This is the author’s version of a work that was submitted/accepted for pub-lication in the following source:

Kamruzzaman, Md., Haque, Md. Mazharul, & Washington, Simon (2014)Analysis of traffic injury severity in Dhaka, Bangladesh. In Transporta-tion Research Board 93rd Annual Meeting Compendium of Papers, Trans-portation Research Board (TRB), Washington, D.C.

This file was downloaded from: http://eprints.qut.edu.au/65897/

c© Copyright 2014 Please consult the authors

Notice: Changes introduced as a result of publishing processes such ascopy-editing and formatting may not be reflected in this document. For adefinitive version of this work, please refer to the published source:

Analysis of traffic injury severity in Dhaka, Bangladesh

Dr Md. KamruzzamanDr Md. Mazharul HaqueProf Simon Washington

Introduction

• Road crashes are disproportionately distributed between developed and developing countries .

About two-thirds of global injuries take place in developing countries.

Similarly, developing countries account for about 85% of the deaths and 90% of annual disability adjusted life years lost.

• However, relatively little knowledge exists on the factors contributing to traffic injury severity in developing countries

• In contrast, ample research conducted in developed countries.

• Planners and policy makers in developing countries often rely on findings from developed countries to formulate safety measures.

Research objectives• Two objectives:

first, to identify factors influencing traffic injury severity in a mega city of a developing country, and

Second, to validate the findings against findings generated from both developed and developing countries.

Research context: Dhaka, Bangladesh

The death was a shockwave in the nation

Drivers do not need to be educated –Shipping Minister

“if a driver can sign his name, can differentiate between a cow and a goat and can understand traffic signals should be eligible for a licence” – Shipping Minister

Protest and petition

The effects of the protest

Pedestrian and driver behaviour

City features

Crash characteristics

Dhaka, the capital of Bangladesh:

14 million people

Non-motorised modes dominate (e.g. walk 20%, rickshaw 40%, bus 30%, car 5%) (Ministry of Environment and Forest and Ministry of Communication, 2010).

highest fatality rate (100 deaths/10,000 motor vehicles) (UNESCAP, 2007)

India (25.3), Sri Lanka (16), Malaysia (5.5), USA (2.1), and UK (1.4) (Ahsan, 2012)

fatalities increased 3.5 times to 3000 deaths/yr

vehicle 2-10/1000 person (India 12, Sri Lanka 25, UK 426, USA 765)

Data

• 5 year crash data ‘07-’11

• Source: Dhaka Metropolitan Police

• 2714 collisions

• First investigation report based on accident reporting form (ARF)

• 13 variables

Selected variablesVariable Categories %Injury severity Property Damage Only 12.27

Injury 18.98Fatality 68.75

Collision type Head-on 4.20Side-swipe 4.86Collision with Pedestrian 59.84Rear-end and others* 31.10

No. of vehicles involved

Single vehicle 63.74Multi-vehicle* 36.26

Location type Non-intersection* 70.74Four-legged intersection 11.79Three-legged intersection 15.7Others (e.g. roundabout, rail-crossing)

1.77

Road geometry Straight and Plain* 96.9Others (e.g. curve, slope) 3.10

Pavement surface texture

Good* 98.89Rough 1.11

Variable Categories %Presence of road divider

Yes* 80.1No 19.9

Roadway classification

Highway 32.98Arterial and feeder road* 67.02

Traffic configuration Two-way 27.19One-way* 72.81Only police 32.54

Traffic control types Only traffic signal 1.62Both police and traffic signal

1.40

Uncontrolled* 64.44Weather condition Good* 99.15

Bad (e.g. rainy, foggy) 0.85Pavement surface condition

Dry* 99.23Skidding (e.g. wet, muddy) 0.77

Time of the day Day* 54.24Night 45.76

N 2,714

Method

• Outcome variable ‘injury severity’ was measured as an ordered category with 3 levels

• Categorical explanatory variables

• An ordered Probit regression model was estimated

• Changes in predicted probabilities have also been computed to better understand the marginal impacts of the explanatory variables on injury severities.

Results: Ordered Probit model estimates Variable Estimate SE z-stat p-value 90% CI

Roadway classification (ref: arterial and feeder road)Highway 0.367 0.061 6.04 <0.001 0.267 0.468

Presence of road divider (ref: yes)No 0.439 0.105 4.19 <0.001 0.267 0.612

Traffic configuration (ref: one-way)Two-way -0.151 0.091 -1.67 0.095 -0.301 -0.002

Traffic control types (ref: uncontrolled)Only police -0.319 0.059 -5.40 <0.001 -0.416 -0.222Only traffic signal -0.070 0.196 -0.36 0.722 -0.392 0.252Both police and traffic signal -0.165 0.212 -0.78 0.437 -0.513 0.184

Time of the day (ref: day)Night 0.311 0.055 5.64 <0.001 0.220 0.402

Collision type (ref: rear-end and others)Head-on 0.164 0.117 1.39 0.164 -0.030 0.357Side-impact -0.405 0.108 -3.76 <0.001 -0.582 -0.228Collision with Pedestrian 1.694 0.061 27.89 <0.001 1.594 1.794

Threshold 1 -0.022 0.126Threshold 2 0.981 0.128Pseudo-R2 0.260

Results: estimates of injury severity probabilities

VariableEstimated probability % Change relative to reference case

Property damage Injury Fatality Property damage Injury FatalityReference case 0.491 0.346 0.163Roadway classification

Highway 0.348 0.382 0.270 -29.06 10.50 65.26Presence of road divider

No 0.322 0.384 0.294 -34.38 11.05 80.09Traffic configuration

Two-way 0.551 0.320 0.129 12.28 -7.43 -21.14Traffic control types

Only police 0.617 0.287 0.097 25.58 -17.10 -40.69Only traffic signal 0.519 0.335 0.147 5.68 -3.24 -10.17Both police and traffic signal 0.557 0.318 0.126 13.34 -8.16 -22.86

Time of the dayNight 0.369 0.379 0.252 -24.79 9.72 54.11

Collision typeHead-on 0.426 0.367 0.207 -13.20 6.16 26.72Side-impact 0.649 0.268 0.083 32.20 -22.45 -49.26Collision with Pedestrian 0.043 0.195 0.762 -91.24 -43.62 366.97

Discussion and conclusionFactors Severity impact Citation

This research Developed countries

Collision with pedestrian Increased Increased Miles-Doan (1996)

Traffic control Mixed Mixed Pitt et al. (1990);

Lee and Abdel-Aty (2005)

Highways Increased Increased Miles-Doan (1996);

Sze and Wong (2007)

Darker period of time Increased Increased Kim et al. (2007);

Klop and Khattak (1999)

Presence of road divider Decreased Decreased Quddus et al. (2002)

One way Increased Decreased Sze and Wong (2007)

Discussion: traffic control schemes

• Mixed in the literature

• Mixed in this research

• Police controlled schemes are the most significant, followed by police and signal, and signal only.

• Driver has a different selection procedure which deserves further investigation

Discussion and conclusionFactors Severity impact Citation

This research Developed countries

Collision with pedestrian Increased Increased Miles-Doan (1996)

Traffic control Mixed Mixed Pitt et al. (1990);

Lee and Abdel-Aty (2005)

Highways Increased Increased Miles-Doan (1996);

Sze and Wong (2007)

Darker period of time Increased Increased Kim et al. (2007);

Klop and Khattak (1999)

Presence of road divider Decreased Decreased Quddus et al. (2002)

One way Increased Decreased Sze and Wong (2007)

Discussion: impacts of one-way road

• Contrasting findings

• 60% crashes involve a pedestrian and a single vehicle

• Two possible explanations:

Presence of barriers in the median prohibit unlawful crossings

Unaware of oncoming vehicles and unable to take evasive action

• Engineering solutions (e.g. railing on footpath) would be helpful

ReferencesAhsan, H.M., 2012. Road Safety in Bangladesh: Key Issues and Countermeasures, Forum. The Daily Star, Dhaka.

Boufous, S., Finch, C., Hayen, A., Williamson, A., 2008. The impact of environmental, vehicle and driver characteristics on injury severity in older drivers hospitalized as a result of a traffic crash. Journal of Safety Research 39, 65-72.

Fatmi, Z., Hadden, W., Razzak, J., Qureshi, H., Hyder, A., Pappas, G., 2007. Incidence, patterns and severity of reported unintentional injuries in Pakistan for persons five years and older: results of the National Health Survey of Pakistan 1990-94. BMC Public Health 7, 152.

Gray, R.C., Quddus, M.A., Evans, A., 2008. Injury severity analysis of accidents involving young male drivers in Great Britain. Journal of Safety Research 39, 483-495.

Kim, J.K., Kim, S., Ulfarsson, G.F., Porrello, L.A., 2007. Bicyclist injury severities in bicycle–motor vehicle accidents. Accid Anal Prev 39, 238–251.

Klop, J.R., Khattak, A.J., 1999. Factors influencing bicycle crash severity on two-lane, undivided roadways in North Carolina. Transportation Research Record: Journal of the Transportation Research Board 1674, 78–85.

Lee, C., Abdel-Aty, M., 2005. Comprehensive analysis of vehicle–pedestrian crashes at intersections in Florida. Accid Anal Prev 37, 775–786.

Miles-Doan, R., 1996. Alcohol use among pedestrians and the odds of surviving an injury: evidence from Florida Law Enforcement Data. Accid Anal Prev 28, 23–31.

Ministry of Environment and Forests, Ministry of Communication, 2010. Developments in EST in Bangladesh, Fifth Regional Environmentally Sustainable Transport (EST) Forum in Asia, Bankok.

Pitt, R., Guyer, B., Hsieh, C., Malek, M., 1990. The severity of pedestrian injuries in children: an analysis of the Pedestrian Injury Causation Study. Accid Anal Prev 22, 549–559.

Quddus, M.A., Noland, R.B., Chin, H.C., 2002. An analysis of motorcycle injury and vehicle damage severity using ordered probit models. Journal of Safety Research 33 445– 462.

Sze, N.N., Wong, S.C., 2007. Diagnostic analysis of the logistic model for pedestrian injury severity in traffic crashes. Accid Anal Prev 39, 1267–1278.

UNESCAP, 2007. THE STATUS PAPER ON ROAD SAFETY PROBLEMS IN BANGLADESH: The Bangladesh Country Paper. United Nations Economic and Social Commission for Asia and the Pacific.

World Health Organization, 2004. World report on road traffic injury prevention, Geneva.

Thank You

Question?