Slide1: Work Package 1 CARE Accident Data George Yannis and Petros Evgenikos National Technical University of Athens Prague, May 11th 2006 Project co-financed by the European Commission, Directorate-General Transport & Energy


Road accident data – The questions: Road accident data – The questions Do we have the data we need? Do we need the data we have?


Slide3: Road Accident Data Usefulness and availability of Data


Slide4: CARE Database Community database on Accidents on the Roads in Europe All accidents & casualties recorded in EU-14 since 1991, in disaggregate form. (≈ 11m road accidents, ≈ 14,5m road accident injuries, ≈ 410k road accident fatalities). Objectives Provide a powerful tool which would make it possible to: i) Identify and quantify road safety problems in Europe ii) Evaluate the efficiency of road safety measures iii) Determine the relevance of Community actions iv) Facilitate the exchange of experience in this field


Slide6: WP1 - CARE Accident Data General Objective Enhancement and Exploitation of CARE accident data Making CARE system a reference point for road safety analysis in Europe Duration 48 months


Road accident data – The questions: Road accident data – The questions Do we have the data we need? Do we need the data we have? WP1 Objectives and Structure


Slide8: Partnership • National Technical University of Athens (NTUA) – EL • Transport Research Laboratory (TRL) – UK • Centre d’ Etudes Technique de l’ Equipement du Sud-Ouest (CETE) - FR • Kuratorium für Schutz und Sicherheit (KuSS) – AT • Institute for Road Safety Research (SWOV) – NL • Road Directorate - Ministry of Transport (DRD) – DK • Közlekedéstudományi Intézet Rt (KTI) - HU • Centrum dopravního výzkumu (CDV) – CZ • Institut Belge pour la Sécurité Routière (IBSR) – BE • Agència de Salut Publica (ASPB) – ES • Vehicle Safety Research Centre (VSRC) - UK


Slide9: Task 1.1 Compatibility improvement of data from new Member States Objectives • Extension of the CARE system by incorporating national level road accident data from the10 new EU Member States (plus Switzerland, Norway) • Development of appropriate transformation rules based on CAREPLUS methodology • Establishment of a broad and compatible road accident database


Slide10: Methodology Establishment of the list of national accident data in the CARE system Collection of national definitions of accident data Establishment of the tabulated (grid) list of accident data definition elements Elaboration of the transformation rules Update of the list of common accident variables and values Update of the Glossary with accident data definitions


Slide11: CARE Database -List of variables


Slide12: The Estonian example Transformation of Estonian accident data - Transformation rules drawn up jointly with the Estonian experts. - Collection of accident variables and values definitions according to the templates proposed. - Final document and transformed data files sent to the European Commission.


Slide13: The Estonian example


Slide14: Estonian transformation rules


Slide15: Task 1.4 Improvement of Accident Data Compatibility in Europe Objective • Establishment of a common road accident data collection system. • Enhancement and further exploitation of the CARE system. • Harmonisation of road accident data at EU level.


Slide16: Necessity for Compatible Statistics Example of CARE data availability Variable: Alcohol level


Slide17: Methodology Collection of national data collection forms, methodologies and data definitions Definition of necessary data for accident analysis Establishment of a basic common accident data collection set and methodology Update of the Glossary with accident data definitions


Slide18: Accident Data Availability (1) Data included in the National Road Accident Databases Road accident database and collection form first introduction (year)


Slide19: Accident Data Availability (2) Definitions of seriously and slightly injured persons


Slide20: Accident Data Availability (3)


Slide21: User Requirements (1) Number of road accident data stakeholders per road accident variable and country * Total number of national road accident data users Scale of significance: H = high importance L = low importance - Not used


Slide22: User Requirements (2) Scale of significance: H = high importance L = low importance - Not used Most important road accident data stakeholders according to the extent and frequency of use of road accident variables


Slide23: Next Steps • Analysis of a specific group of variables included in the national collection forms by each partner, according to the following distribution: • Development of the common framework for road accident data collection. • Examination of the proposed framework by the Member States, joint discussion under the coordination of the European Commission and finalisation of the proposal.


Slide24: Task 1.5 Estimation of the Real Number of Road Accident Casualties Objective • Identification of the underreporting level for each casualty severity (killed, seriously injured, slightly injured) with a uniform methodology. • Identification of the number of casualties according to a new common measurement unit: "hospitalised person".


Slide25: Methodology


Slide26: Literature Review • Descriptions of experience on the underreporting issue in the eight countries through technical literature review: a) Linking accident and medical databases b) ‘Capture-recapture’ techniques. • Execution of studies in eight EU countries (CZ, EL, ES, FR, HU, NL, AT, UK) to apply common methodology to accident data from police and hospitals. Distribution of non-fatal casualties according to the road trauma registry and police files in the Rhône county, 1997–2001


Slide27: Identification of underreporting level Developed methodology to identify the underreporting level: A representative sample of hospitals that receive accident victims will be selected ↓ Medical data regarding road accidents will be collected for a certain period of time ↓ The medical data collected will enter the project database ↓ The medical data will be cross-checked regularly with the police accident records ↓ Definition of underreporting level


Slide28: Critical issues • Approvals from medical authorities. Number of hospitals and length of data collection period upon funding available. • Regular visits to hospitals. Sifting records to identify people injured in road accidents. Examine variation according to type of accident. • Identification based on: casualty’s age, gender, travel mode and accident circumstances (time, location). Allowance for minor errors in the recording of personal details (i.e. small discrepancies in age). • Cross-checking according to “catchment” area of each hospital. • Admission of the casualty to the hospital, length of treatment in medical data and AIS1 score for each body region. 1Abbreviated Injury Scale, ranging from 1 for minor injuries to 6 for injuries that are currently untreatable


Slide29: Matrices Development of appropriate matrices The first 3-dimensional matrix incorporates the length of stay information: The second 3-dimensional matrix incorporates MAIS (the maximum of the AIS scores per body region):


Slide30: Next steps • Execution of the national studies on underreporting in eight countries. • Analysis of data to develop underreporting coefficients relating to alternative bases: hospitalised casualties (or perhaps with a minimum length of stay) and MAIS. • Investigation of strengths and weakness of alternative approaches. • Elaboration of a common definition to be used when estimating the actual number of accidents and casualties: "hospitalised person". • Adoption and finalisation of the common methodology/coefficients/common definition for future use.


Slide31: Task 1.2 Links With In-depth, Exposure, SPI and Other Data Files Objective • Amalgamation of the CARE accident data with other data (i.e. exposure) in road accident analysis. • Establishment of reliable and comparable accident indicators (i.e. fatalities per million veh-kms). • Better description of the road accident phenomenon.


Slide32: Methodology Definition of necessary accident indicators Transformation of accident data Update of the Glossary with accident data definitions


Slide33: Risk/Exposure Data Road safety performance in EU Member States: Car occupant fatality rate per billion vehicle-kilometers in 2002 (Source: ETSC 2006)


Slide34: Road Safety Performance Indicators • Complementary to existing road accident data to measure changes in the operational conditions of traffic system. • Better understanding of road accidents and underlying processes. • Enable policy interventions monitoring and facilitate decision making.


Slide35: Task 1.3 Development of Statistical Reports and Analysis Notes Objective • Production of a comprehensive set of end products with compatible statistics. • Enhancement and further exploitation of the CARE system.


Slide36: Methodology


Slide37: Databases Evaluation Evaluation of several existing databases: (FARS, CARE, UN/ECE, WHO, EUROSTAT, IRTAD, ECMT).


Slide38: Annual Statistical Report 2005 - Basic characteristics of road accidents in the 14 member states of the European Union (except Germany) for the period 1994-2003. - Selection of basic characteristics of fatal road accidents in the EU member. - 55 Tables and 20 figures with the most interesting combination of road accident data DOWNLOAD IT NOW http://www.europa.eu.int/comm/transport/care/studies


Slide39: Basic Fact Sheets - Motorways DOWNLOAD IT NOW http://www.europa.eu.int/comm/transport/ care/studies


Slide40: Basic Fact Sheets 2005


Slide41: Basic Fact Sheets Car occupants Motorcycle and mopeds Motorways


Slide42: Basic Fact Sheets Elderly Pedestrians Children Young people aged (16-24)


Slide43: Next Steps • New version of an Annual Statistical Report and eight Basic Fact Sheets (8) will be prepared by 2007. • Preparation of a CARE user handbook "Guide of CARE analysis". • Preparation of a set of Aggregate Data Files. • Outputs will progressively be enhanced by other SafetyNet results. Data for all 25 countries will be included, as soon as the related work is completed. • All these documents are available at the website of the European Road Safety Observatory (www.erso.eu).


Slide44: WP 1 Care Accident Data