47th Meeting of the
Fire and Rescue Statistics User Group
13th March 2017
Home Office, Marsham Street, London

1. Attendees

(taken in order around the table)

  • Kirsty Bosley
  • Heidi Jones
  • Gill Montague
  • Simon Flood
  • Thomas Roche
  • Brian Martin
  • Dominic Louks
  • Jeremy Fraser-Mitchell
  • Rob Gazzard
  • Andy Mobbs
  • Jason Davies
  • Martin Duggan
  • Julian Parsons
  • Neil Odin
  • Dave Washer
  • Audrey Newton
  • Mike Burroughs
  • Dennis Davis
  • Henry Landis
  • Sheila Pantry
  • Will Wright
  • Ian Gough
  • Celestine Cheong
  • Eliot Mangham
  • Sarah Rich
  • Louise Dowdles
  • Stewart Ross
  • Anna Richardson
  • Isabel Preston
  • Georgina Smalldridge

2. Chair’s Introduction

Kirsty thanked Home Office colleagues for once again hosting the meeting and welcomed the record attendance.

The minutes of the last meeting had been sent to members in advance of the meeting. No changes to the minutes were received and so they were accepted as ready for publication on the FRSUG website.

3. New chair

As Kirsty’s last meeting as chair, members were asked to consider the drive and direction for the group. Overall the view was that this did not need to change substantially. The current membership includes the fire sector (recognised as wider than just the fire and rescue service), statistics users and stakeholders and had government representation, though not leadership. The diversity is positive and statistics providers, producers and users are all represented with no dominant viewpoint.

Given this agreement, volunteers for the chair were requested. Andrew Mobbs of London Fire Brigade offered to stand and was unanimously elected as the next chair.

4. Website list of statistic suppliers – update

On the FRSUG website there is a list of statistical products that are of relevance to FRSUG members. The details on this list need to be updated and so a hard copy list was passed around for update. Sheila Pantry will amend the details accordingly, but members are requested to review the amended list to ensure it is up-to-date and complete.

5. National Reports

Home Office update

Anna Richardson

Resourcing

  • After a short period of the stats team being fully resourced, since the last meeting James Gallucci has left on promotion in December and Fay Graves on maternity leave a few weeks ago. We are in the process of recruiting into the posts – but for the present we are back down to a two person stats team
  • The research team now has two social researchers Stef Bryant and Izzie Preston.
  • Caroline Frampton has left the IRS team and Audrey Newton has rejoined.

Statistics releases

  • Since September:
    • Fire Pensions on 19 October
    • Operational Statistics on 27 October
    • Response times on 19 January
  • Forthcoming releases (provisional dates)
    • Fire Statistics England – April/May
    • Half Year Monitor – May/June
    • Fire Statistics Monitor – summer

Open data

  • As mentioned last time the option being pursued was to publish datasets on individual topics that can’t be internally linked. After further discussion with experts and colleagues we are sticking with this approach as the best way to get as much data out as possible within the confines of data protection.
  • Work is now underway on the first four datasets we plan to release. These were selected on the basis of feedback from users (including the FRSUG), feedback from FRSs on the most common FOIs they receive. we also have an, as yet, unresolved issue with around 5-10% of lower level location data, datasets that do not require geographical detail lower than FRS.
  • The first four data sets are:
    1. Dwelling fires
    2. Other building fires
    3. Domestic Appliance fires
    4. Road vehicle fires
  • Those who are part of the Knowledge Hub open data group (if you wish to join, the easiest way is to contact firestatistics@homeoffice.gsi.gov.uk who will send you an invite) will have seen that Paul Gaught has posted a very rough set of variables for Domestic Appliance Fires and will do similar for Dwelling Fires in next couple of days. We’d appreciate any comments on variables people feel are missing or aren’t that useful.
  • The Fire Minister in a recent speech announced we shall be publishing the first tranche of data in the spring, which will include some/all of the four datasets above, and our plan is to regularly add further datasets after this.

Non-incident data review

  • At the end of last year we started a review of what we have called non-incident data – essentially everything we publish that is not incident data from the IRS.
  • We sent out a survey just before Christmas, supported by CFOA, asking about the usefulness of each of the current data collections – namely: workforce, fire prevention activity, fire safety activity, FRS vehicle accidents and health and safety – as well as suggestions for improvements/new data items.
  • We received a good response – and are currently considering next steps. These may include some limited changes to this year’s Operational Statistics data collection on a voluntary basis, and putting forward potential substantive or mandatory changes to the, soon to be rejuvenated, CLIP Fire group – the gateway process for adding new data items to any local authority data collection. The CLIP group will make sure any new data items are fully considered, not too burdensome and are developed with expert input. The first CLIP meeting is planned for May and Neil Odin will be jointly chairing alongside the Home Office.

Research

  • The research team has been doing some work to produce a more in-depth analysis of trends in fires and fire-related fatalities over the last ten years or so.
  • The aim is to expand on the amount of detail we are able to include in our usual stats releases – and also to put forward some suggestions for factors that may have contributed towards the long term decline in fires and fire-related fatalities.

Scottish update

Stewart Ross and Louise Dowdles

Stewart pointed out the Scottish Fire and Rescue Service (SFRS) roles as both producer and provider of statistics. He is currently reviewing Scottish data to investigate why the long-term decline in dwelling fires seem to be stopping. Using the Cheshire derived fire severity model he is finding an increase in medium and low severity fires while higher severity fires are reducing.

He noted that around 7 per cent of fires were related to smoking materials, but they accounted for 47 per cent of fire fatalities.

Jason Davies confirmed this was true in the West Midlands, with 3 per cent of accidental dwelling fires related to smoking materials, and 50 per cent of fire deaths.

There was discussion around the approaches to smoking with relation to fire safety. London had accepted smoking but aimed to make smokers safer, though are reviewing this approach and considering either preferring electronic cigarettes or advising ceasing smoking.

Louise presented the latest SFRS statistical release, Fire and Rescue Statistics in Scotland, 2015-16 (Incident data) which was released in October 2016. The non-incident statistics are now being reviewed and new guidance is being developed. SFRS 2017 publication dates are:

  • Fire Safety and Organisation Statistics 2016-17: 31/8/17
  • Fire and Rescue Statistics in Scotland, 2016-17: 26/10/17

The SFRS team have met with Home office and Office of the Chief Statistician colleagues to prepare the service for becoming producers of Official Statistics.

Kirsty mentioned recent releases of an HMFSI Local Area Inspection in the Scottish Borders PDF and a Scottish Government Evaluation of Police and Fire Reform: Year 1 Summary Report. PDF The latter considers the benefits and pitfalls of the process and outcomes of the creation of the single services in Scotland. It is intended to provide valuable insight for anyone carrying out similar exercises in public services.

Wales update

Claire Davey provided a written update:

Last month we published bulletins on Grassland fires and Deliberate fires:

and also April to September 2016 data in the form of a headline and StatsWales tables:

Recent data requests have included

Claire has been asked to look into Wales’ current privacy notice in relation to fire data and check that our current uses of the data are covered. This is in preparation for the General Data Protection Regulation (GDPR) which comes into effect on 25 May 2018.

She has presumed that casualties/victims/property owners sign a privacy notice allowing their data to be used, though the view of members at the meeting is that this is not so.

Claire included links to the code of practice and guidance from the Information Commissioners Office (ICO) on privacy notices (including the impact of the GDPR) below. She say they are quite useful although be warned most of the specific examples in the second attachment apply to the private sector.

Discussion

In discussion, members noted that the ICO focus was transferring from a requirement for explicit agreement to share data, more towards a default position of sharing, with signed agreement as the ‘gold standard’ for sharing data.

IRS update

Heidi Jones

The six months since we last met has been spent in further discussion with Home Office IT and commercial colleagues. It is fair to say that this has been a bit frustrating and that it is not easy to work out how things are set up. We are now working with the right people and we are hoping that we will be able to move forward early in the next financial year.

The IT pressures and roadmap in the Home Office are challenging. There is a great deal we could do, we just need to find the most effective way. We need to consider how the IRS fits into the wider Home Office IT portfolio. As an incident level collection there is more in common here than in DCLG.

We will start with an internal options and requirements review as discussed at the last meeting. It is too early to think about new functionality, procurement methods etc.

Changes to the IRS data collected may be needed in the light of inspection and standards, but we are not in a position to look at that yet. Other more routine data change suggestions are still being collated for later discussion with the sector. We always have to consider the business case for data changes, including whether it is reasonable to expect FRSs to collect the new data. We are also looking at guidance changes that might be required.

We are working with Paul on open data, particularly considering open datasets with the potential to reduce the burden on FRS analysts (FOIs etc).

The current IRS system was becoming rather unstable. We commissioned some new functionality which has stopped the system falling over – it has now been stable since mid-December. We are also looking at new ways forward with current contractors.

Our resource situation has changed, Caroline Frampton has left the IRS Helpdesk, Audrey Newton has re-joined the team, bringing back her extensive knowledge of the IRS system and data. I will be working on bringing in technical resource next.

No further meetings have been held with CFOA IRS Working Group, the new chair has not yet been appointed. We have had other discussions with Neil Odin, Shantha Dickinson and Brian Neat.

Finally, Fay Graves, Stef Bryant and Heidi all attended Family Group 4 meeting at Eastleigh in December to talk about the IRS, fire statistics and fire data. These were interesting and useful discussions. We are happy to attend future Family Group meetings if it would be helpful.

6. CFOA update – IDRP

Neil Odin

CFOA has recently adapted to the changes within Home Office but is determined to remain UK FRS focussed. To achieve this, the NFCC (National Fire Chiefs Council) has been set up to avoid policy position across UK diverging. This will be chaired by a full time chair, Roy Wilshire.

  • CFOA remains the corporate body and is a charity. NFCC is the – Policy making forum, will gather chief’s mandate and present lined up viewpoint for UK.
  • Along side the NFCC, Home office is setting up a Professional standards body and an independent inspection body (like the Peel report for the police) so that inspections report in four levels, from outstanding to needing attention.
  • There are 9 subject specific Co-ordinating committees serving the NFCC and these will develop a business plan setting out what is to be delivered and when.
  • The nine are: Workforce, Health, Operations, Prevention, Protection, Sector Improvement, Sector Resources, Finance and finally International issues.
  • The Integrated Data and Research Programme is within ‘sector improvement’.
  • The purpose of this programme was to coordinate data collection and its meaning as well provide governance and steer for Research
  • This month the Integrated Data Research Project (IDRP) is to be reviewed to ensure it continues to be fit for purpose.
  • Priorities for the programme will include: getting services ready for digitisation, Accountability to the community, better able to capitalise on geospatial tech
  • Research needs to be co-ordinated more widely – trying to publish a research database to share what the problems to be addressed are and what projects exist and are planned to address them.
  • NFCC is working with FSRTT and other funding groups – need funding, rigour, ideas, coordination. This will mean that the outcome or recommendations of research are better peer reviewed and then actioned accordingly.
  • Current projects include
  • Project with University of East Anglia looking at ‘Big Data’. Using Exeter data to forecast vulnerable people and consider ways for the FRS to handle big data.
  • Kent – LifeBid project, which raised some IRS challenges
  • NFCC is working globally, with Memorandums of Understanding with AFAC (Australia), ISIFC (USA)
  • Any one with any questions or queries please do look at CFOA website or contact Brian Neat (policy Officer) at brian.neat@hantsfire.gov.uk

7. Stats User Forum

Andy Mobbs

Andy provided a summary of the last Statistics User Forum he attended at the Royal Statistical Society.

The Forum is reviewing their own purpose and administration. The secretarial support has recently changed and papers are now actively distributed, rather than the previous approach where members could access the papers from a shared website.

At the next meeting on 27th March, which Andy has agreed to attend, the Indices of Multiple Deprivation (IMD) and user engagement feedback will be discussed.

Presentations

8. Heritage buildings

Henry Landis presented a summary of the thesis PDF that supported his recent degree. He aimed to use fire statistics to provide evidence for fire losses in heritage buildings. The main technical problem with the analysis was matching IRS records to heritage properties. The incident location data was often inaccurate so that no information could be gleaned from the fire reports. His analysis covered different approaches to geo-locating incidents in use in different FRSs. He found that using Unique Property Reference Numbers, often seen as the future of location data, failed on a number of points including: they were not always on the correct building outline, or the outline itself was wrong, or the outline didn’t represent a building, or the UPRN didn’t exist at all. Other geo-locating methods were also unsatisfactory in various ways.

In an analysis of fire rates in heritage buildings vs dwellings or other buildings, the heritage building fire rate was three times that of dwellings and twice that of other buildings. FRSUG members thought that this is a statistics that should be shared with owners and custodians of heritage properties to alert them to the raised risk.

9. Any Other Business

Julian Parsons highlighted activity that may be of interest to FRSUG. He is leading on commissioning of a study into the efficiency of sprinklers. The analysis will be based around IRS data. There are some very interesting preliminary findings, and also useful information about the limitations in IRS.

Sheila Pantry presented Kirsty with a gift and card from the group thanking her for her chairing of FRSUG over the years.

10. Date of Next Meeting

The proposed date of the next meeting is 5th September 2017. Details to follow.