Data Systems Study (recommendations for the establishment of improved and more timely accident data systems)

Monash University Accident Research Centre - Report #10 - 1989

Authors: A. Drummond & P. Vulcan

Full report in .pdf format [1.9MB]

Abstract:

There is a variety of information collected by a range of organisations, primarily for their own purposes, which is relevant to roads, traffic and safety in Victoria. This report examines the potential for improving the timeliness, quality and level of integration of data available to administrators, practitioners and researchers.

A set of general principles which an improved data system should comply with were formulated, including the principle that accidents reported by police who have attended the scene represent the best overall source of general information on road accidents. Information on casualty accidents should form the core of the system and thus be reliable and comprehensive. Information on property damage only (PDO) accidents may not need to be as detailed, may not need to be as timely and could be limited to a proportion of total PDO accidents

A range of suggestions to improve the timeliness and quality of accident data and the level of integration of data systems. The report concludes by identifying the aim and work required within the four modules which comprise the improved data system.

Executive Summary

There is a variety of information collected by a range of organisations, primarily for their own purposes, which is relevant to roads, traffic and safety in Victoria. Thus, there are a number of different samples of information on the performance of road users which are available from different databases. This report examines the potential for improving the timeliness, quality and level of integration of data available to administrators, practitioners and researchers.

A set of general principles which an improved data system should comply with were formulated;

• the system must recognise the specific needs of the various organisations contributing to it, but should devise ways for some organisations to utilise information collected by others
• accidents reported by police who have attended the scene represent the best overall source of general information on road accidents
• an integrated data system may have to satisfy competing objectives
• information should be of sufficient detail to service the different applications to which the data will be put.

Information on casualty accidents should form the core of the system and thus be reliable and comprehensive. Information on property damage only (PDO) accidents may not need to be as detailed, may not need to be as timely and could be limited to a proportion of total PDO accidents.

Where timeliness of accident data is concerned, the following suggestions have been made;

• strategies to increase the reporting rate in the Police system have been described (page 6)
• the replacement of paper forms through the equipping of police with the technological means for direct, on-site data entry was suggested (page 6)
• potential improvements to the report form were outlined and it was recommended that the form be reviewed to improve timeliness without compromising utility (pages 6-8)
• given the relatively high proportion of accidents reported by drivers attending a Police station, a simplified form, to be filled in by the driver and possibly designed for machine reading, was recommended to reduce Police workload (page 8)
• as police will remain in the front line of reporting, strategies for promoting the importance of this function were suggested (page 9)
• access to the Police tape be made available, given sufficient quality, to allow basic road safety questions to be answered more rapidly (page 9)

• DCA coding should be simplified for property damage accidents to expedite data enhancement and this system be made available to insurance companies (pages 10- 11)
• processing priority should always be given to casualty accidents and non-road accidents should not attract processing resources (page 10)

Where the integration of data systems is concerned, the following suggestions have been made;

• integration through matching/merging is facilitated by a similar record structure across systems, a set of matching variables and a set of agreed matching rules (these are being developed in a separate project) and obtaining common variables from the same source (pages 12-13)
• a pilot project be undertaken to determine the precise procedures, costs and usefulness of property damage accident data collected from insurance companies, taking into account the holding organisation's commercial and legal obligations (pages 13-14)

• reconciliation of Vocational systems should proceed at the data level, given available resources, prior to a move to a common, digitised mapping system in the future (page 15)
• methods to improve the efficiency and timeliness of obtaining blood alcohol concentration data for fatal and injured drivers should be developed (page 16)

Where quality control is concerned, the following suggestions have been made;

• strategies to increase police attendance at casualty accidents should be investigated (page 17-18)
• formal feedback loops should be established to enable changes made to records are communicated to the other elements of the system and ensure consistent datafiles (page 19)

The report concluded by identifying the aim and work required within the four modules which comprise the improved data system, namely;

• the unproved timeliness of data and the earlier provision of access
• the matching/merging of enhanced Police accident and TAC data
• the generation of an independent, expanded property damage only file to supplement the matched RCITAC casualty accident file
• the linking of accident and non-accident RC databases by locational information.

Sponsors: Roads Corporation & Transport Accident Commission