London Fire Brigade Incident risk management toolkit

Transcription

1 London Fire Brigade Incident risk management toolkit Summary The incident risk analysis toolkit (irat) is an umbrella terms to describe a collection of statistical analysis techniques used by the London Fire Brigade (LFB). It is used to find out where most fires happen, and what types of people are most affected. irat was developed during 2005 to support the LFB performance management framework, and it was launched in autumn Its main purpose is to work out where fires are most likely to happen, so fire prevention work can focus on those areas. The tool can be used to analyse the likelihood of any type of fire. However, this work has focused on accidental fires in the home (ADFs) for two reasons. These fires tend to cause the most injuries and fatalities, but can also often be prevented with targeted community safety campaigns and home fire safety visits (HFSVs). In 2008 the data was further refined by factoring in knowledge about which types of people are most likely to start a fire accidentally. This created priority postcodes which meant it was easier to see exactly which parts of London were most in need of HFSVs. Background As part of the Fire Services Act 2004, all fire and rescue authorities must produce an integrated risk management plan (IRMP); in London the IRMP is called the London Safety Plan (LSP). LFB used irat as a tool to better understand what types of people were most at risk in London, and where they lived. Key issues and problems The LFB only had limited resources so any planned work needed to be realistic. At the same time, London is one of the world s largest cities, so any approach needed to suit a huge population. These two points meant that it was vital to accurately target households that were most at risk. The tool needed to be very good at predicting when and where incidents would occur.

2 What they did Mapping where accidental fires in the home are most likely to occur The number of fires in any one area can be correlated with the demographic, social and environmental characteristics of that area. This process is particularly useful because as areas change, experts can predict how that will affect the number of incidents. In London this was done by ward, and each one was given a rank from 1 to 625, where 1 was the ward that had the highest number of accidental fires relative to the number of households in that area. This list of 625 wards was then organised into five groups, as shown in the following table: The best way to collect this data is to compare the number of fires in a given area with the total number of households in that area. Table 1: Reducing the list of 625 wards into five bands, showing the number of accidental fires per 10,000 households Fires per 10,000 households Well above average 25 < ~ 300 Above average 20 < and 25 ~ 400 Average 15 < and 20 ~ 500 Below average 10 < and 15 ~ 666 Well below average 10 ~ 1000 Number of households per fire The table shows that wards in the top band are almost twice as likely to have accidental fires than the average ward. Those in the bottom band are half as likely.

3 The map below shows where the risk areas are across London:

4 Lifestyle profiling A commercial product called Mosaic was used to help find out what types of people may be more at risk from accidental house fire. This was particularly useful because although LFB collects detailed information about people involved in fatal fires, it does not hold comprehensive information about those who have an accidental fire but are not injured. Nor does it have information on those who do not call out the emergency services when they have a fire, and instead put it out themselves. Mosaic categorises residents into 11 lifestyle groups, which are shown in the table below. This shows that accidental fires are not random events, and certain types of people are more likely to have one. Table 2: Percentage of Mosaic groups compared to accidental fires in the home Mosaic group Percentage of these types of households in London as a whole Percentage of ADFs compared to London as a whole Percentage of fires causing injuries, compared to London as a whole Group A Group B Group C Group D Group E Group F Group G Group H Group I Group J Group K Sources: Mosaic group data from Mosaic postcode directory 2006 ADF and casualty data from LFB incident recording system 2004/ /07

5 This information is reinterpreted in the diagram below, which makes it clear which groups are more likely to have an accidental fire. The taller the bar above the centre line, the more at risk the lifestyle group is. From this analysis the groups that would most benefit from safety visits are I, F and J. These lifestyle groups are defined as follows : Group I This group is made up mainly by the elderly, who rely on state benefits for income. Their properties are likely to have been designed by councils or housing associations. They may live in retirement homes, sheltered accommodation, or in bungalows often situated within larger council estates. Group J This group consists mostly of pensioners who own their homes and who have some source of income beyond the basic state pension. Many have quite active lifestyles. The Mosaic data is updated each year and therefore provides a more accurate picture of London s demographic than the 2001 Census. The data is available at postcode and household level, which means targeting can be refined to very specific areas, streets and homes. Group F This group is characterised by those who rely on the council for their accommodation, use public transport, and often receive state benefits for most of their income. Few have rewarding or well-paid jobs, and the lack of stability within families can lead to high levels of anti-social behaviour, particularly among children.

6 Priority postcodes Organising the areas into priority postcodes means putting together the two different types of information previously described the wards where fires are most likely, and the types of people most likely to be involved. The most at-risk groups of people who live in the most at-risk areas are therefore the ones that would benefit most from HFSVs. Focusing on priority groups is vital, as the LFB cannot possibly visit each of the 3.5 million homes in London. For the period 2008/09, around 630,000 homes were identified as being in a priority postcode. Map showing an example of priority postcodes in Southwark

7 Key outcomes The end product was proactive and riskbased. The tool was easy to understand and its outputs were simple, focusing on those most at risk from fire in London. The strategy was completely evidence-based with data supporting each part of the process. Resources Few additional resources were needed. These included: use of existing information management staff off-the-shelf statistical analysis software off-the-shelf mapping software for presentation. Using the tool saved time because it meant there was no need for local work to prioritise areas. Who was involved? the director of corporate services the assistant commissioner - community safety the head of information management the risk information team (four people). Barriers and how they were overcome The basic premise of irat is that it focuses on areas where fires are most likely, not necessarily where there are the highest total number of fires. Initially, some staff found this difficult to understand. Put simply, a block of 10 houses where five fires were likely was more deserving of attention than a block of 100 houses where 20 fires were expected. However, using the priority postcode groupings made the process clearer. Critical success factors The key to the success of the toolkit is the fact that it is evidence-based. Every part of the process can be backed up with evidence, so when a decision is made on where to channel fire prevention resources, it is clear to observers that there is sound theory behind it. How could it be done better? The current process is based around assumptions rather than facts. It finds priority areas based on the kinds of people and places that are likely to have most fires. It would be better if the process was instead based around statistics that show where actual fires have happened, regardless of whether or not these were reported to LFB. However, collecting this kind of information would be a far greater drain on resources. Key contact Andy Mobbs LFB risk information manager Tel: ext andrew.mobbs@london-fire.gov.uk For more information, contact: knowledge@idea.gov.uk Local Government Improvement and Development Layden House, Turnmill Street, London EC1M 5LG Telephone Facsimile info@local.gov.uk For a copy in Braille, Welsh, larger print or audio, please contact us on We consider requests on an individual basis. Local Government Improvement and Development October 2010 L10-757