A step towards zero: understanding preventable residential fire fatalities

More than one residential fire-related death occurs every week in Australia – and these deaths are preventable. New research shines a light on the people most at risk.

By Lucinda Coates, Andrew Gissing and Dr Katharine Haynes from Risk Frontiers/Macquarie University and Bushfire and Natural Hazards CRC, Geoff Kaandorp and Julie Harris from the Metropolitan Fire and Emergency Services Board, Jonathan van Leeuwen, Ashley Avci, Jacob Evans, Steven George and Robin van den Honert from Risk Frontiers/Macquarie University. This article was first published in Issue One 2020 of Fire Australia. 

On average, more than one preventable fire-related death occurs in a residential context every week in Australia. That equates to approximately the same number of deaths every three years as occurred during the Black Saturday bushfires (173). At least 900 people have died from July 2003 to June 2017 – deaths that, overwhelmingly, could have been avoided.

Preventable fires are fires where individuals, fire services or other stakeholders may have been able to identify the risks related to a person and/or a physical environment and take actions or develop intervention strategies which, if applied, may have reduced the risk of a fire taking place. Deaths from residential fires have significant social, economic and emotional impacts on individuals, families and communities, and on the firefighters and other emergency service volunteers and employees who attend these incidents.

The Bushfire and Natural Hazards CRC study, undertaken by Risk Frontiers, the Metropolitan Fire and Emergency Services Board (MFB) and Macquarie University for AFAC, draws on 14 years of data to provide an update on the evidence around the extent of preventable residential fire fatalities in Australia and identify those people most at risk of dying in residential fires.

Undertaken through the CRC’s Tactical Research Fund, the Strategic analysis of preventable residential fire fatalities project analysed records from the National Coronial Information System (NCIS) database, supported by the analysis of publicly available coronial reports. Single variable, contingency table and machine-learning analysis was used to determine those most at risk of dying in a preventable residential fire.

The study revealed that those most at risk include:

• older people – people aged over 65 represented 36% of total fatalities
• young children aged zero to four – 8% of all fatalities
• people with a disability – 62%
• Aboriginal and Torres Strait Islander people – over-represented by a factor of 2.5
• people who smoke – 65% were smokers
• people with medications (34%) or alcohol (33%) present in their blood
• males – represented 64% of all fatalities, particularly those aged over 45
• people who lived alone – 45%
• people who lived in the most socially and financially disadvantaged locations.

The results from this study highlight that reducing residential fire fatality risk is complex. The presence of a single risk factor on its own is unlikely to significantly increase a person’s risk of dying in a residential fire. Rather, it is the combination of a range of risk factors surrounding the person – their behaviours, their residential environment, any disability or disadvantage that they are experiencing and other external factors – that is likely to impact their overall level of risk of having a fire that results in their death.

The research shows no declining trend

The research builds on a 2005 AFAC study, which found that the most at-risk groups for residential fire fatalities in Australia included males, those aged 65 or over, children under four and adults who had consumed alcohol. One of the project aims was to update the AFAC study and to confirm, at a national level, findings from Aufiero et al. (2011), which found that, in metropolitan Melbourne, older people and people with a disability were at higher risk and that many residential fire victims were recipients of community-funded care programs.

Australian records from 1 July 2003 to 30 June 2017 were accessed in the NCIS by a variety of searches. After refinement of the applicable dataset, relevant structured and non-structured data from the NCIS (comprising the summary page, police, autopsy and toxicology reports and coroner’s findings) were coded for 41 fields and entered into a specially constructed database. Once complete, the data was statistically analysed.

Between 2003 and 2017 there was no clear declining trend in fire fatalities.

The data shows those most at risk

Older people

By age cohort, those aged over 65 are the group most at risk of dying in a residential fire, a risk that increases with age. The data indicates that the other factors that increase risk in older people include smoking, having a disability, the presence of alcohol and/or medications in their blood, living alone and requiring support to live at home. Where these factors are present in combination, an older person’s risk increased significantly.

Young children

Children aged under four had the largest number of deaths of any five-year age group. The cause of fire was more often lighters or matches, which may indicate that a significant number of fires were lit by children during fire play. The link to social and financial disadvantage was particularly significant in this cohort, with almost half of deaths in the zero to four age bracket occurring in locations in the top 10% of greatest socio-economic disadvantage, and 87% of fatalities occurring in the top 40% of locations of greatest disadvantage.

People with a disability

Within the data, 47% of decedents were identified as having at least one disability present (physical disabilities: 46%, mental health disorders: 28%, and neurological disorders: 10%). The data suggests that people with a disability often died between the hours of 8 am and 12 pm. This contrasts with the overall data, where fatal fires more often occurred overnight during sleeping hours. This may indicate that for people with a disability, their disability rather than being asleep may have contributed to their inability to safely escape the fire. Similarly, people with a disability more often had a working smoke alarm.

Aboriginal and Torres Strait Islander people

More than 8% of decedents were identified as Aboriginal, Torres Strait Islander, or both. Approximately 3% of the Australian population identifies as Aboriginal or Torres Strait Islander, meaning that this cohort is over-represented in the data by a factor of 2.5. Aboriginal and Torres Strait Islander people comprised 12% of fatalities under 65 years of age and 3% of people over 65 years, likely reflecting the younger age structure of the Aboriginal and Torres Strait Islander population.

The residence

Freestanding houses/villas were the housing type where the majority (67%) of fatal fires occurred. However, these freestanding houses comprise 78% of the housing stock in Australia, so other housing types may be over-represented in the fatality data. Similarly, owner-occupiers were the most commonly identified property tenure (53%), but owner-occupiers account for approximately 67% of all property tenures in Australia. This indicates that other tenure types, such as private and public rentals, may be over-represented in fire fatalities.

The location

Most fatal residential fires occurred in major cities, but there was over-representation of deaths in regional and remote areas. The analysis of the fatality data in relation to areas of relative socio-economic advantage and disadvantage shows that most fatalities occurred where there is relatively greater socio-economic disadvantage. Fatal preventable residential fires start most commonly in the living room or bedroom. They are not necessarily large or severe fires, with approximately half of fatal fires burning one room or less of the structure.

Seasonality

Most fatal residential fires occur during the winter months. They occur most commonly between the hours of 8 pm to 8 am and, particularly, from midnight to 4 am.

Smoke alarms

In a large majority of cases (66%), it is unknown if a smoke alarm was present, despite smoke alarms being a requirement by law in all residential properties. The extent that the presence of a smoke alarm was noted by coroners is low considering their importance and that the absence of a smoke alarm may have had an impact on the fatality outcome (e.g. by providing an earlier warning to the fire victim).

Smoking

People who smoke are over-represented to a large extent in residential fire fatalities. Of cases where the smoking status of the decedent was known, 65% of people were smokers. During the study period, smoking rates in Australia decreased significantly and reduced-fire-risk cigarettes were mandated in Australia in 2010. In the 2004–05 financial year, 23% of Australians were smokers. By 2014-15 this had decreased to 16%. The fatality data does not reflect any decline in the number of smokers who died over the course of the study period. It is unclear why this is the case.

Smoking materials are a major cause of ignition of fatal residential fires. For those cases where the fire cause was known, over a quarter were caused by smoking materials, with just over a third of those relating to smoking in bed. There was a strong link between smoking materials as the cause of fire and the residence being in a relatively disadvantaged area, with 49% of fires caused by smoking materials occurring in the top 25% of the most disadvantaged locations.

Working toward zero

MFB’s Acting Chief Executive Officer/ Chief Officer David Bruce said the research provides valuable insights that agencies can use to inform practice.

“The majority of fatal house fires are preventable, which is concerning. As firefighters, our job is not just to extinguish fires but also to stop them from occurring in the first place.

“Studies like this one provide vital intelligence to assist fire and rescue services better understand why these incidents are happening and who is most at risk. This enables us to develop evidence-based policies and practices to hopefully reduce the number of fatal fires. Even one person dying in a house fire is one too many,” Mr Bruce said.

This research will inform future fire safety campaigns nationally, with fire and emergency services across Australia using the data to develop a national residential fire strategy, to reduce preventable residential fire fatalities toward zero.

The groups identified as most at risk are also the groups that are the most difficult to reach in general fire safety campaigns. Lifesaving information can now be better targeted to the areas it is needed most.

Find out more about this research here.