ROAD SAFETY STATISTICS

Data requirements
• Reference:

• The Way Forward-Transportation Planning and

Road Safety
– Editors…
– Geetam Tiwari, Dinesh Mohan, Nicole Muhlrad
– (McMillan Book Company)
 Types and Sources of Data
OBJECTIVES
• To specify the data on injuries necessary to design interventions that prevent / minimise it.
• To locate and compile the relevant data for comparison and analysis.
 
WHY DATA ARE IMPORTANT?
• Reliable data on injuries are necessary to:
• assess the nature and extent of injuries in a population;
• identify groups that are most at-risk for specific injuries;
• establish priorities for intervention;
• allocate appropriate resources for injury control programmes;
• design countermeasures;
• evaluate the effectiveness of laws, technologic changes, environmental modifications and
educational campaigns in reducing injuries; and
• convince the public and policy makers of the importance of certain injuries and the need
for appropriate action.
 SOURCES AND LIMITATIONS OF INJURY
DATA
• Summary statistics for cross-country comparisons are
available from a variety of sources. Most of these
summaries, however, are based on statistics submitted by
government agencies within the individual countries. The
completeness and reliability of the data can vary widely.
• Data sources on injuries within countries are usually quite
numerous. The value and limitations of commonly - used
sources of injury information - death certificates, hospital
records, police data, and special surveillance systems - are
discussed below.
 Vital Statistics Data
Official registries of deaths are often the most accurate source of information on
serious injuries occurring in a population. However, in some countries, it is legal
to cremate or bury the dead without filing a report. Also, mortality data· may not
be comparable across countries because the persons completing the reports of
death may have extensive training (e.g. medical examiners) or none at all (e.g.
village leaders in remote communities).
Hospital Records
Medical records are vitally important in identifying causes of injury morbidity. For
example, house fires are the most common cause of bum-related deaths; scalds
are the most common cause of medically ­treated bums. Also, medical records
are usually the only source of detailed information about the nature, severity
and costs of all types of injuries in a population. Estimates of the cost of injuries
can be based on hospital records of the treatment received, length of stay and
the condition of the patient at the time of discharge from the hospital.
 Hospital Data-Serious Limitations
• Patients treated at an individual hospital are often not representative of the total
population of injury victims in a community. Geographic, economic, clinical and other
factors result in ‘selection bias’ when hospital-based, rather than population-based,
data are used to analyse injury cases. For example, a hospital with a neurosurgeon on
the staff will likely treat many more head-injured patients than hospitals without
neurosurgeons. Head injuries may therefore appear to be a much more common
problem than other injuries in the community. Similarly, private hospitals that cater to
upper-income patients are likely to admit more car-crash victims than occupationally
injured industrial workers. Government hospitals treating poor people in LICs will see
more people who are injured as pedestrians and bicyclists than as car occupants. A
hospital on the outskirts of a city will receive more patients with agricultural injuries
than hospitals in the city centre.
• Because it is often not possible to identify the catchment area of a particular hospital,
population-based incidence rates cannot be calculated.
• Medical records generally have little or no information on the circumstances of the
injury or the occupation of the victim.
 Surveillance
• By surveillance is meant an on-going, systematic
programme of data collection and analysis. The
purpose of surveillance is to provide information
about the incidence and severity of injuries in a
population; to identify new injury problems
early, so that interventions can be put in place;
to determine priorities for action, both in terms
of injury problems and high-risk groups; and to
help evaluate preventive measures.
 SURVEILLANCE PROGRAMMES
• Meaningful surveillance programmes - those that contain
accurate, comprehensive data - require enormous investments
of time, money and personnel to both establish and maintain.
Instead of a surveillance system, a time-limited (e.g. one-
month) survey of injuries can be conducted as a baseline and
then repeated after 6 months or I year. Repeated, cross-
sectional surveys can often provide all the information
necessary for setting priorities and targeting interventions. In
addition, special areas of interest can be examined more
easily. The population census, conducted once every 10 years
in many countries, is a repetitive cross-­sectional survey.
 Examples of Surveillance efforts
• Examples of surveillance programmes are
hospital registries that record medical data on
specific conditions. Among the national
surveillance systems established at great effort
and expense to specifically gather injury data
are NEISS (USA), PORS (the Netherlands) and
HASS (United Kingdom).
• The National Electronic Injury Surveillance
System (NEISS) is a probability sample of
hospital emergency departments in the US
and its territories.
• The Fatal Accident Reporting System (FARS)
was developed by the US National Highway
Traffic Safety Administration (NHTSA)
 NEISS
• The National Electronic Injury Surveillance System (NEISS) is a
probability sample of hospital emergency departments in the US
and its territories.
• It is used by the US Consumer Product Safety Commission (CPSC) to
provide national estimates of the number and severity of injuries
associated with consumer products and treated in hospital
emergency departments.
• The 5,939 US hospitals having emergency departments and / or
emergency visits were stratified by hospital size and ordered by
geographic location.
• Information from these and other sources (such as death
certificates and consumer complaints) guides the CPSC in setting
priorities for action or in-depth epidemiologic investigation.
 Need for Data Collection
• Some of the key reasons for collecting crash and
injury data are to:
1) overview the problem,
2) monitor trends,
3) identify high risk/problem groups,
4) identify high risk, hazardous locations,
5) enable objective planning and resource
management,
6) evaluate effectiveness and monitor achievement of
targets,
7) make international comparisons and
8) provide evidence for prosecution.
 Information Sources
• Potential sources of data include:
1) Police crash data
2) Hospital and medical data
3) Insurance data 
4) Vehicle operators 
5) Special surveys
 Questionnaires Broad Categories
• For the road safety professional the
database should at least be able to answer
the following questions:
– Where did crashes occur?
– When did crashes occur?
– Who was involved and who was injured?
– What was result of the crashes?
– What environmental conditions?
– How or why did the crashes occur?
 Road Accident Recording Form
Form No:

Filled by: Date filled:

Police report available: Yes/ No If yes, FIR No.

1. Time of accident: (24hr) 2. Date (DD) (MM) (YY)

3. Day: 4. Holiday: 5. Hit & Run: 6. Accident severity:

7. Number of fatalities: 8. No. injured: 9. Number vehicles:

10. Collision type: 11. Collision spot: 12. Type of road:

13. Divider: 14. Location: 15. City/Town/Village name:

16.RoadCategory 17. Distance Km m 18. From

Road Road
1 3

Road 2

19. Name Road 1: 20. Name Road 2:

21. Name Road 3: 22. Landmark:

15 08/27/2023 Department of Civil Engineering

Brief description of accident:
 Form No.

23. Type: 24. Manoeuvre: 25. Loading: 26.Disposition

27. Mechanical Failure 28. Impact- Vehicle/Object

Vehicle 1: 29. Make/Model

30. Type: 31. Manoeuvre: 32. Loading: 33.Disposition

34. Mechanical Failure 35. Impact- Vehicle/Object

Vehicle 2: 36. Make/Model

37. Type: 38. Manoeuvre: 39. Loading: 40.Disposition

41. Mechanical Failure 42. Impact- Vehicle/Object

43. Make/Model:
Vehicle 3:

Victim 1: 44. Type: 45. Occupant veh. 46.Road user: 47. Age:

48. Sex: 49. Injury: 50. Pedestrian/vehicle impact

Injury details 51. Mode of treatment 52. Days in Hospital

53. Injury 1 54. Injury 1 severity

55. Injury 2 56. Injury 2 severity
57. Injury 3 58. Injury 3 severity
59. Injury 4 60. Injury 4 severity
61. Injury 5 62. Injury 5 severity
63. Injury 6 64. Injury 6 severity

65. Most Severe Injury 66. ISS

16 08/27/2023 Department of Civil Engineering

 CODING INSTRUCTIONS FOR ROAD ACCIDENT STUDY

Accident on a straight road Accident at a

crossing
1. Form No: __ __ __ __ __ __ __ __ __
ROAD NO 2

Case Number in serial order

ROAD NO 1 ROAD NO 1
City / Village/ Location No.
Thana No.
ROAD NO 3
ROAD NO 2

For Items 19-21 use codes

as above. Road 3 is not necessary for accidents at crossings

Police report available: 0=No, 1=Yes

3. Day: Monday =1, Tuesday=2, Wednesday=3, Thursday=4,Friday=5, Saturday=6, Sunday=7, Unknown=9.

4. Holiday:0 = No, 1 = Yes, 9=Unknown

5. Hit & Run: 1 = Yes; 0 = No, 9=Unknown

6. Accident severity: 1=Damage Only(DO), 2= Injury, 3=Fatal, 9= Unknown

10. CollisionType: *1 = Hit pedestrian 2 = Vehicles head on 3 = Vehicle hit from back 4 = Vehicle hit from side 5 = Overtun
6 = Vehicle hit fixed object 7 = Run off the road 8= Others 9 = Unknown

11. Collision spot: 1 = On straight road 2 = Road junction 3 = Other 9 = Unknown

12. Type of Road: Type of Surface. 0= Unmetalled 1=Metalled (Black topped/Concrete), 2=Others, 9=Unknown

*In case of more than one type reported in a series of events, the first collision/event will be considered.

13. Divider: Whether divider was present on the road? 0=No, 1=Yes, 9= Unknown

14. Location: 1=Urban, 2=Rural, 3=Semi-Urban, 4=Other, 9=Unknown

17 08/27/2023 Department of Civil Engineering

 16. Road Category: ty/Rural road, 2= State Highway, 3= National Highway, 4= Other, 9= Unknown ( no other road categories
except NH and SH are given as they 1=Ci are the only ones who are appropriately marked on the milepost. For other roads their categories
may not be obvious during data collection at the field.
17. Distance: Km post. In the absence of Km post - from the nearest urban centre.

VEHICLE
23. Road User Type: 1 = Multi-Axle Heavy Goods vehicle 2 = 2-Axle Heavy Goods vehicle 3=Light Goods Vehicle
(V1, V2 etc.) 4 =Bus 5 = Car/van/jeep/taxi 6 = Three wheeler scooter rickshaw 7 = Motorcycle/scooter/moped
8 = Tractor 9 = Cycle rickshaw 10 = Thela 11 = Animal drawn vehicle 12=Bicycle
13= Pedestrian 14 = Other 99 = Unknown
24. Manoeuver 1 = Proceeding straight 2 = Turning 3 = Reversing 4 = Overtaking 5=Parked/Stopped 6 = Other
of vehicle crash time: 9 = Unknown
25. Loading: 1=Normal; 2= Overloaded; 3= Others; 9= Unknown
26. Disposition of vehicle: 0=Not Roadworthy (needs to be towed away); 1= Roadworthy ( can drive away ), 9 = Unknown
27. Mechanical Failure: 1=Yes; 0=No; 9= Unknown
28. Impact-Vehicle/Object: Vehicle number ( If another vehicle impacted this vehicle, enter the appropriate number of that vehicle
from section 23) 11=Pedestrian 12=Tree, 13=Kerb/Median; 14= Pole; 15= Other; 99=Unknown

VICTIM
44. Type: 1= Passenger, 2= Driver, 3= Pedestrian, 9= Unknown
45. Occupant Vehicle: Which of the above vehicles (vehicle1/2/3 etc.)? OR Pedestrian =0
46. Road User: Occupant code of Vehicle type/Pedestrian from code 23.
48. Sex : 1 = Male 2 = Female
49. Injury : 0 = No injury 1= Injured 2 = Fatal 9 = Unknown
50. Pedestrian/Vehicle Impact: Enter Vehicle number OR 11= Flying Object, 12= Fall, 13=Others, 99= Unknown

INJURY
51. Mode of Treatment: None 0 First Aid only 1 Discharged after casualty ward treatment 2 Admitted to the hospital 3
Other 8 Unknown 9
52. Number of Days in Hospital: Days, Unknown- 999
53. Injury: From AIS code
54. Injury Severity: -do-
65. Most Severe Injury: -do-
66. ISS: -do-

18 08/27/2023 Department of Civil Engineering

 AIMS AND OBJECTIVES OF TRAFFIC
SAFETY ANALYSIS
• To find solutions for such problems which are not readily available as
innovative work needs to be done to arrive at new policies and design
• To show that basic scientific understanding and methods of scientific
analysis can be shared internationally but priorities in safety policies may
need to be location specific and context sensitive
• To demonstrate that control of traffic deaths and injuries must follow the
same principle as control of any other health problem
• Providing solutions based on low cost treatments for different arterials
• Identification of safety measures for short term planning and long term
planning based on the nature and severity of crash
Accident Rates
 DIMENSIONS OF ROAD TRAFFIC INJURY
PROBLEM
01

15 DEATH
HOSPITALIZATION
MEDICAL TREATMENTS

50

As a rough rule of thumb, for every traffic death , there are of the order of 15 injuries severe
enough to require hospitalization and more than a further 50 injuries requiring medical
treatment but not hospitalization
TRANSPORT SAFETY, PERSONAL SAFETY
AND MOTORIZATION

HIGH
TRANSPORT SAFETY

RISK OF PERSONAL SAFETY

DEATH

LOW

LOW MOTORIZATION HIGH

PROPORTION OF GLOBAL ROAD TRAFFIC
DEATHS AND STAGE OF MOTORIZATION
GLOBAL ROAD TRAFFIC DEATHS %

Motorized Countries 14

Motorizing Countries 86

Asia/ Pacific 44

Latin America/ Caribbean 13

Central & Eastern Europe 12

Africa 11

Middle East 06
 PROJECTED PROPORTIONATE INCRAESES (DECREASES) IN
THE ABSOLUTE NUMBER OF ROAD TRAFFIC DEATHS IN
SELECTED REGIONS (2020 AGAINST THE BASE YEAR 2000)

INCREASE IN 2020 VS 2000 (%)

South Asia 145
East Asia and Pacific 80
Sub- Saharan Africa 80
Middle East and North Africa 70
Latin America/ Caribbean 50
Central & Eastern 20
Motorized countries (30)
PERCENTAGE OF ROAD USERS KILLED IN VARIOUS MODES OF
TRANSPORT AS POPULATION OF ALL FATALITIES
COUNTRIES PEDESTRIANS RIDERS FOUR WHEELERS OTHERS
AUSTRALIA 18% 10% 65% 7%
JAPAN 28% 21% 41% 10%
MALYSIA 15% 57% 18% 10%
NETHERLANDS 10% 12% 56% 22%
NORWAY 15% 12% 65% 8%
THAILAND 10% 73% 13% 4%
USA 12% 5% 80% 3%
DEHLI, INDIA 42% 28% 14% 16%
BANDUNG, INDONESIA 33% 41% 17% 9%
COLUMBO, SIRILANKA 39% 35% 14% 12%
KARACHI, PAKISTAN 48% 47% 4% 1%

By comparing the ratios in between developed and developing countries, it is quite

evident that large number of four wheelers are involved in traffic crashes of
developed countries while scenario is completely in developing countries (high
percentage of riders and Pedestrians)
 Essential Information

• NAME

• ADDRESS

• OCCUPATION

• TIME OF INCIDENT

• TIME OF ARRIVAL

• ARRIVED
 VEHCILE INVOLVEMENT

• Motorbike 1
• Minivan/ coaster 2
• Bus/ mini bus/ coach 3
• Truck 4
• Taxi 5
• Bicycle 6
INTERACTION BETWEEM VEHICLES
• Car 7
• Water/ oil tanker 8
• Rickshaw 9
• Dumper 10
• Trailer 11
• Loading Pick up 12
• Others 13
 ROAD USERS (VICTIM)

• Rider 1

• Pillion 2

• Driver (4 wheeler) 3

• Passenger 4

• Pedestrian 5

• Driver (3 wheeler) 6

• others 7

• Driver (more than 4) 8

 TYPE OF COLLISION

• Head on
• Rear End
• Hit Object
• Merging
• Side Swipe
• Right Angle
• Unknown

NEED OF ACCIDENT INVESTIGATION

 LOCATION DETAILS

• Intersection
• Mid block
• U turn
• Fly over
• Bridge
• Underpass
• Unknown
 DISPOSAL

• Discharged 1

• Admitted to ward 2
Minor = 1
• Detained for 3
Serious = 2,3,5,6,7
• Expired 4 Fatal = 4
• Transferred to 5

• Referred from 6

• LAMA 7
 Data Analysis

• Data are explored with focus on following questions

-What time of day the accidents result in most injuries?
-Are road alignment and profile factors in accident injuries?
-To what extent is speed limit a factor in injuries?
-Do seasonal factors –spring/summer/winter explain injuries?
-Whether manner of collision is a factor in the injury?
-Whether region affects the probability of injury?
-Do number of vehicles involved explain the injury?
-Do accidents on interstate highways result in more injuries
 Methodology

• Started with data organization, exploratory

analysis and descriptive statistics
• Selected variables for modeling the
relationships.
• Regression to prediction and Chi-square tests
group membership
• Evaluation of model results
• Recommendations from the analysis
Exploratory Analysis
Exploratory Analysis…
TREND LINE-Multiple Linear Regression
Equation
 Chi-square test
• Chi-square analysis methods are approximate
methods that are among the most commonly
used of all statistical techniques.
• The method introduced will be used to
examine “frequency” or “count” data.
• Does a pattern of frequencies of a single
categorical variable significantly differ from an
expected pattern
 Procedure for Chi-square test
e.g. that the number of accidents
1. Formulate the statistical is independent of the age of the
hypothesis driver.
2. Select the significance α = 0.1, 0.05 or 0.01
level
 ( Fo  Fe ) 2 
3. Calculate the test statistic   
2

 Fe 
4. Determine the degree of Fe = F r Fc / N
freedom df = (R-1)(C-1)
5. Compare computed test
If calculated 2 is greater than 2 table
statistic against a value, reject Ho
tabled/critical value