Making roads safer
The former Secretary General of the International Transport Forum, Jack Short argues that the extent of the traffic injury problem in Ireland requires increased policy attention.
The success of road safety policies is usually measured by the extent to which the number of fatalities fall. In Ireland, as in most developed countries, road traffic fatalities have reduced steadily, and at times dramatically, from the high of 640 in 1972 to 166 in 2015.
Danger on the roads cannot be summarised by fatality totals alone. There are also many material damage and injury collisions which get less policy attention but which have large costs for individuals and for society. Official data show an average of over 8,000 people were injured and an additional 20,000 material damage collisions occured from 2005 to 2011.
Attention in many countries is turning to the problem of traffic injuries. This is not because the fatality problem is resolved, far from it, but because of the serious nature of the injury problem. Crucially, many injuries are not counted in official figures. Moreover, the number of injuries has often not declined as fast as the number of fatalities, implying that the achievement has been more a reduction in the consequences of collisions than in their number.
In the decade to 2012, for example, fatalities per capita fell by 64 per cent in Ireland while injuries fell only by 26 per cent. A safe road system implies reducing the number and severity of all types of collisions.
Significance
A recurring problem is that data on traffic injuries is not reliable because of varying definitions and recording methods. As Table 1 and Figure 1 shows for a selection of countries the number of injuries can vary from 15 injuries per fatality to over 150. Ireland’s official data shows there are roughly 45 injuries per fatality.
The official injury data in Ireland is collected by An Garda Síochána and published by the Road Safety Authority (RSA). These are not the only sources and hospitals, A&E units, doctors, the Injuries Board and the insurance industry all record instances of traffic injuries.
Distinguishing injuries by severity is obviously important. Recently, it was agreed that serious injuries in the EU will henceforth be defined as those with a score of three or more on a scale known as the maximum abbreviated injury scale (MAIS). This is a clinical measure of the risk of death, which can be derived from the clinical coding carried out in hospitals.
Findings
The research focused on three sources where detailed data was available, Gardaí, hospitals and the Injuries Board. The data was first analysed separately and then together using a technique known as record linkage. This technique uses probabilistic methods to find matches between the different data sets on the basis of common variables like age, gender, date of incident, type of accident or county of residence.
Two sets of record linkages were undertaken. Firstly, Gardai and hospital data was linked for the period 2005 to 2011 and secondly, Gardai, hospital and Injuries Board data was linked for the years 2010-2011.
The linkage between Gardai and hospital data over the period 2005-2011 shows that less than 20 per cent of Gardai-reported injuries are matched with hospital patients and less than 30 per cent of hospital patients are matched with a Gardai-reported injury.
Matching rates for different population subgroups vary widely. Persons under 14 and over 65 are more likely to be matched than other age groups. Car drivers and passengers in hospital are more likely to be linked to a police-recorded collision than other road users. On the other hand, hospitals have far more cyclist injuries and more young people than in the Gardai data. The case of cyclists is striking. In 2011, almost 12 per cent of hospital bed nights accounted for by transport accident victims were taken by cyclists, compared to their less than 1 per cent share in traffic.
Using the clinical severity measure, the number of serious injuries in 2011 was about double the Gardai recorded total. Linkage involving the Injuries Board dataset shows lower matching rates than between hospitals and Gardai, indicating a large number of injuries in addition to those identified by police and hospital data.
Less than 10 per cent of Injuries Board cases are matched with a hospital case, illustrating that the Injuries Board dataset generally contains injuries that are less serious clinically, not involving a hospital stay. A higher share, though still only about a sixth of the Injuries Board cases, are matched with Gardai data.
Since the number of injuries is understated, so too is the cost to society. A recalculation of the official estimates of the social costs of injuries, as published by RSA, shows that the costs of injuries should be over €500 million more than the official estimate of €283 million for 2011.
Implications
The extent of the traffic injury problem shows that the subject needs to be given increased policy attention. Injuries need to be integrated into the road safety policy strategy.
There are still many unknowns and a key starting point for a fuller understanding of the injuries problem is improved data and better coordination of the different sources. Because injuries are complicated, new data series involving hospital and other sources are needed. Serious injuries should be defined using clinical criteria, and Gardai should not have to assess injury severity. The insurance industry has valuable data and could make a useful contribution here.
Since official data sources seriously understate the extent of cyclist and motorcyclist injuries, increased efforts are needed to ensure safer travel for vulnerable road users. Specifically, the national cycling strategy needs to do more to reduce risks to the growing number of cyclists and specific injuries should be the subject of particular policy attention.
In conclusion, a deeper understanding of the injury problem will be vital to safety policy in the future, and improved data, a higher policy profile and additional research can all contribute to this aim.
