Accident Analysis and Prevention 43 (2011) 804–812 Contents lists available at ScienceDirect Accident Analysis and Prevention journal homepage: www.elsevier.com/locate/aap Characteristics of single vehicle rollover fatalities in three Australian states (2000–2007) B. Fréchède a,b , A.S. McIntosh a,∗ , R. Grzebieta c , M.R. Bambach c a School of Risk and Safety Sciences, The University of New South Wales, UNSW, Sydney, NSW 2052, Australia b Université de Lyon, F-69622, Lyon, France; INRETS, UMR T9406, LBMC, F-69625, Bron; Université Lyon 1, Villeurbanne. LBMC, 43 Bd du 11 Novembre 1918, 69622 Villeurbanne, France c NSW Injury Risk Management Research Centre, The University of New South Wales, UNSW, Sydney NSW 2052, Australia article info Article history: Received 21 May 2010 Received in revised form 28 September 2010 Accepted 24 October 2010 Keywords: Rollover crashes Trauma Motor vehicle safety Occupant injury abstract An analysis of 2000–2007 single vehicle rollover fatalities in three Australian states was carried out using data from the Australian National Coroners Information System. In this paper, successive selection criteria were applied to the initial dataset to analyse: - Occupant fatalities in single passenger car crashes (1743 cases), - Occupant fatalities in single passenger car rollovers (474 cases), overall, rollovers accounted for 35% of all occupant fatalities in a single vehicle transport injury event. For these fatalities, the occupant was ejected or stayed contained in equal proportions. However, results showed strong disparities between the more urban and densely populated states of New South Wales and Victoria, compared to the Northern Territory in terms of crash type distribution and containment of the occupant. Differences were also found in rollover initiation, speed at initiation and number of turns. Overall, the strongest association of fatal neck/thoracic spine injuries with head injuries was found for the contained, restrained occupant. This analysis of single vehicle rollover fatalities is consistent with previous findings. It also shows that in Australia, strategies for rollover injury risk mitigation will need to take into account a broad range of characteristics to be effective. © 2010 Elsevier Ltd. All rights reserved. 1. Introduction 1.1. Context According to recent Australian data, in one in every three to four vehicle fatalities the vehicle was found to have rolled over (Grzebieta et al., 2008). In the USA this proportion is slightly higher, with one in three fatalities having been reported to occur in a vehi- cle crash involving a rollover (NHTSA, 2008). In comparison, there appears to be fewer occupants killed proportionally in European crashes involving a rollover: one in five to seven fatalities in the UK (Gugler et al., 2004; Cuerden et al., 2009). Despite concerns about the high risk of severe and fatal injury from rollover crashes, the patterns of injury and risk factors have not been adequately described in Australia. The Australian roads are different to US and mainland European roads in terms of urbani- sation, sealed and unsealed shoulders, roadside hazards, distances ∗ Corresponding author. Tel.: +61 2 9385 5348; fax: +61 2 9385 6190. E-mail address: a.mcintosh@unsw.edu.au (A.S. McIntosh). travelled, etc. For example, Australians drive on the left hand side of the road. Road travel is longer in Australia compared to Europe but is similar to the US and hence fatigue may play a greater role in Australia than in Europe. There are also many more unsealed road shoulders in rural areas compared to Europe or the US. Equip- ment, such as roadside barriers, also differs between these regions, which may result in different contributions to either the initiation or roll phases in a rollover. Understanding these factors is a prereq- uisite for a Safe Systems framework (ATC, 2008; OECD/ITF, 2008) and provides a perspective on Australian crashes for researchers comparing different driving environments. In Australia, studies by Treacy et al. (2002) and McLean et al. (2005) presented rollover characteristics in two Australian states: the Northern Territory (NT) and South Australia. These studies revealed possible differences in injury patterns between these two states, but this comparison was confounded by differences in study aims and methods. None of these authors compared their analyses to US or European results. Population density, road environments, and speed limits differ within and between Australian States. For example, in the authors’ State of New South Wales (NSW), the population density in 2006 was reported to be 8.5 persons per 0001-4575/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.aap.2010.10.028