Intended for healthcare professionals


Deaths of children in house fires

BMJ 1995; 311 doi: (Published 25 November 1995) Cite this as: BMJ 1995;311:1381
  1. Ian Roberts
  1. Director Child Health Monitoring Unit, Institute of Child Health, University of London, London WCIN IEH

    Fanning the flames of child health advocacy?

    You are never at your best at 4 am, and when the intensive care unit calls to say that the 7 year old girl with smoke inhalation is showing desaturation despite 100% oxygen it takes you a while to recall the ventilator settings. Admitted earlier that day, she was lucky to escape with her life after a house fire in one of the poorer parts of town. Few paediatricians are surprised by the steep social class gradient in childhood mortality and morbidity resulting from residential fires.1 The link with poverty can be made intuitively on the basis of repeated exposure to scenarios like this.

    Poverty is both a private trouble and a public issue. Between 1979 and 1987 the number of people living in relative poverty (with less than half the national mean household income) increased from five million to 10 million.2 But in some cases the links are less obvious, and bedside intuition may be unreliable. A principal task when monitoring population child health is to link the private troubles in children's lives with the public issues that engender them.3

    Between 1983 and 1992 an average of 69 childhood deaths from residential fires occurred each year.4 The risk of death from a residential fire is related to two sets of determinants: risk factors for the occurrence of a fire and risk factors for death once a fire has occurred. In Britain fire safety and crime are both the responsibility of the Home Office. This means that questions on residential fires are included in the British crime survey, which is financed by the Home Office.5 Data from the 1994 survey show that the risk of fire is strongly related to the type of housing. The risk is greatest for those living in the poorest council housing and in temporary accommodation. Single parent families also have a significantly increased risk.

    The number of fires attended by fire brigades in Britain increased from 57000 in 1983 to 65000 in 1992, but these figures must be interpreted with caution.6 Only about 15% of fires are attended by brigades, so that the increase may simply reflect changes in the proportion of fires that are attended. However, estimates from the survey also show an increase in house fires (whether or not attended by brigades), from 473000 in 1991 to 814000 in 1993. The number of non-fatal casualties caused by fires has also increased, from 7137 in 1983 to 11402 in 1992. These increases are not surprising, since the prevalence of the main risk factors for fire--inadequate housing, single parenthood, and family poverty--also increased over this period. The number of families declared homeless doubled between 1980 and 1991, with the number of households in temporary accommodation increasing nearly fivefold.7 Single parent families currently make up 21% of families with children, compared with 12% in 1981.8

    Although the number of fires seems to be increasing, there has been no discernible increase in the death rate from residential fires, which suggests that house fires may have become more survivable. One of the most important risk factors for death in the event of a fire is the absence of a smoke alarm (odds ratio=3.4 (95% confidence interval 2.1 to 5.6)).9 Over the past five years the proportion of households in Britain with smoke alarms has increased substantially. In the 1992 British crime survey 45% of householders said that they had a smoke alarm compared with only 8% in 1988. The households at greatest risk of fire, however, are the least likely to have smoke alarms. In a survey by the Office of Population Censuses and Surveys 46% of single parent families had a smoke alarm fitted where they lived compared with 61% of two parent families. Use of smoke alarms is lowest among those with low household incomes and those living in rented accommodation.10

    Inadequate housing, homelessness, and family poverty are structural issues but are no less amenable to intervention than the health conditions they engender. The way that they differ is in the type of intervention required. The primary strategy for overcoming structural barriers to child health is public health advocacy.11 Advocacy is structural therapeutics. Yet in comparison with our North American counterparts, British paediatricians show a considerable reluctance to prescribe. This year alone, the American Academy of Pediatrics has published position statements on family support programmes, child labour, alcohol and substance misuse, violence in the media, and bicycle helmets. Over the same period the British Paediatric Association published a solitary position statement on breast feeding. North American efforts build on a strong historical tradition of advocacy. In the late 19th century Abraham Jacobi and August Caille, founders of the American Pediatric Society, were pivotal in combating structural barriers to the control of diphtheria.12 Jacobi urged that fumigation of houses be publicly financed and that incinerators to burn the linen of those who died of diphtheria be built with public funds. Caille believed that overcrowding and poor oral hygiene were causes of diphtheria and called for improved social housing and the provision of free dental care.

    How would Jacobi have responded to the problem of deaths in residential fires? Almost certainly he would have called for better provision of social housing, an increase in benefit support for single parent families, programmes giving away smoke detectors, and legislation on the installation and maintenance of smoke alarms in rental accommodation, with responsibility placed on landlords. Jacobi used clinical intuition to link private troubles with public issues. Today we have modern epidemiology. But without the political will to use epidemiological evidence to influence public policy the potential of this advance will never be realised.


    View Abstract