975 to 1997, in Transportation Research Record of the National Academies: Journal
of the Transportation Research Board, vol. 1746, no. 1, pp.47-58.
McLaughlin, S. B., J. M. Hankey, S. G. Klauer, and T. A. Dingus, 2009. Contributing Factors to
Runoff-road Crashes and Near-crashes, Report No. DOT-HS-811-079, U.S. Department of
Transportation.
Montella, A., and M. Pernetti, 2010. In-depth Investigation of Run-off-the-road Crashes on the
Motorway Naples-Candela, Proceedings of the 4th International symposium on highway geometric
design, pp. 2-5.
Spainhour, L. K., and A. Mishra, 2008. Analysis of Fatal Run-Off-The-Road Crashes Involving
Overcorrection. Transportation Research Record: Journal of the Transportation Research Board,
No. 2069, TRB, National Research Council, Washington, D.C., pp. 1-8.
Sullivan, J. M., and M. J. Flannagan, 2002. The Role of Ambient Light Level in Fatal Crashes:
Inferences from Daylight Saving Time Transitions," Accident Analysis and Prevention, vol 34, no. 4,
pp.487-498.
51