Download:
pdf |
pdf
USFA Response to the
OMB 2013 NFIRS Terms of
Clearance
USFA Response to the
OMB 2013 NFIRS Terms of Clearance
A System Review and Assessment of Data Quality
September 2014
National Fire Data Center
U.S. Fire Administration
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
TABLE OF CONTENTS
Introduction .................................................................................................................................... 1
About NFIRS .................................................................................................................................... 2
NFIRS Enhancements................................................................................................................... 8
NFIRS Training ............................................................................................................................. 9
Uses of NFIRS ............................................................................................................................. 10
NFIRS and the NFPA Survey ....................................................................................................... 11
Incident Reporting and Submission Process ................................................................................. 13
Incident Reporting ..................................................................................................................... 13
Submission to the National Production Database .................................................................... 14
Public Data Release and Data Review ....................................................................................... 14
Data Quality and USFA Interaction with States ........................................................................ 16
Data Quality Checks by the NFDC ............................................................................................. 16
Key Data Considerations for the User ........................................................................................... 17
Unknown Entries ....................................................................................................................... 17
Fires vs NFIRS Record Counts .................................................................................................... 18
Counting Fires vs Counting Fire‐Related Statistics .................................................................... 18
Confined vs Nonconfined Fires ................................................................................................. 18
Mutual Aid ................................................................................................................................. 19
Types of Fires ............................................................................................................................. 19
Property Definitions .................................................................................................................. 19
Multi‐year and Trend Analyses ................................................................................................. 20
Cause ......................................................................................................................................... 20
Smoke Alarms and Smoke Alarm Performance ........................................................................ 23
Dollar Loss Data ......................................................................................................................... 23
Structures, Buildings, and Nonbuildings ................................................................................... 23
NFIRS Data Quality ........................................................................................................................ 24
State‐based Data Quality .......................................................................................................... 24
Data Quality of Key Data Elements ........................................................................................... 27
NFIRS Data Element Quality .................................................................................................. 30
NFPA Survey .................................................................................................................................. 64
Sample Selection ....................................................................................................................... 64
Data Collection .......................................................................................................................... 65
Estimation Methodology ........................................................................................................... 66
Fire Experience of Nonrespondents .......................................................................................... 67
Resources ...................................................................................................................................... 68
i
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
LIST OF FIGURES
Figure 1. NFIRS Fire Department Participation 1980‐2011, Fire Incidents only ............................. 4
Figure 2. Total Reported Fire Incidents, NFIRS Public Data Release 2003‐2011 ........................... 7
Figure 3. NFIRS Incident Reporting, Submission, and Dissemination Process ............................. 15
Figure 4. 3‐Year NFIRS Data Quality by State Relative to National Average Data Quality
Measure, 2009‐2011 ............................................................................................................... 25
Figure 5. Annual NFIRS Data Quality by State Relative to National Average Data Quality
Measure, 2009‐2011 ............................................................................................................... 26
LIST OF TABLES
Table 1. NFIRS Modules .................................................................................................................. 5
Table 2. NFIRS Fire Incident Data Reporting by Version (percent) ................................................. 7
Table 3. Mid‐Level Structure Fire Cause Groupings ..................................................................... 20
Table 4. NFIRS Data Elements Used in USFA Analyses (Fires) ...................................................... 28
Table 5. NFIRS Data Elements Used in USFA Analyses (Civilian Casualties) ................................. 29
Table 6. Data Element Quality and Usability Summary for Common NFIRS Data Elements Type of
Fire, 2009‐2011 ....................................................................................................................... 32
Table 7. Data Element Quality and Usability Summary for Common NFIRS Data Elements Type of
Fire with Deaths or Injuries, 2009‐2011 ................................................................................. 33
Table 8. Data Element Quality and Usability Summary for Common NFIRS Data Elements Type of
Fire with Deaths or Injuries, 2009‐2011 ................................................................................. 34
Table 9. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Nonbuilding Structure Fires, 2009‐2011 ................................................. 37
Table 10. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Buildings and Mobile Property Structures, 2009‐2011 ........................... 38
Table 11. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Vehicle Fires, 2009‐2011.......................................................................... 40
Table 12. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Outside Fires, 2009‐2011 ......................................................................... 42
Table 13. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Fatal Fires, 2009‐2011 ............................................................................. 43
Table 14. Data Quality for Common NFIRS Data Elements Used in USFA Analyses Distribution of
Fatal Fires by Reported Deaths, 2009‐2011 ........................................................................... 45
Table 15. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Buildings and Mobile Property Structures Fatal Fires, 2009‐2011 .......... 46
Table 16. Data Quality for Common NFIRS Data Elements Used in USFA Analyses Distribution of
Buildings and Mobile Property Structures Fatal Fires by Reported Deaths, 2009‐2011 ........ 48
Table 17. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Fires with Injuries, 2009‐2011 ................................................................. 48
Table 18. Data Quality for Common NFIRS Data Elements Used in USFA Analyses Distribution of
Reported Fires with Injuries by Reported Injuries, 2009‐2011 .............................................. 51
ii
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 19. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Buildings and Mobile Property Structures Fires with Injuries, 2009‐2011
................................................................................................................................................. 51
Table 20. Data Quality for Common NFIRS Data Elements Used in USFA Analyses Distribution of
Buildings and Mobile Property Structures with Injuries by Reported Injuries, 2009‐2011 ... 53
Table 21. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Distribution of Reported Property Dollar Loss, 2009‐2011 ..................................... 54
Table 22. Data Quality for Common NFIRS Data Elements Used in USFA Analyses Distribution of
Reported Contents Dollar Loss in Buildings and Mobile Property Structures Fires, 2009‐2011
................................................................................................................................................. 55
Table 23. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Deaths, 2009‐2011 ................................................................................... 56
Table 24. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Deaths in Buildings and Mobile Property Structures, 2009‐2011 ........... 58
Table 25. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Injuries, 2009‐2011 .................................................................................. 60
Table 26. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA
Analyses Reported Injuries in Buildings and Mobile Property Structures, 2009‐2011 .......... 62
iii
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
INTRODUCTION
Federal agencies conduct or sponsor a wide variety of information collections to gather data
from businesses, individuals, schools, hospitals, and state, local, and tribal governments.
Information collections that employ surveys are frequently used for general‐purpose statistics
as well as for program evaluations or research studies that answer more specific research
questions. Data collected by Federal agencies are widely used to make informed decisions and
to provide necessary information for policy makers and planners. The collection of this
information can take many forms and is accomplished in a variety of ways.
The Paperwork Reduction Act of 1995 (PRA) requires agencies to submit requests to collect
information from the public to the Office of Management and Budget (OMB) for approval.
These requests, variously known as Information Collection Requests (ICRs), PRA submissions, or
“OMB clearance packages”, are required for any survey used for general purpose statistics,
program evaluations, or research studies. The purpose of the PRA is to ensure that the public is
not overburdened by the federal data collection. In a given period, the OMB may focus on the
design, methodology, practical utility of data to the federal government and other issues. In the
current clearance period, the OMB has requested that the U.S. Fire Administration (USFA)
assess and document the quality of the information from the National Fire Incident Reporting
System (NFIRS) and the National Fire Protection Association (NFPA) and make it more
accessible and useable for NFIRS users.
Although NFIRS is by design a voluntary incident reporting system and not a survey, the OMB
considers the system to fall under the PRA. In the past, OMB has questioned why NFIRS is a
voluntary census of incidents for all departments rather than a statistical sample. USFA has
investigated the possibility of sampling and the issues surrounding it. While there are certain
advantages to a statistical sampling methodology, USFA has not undertaken a sampling
approach to fire incident reporting for several reasons. First and foremost, Public Law 93‐498
directed the then newly created USFA to develop a standardized incident data reporting
method and to assist local and State agencies in reporting incident data to this system.1
Because NFIRS is used at the local, state, and Federal levels, abandoning it for a statistical
sampling method would adversely impact state and local fire department incident reporting
and the NFIRS standard, which is also used internationally. Additionally, much of the cost
burden of the current NFIRS reporting is carried by the state NFIRS operations. USFA has relied
heavily on states as cooperative partners in bearing the costs and resources of maintaining the
system, and the states have relied on USFA to shoulder the development costs. Switching to a
1 http://uscode.house.gov/view.xhtml?req=%28%22national+fire+data+center%22%29&f=treesort&fq=
true&num=1&hl=true&edition=prelim&granuleId=USC‐prelim‐title15‐section2208
1
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
sampling method would mean USFA would need to acquire additional funding and personnel to
design and maintain a new sampling system, leaving the state and local entities to wholly cover
the cost of an incident reporting system.
In addition to completing two separate studies showing that there is no evident systematic non‐
response bias to the NFIRS system, under the PRA process USFA has also made exceptional
progress working with our state and local partners in addressing previous OMB PRA terms of
clearance instructions to increase the use of the voluntary NFIRS system:
“The agency is instructed to continue efforts to improve response rates to NFIRS. The
agency should also investigate the possibility of using additional studies to determine
what characteristics of a fire department might make it less willing to participate in
NFIRS and determine whether there is a systematic nonresponse bias to the system
that should be disclosed in the published reports based on this system.” (2003)
“The agency is instructed to continue efforts to increase the utilization of NFIRS. The
agency must provide a report to OMB on the bias in NFIRS due to non‐response. This
report is a condition of future OMB approvals.” (2006)
“The agency is instructed to continue efforts to increase the utilization of NFIRS. The
agency must provide a report to OMB on the bias in NFIRS due to non‐response. This
report is a condition of future OMB approvals.” (2009)
The current terms of clearance require that:
“…FEMA will engage in efforts over the clearance period to assess and document the
quality of the information from NFPA and NFIRS and make this more accessible and
useable for NFIRS users.” (2013)
The following is in response to the 2013 Terms of Clearance. It is a review of the NFIRS system,
the many robust data quality checks and mechanisms which are an integral part of the system,
and an assessment of the data quality both at the state level and at the data element level. The
data element assessment is of the most commonly used data elements in NFIRS data analyses.
NFIRS data from the three most recent years available at the time of this document’s
production (2009‐2011) are reviewed. Although the USFA has no authority over the NFPA
survey, a section drawn from published NFPA documents covering the NFPA survey
methodology is also included.
ABOUT NFIRS
The NFIRS was established in 1975 as one of the first programs of the National Fire Prevention
and Control Administration, which later became the USFA. The basic concept of NFIRS has not
2
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
changed since the system’s inception. All states and all fire departments within them have been
invited to participate on a voluntary basis. Participating fire departments report a common core
of information on an incident and any casualties that ensue by using a common set of
definitions. Detailed incident data are reported locally. Local agencies forward the completed
NFIRS modules to the state agency responsible for NFIRS data. The state agency combines the
information with data from other fire departments into a statewide database and then
transmits the data to the National Fire Data Center (NFDC) at the USFA. Data on individual
incidents and casualties are preserved incident by incident at local, state, and national levels.
From an initial six states in 1976, NFIRS has grown in both participation and use. Over the life of
the system, all 50 states, the District of Columbia, and more than 40 major metropolitan areas
have reported to NFIRS. As well, more than 30,000 fire departments have been assigned
participating NFIRS fire department identification (FDID) numbers by their states. Once limited
to fire incidents only, NFIRS now encompasses all incidents to which the fire department
responds: fire, emergency medical services (EMS), hazardous materials (hazmat), and others.
Approximately 1 million fire incident records and 22 million non‐fire incident records are added
to the database each year. NFIRS is the world’s largest collection of incidents to which fire
departments have responded.
Between 1985 and 1999, the level of participation remained relatively constant: A few states
came in or left the system each year, and at least 39 states reported to NFIRS. Most years also
included participation from the District of Columbia. The number of fire departments
participating within the states remained relatively constant as well, with a slight dip in
participation during the system migration from version 4.1 to 5.0 in 1999. In 2000, the number
of states increased to 43, the Department of Defense adopted NFIRS reporting, and fire
department participation began to bounce back from the version 5.0 transition low. Since 2000,
state and fire department participation has been steadily increasing. In 2003, NFIRS reached a
milestone with participation by all 50 states. The following year, NFIRS achieved another
significant goal: NFIRS not only achieved the national goal of 100 percent state participation,
including the District of Columbia, but also for the first time, the Native American Tribal
Authorities submitted data.
NFIRS continues to grow and mature. As of 2007, a new level of participation had been
achieved: all 50 states, the District of Columbia, Native American Tribal Authorities, Northern
Mariana Islands, and Puerto Rico all participated in NFIRS for a total of 54 state, district, tribal
authority, and commonwealth entities. However, the Northern Mariana Islands and Puerto Rico
are no longer reporting incident data to NFIRS. Fire departments reporting fire incidents grew
to 20,680 in 2011 (Figure 1). Across participating entities, 69 percent of U.S. fire departments
3
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
reported fire incidents to NFIRS in 2011.2 The percentage of fire departments participating in
NFIRS varies from state to state, with some states not participating at all in some years. With
over two‐thirds of all fire departments nationwide reporting fire incidents to NFIRS 5.0,
however, the reporting departments represent a very large dataset that enables USFA to make
reasonable estimates of various facets of the fire problem. Although some states do require
their departments to participate in the state system, participation in NFIRS is voluntary.
Additionally, if a fire department is a recipient of a Fire Act Grant, participation is required.3
Figure 1. NFIRS Fire Department Participation
1980‐2011, Fire Incidents only
20,000
15,000
10,000
5,000
0
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Number of Participating Fire Departments
25,000
Source: NFIRS.
Note:
1999‐2008 includes participation from NFIRS 4.1 and NFIRS 5.0.; 2009 and later includes participation only
from NFIRS 5.0
Corresponding to increased participation, the numbers of fires, deaths, and injuries, as well as
estimates of dollar loss reported to NFIRS, also have grown; an estimated 71 percent of all U.S.
fires to which fire departments responded in 2011 were captured in NFIRS.
2 For 2011, NFPA estimated that there were 30,145 fire departments in the U.S.
NFPA, U.S. Fire Department Profile Through 2011, http://www.nfpa.org/assets/files//PDF/OS.FDProfile.pdf, October 2012.
3 From the Assistance to Firefighters Grant Guidance and Application Kit (June 2012), if the applicant is a fire department, the
department must agree to provide information, through established reporting channels, to NFIRS for the period covered by the
assistance. If a fire department does not currently participate in the incident reporting system and does not have the capacity
to report at the time of the award, the department must agree to provide information to the system for a 12‐month period that
begins as soon as the department develops the capacity to report. See http://www.fema.gov/library/viewRecord.do?id=6007
(fg_2012_afg_program_guidance.pdf).
4
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
There are, of course, many problems in assembling a real‐world database, and NFIRS is no
exception. Although NFIRS does not represent 100 percent of incidents reported to fire
departments each year, the enormous dataset and good efforts by the fire service result in a
huge amount of useful information. Because of advances in computer technology and data
reporting techniques over the past 35 years and improvements suggested by participants,
NFIRS has been revised periodically. The newest revision, NFIRS 5.0, became operational in
January 1999.
NFIRS 5.0 captures information on all incidents, not just fires, to which a fire department
responds. In addition to many data coding improvements, version 5.0 provides 11 modules that
recognize the increasingly diverse activities of fire departments today. These modules,
together, contain 567 data elements or fields.
The Basic Module is the main module, which is completed for every incident. The other
modules are filled out, when appropriate, to provide additional information on an incident. All
11 modules are listed in Table 1 below:
Table 1. NFIRS Modules
Module
Description
Basic Module
Fire Module
Structure Fire Module
Civilian Fire Casualty Module
Fire Service Casualty Module
EMS Module
Hazardous Materials Module
Wildland Fire Module
Apparatus/Resources Module
Personnel Module
Arson Module
General information for each incident
Fire incident information
Information on structure fires
Fire‐related injuries or deaths to civilians
Injuries or deaths to firefighters
Medical incidents
Hazardous materials incidents
Wildland or vegetation fires
Apparatus‐specific information
Personnel associated with apparatus
Intentionally set fire information
Source: NFIRS.
Data from the modules are grouped together each calendar year to create the Public Data
Release (PDR) files in delimited text (.txt) format that are then released annually into the public
domain. For NFIRS data submitted prior to 2012, the PDR files were released in dBase (.dbf)
format. The Apparatus/Resources and Personnel Modules are excluded from the PDR because
they are intended for local fire department use only, and the PDR dataset’s main utility is
intended for national analyses. The PDR files consist of a subset of the data fields contained
within the NFIRS national production database. For example, data elements with sensitive or
identifying information are removed as are data elements that are wholly used for maintenance
or production purposes. The PDR data structure has been considerably simplified from the
5
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
production database’s schema for ease of use. The PDR files from 2004 to the present only
include fire and hazmat incidents and their related data tables. Prior to 2004, all incidents were
included in the PDR files.
In its basic form, the NFIRS PDR files have a relational data structure where data from each
incident module is represented by a row in a data table. The primary tables (basic incident and
incident address) contain most of the Basic Module data. There is exactly one record in the
basic incident table for every incident reported to NFIRS. All other modules, represented by
data tables with similar names (such as fire incident or civilian casualties), have records that are
linked to the basic incident table through unique incident identification key fields (state, fire
department ID, incident date, incident number, and exposure number). Some module data are
split across several tables (e.g., basic incident, incident address, and basic aid tables); one table
(fire incident) combines data from two modules (Fire Module and Structure Fire Module). Some
tables, such as fire incident, will only have one record for each relevant incident in the basic
incident table, while tables such as civilian casualty may have several records linked to a single
incident in the case where multiple injuries or deaths occur in the same incident.
The current version of NFIRS, NFIRS 5.0, is the result of a collaborative effort between USFA and
state and local users and incorporates many improvements. The design of NFIRS 5.0 makes the
system easier to use than previous NFIRS versions because it captures only the data required to
profile the extent of the incident. Some fires, for example, require just basic information to be
recorded, whereas others require considerably more detail.
State participation is voluntary, and each state specifies NFIRS reporting requirements for its
fire departments. States have the flexibility to adapt their state reporting systems to their
specific needs. As a result, the design of a state’s incident reporting system varies from state to
state. NFIRS 5.0 was designed so that data from state systems can be converted to a single
format that is used at the national level to aggregate and store NFIRS data.
The proportion of 5.0 data has steadily increased since the introduction of NFIRS 5.0 in 1999
(Table 2). The proportion of 5.0 data rose to 99 percent by the 2008 dataset. Prior to 2009,
NFIRS 4.1 data in its converted form had been accepted by the system; however, USFA only
uses native 5.0 data in its NFIRS‐based analyses and quality checks.4 For this reason, USFA
multi‐year analyses do not include NFIRS 5.0 data prior to 2003 as prior to 2003 the proportion
of 5.0 data in the NFIRS PDR was less than 80 percent. Since Jan. 1, 2009, NFIRS 4.1 data are no
longer accepted by the system.
4 Although, beginning in 2009, NFIRS does not accept version 4.1 data, a few future‐dated NFIRS Version 4.1 records from past
years do appear in the database from user entry error; these records do not belong in the data year in which they were
submitted and should not be included in analyses.
6
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 2. NFIRS Fire Incident Data Reporting by Version
(percent)
Year
NFIRS 4.1, 5.0 format NFIRS 5.0
1999
92
8
2000
77
23
2001
48
52
2002
31
69
2003
19
81
2004
11
89
2005
5
95
2006
5
95
2007
2
98
2008
1
99
2009 – current
0
100
Source: NFIRS.
Incidents submitted to the National database and reflected in the PDR declined initially as
NFIRS 4.1 acceptance was phased out, but increased as departments fully adopted NFIRS 5.0. It
is important to remember that the PDR is a one‐time snapshot of the incident data submitted
by the July 1 deadline. Additional data may be submitted to the national database after this
deadline.
Figure 2. Total Reported Fire Incidents, NFIRS Public Data Release
2003‐2011
Reported Fire Incidents
1,400,000
1,211,630 1,219,301
1,157,845
1,126,792
1,200,000
1,000,000
873,090
1,194,982 1,214,662
1,111,902
934,640
4.1 NFIRS Data
800,000
5.0 NFIRS Data
600,000
400,000
Total Reported
Fire Incidents in
NFIRS Public
Release Data
200,000
‐
2003
2004
2005
2006
2007
2008
2009
2010
2011
Source: NFIRS.
Note:
Includes all incident records in the NFIRS PDR less any incidents with fatal data quality errors.
7
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
NFIRS Enhancements
Under the USFA Reauthorization Act of 2008, the U.S. Congress authorized and funded USFA to
develop a Web based data entry tool enhancement to NFIRS. This upgrade to the system began
in October 2008. In 2010, a data warehouse for generating output reports for use in analyses
was developed with additional funding provided by Congress for NFIRS enhancements. These
improvements make reporting and accessing the NFIRS data much easier for fire departments.
In July 2010, USFA completed and deployed the new Web‐based data entry tool. The Data Entry
Browser Interface (DEBI) is a one‐purpose tool for use by the fire service to document incident
information within NFIRS. While the functionality is the same as the NFIRS client Data Entry
Tool that has been available for many years, DEBI allows entry of incidents using a standard
Web browser, eliminating the need to download, install and configure client software.
The development of a flexible NFIRS data warehouse with comprehensive data mining
capabilities was completed in July 2011. It is scheduled for deployment to national, state and
fire department NFIRS users in three phases beginning in summer 2014. The data warehouse
will allow NFIRS users to access and report on nationally reported data with significantly
increased functionality over the current report generation tool. The data have been
transformed into a custom schema that greatly increases the speed of report generation and
data access. NFIRS users will be able to generate reports using data from other departments
and states, which was not previously possible.
The data warehouse includes a suite of 30 NFIRS data quality reports. The reports include
functionality to:
Track and assign data quality measures for critical data elements in each module and
provide overall data quality ratings by state and department;
Identify incidents which, though valid, have serious data quality issues and list the
problems found with the incidents;
Identify incidents with outlier dates and times for incident response and duration;
Identify incidents with dollar loss data quality issues;
Identify incidents that should have been marked as invalid;
Track departments and incidents with abnormal fires under investigation percentages
and durations;
Identify incidents with mutual aid field data quality issues;
Identify near duplicate incidents; and
Identify incidents with missing casualty modules.
8
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
More detailed information regarding the NFIRS enhancements is available at
http://www.usfa.fema.gov/fireservice/nfirs/enhancements/.
NFIRS Training
To promote best practices, USFA offers several NFIRS training courses for fire departments and
state‐level departments that manage NFIRS. The courses are available at the National Fire
Academy (NFA), online, and within localities (i.e., field deliveries). These training courses
include the NFIRS Program Management course (six‐day NFA course), the NFIRS 5.0 Self‐Study
course (online), the Introduction to NFIRS 5.0 course (two‐day NFA course or field delivery
course), and the NFIRS Data Analysis and Problem Solving Techniques course (two‐day NFA
course or field delivery course).
The NFIRS Program Management course teaches the participants the full duties of NFIRS
program management and enables participants to promote, support, and manage NFIRS
incident reporting successfully. This six‐day course is offered as a resident course at the
NFA and is offered as a field delivery course. The foundation of the training is built on
the use of the participants’ own data in the national database. This allows the
participants to see the quality of their data and the impact it has on analysis and
decision‐making. The course teaches the participants the five roles of NFIRS program
management: communicating, administrating, planning, training, and operating. The
participants learn
o the NFIRS rules to ensure that data are coded according to the NFIRS standard,
o to use tools (e.g., web‐based reports, queries, Excel, and pivot tables) to identify
data quality issues (e.g., invalid incidents, incomplete incidents, outdated
incidents),
o how to use the data to recognize problems in their communities,
o how to effectively present the data to decision‐makers and other users of fire
data,
o how to administer different training methods (e.g., train‐the‐trainer, initial, and
refresher) to ensure that accurate, complete, and timely data are available, and
o how to develop a local analytical tool from the data in the national database by
exporting data from the national database,
importing the data to a local database, or
building a local interactive tool to display, troubleshoot, query, and
analyze the data.
The Introduction to NFIRS Self‐Study (online) course provides an overview of the
incident reporting system, its modules, and rules for documenting incidents.
9
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
The Introduction to NFIRS 5.0 course emphasizes how to properly document incidents
using standardized NFIRS data elements and codes for achieving uniformity in incident
reporting.
The NFIRS Data Analysis course teaches the participants how to better evaluate the
reported data, how to use the data to identify problems, evaluate resources, and
measure services provided. By using their own data in the national database,
participants gain a direct correlation to the training and its immediate value to their
departments and communities.
This training program is designed specifically to support local fire service organizations and
assists them in providing data both to management and decision‐makers, to the state uniform
fire reporting system, and to NFIRS nationwide. Additional information on NFIRS training
courses can be found at http://www.usfa.fema.gov/fireservice/nfirs/ training/.
To assist fire departments who use NFIRS to improve data quality and reliability, USFA publishes
short, informative notices called NFIRSGrams. By addressing frequently asked questions and
common mistakes made when completing incident forms, these bulletins provide coding help
to fire department personnel using NFIRS. NFIRSGrams also help NFIRS users to better
understand their impact on the quality of the information from NFIRS at the local, state, and
national levels.
USFA’s NFIRS Support Center also offers a consolidated national help desk to provide technical
support to fire departments and NFIRS State Program Managers regarding all aspects of NFIRS.
Uses of NFIRS
NFIRS data are used extensively at all levels of government for major fire protection decisions.
At the local level, incident and casualty information is used for setting priorities and targeting
resources. The reported data are particularly useful for designing fire prevention and
educational programs and EMS‐related activities specifically suited to the real emergency
problems local communities face.
At the state level, NFIRS is used in many capacities. One valuable contribution is that some state
legislatures use these data to justify budgets and to pass important bills on fire‐related issues
such as sprinklers, fireworks, and arson.
Many federal agencies, in addition to USFA, make use of NFIRS data. NFIRS data are used, for
example, by the Consumer Product Safety Commission (CPSC) to identify problem products and
to monitor corrective actions. The Department of Transportation uses NFIRS data to identify
fire problems in automobiles, which has resulted in mandated recalls. The Department of
10
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Housing and Urban Development uses NFIRS to evaluate safety of manufactured housing
(mobile homes).
The USFA uses the data to design prevention programs, to order firefighter safety priorities, to
assist in the development of training courses at the National Fire Academy, and for a host of
other purposes. Thousands of fire departments, scores of states, and hundreds of industries
have used the data. The potential for even greater use remains. The USFA report, Uses of
NFIRS: The Many Uses of the National Fire Incident Reporting System, further describes the uses
of the data and is available online at
http://www.usfa.fema.gov/downloads/pdf/publications/nfirsuse.pdf.
NFIRS and the NFPA Survey
There are two major sources of data about the U.S. fire problem: USFA’s NFIRS and the NFPA
Survey of Fire Departments for U.S. Fire Experience. NFPA is an international nonprofit
organization whose mission is to reduce the burden of fire and other hazards.
While the USFA provides fire departments with NFIRS as a method to report fire data to
understand the details of the U.S. fire problem – how fires start, where they occur, when they
occur, what (if any) equipment is involved, and other associated elements of information – the
NFPA surveys fire departments each year to determine the nature and characteristics of fire
departments across the U.S. and uses the data collected in the survey to estimate the
magnitude of the fire problem. NFPA’s Survey of Fire Departments for U.S. Fire Experience is
based on a stratified random sample of U.S. fire departments.5 The sample of departments is
stratified by size of community protected, and a ratio estimation methodology is used to
develop national level summary estimates on fire loss statistics (the total numbers of reported
fires, fire deaths, fire injuries and direct dollar loss) as well as summary estimates of fires and
losses by major incident types (i.e., structure, vehicle, outside and other).
Thus, overall estimates of the fire problem come from NFPA’s annual Survey of Fire
Departments for U.S. Fire Experience. As noted, this survey produces national level summary
estimates on fire loss statistics as well as summary estimates of fires and losses by major
incident types (i.e., structure, vehicle, outside and other). The summary estimates by major
incident type are further broken down to the next tier – e.g. residential structures, highway
vehicles, etc. The raw NFPA survey data are of a proprietary nature and not available to the
public, USFA, or various other national fire data analysts.
5 For detailed information regarding NFPA’s survey methodology, see NFPA’s annual report on Fire Loss in the United States:
http://www.nfpa.org/~/media/Files/Research/NFPA%20reports/Overall%20Fire%20Statistics/osfireloss.pdf.
11
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
All nationally‐based estimates on subsets of fire data, however, are derived by using the NFPA
estimates to scale up the raw NFIRS subset data, which is a standard statistical technique. These
“national estimates” are not the raw totals from NFIRS. The estimates are based on a method of
apportioning the NFPA estimates for total fires, structure, vehicle, outside and other fires.6
Generally, these national estimates are derived by computing a percentage of fires, deaths,
injuries, or dollar loss in a particular NFIRS category and multiplying it by the corresponding
total estimate from the NFPA annual survey.7 In analyses, the unknown and missing data values
should not be ignored. The approach taken by USFA in presenting the data is to provide not
only the “raw” percentages of each category, but also the “adjusted” percentages computed
using only those incidents for which data were provided.
One problem with this approach is that the proportions of fires and fire losses differ between
the large NFIRS dataset and the NFPA survey sample.8 Nonetheless, to be consistent with
approaches being used by other fire data analysts, the NFPA estimates of fires, deaths, injuries
and dollar loss are used as a starting point. The details of the fire problem below the national
level are based on proportions from NFIRS. Because the proportions of fires and fire losses
differ between NFIRS and the NFPA estimates, from time to time, this approach leads to
inconsistencies. These inconsistencies will remain until all estimates can be derived from NFIRS
data alone.
Ideally, one would like to have all of the data for the various components come from one
consistent data source—NFIRS. One of the critical pieces of data necessary to do so is missing:
the overall population protected by all reporting fire departments. This “residential population
protected” is not reported to NFIRS, nor are the data easy to come by, especially where a
county or other jurisdiction is served by several fire departments that each report their fires
independently.
Other issues—such as full reporting because of reporting deadlines, data access, budgetary
considerations, and the like—add a layer of complexity to using the NFIRS data to create
estimates. Through the years, a number of ad‐hoc studies have been undertaken to identify
NFIRS nonresponse bias, but none have identified major reporting issues. Most of the NFIRS
6 The foundation of computing national estimates is based on “The National Estimates Approach to U.S. Fire Statistics” by Hall,
J. and Harwood, B.: http://www.nfpa.org/~/media/6906FADB2CE149488FB5103F4A750A05.ashx.
7 The NFPA summary estimates are used for the overall U.S. fire losses; fire losses from structure, vehicle, outside and other
fires; and as the basis for USFA’s estimates of residential and nonresidential building fires and losses. The alternative approach
for these summary numbers is to use the relative percentage of fires (or other loss measures) from NFIRS and scale up (multiply
by) the NFPA estimate of total fires.
8 For additional information regarding the differences in proportions of fires and losses between NFIRS data and the NFPA
survey, see the section entitled Differences Between NFIRS Data and NFPA Survey Data in “USFA’s Data Sources and
Methodology Documentation,” March 2014,
http://www.usfa.fema.gov/downloads/pdf/statistics/data_sources_methodology.pdf.
12
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
data exhibit stability from one year to the next. Results based on the full data set are generally
similar to those based on part of the data, another indication of data reliability. The dataset is
so large—on average over the past 5 years about 67 percent of reported fires9—and reasonably
distributed geographically and by size of community that it is used as input to developing
national estimates.10
INCIDENT REPORTING AND SUBMISSION PROCESS
Over 20,000 fire departments11 and state governments actively work on improving the
information they enter into NFIRS. Their data, based on their own needs for high quality fire
and other incident data, is used to plan, fund, and implement effective local, state, and regional
fire and emergency services programs.
At the same time, the NFDC engages in efforts to both document and communicate to users the
content and quality of fire information products resulting from NFIRS data. The NFDC
accomplishes these efforts through an ongoing and robust data quality assessment process,
detailed analysis of NFIRS data and production of reports, training and informational bulletins,
data analysis tools, and a congressionally funded NFIRS enhancement project.
In addition to the suite of data quality reports that will be available to NFIRS data users
beginning summer 2014, the NFIRS data are quality‐checked during data entry, data submission
to the national database, and prior to the creation of the PDR. The reporting, submission, and
subsequent dissemination process is shown in Figure 3.
Incident Reporting
The NFIRS reporting format is mostly consistent with the NFPA Standard 901, "Uniform Coding
for Fire Protection" 2001 version. The current version of NFIRS, version 5.0, expands the
reporting of data beyond fires to include the full range of fire department activity on a national
scale. It is an all‐incident reporting system.
Within the NFIRS participating states, participating local fire departments fill out incident,
casualty and optional reports for fires and other incident types as they occur. The majority of
the data are reported electronically through third party software, the NFIRS data entry tools, or
the reporting department’s own system. In a very few departments, the data may be written by
hand on paper forms and entered electronically at a later time. They forward the completed
9 The data reported in NFIRS is continually growing: between 2007 and 2011, NFIRS data represented 67, 61, 57, and 59 percent
of reported fires, fire deaths, fire injuries, and direct dollar loss when compared to the statistical estimates from NFPA’s Annual
Survey of Fire Departments. Between 2009 and 2011 these percentages rose to 70, 64, 58, and 62 respectively.
10 See USFA’s NFIRS Representativeness Study (October 2008) previously submitted to the OMB.
11 NFIRS; see also See Figure 1.
13
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
incidents electronically (or via paper forms if the department has no other means) to their state
office where the data are validated and consolidated into a single electronic database. Each fire
department is responsible for the data they report and each department is encouraged to
ensure that their data are complete and accurate. The extent of data quality checks varies from
department to department.
At the state level, the data from the participating fire departments are validated. Data are
validated automatically by the application on import into the NFIRS database. Data are checked
to make sure required fields are present and that field values are within acceptable ranges. Log
files are generated with validation results which are available to the user via either email (Bulk
Import Utility) or access to the logs stored on their local computer (Data Entry Tool client
software). The extent of data quality checks varies from state to state. Incident data that cause
critical errors and fail validation checks are sent back to the local fire department for correction
and resubmission.
Submission to the National Production Database
Periodically, the aggregated statewide data are sent to the NFDC to be released and included in
the national production database. When and how states send their data depends on the
individual state – some states (and their departments) use USFA’s data entry tool to enter,
store, and manage their data on USFA’s federal server warehouse. Other states, typically the
larger states, keep their data locally and report their incident data en masse, quarterly or
yearly, just prior to the annual federal reporting deadline. Submission guidelines12 call for
quarterly reporting at a minimum during the year and an annual deadline for states on July 1.
Regardless of the submission or entry method, all state data submitted to the federal servers
are stored in individual state partitions. From the state partitions, the states manage their final
datasets prior to releasing their data to the national production database. All state data belongs
to the individual states with the state having the sole responsibility for its content. The NFDC is
the custodian of the data and does not have the authority to make changes to the state data in
the state partitions.
Public Data Release and Data Review
After the data submission deadline, a “snapshot” of the national production database is taken.
It is from this snapshot that the PDR is created. During the PDR process, the released incident
data are checked for fire death and reported dollar loss consistency. Because deaths and total
dollar loss are important metrics of the U.S. fire problem, incidents with more than 20 million in
12 NFIRS Reporting Guidelines: https://www.nfirs.fema.gov/system/guidelines.shtm.
14
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
dollar loss or more than 5 deaths are sent back to the state for verification.13 Once the PDR is
finalized, it is packaged on a CD with documentation and made available to the public and other
government agencies and outside organizations by request.
Internally, once the annual PDR is completed, USFA assigns data quality ratings based on a
calculated formula to each individual department, each state, and nationally overall. The data
quality rating is calculated using quality measures of the fields used in the creation of annual
fire causes and is intended to help track the quality of the data used by USFA to assign these
causes each year.
Figure 3. NFIRS Incident Reporting, Submission, and Dissemination Process
13 Injuries are also an important metric; however, injuries totals are not checked as it is possible to have a large number of
injuries at an incident making checking this field at a national level difficult.
15
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Data Quality and USFA Interaction with States
Three criteria are used in the monitoring of the data in NFIRS during the year: (1) the data are
complete, (2) the data are accurate, and (3) the data are current. These criteria are monitored
by creating reports from the database that show the number of reporting fire departments, the
number of incidents by state, the number of invalid incidents, and the number of unreleased
incidents. The USFA provides the reports to the state NFIRS program managers and work with
them to resolve any data issues. Technical assistance (e.g., telephone support or site visits) is
provided to states to help address any data quality and data reporting needs.
Data quality is an area of great importance. Audits of the data are performed during the year to
identify any inconsistences in the data. The audits focus on three criteria: gaps in reporting,
critical errors in the data, and outliers in the data. In particular, USFA works closely with states
to monitor the quality of data coming from third party vendor software. Each state is
responsible for enforcing that the NFIRS third party software sold by vendors in their state is
compliant with NFIRS standards. USFA will assist states in monitoring vendor data quality issues
or will contact vendors directly to discuss an issue at a state’s request. Other data quality issues
are questionable high dollar loss incidents and questionable high numbers of fire deaths.
Annually, the USFA queries the database for questionable values (i.e., outliers) and verifies the
values with state‐level NFIRS program managers and local‐level NFIRS program managers. The
data quality steps are important to ensure that the data meet USFA’s three criteria before the
data are released in the NFIRS PDR format.
Data Quality Checks by the NFDC
The PDR is further quality checked by the NFDC staff and statisticians upon receipt and prior to
release to the general public. The NFDC staff and statisticians double‐check for such items as:
Missing required modules,
Null values in required data elements,
Invalid values or codes that are not in the current National Fire Incident Reporting
System Complete Reference Guide (CRG),
Large outliers, and
Duplicates in multi‐entry data elements
Any issues are reported to the NFIRS Program Manager who in turn reports them to the NFIRS
support contractor. The support contractor investigates and incorporates any needed changes
and updates into the NFIRS validation rules or the PDR generation procedures. If the record
clearly contains outliers, it is generally USFA’s practice to recommend excluding it from
analyses. Before excluding such records, however, as a data quality check, a quick internet
16
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
search is conducted to see if some unusual fire did occur. As described above, USFA follows up
with the fire department that submitted the incident record for data verification.
Although invalid values and duplicates in multi‐entry fields are reviewed, these values are not
necessarily changed on the PDR as each major data user (e.g., the NFDC itself, CPSC, NFPA and
others) has its own method of analyzing the fire data in the PDR.
Additionally, the NFDC staff and statisticians, in their analytical database based on the PDR
data, check the following:
Verify all incidents are version 5.0;
Check for null or missing values in Property Use when:
o mutual or automatic aid is received,
o other aid is given, or
o no aid is given or received.
If null values occur, the incidents are deleted from the analytic database; and
For fires, check for null values in the AID field. If null/missing values occur, the incidents
are also deleted from the analytic database.
USFA’s analytic contractor performs additional data quality checks as it loads the PDR data into
its NFIRS database. The analytic contractor checks for:
Potential duplicate entries—entries whose unique identifying information is identical
save for a leading or trailing blank or zero, and
Orphan records—records in the secondary files that do not have a parent record in the
main data file.
Any potential duplicate entries and orphan records are reported back to the NFDC and to the
NFIRS Program Manager. The NFDC recommends if records under either of these instances
should be deleted from the NFDC analytic database. Major data users are notified of the NFDC
decisions and may or may not implement them in their own datasets.
KEY DATA CONSIDERATIONS FOR THE USER
Unknown Entries
In a small number of incident or casualty reports sent to NFIRS, many data items are either not
reported (i.e., null, blank or missing values) or are reported as “unknown”. For most coded
fields, ‘U’, ‘UU’, and ‘UUU’ are codes available in NFIRS 5.0 to indicate “unknown”. In some
cases, even after the best attempts by fire investigators to document the fire, the information is
17
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
truly unknown. In other cases, the information reported as unknown in the initial NFIRS report
is not updated after the fire investigation is completed.
In analyses, the unknowns should not be ignored. The approach taken by USFA in presenting
the data is to provide not only the “raw” percentages of each category, but also the “adjusted”
percentages computed using only those incidents for which data were provided.
Null and blank values differ from entries coded as unknown. Null and blank values are
considered unreported data and differ in meaning and substance from “unknown” data. In data
elements where information is required, a null or blank value may invalidate the record.
Unknown entries are of the highest concern for data quality. For some data elements, the
number of incidents with null, blank, or unknown entries can be larger than the number of
incidents for which data were provided. Through the various USFA and NFDC training initiatives
and efforts by various fire organizations (e.g., the National Association of State Fire Marshals
(NASFM)), fire departments are encouraged to reduce the number of unknown entries by fully
documenting the fire incident.
Fires vs NFIRS Record Counts
It is important to underscore that the raw NFIRS record counts are not the total number of fires
reported by fire departments each year. NFIRS is a large but not randomly selected sample of
fires reported to fire departments and, as such, analyses address the relative proportions
(percentages) or apply the national estimates methodology to produce estimates of the aspect
of the fire problem that is under analysis.
Counting Fires vs Counting Fire‐Related Statistics
When the data element in question is a fire‐related data element (i.e., captured under the fire
modules—fire, structure fire, wildland, etc.), fires are counted. When the data element in
question is in the casualty modules, casualties are counted. Even the most seasoned fire data
analyst may overlook this distinction from time to time. The proper phrasing of analysis for
casualties counted from fire data elements is “fires with casualties/injuries/deaths” or an
equivalent statement.
Confined vs Nonconfined Fires
Confined fires are generally small, low loss fires and are allowed abbreviated reporting. This
limited reporting can result in an increase in the proportion of null or missing values. USFA
generally separates the analyses into a confined fires version and a nonconfined fires version
and recommends that others do the same. The resulting analysis can be very generic but there
are instances where this is reasonable. While the NFPA survey includes a category for confined
18
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
fires, NFPA does not publish estimates of confined fires. It is unclear what the effect of this has
on estimates derived from NFIRS datasets that include confined fires.
Mutual Aid
Some records in NFIRS refer to aid provided to another fire department, either mutual aid given
to an outside fire service entity upon request of the outside entity or automatic aid given
through mutual‐aid agreements. To isolate individual fire incidents, only records of the primary
fire department are included. This is achieved by excluding records reflecting aid provided, as in
essence, not excluding aid incidents when analyzing incidents may result in the double counting
of those incidents where both the giving and receiving departments report to NFIRS.
This exclusion is also a data quality consideration. The fire department receiving aid is
considered the “owner” of the incident and this fire department is responsible for providing the
incident data. The aid giving department’s incident record is generally only a record of having
given aid and most, if not all, data elements are not required or submitted. Including these
records would result in an unacceptably large number of unknown entries.
Mutual‐aid given incidents are excluded from all analyses with one major exception: when
counting firefighter casualties.
Types of Fires
The general categories of fire incidents are broadly defined by the type of incident, with the
four major incident types of structure, vehicle/mobile properties, outside, and other. Structure
fires are further broken down into residential and nonresidential structure fires based on the
property definitions (see below) as well as by type of structure – building and nonbuilding fires.
Type of incident, property use, and type of structure (for structure fires) are required elements
with near 100 percent compliance.
Property Definitions
The general categories of property use are defined by the property use data element. Property
use is a required field; blank and null values are not expected (but do occur occasionally under
specific rare circumstances14). Null values in property use receive special treatment as the
requirements for certain types of incidents changed in 2006. These specifications are discussed
more fully in USFA’s National Fire Incident Reporting System Version 5.0 Fire Data Analysis
Guidelines and Issues. These guidelines discuss specific data elements and how USFA analyzes
and interprets the results of the analyses.
14 The entry rules that allow blank and null property types to occur will be eliminated as of January 2015.
19
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Multi‐year and Trend Analyses
It is important to note that NFIRS data may fluctuate from year to year, resulting in variability. It
is possible that any given year may be an anomalous year for a subset of fire data or for the
data overall. Statistically rare, but real‐world incidents do occur. Large conflagrations such as
the various California wildfires, large petrochemical plant fires such as the 1989 Houston Ship
Channel fire, and large multi‐fatality, multi‐injury fires such as the 1980 MGM Grand Hotel fire
or the 2003 fire at The Station Nightclub can have one‐time effects on fire analyses for that
year.
For these reasons—yearly fluctuation and single‐event spikes—it is often preferable to
aggregate several years’ data for analyses. USFA uses three‐year averaged data and, where
possible, analyzes trends of five or more years’ data. Trends are usually described by the
change in the linear best fit. Moving averages are another type of trend analyses available.
Cause
The cause of a fire is often a complex chain of events. To make it easier to grasp the “big
picture”, USFA originally developed a cause hierarchy for structure fires, where the majority of
fire losses occur. 15 The cause for other incident types is based on the distributions of the NFIRS
cause of ignition data element. This data element captures a very broad sense of the cause of
the fire.
The hierarchy schema provides three levels of cause descriptions: a set of more detailed causes,
a set of mid‐level causes, and a set of high‐level causes. The mid‐level categories of fire causes
such as heating, cooking and playing with heat source are used by the USFA. Fires are assigned
to one of the 16 mid‐level cause groupings using a hierarchy of definitions, as shown in Table 3.
A fire is included in the highest category into which it fits on the list. If it does not fit the top
category, then the second one is considered, and if not that one, the third and so on.
In principle, it is the cause of the fire which results in deaths, injuries and dollar loss that should
be analyzed, not numbers of deaths and injuries associated with fire causes.
Table 3. Mid‐Level Structure Fire Cause Groupings
Cause Category
Exposure
Intentional
Cause under investigation
Definition
Caused by heat spreading from another hostile fire.
Cause of ignition is intentional or fire is deliberately set.
Cause is under investigation and a valid NFIRS Arson Module is
present.
15 The structure fire cause hierarchy and specific definitions in terms of the NFIRS 5.0 codes may be found at http://www.usfa.
fema.gov/fireservice/nfirs/tools/fire_cause_category_matrix.shtm. The hierarchy involves a large number of subcategories that
are later grouped into the 16 mid‐level cause categories, then the eight high‐level cause groupings.
20
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Cause Category
Playing with heat source
Natural
Other heat
Smoking
Heating
Cooking
Appliances
Electrical malfunction
Other equipment
Open flame, spark (heat from)
Other unintentional, careless
Equipment misoperation, failure
Unknown
September 2014
Definition
Includes all fires caused by individuals playing with any materials
contained in the categories below as well as fires where the factors
contributing to ignition include playing with heat source. Children
playing with fire are included in this category.
Caused by the sun’s heat, spontaneous ignition, chemicals, lightning,
static discharge, high winds, storms, high water including floods,
earthquakes, volcanic action, and animals.
Includes fireworks; explosives; flame/torch used for lighting; heat or
spark from friction; molten material; hot material; heat from hot, or
smoldering objects.
Cigarettes, cigars, pipes, and heat from undetermined smoking
materials.
Includes confined chimney or flue fire, fire confined to fuel
burner/boiler malfunction, central heating, fixed and portable local
heating units, fireplaces and chimneys, furnaces, boilers, water
heaters as source of heat.
Includes confined cooking fires, stoves, ovens, fixed and portable
warming units, deep fat fryers, open grills as source of heat.
Includes televisions, radios, video equipment, phonographs, dryers,
washing machines, dishwashers, garbage disposals, vacuum cleaners,
hand tools, electric blankets, irons, hairdryers, electric razors, can
openers, dehumidifiers, heat pumps, water cooling devices, air
conditioners, freezers, and refrigeration equipment as source of heat.
Includes electrical distribution, wiring, transformers, meter boxes,
power switching gear, outlets, cords, plugs, surge protectors, electric
fences, lighting fixtures, electrical arcing as source of heat.
Includes special equipment (radar, x‐ray, computer, telephone,
transmitters, vending machine, office machine, pumps, printing press,
gardening tools, agricultural equipment), processing equipment
(furnace, kiln, other industrial machines), service, maintenance
equipment (incinerator, elevator), separate motor or generator,
vehicle in a structure, unspecified equipment.
Includes torches, candles, matches, lighters, open fire, ember, ash,
rekindled fire, backfire from internal combustion engine as source of
heat.
Includes misuse of material or product, abandoned or discarded
materials or products, heat source too close to combustibles, other
unintentional (mechanical failure/malfunction, backfire).
Includes equipment operation deficiency, equipment malfunction.
Cause of fire undetermined or not reported.
Source: USFA.
The percentage of unknown fire causes has seen a steady increase since the introduction of
NFIRS version 5.0. This increase may be due, in part, to the fact that the original cause hierarchy
does not apply as well to NFIRS 5.0 data. 16 While the cause hierarchy was revised to
16 See the full description of the cause hierarchy in Fire in the United States 1995‐2004, 14th edition
21
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
incorporate the 5.0 data as best possible, the result is that where the fit is imperfect, many
incidents are assigned to the unknown cause category.
Further, with the current NFPA 921 Guide for Fire and Explosion Investigations, there may be
reluctance on the part of some fire departments to enter anything but “unknown” in causal
fields unless these elements can be determined with certainty.17 NFPA 921 sets guidelines for
scientific‐based investigation and analysis of fire and explosion incidents and is considered the
foremost guide for rendering accurate opinions as to incident origin, cause, responsibility, and
prevention. The issues surrounding NFPA 921’s role in fire departments’ willingness to
determine fire cause are complex, but if the investigation does not or cannot meet these
guidelines, often the preliminary cause of “under investigation” remains in NFIRS and, under
the cause hierarchy, those fires are assigned to the unknown cause category.
The NASFM recently studied the problem of the large numbers of unknown data in the causal
data elements.18 This report identified five recommendations to solve the issue of unreported
fire causal information that have resonance for data quality in general:
Whether a cause is determined or remains undetermined after investigation, fire
departments must “Close the Loop” by updating the codes in the NFIRS incident report.
Codes need to be updated in the system once a cause is determined. “Under
Investigation” reports should always be revisited and updated after the investigation.
Clear the “Litigation Cloud” by addressing the liability concerns for cause determination
through a multi‐pronged approach. Several approaches to this area are suggested.
Improve training for chiefs, officers and front‐line personnel on the concepts and
reasons behind the need for reporting, as well as how fire incident data can be used to
advance fire prevention and suppression goals is needed.
Improve NFIRS by developing and implementing the next generation of NFIRS – what is
commonly referred to as NFIRS Version 6 – with input from stakeholders who are tasked
with inputting the data at the local level, as well as those who analyze and use the data
at all levels.
Improve quality assurance and quality control in fire incident reporting throughout the
system. Specific systemic changes to improve quality assurances and quality control
include designating a “Data Champion” to be responsible for NFIRS quality
17 National Association of State Fire Marshals Fire Research & Education Foundation, Conquering the “Unknowns” Research and
Recommendations on the Chronic Problem of Undetermined and Missing Data in the Causal Factors Sections of the National Fire
Incident Reporting System, final report for Award No. EMW‐2011‐FP‐00356 Assistance to Firefighters Fire Prevention & Safety
Grant Program Federal Emergency Management Agency. Undated.
http://www.firemarshals.org/pdf/NASFMFoundationFinalReportConqueringtheUnknowns.pdf.
18 Ibid.
22
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
control/quality assurance at the local level; providing a mechanism for departments to
report “no incidents” periodically; emphasizing the importance of dedicated State NFIRS
Program Managers to work with departments in their state; adopting a Standard
Operation Procedure or Standard Operating Guideline (SOP/SOG) on completing
incident reports; and revitalizing the National Fire Information Council (NFIC) with a
focus on developing strategies and training to improve the quality of the nation’s NFIRS
data.
The NASFM report notes that there will always be fires whose cause cannot legitimately be
determined even after investigation but that for many incidents, however, definite steps can be
taken toward reducing the level of “undetermined” or unreported responses in the causal
factors section of NFIRS. “To achieve this will require confronting some difficult, thorny issues
that do not have clear solutions. But, if you don’t write it down, it didn’t happen, and we may
never be able to quantify what has been lost by not having sufficient data on the causes of
fires.”19
Smoke Alarms and Smoke Alarm Performance
Smoke alarm data are reported at the fire incident level (not the casualty level). Smoke alarm
data are analyzed for presence, operation, and effectiveness; other smoke alarm data elements
are not analyzed at this time. Smoke alarm performance is analyzed for nonconfined fires only;
confined fires have abbreviated reporting and the various smoke alarm data are not required.
From a data quality perspective, including confined fire smoke alarm data (which is mostly null)
with the more robust nonconfined fire smoke alarm data degrades the data quality.
Dollar Loss Data
It is difficult to estimate dollar loss. Insurance claim data are generally not available to the
public and fire departments rarely have the time to research the actual value of a property. It is
not unusual for the property or contents loss to be unreported. Even when reported, there are
often inconsistencies. For example, there are many reported fires where the fire spread
suggests damage but property loss is not reported or seems low (or high) for the extent of
reported fire spread.
Structures, Buildings, and Nonbuildings
NFIRS 5.0 allows for the differentiation of structures between buildings and nonbuildings. In
NFIRS, a structure is a built object and can include nonbuildings such as platforms, tents,
19 Ibid.
23
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
connective structures such as bridges or fences, telephone poles, and various other structures
in addition to buildings.
Structures are split into building and nonbuilding structures for purposes of data quality. While
most structures are buildings (analyses of NFIRS structure fires between 2009 and 2011 show
that 93 percent of structure fires occur in buildings), the distinction between buildings and
nonbuildings is particularly important when determining the effectiveness of non‐behavior‐
based fire safety mechanisms such as smoke alarms and residential sprinklers. These important
components of early fire detection apply to buildings and not necessarily to these other types
of structures.
NFIRS DATA QUALITY
Two major assessments are used by USFA when monitoring the quality of the incident data
reported by the participating states: the overall quality of the data submitted by the state as a
unit and the overall quality of key data elements across all participating states and fire
departments.
State‐based Data Quality
Based on USFA’s state data quality rating, 12 states – Nebraska, Massachusetts, Alaska, North
Dakota, Montana, Idaho, South Dakota, Utah, Colorado, California, Florida, and Washington –
and the District of Columbia have demonstrated consistent high quality data (Figure 4).
Nebraska and the District of Columbia have vied for the top‐ranked state for data quality over
the 2009‐2011 period.
As a group, the states with the best data quality ratings relative to the national average are
those in the central Midwest through the Northwest and California, with the mid‐Atlantic and
southeastern states (with the exception of Florida) having the lowest ratings (Figure 5). The
reasons for this consistent pattern are not clear.
24
6.00
6.0
5.00
3.0
4.00
0.0
3.00
‐3.0
2.00
‐6.0
1.00
‐9.0
0.00
Relative Data Quality
9.0
September 2014
USFA Response to the OMB 2013
2013 NFIRS Terms of Clearance
Figure 4. 3‐Year NFIRS Data Quality by State Relative to
National Average Data Quality Measure, 2009‐2011
Delaware
New Jersey
Maryland
Louisiana
Hawaii
Arizona
Connecticut
Alabama
Kentucky
Indiana
Rhode Island
Pennsylvania
Mississippi
Texas
Missouri
South Carolina
Georgia
Ohio
Arkansas
Nevada
New Hampshire
West Virginia
North Carolina
Native American Tribal Authority
Oklahoma
New Mexico
New York
Tennessee
Wisconsin
Minnesota
Oregon
Iowa
Virginia
Illinois
Vermont
Wyoming
Maine
Michigan
Kansas
Washington
Florida
California
Colorado
Utah
South Dakota
Idaho
Montana
Alaska
Massachusetts
North Dakota
Nebraska
District of Columbia
25
USFA Response to the OMB 2013
2013 NFIRS Terms of Clearance
September 2014
Figure 5. Annual NFIRS Data Quality by State
Relative to National Average Data Quality Measure, 2009‐2011
2009
Good Average Poor
2010
Good Average Poor
2011
Good Average Poor
26
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Data Quality of Key Data Elements
Although NFIRS contains hundreds of data elements, only a few are used in producing USFA’s
topical and analytic reports. Most of the elements used in USFA’s analyses are required to be
completed for each fire incident type. For small confined fires, outside rubbish fires with no
value, and other unclassified fires, however, only the most basic incident information is
required.
A complete list of NFIRS data elements is documented in the NFIRS 5.0 Complete Reference
Guide.20 Table 4 identifies the NFIRS data elements that are used most often in fire data
analyses produced by USFA. Not all types of fires require the same data elements. Table 4
displays the data elements and the data element’s required completion by type of fire. Table 5
identifies the NFIRS data elements that are used most often in the analyses of casualties
produced by USFA. Because of the limited reporting required of confined fires and other
unclassified fires, these fires are not included here or in the data quality tables that follow.
Incident type, incident date, and alarm time are integral elements of the incident and are not
subject to quality review: either these elements exist and are within valid ranges or the incident
record is not accepted into the national fire database. Similarly, deaths, injuries, and contents
and dollar loss can only be “sanity checked” as there are not specific and definable correct
entries, only “reasonable” ones. These elements are shown in the detailed tables as
distributions. Contents loss is only shown for buildings and mobile property structure fires.
Four types of fires are defined for the purposes of data quality analysis. These fire types are
largely based on the NFIRS modules that departments are required to complete.
Nonbuilding structure fires (also called ‘special structure’ fires)—fires in or of structures
that are not buildings or used as buildings, e.g. bridges or fences. For nonbuilding
structure fires, the fire module and the first element in the structure fire module are
required.
Buildings and mobile property structures—fires in structures that are buildings or in
mobile properties are used as structures, e.g. manufactured or mobile homes. For
buildings and mobile property structure fires, both the fire and structure fire modules
are required.
Vehicle fires—fires in vehicles or other mobile property, e.g., trucks, planes, trains. For
vehicle fires, the fire module is required.
20 “NFIRS 5.0 Complete Reference Guide,” USFA, January 2013: http://www.nfirs.fema.gov/documentation/reference/.
27
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Outside fires—fires that occur outdoors that may be open fires, grass fires, crop fires,
other vegetation fires, and the like. For outside fires, the fire module or the wildland fire
module is required.
Table 4. NFIRS Data Elements Used in USFA Analyses (Fires)
Data Element Description
Incident Type
Property Use
Incident Date
The actual situation found on scene
when emergency personnel arrived.
The actual use of the property where
the incident occurred, not the overall
use of mixed use properties of which
the property is part.
The month, day, and year of incident.
The actual month, day, year, and time
of day (hour, minute and seconds)
Alarm Time
when the alarm was received by the
fire department.
A civilian fire death resulting from the
incident or during the mitigation of the
incident (includes emergency
Deaths
personnel who are not part of the fire
department, such as police officers or
utility workers).
A civilian fire injury resulting from the
incident or during the mitigation of the
incident (includes emergency
Injuries
personnel who are not part of the fire
department, such as police officers or
utility workers).
Property Loss The total property dollar loss.
Contents Loss The total property contents dollar loss.
Area of Fire
The primary use of the area where the
Origin
fire started within the property.
The source of heat that ignited the
Heat Source
Item First Ignited to cause the fire.
The use or configuration of the item or
material first ignited by the heat
Item First
source. The item that had sufficient
Ignited
volume or heat intensity to extend to
uncontrolled or self‐perpetuating fire.
Required Data Element
Buildings
All
Nonbuilding and Mobile
Vehicles
Firesa Structures
Property
Structuresb
Outsidec
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Nod
Yes
Yes
Yes
Yes
Nod
Yes
Yes
Yes
Yes
Nod
Yes
Yes
Yes
Yes
28
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Required Data Element
Buildings
All
Nonbuilding and Mobile
Vehicles
Firesa Structures
Property
Structuresb
Data Element Description
Cause of
Ignition
Factors
Contributing
to Ignition
Equipment
Involved in
Ignition
Fire Spread
Presence of
Detectors
Detector
Operation
Detector
Effectiveness
Presence of
Automatic
Extinguishing
System (AES)
September 2014
General description of why the heat
source and the combustible material
were able to combine to initiate the
fire. This is the best determination of
the firefighter at the scene and may be
changed later as a result of further
investigation or other information.
The contributing factors that allowed
the heat source and combustible
material to combine to ignite the fire.
The piece of equipment that provided
the principal heat source to cause the
ignition if the equipment
malfunctioned or was used improperly.
The extent of fire spread in terms of
how far the flame damage extended.
The existence of fire detection
equipment within its designed range of
the fire.
The operation and effectiveness of the
detector relative to the area of fire
origin.
The effectiveness of the fire detection
equipment in alerting occupants.
The existence of an AES within the
AES’s designed range of a fire.
Outsidec
Nod
Yes
Yes
Yes
Yes
Nod
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
Yes
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Yes
No
No
Source: NFIRS 5.0 Complete Reference Guide, USFA, January 2013.
Notes: a. All Fires: Includes small confined fires, outside rubbish fires with no value, and other unclassified fires which have less stringent
reporting requirements.
b. Buildings and Mobile Property Structures: Does not include confined building fires (small fires confined to the object of origin)
as for these fires only the most basic information about the incident is required.
c. Outside: Does not include outside rubbish fires (fires with no value) or other, unspecified fires as for these fires only the most
basic information about the incident is required.
d. Area of Fire Origin, Heat Source, Item First Ignited, Cause of Ignition, and Factors Contributing for ignition are not required for
small confined fires, outside rubbish fires with no value, and other unclassified fires as these types of fires have less stringent
reporting requirements.
Table 5. NFIRS Data Elements Used in USFA Analyses (Civilian Casualties)
Data Element
Description
Gender
Age or Date of Birth
The gender of the injured person.
The casualty’s age in years or, if the casualty is an infant, the age
in months OR the month, day, and year of birth of the casualty.
Required
Data
Yes
Yes
29
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Race
September 2014
Description
The identification of the race of the casualty based on U.S. Office
of Management and Budget (OMB) designations.
Ethnicity
Identifies the ethnicity of the casualty. Ethnicity is an ethnic
classification or affiliation. Ethnicity designates a population
subgroup having a common cultural heritage, as distinguished by
customs, characteristics, language, or common history, amongst
other attributes. Currently, Hispanic/Latino is the only OMB
designation for ethnicity.
Severity
The relative severity or seriousness of the injury on a scale from
“least serious” (minor) to “most serious” (death).
Cause of Injury
The physical event that caused the injury.
Human Factors
The physical or mental state of the person before becoming a
Contributing to Injury casualty.
Factors Contributing The most significant factors contributing to the injury of the
to Injury
casualty.
Activity When Injured The action or activity in which the person was engaged at the
time of the injury.
Primary Apparent
The casualty’s most serious apparent injury.
Symptom
Primary Area of Body The part of the body that sustained the most serious injury.
Injured
Required
Data
No
No
Yes
No
No
No
No
No
No
Source: “NFIRS 5.0 Complete Reference Guide”, USFA, January 2013.
Note:
In the NFIRS paper forms and the NFIRS CRG, the full data element name is Age or Date of Birth; in the civilian
fire casualty file in the PDR, the data element is Age. All other data elements retain their names in the PDR.
NFIRS Data Element Quality
In the sections that follow, data elements typically used in USFA analyses are assessed for
overall data quality and usefulness for the period 2009‐2011. The first section is an overall
summary of key data elements. The second section presents the detailed assessment by year,
type of fire, and data element of the data quality and data usability of data elements commonly
used in USFA’s NFIRS data analyses
Summary Tables
In the summary tables that follow, data elements typically used in USFA analyses are assessed
for overall data quality and usefulness for the period 2009‐2011. The summary tables are
organized as follows:
Table 6, by type of fire – nonbuilding structures, buildings and mobile property
structures, vehicles, and outside fires. Detailed breakouts of these summary tables by
type of fire and year are found in Table 9 through Table 12.
Table 7, fires with casualties – by type of casualty (deaths, injuries) and general type of
fire (all fires or building and mobile property fire, where the majority of fire casualties
30
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
occur). Detailed breakouts of these summary tables by type of fire and year are found in
Table 13, Table 15, Table 17, and Table 19. And lastly,
Table 8, by type of casualty and general type of fire (all fires or building and mobile
property fire, where the majority of fire casualties occur). Detailed breakouts of these
summary tables by type of fire and year are found in Table 23 through Table 26.
Data quality for each data element is indexed on the total proportion of valid entries for that
element. Data usability for each data element is indexed on the proportion of valid known
entries for that element. Both have a maximum value of 100. Ranges of these indices are shown
rather than averages. It is possible that a data element has a high data element quality index
but a moderate data element usability index – while entries are valid and complete, the
difference in the indices reflects the amount of data coded as unknown.
Required data elements, shown in Table 4 above, have an overall data quality index that is
either a perfect value of 100 or very near. The exceptions are outside fires, all fires with injuries,
and the presence of detectors and automatic extinguishing systems in fires other than buildings
and mobile property structures. Incident data for many types of outside fires can be reported
via the wildland module in lieu of the fire module. The wildland module does not have the full
set of required data elements contained in the fire module. As a result, many required fire
module data elements have no entries. The overall result is a lower data quality index.
A similar situation occurs with the data quality for fires with injuries. Fire injuries occur
frequently and across all types of fires. The data quality index is in the low 90s because of the
number of injuries that occur in outside and other fires where alternate reporting methods are
allowed and reporting requirements are somewhat less stringent. In addition, determining
many of the fire‐related data elements for outside fires is difficult because of the nature of the
fires.
The lower data quality index for the presence of detectors and automatic extinguishing systems
in fires other than buildings and mobile property structures is a result of deaths and injuries in
incidents, such as vehicle and outside fires, where the structure module is not required (these
data variables only apply to buildings and mobile structures).
31
USFA Response to the OMB 2013
2013 NFIRS Terms of Clearance
September 2014
Table 6. Data Element Quality and Usability Summary for Common NFIRS Data Elements
Type of Fire, 2009‐2011
Type of Fire
Data Element
Nonbuilding Structure
Buildings and Mobile
Property Structure
Vehicle
Outside
Quality
Usability
Quality
Usability
Quality
Usability
Quality
Usability
Index Range Index Range Index Range Index Range Index Range Index Range Index Range Index Range
Property Use
Area of Fire Origin
Heat Source
Item First Ignited
Cause of Ignition
Factors Contributing to
Ignition
Equipment Involved in
Ignition
Fire Spread
Presence of Detectors
Detector Operation
Detector Effectiveness
Presence of Automatic
Extinguishing System (AES)
100
100
100
100
100
99.2 ‐ 99.5
84.6 ‐ 84.9
64.6 ‐ 65
64.1 ‐ 64.6
82.7 ‐ 84.2
100
100
100
100
100
99.3 ‐ 99.4
86.3 ‐ 86.5
62.8 ‐ 63.7
62.6 ‐ 63.3
86.4 ‐ 86.6
100
100
100
100
100
98.4 ‐ 98.6
86.2 ‐ 86.3
49.4 ‐ 49.6
41.6 ‐ 43.5
77.7 ‐ 78
100
77.4 ‐ 79.8
90.1 ‐ 91.1
77.4 ‐ 79.7
99.9 ‐ 100
98.1 ‐ 98.5
63.4 ‐ 65.7
45.2 ‐ 47.7
50.7 ‐ 53.8
68.6 ‐ 69.6
100
73.6 ‐ 74.4
100
73 ‐ 73.4
100
70.1 ‐ 70.2
87.5 ‐ 88.6
64.7 ‐ 65.5
36.1 ‐ 40.9
–
–
–
–
35.4 ‐ 40.2
–
–
–
–
41.6 ‐ 42.7
94.7 ‐ 95.4
95.4 ‐ 95.6
100
100
40.3 ‐ 41.4
94.7 ‐ 95.4
68.7 ‐ 69.2
81.3 ‐ 83.2
90.2 ‐ 90.5
27.9 ‐ 29.3
–
–
–
–
27.6 ‐ 28.9
–
–
–
–
26.4 ‐ 27.2
–
–
–
–
25.3 ‐ 26.3
–
–
–
–
–
–
95.1 ‐ 95.2
86.6 ‐ 86.8
–
–
–
–
32
USFA Response to the OMB 2013
2013 NFIRS Terms of Clearance
September 2014
Table 7. Data Element Quality and Usability Summary for Common NFIRS Data Elements
Type of Fire with Deaths or Injuries, 2009‐2011
Type of Fire
Fatal Fires
Data Element
Fatal Fires in Buildings and
Mobile Property Structure
Fires with Injuries
Buildings and Mobile
Property Structure Fires
with Injuries
Quality
Usability
Quality
Usability
Quality
Usability
Quality
Usability
Index Range Index Range Index Range Index Range Index Range Index Range Index Range Index Range
Property Use
Area of Fire Origin
Heat Source
Item First Ignited
Cause of Ignition
Factors Contributing to
Ignition
Equipment Involved in
Ignition
Fire Spread
Presence of Detectors
Detector Operation
Detector Effectiveness
Presence of Automatic
Extinguishing System (AES)
100
99.2 ‐ 99.5
99.1 ‐ 99.5
99.1 ‐ 99.5
99.1 ‐ 99.5
99.4 ‐ 99.8
79.1 ‐ 79.5
43.3 ‐ 44.3
43.1 ‐ 45.3
85.6 ‐ 87
100
100
100
100
100
99.8 ‐ 99.8
79.9 ‐ 80.4
42.9 ‐ 43.3
42.3 ‐ 45.2
86.1 ‐ 86.8
100
90.6 ‐ 91.9
90.6 ‐ 91.9
90.6 ‐ 91.9
90.6 ‐ 92
99.5 ‐ 99.6
85.9 ‐ 86.8
66.2 ‐ 67.7
65.4 ‐ 66.1
83.2 ‐ 84.8
100
100
100
99.9 ‐ 100
100
99.4 ‐ 99.6
94 ‐ 94.4
71.4 ‐ 72.4
70.6 ‐ 71.6
91.9 – 92
99.2 ‐ 99.5
54.5 ‐ 56.4
100
51.4 ‐ 54.9
90.6 ‐ 91.9
70.2 ‐ 71.4
100
75.4 ‐ 77.1
36.5 ‐ 38.8
78.5 ‐ 79.2
78.7 ‐ 79.5
100
100
34 ‐ 36.6
78.5 ‐ 79.2
46.6 ‐ 48.7
61.3 ‐ 65.5
65.8 ‐ 68.4
38.8 ‐ 42
97.9 ‐ 98.2
98.2 ‐ 98.4
100
62 ‐ 71.2
35.9 ‐ 39.3
97.9 ‐ 98.2
58.3 ‐ 60.1
61.4 ‐ 65.5
42.3 ‐ 46.8
43.3 ‐ 44.5
71 ‐ 74.6
71.3 ‐ 74.7
100
100
42.1 ‐ 43.6
71 ‐ 74.6
54.5 ‐ 58.2
81 ‐ 85.5
87.8 ‐ 88.9
48.3 ‐ 49.4
98.6 ‐ 98.7
98.7 ‐ 98.8
100
100
46.8 – 48
98.6 ‐ 98.7
75.3 ‐ 77.1
80.3 – 85
87.3 ‐ 88.5
78.5 ‐ 79.2
71.6 ‐ 72.3
97.8 ‐ 98.3
89.4 ‐ 90.1
71 ‐ 74.3
68.7 ‐ 71.4
98.1 ‐ 98.5
94.4 ‐ 95.5
33
USFA Response to the OMB 2013
2013 NFIRS Terms of Clearance
September 2014
Table 8. Data Element Quality and Usability Summary for Common NFIRS Data Elements
Type of Fire with Deaths or Injuries, 2009‐2011
Type of Casualty
Data Element
Deaths in All Fires
Deaths in Buildings and
Mobile Property Structures
Injuries in All Fires
Injuries in Buildings and
Mobile Property Structures
Quality
Usability
Quality
Usability
Quality
Usability
Quality
Usability
Index Range Index Range Index Range Index Range Index Range Index Range Index Range Index Range
Gender
Age or Date of Birth
Race
Ethnicity
Severity
Cause of Injury
Human Factors Contributing to
Injury
Factors Contributing to Injury
Activity When Injured
Primary Apparent Symptom
Primary Area of Body Injured
100
97.7 ‐ 98.5
62.3 ‐ 64.6
39.4 ‐ 42.9
100
70.1 ‐ 72.3
100
97.7 ‐ 98.5
58.4 ‐ 60.6
39.4 ‐ 42.9
100
57.3 ‐ 60.2
100
98.4 ‐ 98.8
64.8 ‐ 66.5
41 ‐ 44.5
100
70.4 ‐ 73.1
100
98.4 ‐ 98.8
61.8 ‐ 63.4
41 ‐ 44.5
100
57.9 ‐ 60.9
100
97.7 ‐ 98.2
57.2 ‐ 60
39.6 ‐ 42
100
74.4 ‐ 75.7
100
97.7 ‐ 98.2
53.7 ‐ 56.4
39.6 ‐ 42
95.9 ‐ 96.3
71.7 ‐ 73.5
100
97.6 ‐ 98
56.7 ‐ 60
39.5 ‐ 42.1
100
73.6 ‐ 75.8
100
97.6 ‐ 98
53.1 ‐ 56.4
39.5 ‐ 42.1
95.7 ‐ 96.3
70.8 ‐ 73.4
53.4 ‐ 55.4
43.1 ‐ 48
53.7 ‐ 54.4
53.3 ‐ 54.9
47.8 ‐ 50.1
53.4 ‐ 55.4
43.1 ‐ 48
31.1 ‐ 32.8
45 ‐ 45.7
47.8 ‐ 50.1
53.4 ‐ 57.6
41.2 ‐ 47.7
55 ‐ 56.2
54.1 ‐ 57.2
47.3 ‐ 51.7
53.4 ‐ 57.6
41.2 ‐ 47.7
31.7 ‐ 34.4
46.3 ‐ 48.1
47.3 ‐ 51.7
59.5 ‐ 60.8
49.4 ‐ 52.3
65.8 ‐ 67.4
64.8 ‐ 68.8
58.1 ‐ 60.6
59.5 ‐ 60.8
49.4 ‐ 52.2
58.1 ‐ 59.9
63.7 ‐ 67.6
58.1 ‐ 60.6
58.5 ‐ 60.2
48.6 ‐ 51.5
65 ‐ 67.8
64.9 ‐ 69.6
57.9 ‐ 60.7
58.5 ‐ 60.2
48.6 ‐ 51.4
57.6 ‐ 60.5
63.7 ‐ 68.2
57.9 ‐ 60.6
Note: Age or Date of Birth–In the NFIRS paper forms and the NFIRS CRG, the full data element name is Age or Date of Birth; in the civilian fire casualty file in the PDR, the data
element is Age.
34
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Detailed Data Quality and Usability of Commonly Used Data Elements in Analyses
The following tables present the detailed assessment of data quality and data usability of data
elements commonly used in USFA’s NFIRS data analyses. For coded data elements, four data
quality measures are assessed based on the proportion of fires that fall under the measure for
the data element in question. These measures are:
Valid Known – the number/percent of data element entries that are valid according to
the NFIRS CRG,
Valid Unknown – the number/percent of data element entries that are coded as
unknown (“U”, “UU”, or “UUU”), where unknown is a valid entry according to the NFIRS
CRG,
Invalid – the number/percent of data element entries whose entries are not valid
according to the NFIRS CRG, and
Null or No Entry – the number of entries where no data has been entered whether or
not the data element is required to be completed.
The two quality indices, data element quality and data element usability, are also included in
each these tables. As noted earlier, it is possible that a data element with a high data quality
index may have a moderate, or even low, usability index. For example, in Table 9 the data
element “heat source” has a data quality index of 100 – all entries in heat source for
nonbuilding fires are valid entries, whether the actual heat source is explicitly specified or it is
coded as “unknown”. The usability index, however, is 64.6 because a very large proportion of
those entries are coded as “unknown”, making definitive analyses using this data element
problematic.
For data elements that have direct entry, specifically dollar losses and numbers of casualties,
distributions of entries are shown.
Table 9 through Table 12 are detailed breakouts by year of common NFIRS data elements by
the four major types of fire. These tables are also summarized in Table 6 above.
Table 13 through Table 20 are paired tables of detailed breakouts by year of common NFIRS
data elements for fires with casualties by type of casualty (deaths, injuries) and general type of
fire (all fires or building and mobile property fire). Each detailed table is followed by a
distribution of the number of casualties (e.g., reported deaths or injuries) by year. The detailed
tables are also summarized in Table 7 above.
Table 21 is the distribution of reported property loss by year for each of the four major fire
types.
35
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 22 is the distribution of reported contents loss by year for Building and Mobile Property
Structure Fires, where contents losses are typically found.
Table 23 through Table 26 are detailed breakouts by year of common NFIRS data elements for
reported casualties (deaths, injuries) by general type of fire (all fires or building and mobile
property fire). These tables are also summarized in Table 8 above.
36
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 9. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Nonbuilding Structure Fires, 2009‐2011
Data Element
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
2010
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Number
of
Reported
Fires
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Invalid
Null or
No Entry
8,254
100.0
8,254
100.0
8,254
100.0
8,254
100.0
8,254
100.0
8,254
100.0
8,254
100.0
8,213
99.5
6,979
84.6
5,334
64.6
5,330
64.6
6,825
82.7
6,072
73.6
3,320
40.2
41
0.5
1,275
15.4
2,920
35.4
2,924
35.4
1,429
17.3
2,179
26.4
60
0.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
4,874
59.1
9,381
100.0
9,381
100.0
9,381
100.0
9,381
100.0
9,381
100.0
9,381
100.0
9,381
100.0
9,311
99.3
7,947
84.7
6,100.0
65
6,010
64.1
7,841
83.6
6,984
74.4
3,454
36.8
70
0.7
1,434
15.3
3,281
35
3,371
35.9
1,540
16.4
2,397
25.6
88
0.9
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5,839
62.2
9,996
100.0
9,996
100.0
9,996
100.0
9,996
100.0
9,996
100.0
9,918
99.2
8,486
84.9
6,479
64.8
6,422
64.2
8,413
84.2
78
0.8
1,510
15.1
3,517
35.2
3,574
35.8
1,583
15.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Quality
Index
Usability
Index
100
99.5
100
84.6
100
64.6
100
64.6
100
82.7
100
73.6
40.9
40.2
100
99.3
100
84.7
100
65
100
64.1
100
83.6
100
74.4
37.7
36.8
100
99.5
100
84.6
100
64.6
100
64.6
100
82.7
37
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
September 2014
Number
of
Reported
Fires
9,996
100.0
9,996
100.0
Number and Percent of Reported Fires
Valid
Known
7,361
73.6
3,541
35.4
Valid
Unknowna
2,635
26.4
73
0.7
Null or
Invalid
No Entry
0
0
0
0
0
0
6,382
63.8
Quality
Index
Usability
Index
100
73.6
36.1
40.2
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
Table 10. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Buildings and Mobile Property Structures21, 2009‐2011
Data Element
Number
of
Reported
Fires
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Null or
Invalid
No Entry
167,249
100.0
167,249
100.0
167,249
100.0
167,249
100.0
167,249
100.0
167,249
100.0
167,249
100.0
167,249
100.0
167,249
100.0
59,248
100.0
166,258
99.4
144,258
86.3
104,991
62.8
105,813
63.3
144,432
86.4
122,202
73.1
69,307
41.4
158,389
94.7
114,842
68.7
48,140
81.3
991
0.6
22,986
13.7
62,253
37.2
61,431
36.7
22,812
13.6
45,036
26.9
2,216
1.3
0
0
44,631
26.7
11,108
18.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
0
5
0
5
0
5
0
11
0
95,726
57.2
8,860
5.3
7,776
4.6
0
0
34,059
30,726
3,333
0
0
100.0
90.2
9.8
0
0
Quality
Index
Usability
Index
100
99.4
100
86.3
100
62.8
100
63.3
100
86.4
100
73.1
42.7
41.4
94.7
94.7
95.4
68.7
100
81.3
100
90.2
21 Does not include confined building fires (small fires confined to the object of origin) as for these fires, only the most basic
information about the incident is required.
38
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Presence of Automatic
Extinguishing System (AES)
Percent
2010
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
September 2014
Number
of
Reported
Fires
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Null or
Invalid
No Entry
167,249
144,790
14,134
0
8,325
100.0
86.6
8.5
0
5
187,711
100.0
187,711
100.0
187,711
100.0
187,711
100.0
187,711
100.0
187,711
100.0
187,711
100.0
187,711
100.0
187,711
100.0
68,466
100.0
186,469
99.3
162,375
86.5
119,575
63.7
118,632
63.2
162,602
86.6
137,805
73.4
75,692
40.3
178,380
95
129,825
69.2
56,956
83.2
1,242
0.7
25,327
13.5
68,127
36.3
69,062
36.8
25,092
13.4
49,893
26.6
2,382
1.3
0
0
49,128
26.2
11,507
16.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
9
0
17
0
17
0
13
0
109,637
58.4
9,331
5
8,758
4.7
3
0
39,608
35,836
3,770
0
2
100.0
90.5
9.5
0
0
187,711
162,868
15,551
0
9,292
100.0
86.8
8.3
0
5
188,976
100.0
188,976
100.0
188,976
100.0
188,976
100.0
188,976
100.0
188,976
100.0
187,852
99.4
163,235
86.4
119,536
63.3
118,283
62.6
163,557
86.5
137,993
73
1,123
0.6
25,740
13.6
69,439
36.7
70,692
37.4
25,418
13.5
50,979
27
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
1
0
1
0
4
0
Quality
Index
Usability
Index
95.1
86.6
100
99.3
100
86.5
100
63.7
100
63.2
100
86.6
100
73.4
41.6
40.3
95
95
95.4
69.2
100
83.2
100
90.5
95.1
86.8
100
99.4
100
86.4
100
63.3
100
62.6
100
86.5
100
73
39
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number
of
Data Element
Reported
Fires
Equipment Involved in Ignition
188,976
Percent
100.0
Fire Spread
188,976
Percent
100.0
Presence of Detectors
188,976
Percent
100.0
Detector Operation (when present)
69,324
Percent
100.0
Detector Effectiveness (when
40,302
present and operating)
Percent
100.0
Presence of Automatic
188,976
Extinguishing System (AES)
Percent
100.0
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Null or
Invalid
No Entry
76,291
40.4
180,321
95.4
130,179
68.9
57,689
83.2
2,267
1.2
0
0
50,391
26.7
11,634
16.8
0
0
0
0
0
0
0
0
110,418
58.4
8,655
4.6
8,406
4.4
1
0
36,380
3,922
0
0
90.3
9.7
0
0
163,858
16,131
0
8,987
86.7
8.5
0
4.8
Quality
Index
Usability
Index
41.6
40.4
95.4
95.4
95.6
68.9
100
83.2
100
90.3
95.2
86.7
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
Table 11. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Vehicle Fires, 2009‐2011
Data Element
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
2010
Property Use
Percent
Number
of
Reported
Fires
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Invalid
Null or
No Entry
149,471
100.0
149,471
100.0
149,471
100.0
149,471
100.0
149,471
100.0
149,471
100.0
149,471
100.0
147,006
98.4
128,922
86.3
74,076
49.6
64,987
43.5
116,136
77.7
104,722
70.1
43,260
28.9
2,465
1.6
20,548
13.7
75,394
50.4
84,483
56.5
33,334
22.3
44,744
29.9
585
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
1
0
1
0
5
0
105,626
70.7
155,267
100.0
152,854
98.4
2,412
1.6
1
0
0
0
Quality
Index
Usability
Index
100
98.4
100
86.3
100
49.6
100
43.5
100
77.7
100
70.1
29.3
28.9
100
98.4
40
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
September 2014
Number
of
Reported
Fires
155,267
100.0
155,267
100.0
155,267
100.0
155,267
100.0
155,267
100.0
155,267
100.0
149,333
100.0
149,333
100.0
149,333
100.0
149,333
100.0
149,333
100.0
149,333
100.0
149,333
100.0
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Invalid
Null or
No Entry
133,863
86.2
76,749
49.4
65,395
42.1
120,795
77.8
109,026
70.2
42,792
27.6
21,400
13.8
78,514
50.6
89,859
57.9
34,457
22.2
46,231
29.8
508
0.3
0
0
0
0
0
0
2
0
0
0
0
0
4
0
4
0
13
0
13
0
10
0
111,967
72.1
147,178
98.6
128,838
86.3
73,884
49.5
62,185
41.6
116,504
78
104,707
70.1
42,343
28.4
2,155
1.4
20,495
13.7
75,449
50.5
87,148
58.4
32,829
22
44,625
29.9
456
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
106,534
71.3
Quality
Index
Usability
Index
100
86.2
100
49.4
100
42.1
100
77.8
100
70.2
27.9
27.6
100
98.6
100
86.3
100
49.5
100
41.6
100
78
100
70.1
28.7
28.4
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
41
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 12. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Outside22 Fires, 2009‐2011
Data Element
2009
Property Use
Percent
Area of Fire Originb
Percent
Heat Source
Percent
Item First Ignitedb
Percent
Cause of Ignitionc
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
2010
Property Use
Percent
Area of Fire Originb
Percent
Heat Source
Percent
Item First Ignitedb
Percent
Cause of Ignitionc
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
2011
Property Use
Percent
Area of Fire Originb
Percent
Heat Source
Percent
Number
of
Reported
Fires
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Invalid
Quality
Null or
Index
No Entry
234,258
100.0
234,258
100.0
234,258
100.0
234,258
100.0
234,258
100.0
234,258
100.0
234,258
100.0
229,872
98.1
153,255
65.4
111,835
47.7
125,985
53.8
163,149
69.6
153,410
65.5
59,328
25.3
4,386
1.9
30,992
13.2
100.0,183
42.8
58,262
24.9
71,081
30.3
51,566
22
2,684
1.1
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
50,011
21.3
22,240
9.5
50,011
21.3
28
0
29,282
12.5
172,243
73.5
257,951
100.0
257,951
100.0
257,951
100.0
257,951
100.0
257,951
100.0
257,951
100.0
257,951
100.0
253,969
98.5
169,366
65.7
116,539
45.2
135,430
52.5
176,886
68.6
168,525
65.3
67,968
26.3
3,982
1.5
36,261
14.1
118,306
45.9
70,187
27.2
81,018
31.4
60,224
23.3
2,225
0.9
0
0
0
0
0
0
0
0
1
0
0
0
4
0
0
0
52,324
20.3
23,106
9
52,334
20.3
46
0
29,202
11.3
187,754
72.8
274,379
100.0
274,379
100.0
274,379
100.0
269,651
98.3
174,031
63.4
124,467
45.4
4,727
1.7
38,319
14
122,550
44.7
1
0
0
0
0
0
0
0
62,029
22.6
27,362
10
Usability
Index
100
98.1
78.6
65.4
90.5
47.7
78.7
53.8
99.9
69.6
87.5
65.5
26.4
25.3
100
98.5
79.8
65.7
91.1
45.2
79.7
52.5
100
68.6
88.6
65.3
27.2
26.3
100
98.3
77.4
63.4
90.1
45.4
22 Does not include outside rubbish fires (fires with no value) or other, unspecified fires as for these fires, only the most basic
information about the incident is required.
42
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Item First Ignitedb
Percent
Cause of Ignitionc
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
September 2014
Number
of
Reported
Fires
274,379
100.0
274,379
100.0
274,379
100.0
274,379
100.0
Number and Percent of Reported Fires
Valid
Known
139,019
50.7
188,816
68.8
177,608
64.7
71,994
26.2
Valid
Unknowna
Invalid
73,331
26.7
85,562
31.2
62,812
22.9
2,007
0.7
0
0
0
0
0
0
4
0
Quality
Null or
Index
No Entry
62,029
22.6
1
0
33,959
12.4
200,374
73
Usability
Index
77.4
50.7
100
68.8
87.6
64.7
26.9
26.2
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
b. For vegetation, crop, and other specific types of outside fires, data can be reported via the fire or wildland fire modules. Area of
Fire Origin and Item First Ignited are found in the fire module only.
c. For vegetation, crop, and other specific types of outside fires, data can be reported via the fire or wildland fire modules. The
general cause data element is Cause of Ignition in the fire module and Wildland Fire Cause in the wildland fire module.
Table 13. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Fatal Fires, 2009‐2011
Data Element
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Number
of
Reported
Fires
1,512
100.0
1,512
100.0
1,512
100.0
1,512
100.0
1,512
100.0
1,512
100.0
1,512
100.0
1,512
100.0
1,512
100.0
424
100.0
Number and Percent of Reported Fires
Valid
Known
1,508
99.7
1,199
79.3
670
44.3
685
45.3
1,316
87.0
853
56.4
554
36.6
1,189
78.6
708
46.8
275
64.9
Valid
Unknowna
4
0.3
306
20.2
835
55.2
820
54.2
189
12.5
652
43.1
33
2.2
0
0
486
32.1
149
35.1
Quality
Null or
Index
No Entry
Invalid
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7
0.5
7
0.5
7
0.5
7
0.5
7
0.5
925
61.2
323
21.4
318
21
0
0
Usability
Index
100
99.7
99.5
79.3
99.5
44.3
99.5
45.3
99.5
87
99.5
56.4
38.8
36.6
78.6
78.6
78.9
46.8
100
64.9
43
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
2010
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
September 2014
Number
of
Reported
Fires
170
100.0
1,512
Number and Percent of Reported Fires
Valid
Known
116
68.2
1,083
Valid
Unknowna
54
31.8
105
Quality
Null or
Index
No Entry
Invalid
0
0
0
Usability
Index
0
100
68.2
78.5
71.6
100
99.4
99.2
79.5
99.1
44.3
99.1
43.1
99.1
85.6
99.2
54.5
36.5
34
79.2
79.2
79.5
48.7
100
61.3
100
68.4
79.2
72.3
100
99.8
99.3
79.1
99.4
43.3
99.4
43.1
0
324
100.0
71.6
6.9
0
21.4
1,637
100.0
1,637
100.0
1,637
100.0
1,637
100.0
1,637
100.0
1,637
100.0
1,637
100.0
1,637
100.0
1,637
100.0
481
100.0
1,627
99.4
1,301
79.5
726
44.3
706
43.1
1,402
85.6
892
54.5
557
34
1,296
79.2
798
48.7
295
61.3
10
0.6
322
19.7
897
54.8
917
56
221
13.5
731
44.7
41
2.5
0
0
505
30.8
186
38.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
14
0.9
14
0.9
14
0.9
14
0.9
14
0.9
1,039
63.5
341
20.8
334
20.4
0
0
193
132
61
0
0
100.0
68.4
31.6
0
0
1,637
1,183
113
0
341
100.0
72.3
6.9
0
20.8
1,690
100.0
1,690
100.0
1,690
100.0
1,690
100.0
1,686
99.8
1,337
79.1
731
43.3
728
43.1
4
0.2
342
20.2
948
56.1
951
56.3
0
0
0
0
0
0
0
0
0
0
11
0.7
11
0.7
11
0.7
44
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number
of
Data Element
Reported
Fires
Cause of Ignition
1,690
Percent
100.0
Factors Contributing to Ignition
1,690
Percent
100.0
Equipment Involved in Ignition
1,690
Percent
100.0
Fire Spread
1,690
Percent
100.0
Presence of Detectors
1,690
Percent
100.0
Detector Operation (when present)
478
Percent
100.0
Detector Effectiveness (when
222
present and operating)
Percent
100.0
Presence of Automatic
1,690
Extinguishing System (AES)
Percent
100.0
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Quality
Null or
Index
No Entry
Invalid
1,458
86.3
929
55
606
35.9
1,327
78.5
788
46.6
313
65.5
221
13.1
750
44.4
35
2.1
0
0
542
32.1
165
34.5
0
0
0
0
0
0
0
0
0
0
0
0
11
0.7
11
0.7
1,049
62.1
363
21.5
360
21.3
0
0
146
76
0
0
65.8
34.2
0
0
1,217
112
0
361
72
6.6
0
21.4
Usability
Index
99.4
86.3
99.4
55
38
35.9
78.5
78.5
78.7
46.6
100
65.5
100
65.8
78.6
72
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
Table 14. Data Quality for Common NFIRS Data Elements Used in USFA Analyses
Distribution of Fatal Fires by Reported Deaths, 2009‐2011
Deaths per
Incident
1
2
3
4
5
more than 5
Total Fires
2009
Deaths Percent
1,284
84.9
172
11.4
36
2.4
10
0.7
8
0.5
2
0.1
1,512
100.0
2010
Deaths
1,409
164
37
13
9
5
1,637
Percent
86.1
10.0
2.3
0.8
0.5
0.3
100.0
2011
Deaths Percent
1,480
87.6
148
8.8
36
2.1
13
0.8
7
0.4
6
0.4
1,690
100.0
Source: NFIRS.
45
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 15. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Buildings23 and Mobile Property Structures Fatal Fires, 2009‐2011
Data Element
Number and Percent of Reported Fires
Number of
Quality
Reported Valid
Valid
Null or
Index
Invalid
Fires
Known
Unknowna
No Entry
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when
present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
2010
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
1,206
100.0
1,206
100.0
1,206
100.0
1,206
100.0
1,206
100.0
1,206
100.0
1,206
100.0
1,206
100.0
1,206
100.0
423
1,203
99.8
966
80.1
522
43.3
545
45.2
1,047
86.8
662
54.9
474
39.3
1,183
98.1
705
58.5
274
3
0.2
240
19.9
684
56.7
661
54.8
159
13.2
544
45.1
32
2.7
0
0
481
39.9
149
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
700
58
23
1.9
20
1.7
0
100.0
274
64.8
116
35.2
54
0
0
0
104
100.0
1,206
42.3
1,078
19.7
103
0
0
38
25
100.0
89.4
8.5
0
2.1
1,316
100.0
1,316
100.0
1,316
100.0
1,316
100.0
1,313
99.8
1,058
80.4
567
43.1
557
42.3
3
0.2
258
19.6
749
56.9
759
57.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Usability
Index
100
99.8
100
80.1
100
43.3
100
45.2
100
86.8
100
54.9
42.0
39.3
98.1
98.1
98.4
58.5
100
64.8
62
42.3
97.9
89.4
100
99.8
100
80.4
100
43.1
100
42.3
23 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value) as for these fires, only the most basic information about the incident is required.
46
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when
present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when
present)
Percent
Detector Effectiveness (when
present and operating)
Percent
September 2014
Number and Percent of Reported Fires
Number of
Quality
Reported Valid
Valid
Null or
Index
Invalid
Fires
Known
Unknowna
No Entry
1,316
100.0
1,316
100.0
1,316
100.0
1,316
100.0
1,316
100.0
1,133
86.1
677
51.4
473
35.9
1,289
97.9
791
60.1
183
13.9
639
48.6
38
2.9
0
0
501
38.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
805
61.2
27
2.1
24
1.8
479
294
185
0
0
100.0
61.4
38.6
0
0
294
132
61
0
101
100.0
44.9
20.7
0
34.4
1,316
1,177
110
0
29
100.0
89.4
8.4
0
2.2
1,338
100.0
1,338
100.0
1,338
100.0
1,338
100.0
1,338
100.0
1,338
100.0
1,338
100.0
1,338
100.0
1,338
100.0
1,335
99.8
1,069
79.9
574
42.9
587
43.9
1,157
86.5
700
52.3
493
36.8
1,314
98.2
780
58.3
3
0.2
269
20.1
764
57.1
751
56.1
181
13.5
638
47.7
35
2.6
0
0
536
40.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
810
60.5
24
1.8
22
1.6
476
312
164
0
0
100.0
65.5
34.5
0
0
312
146
76
0
90
100.0
46.8
24.4
0
28.8
Usability
Index
100
86.1
100
51.4
38.8
35.9
97.9
97.9
98.2
60.1
100
61.4
65.6
44.9
97.8
89.4
100
99.8
100
79.9
100
42.9
100
43.9
100
86.5
100
52.3
39.4
36.8
98.2
98.2
98.4
58.3
100
65.5
71.2
46.8
47
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number and Percent of Reported Fires
Number of
Quality
Reported Valid
Valid
Null or
Index
Invalid
Fires
Known
Unknowna
No Entry
Data Element
Presence of Automatic
Extinguishing System (AES)
Percent
1,338
1,205
110
0
23
100.0
90.1
8.2
0
1.7
Usability
Index
98.3
90.1
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
Table 16. Data Quality for Common NFIRS Data Elements Used in USFA Analyses
Distribution of Buildings24 and Mobile Property Structures Fatal Fires by Reported Deaths, 2009‐2011
Deaths per
Incident
1
2
3
4
5
more than 5
Total Fires
2009
2010
Deaths Percent
1015
84.2
141
11.7
33
2.7
10
0.8
5
0.4
2
0.2
1,206
100.0
Deaths
1127
134
31
11
8
5
1,316
Percent
85.6
10.2
2.4
0.8
0.6
0.4
100.0
2011
Deaths Percent
1168
87.3
120
9.0
30
2.2
10
0.7
5
0.4
5
0.4
1,338
100.0
Source: NFIRS.
Table 17. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Fires with Injuries, 2009‐2011
Data Element
Number
Number and Percent of Reported Fires
of
Quality
Valid
Null or
Reported Valid
Index
Invalid
a
Known
Unknown
No Entry
Fires
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
7,599
100.0
7,599
100.0
7,599
100.0
7,599
100.0
7,599
100.0
7,571
99.6
6,526
85.9
5,046
66.4
5,014
66.0
6,322
83.2
28
0.4
356
4.7
1,836
24.2
1,868
24.6
560
7.4
0
0
0
0
0
0
0
0
0
0
0
0
717
9.4
717
9.4
717
9.4
717
9.4
Usability
Index
100
99.6
90.6
85.9
90.6
66.4
90.6
66.0
90.6
83.2
24 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value) as for these fires, only the most basic information about the incident is required.
48
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number
Number and Percent of Reported Fires
of
Quality
Data Element
Valid
Valid
Null or
Reported
Index
Invalid
Known
Unknowna
No Entry
Fires
Factors Contributing to Ignition
7,599
5,419
1,463
0
717
90.6
Percent
100.0
71.3
19.3
0
9.4
Equipment Involved in Ignition
7,599
3,228
99
0
4,272
43.8
Percent
100.0
42.5
1.3
0
56.2
Fire Spread
7,599
5,393
0
0
2,206
71.0
Percent
100.0
71.0
0
0
29
Presence of Detectors
7,599
4,139
1,279
0
2,181
71.3
Percent
100.0
54.5
16.8
0
28.7
Detector Operation (when present)
2,722
2,187
535
0
0
100
Percent
100.0
80.3
19.7
0
0
Detector Effectiveness (when
1,631
1,424
207
0
0
present and operating)
100
Percent
100.0
87.3
12.7
0
0
Presence of Automatic
7,599
5,219
174
0
2,206
Extinguishing System (AES)
71.0
Percent
100.0
68.7
2.3
0
29
2010
Property Use
8,164
8,126
38
0
0
100
Percent
100.0
99.5
0.5
0
0
Area of Fire Origin
8,164
7,012
426
0
726
91.1
Percent
100.0
85.9
5.2
0
8.9
Heat Source
8,164
5,407
2,031
0
726
91.1
Percent
100.0
66.2
24.9
0
8.9
Item First Ignited
8,164
5,336
2,100.0
0
728
91.1
Percent
100.0
65.4
25.7
0
8.9
Cause of Ignition
8,164
6,843
593
0
728
91.1
Percent
100.0
83.8
7.3
0
8.9
Factors Contributing to Ignition
8,164
5,733
1,705
0
726
91.1
Percent
100.0
70.2
20.9
0
8.9
Equipment Involved in Ignition
8,164
3,436
99
0
4,629
43.3
Percent
100.0
42.1
1.2
0
56.7
Fire Spread
8,164
5,959
0
0
2,205
73
Percent
100.0
73
0
0
27
Presence of Detectors
8,164
4,688
1,284
0
2,192
73.1
Percent
100.0
57.4
15.7
0
26.8
Detector Operation (when present)
3,321
2,841
480
0
0
100
Percent
100.0
85.5
14.5
0
0
Detector Effectiveness (when
2,091
1,859
232
0
0
present and operating)
100
Percent
100.0
88.9
11.1
0
0
Presence of Automatic
8,164
5,701
225
0
2,238
Extinguishing System (AES)
72.6
Percent
100.0
69.8
2.8
0
27.4
Usability
Index
71.3
42.5
71.0
54.5
80.3
87.3
68.7
99.5
85.9
66.2
65.4
83.8
70.2
42.1
73
57.4
85.5
88.9
69.8
49
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
September 2014
Number
Number and Percent of Reported Fires
of
Quality
Valid
Valid
Null or
Reported
Index
Invalid
Known
Unknowna
No Entry
Fires
8,121
100.0
8,121
100.0
8,121
100.0
8,121
100.0
8,121
100.0
8,121
100.0
8,121
100.0
8,121
100.0
8,121
100.0
3,347
100.0
8,083
99.5
7,052
86.8
5,498
67.7
5,371
66.1
6,883
84.8
5,802
71.4
3,540
43.6
6,056
74.6
4,724
58.2
2,853
85.2
38
0.5
412
5.1
1,966
24.2
2,093
25.8
581
7.2
1,662
20.5
72
0.9
0
0
1,342
16.5
494
14.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
657
8.1
657
8.1
657
8.1
657
8.1
657
8.1
4,509
55.5
2,065
25.4
2,055
25.3
0
0
2,105
1,871
234
0
0
100.0
88.9
11.1
0
0
8,121
5,802
236
0
2,083
100.0
71.4
2.9
0
25.6
Usability
Index
100
99.5
91.9
86.8
91.9
67.7
91.9
66.1
92
84.8
91.9
71.4
44.5
43.6
74.6
74.6
74.7
58.2
100
85.2
100
88.9
74.3
71.4
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are
distinctly different from blanks and no entries as the lack of information only implies that the information was not given and does
not imply that the element is truly “unknown”.
50
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 18. Data Quality for Common NFIRS Data Elements Used in USFA Analyses
Distribution of Reported Fires with Injuries by Reported Injuries, 2009‐2011
Injuries per
Incident
1
2
3
4
5
more than 5
Total Injuries
2009
2010
Deaths Percent
6,378
83.9
832
10.9
212
2.8
84
1.1
50
0.7
43
0.6
7,599
100.0
Deaths
6,763
973
263
76
37
52
8,164
2011
Percent
82.8
11.9
3.2
0.9
0.5
0.6
100.0
Deaths Percent
6,738
83.0
942
11.6
254
3.1
97
1.2
26
0.3
64
0.8
8,121
100.0
Source: NFIRS.
Table 19. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Buildings25 and Mobile Property Structures Fires with Injuries, 2009‐2011
Data Element
Number
of
Reported
Fires
2009
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Invalid
Quality
Null or
Index
No Entry
5,287
100.0
5,287
100.0
5,287
100.0
5,287
100.0
5,287
100.0
5,287
100.0
5,287
100.0
5,287
100.0
5,287
100.0
2,722
100.0
1,631
5,268
99.6
4,992
94.4
3,795
71.8
3,787
71.6
4,860
91.9
4,076
77.1
2,528
47.8
5,215
98.6
3,983
75.3
2,187
80.3
1,424
19
0.4
295
5.6
1,492
28.2
1,500
28.4
427
8.1
1,211
22.9
87
1.6
0
0
1,242
23.5
535
19.7
207
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2,672
50.5
72
1.4
62
1.2
0
0
0
100.0
87.3
12.7
0
0
Usability
Index
100
99.6
100
94.4
100
71.8
100
71.6
100
91.9
100
77.1
49.4
47.8
98.6
98.6
98.8
75.3
100
80.3
100
87.3
25 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value) as for these fires, only the most basic information about the incident is required.
51
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Presence of Automatic
Extinguishing System (AES)
Percent
2010
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
Equipment Involved in Ignition
Percent
Fire Spread
Percent
Presence of Detectors
Percent
Detector Operation (when present)
Percent
Detector Effectiveness (when
present and operating)
Percent
Presence of Automatic
Extinguishing System (AES)
Percent
2011
Property Use
Percent
Area of Fire Origin
Percent
Heat Source
Percent
Item First Ignited
Percent
Cause of Ignition
Percent
Factors Contributing to Ignition
Percent
September 2014
Number
of
Reported
Fires
5,287
Number and Percent of Reported Fires
Valid
Known
5,047
100.0
95.5
5,759
100.0
5,759
100.0
5,759
100.0
5,759
100.0
5,759
100.0
5,759
100.0
5,759
100.0
5,759
100.0
5,759
100.0
3,123
100.0
Valid
Unknowna
Invalid
161
0
Quality
Null or
Index
No Entry
Usability
Index
79
3
0
1.5
5,729
99.5
5,412
94
4,112
71.4
4,068
70.6
5,297
92
4,342
75.4
2,696
46.8
5,680
98.6
4,443
77.1
2,656
85
30
0.5
347
6
1,647
28.6
1,690
29.3
461
8
1,417
24.6
89
1.5
0
0
1,246
21.6
467
15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
2,974
51.6
79
1.4
70
1.2
0
0
1,946
1,718
228
0
0
100.0
88.3
11.7
0
0
5,759
5,439
213
0
107
100.0
94.4
3.7
0
1.9
5,852
100.0
5,852
100.0
5,852
100.0
5,852
100.0
5,852
100.0
5,852
100.0
5,818
99.4
5,518
94.3
4,237
72.4
4,164
71.2
5,382
92
4,448
76
34
0.6
334
5.7
1,615
27.6
1,688
28.8
470
8
1,404
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
98.5
95.5
100
99.5
100
94
100
71.4
99.9
70.6
100
92
100
75.4
48.3
46.8
98.6
98.6
98.7
77.1
100
85
100
88.3
98.1
94.4
100
99.4
100
94.3
100
72.4
100
71.2
100
92
100
76
52
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number
of
Data Element
Reported
Fires
Equipment Involved in Ignition
5,852
Percent
100.0
Fire Spread
5,852
Percent
100.0
Presence of Detectors
5,852
Percent
100.0
Detector Operation (when present)
3,154
Percent
100.0
Detector Effectiveness (when
1,977
present and operating)
Percent
100.0
Presence of Automatic
5,852
Extinguishing System (AES)
Percent
100.0
Number and Percent of Reported Fires
Valid
Known
Valid
Unknowna
Invalid
Quality
Null or
Index
No Entry
2,808
48
5,778
98.7
4,492
76.8
2,672
84.7
69
1.2
0
0
1,285
22
482
15.3
0
0
0
0
0
0
0
0
2,975
50.8
74
1.3
75
1.3
0
0
1,750
227
0
0
88.5
11.5
0
0
5,539
216
0
97
94.7
3.7
0
1.7
Usability
Index
49.2
48
98.7
98.7
98.8
76.8
100
84.7
100
88.5
98.4
94.7
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are distinctly
different from blanks and no entries as the lack of information only implies that the information was not given and does not imply
that the element is truly “unknown”.
Table 20. Data Quality for Common NFIRS Data Elements Used in USFA Analyses
Distribution of Buildings26 and Mobile Property Structures with Injuries by Reported Injuries, 2009‐2011
Deaths per
Incident
1
2
3
4
5
more than 5
Total Injuries
2009
Deaths Percent
4,272
80.8
672
12.7
184
3.5
76
1.4
42
0.8
41
0.8
5,287
100.0
2010
Deaths
4,599
785
224
71
34
46
5,759
Percent
79.9
13.6
3.9
1.2
0.6
0.8
100.0
2011
Deaths Percent
4,690
80.1
764
13.1
227
3.9
90
1.5
24
0.4
57
1.0
5,852
100.0
Source: NFIRS.
26 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value) as for these fires, only the most basic information about the incident is required.
53
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 21. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Distribution of Reported Property Dollar Loss27, 2009‐2011
Reported Property
Dollar Loss
2009
No Loss Reported
$0
$100 and under
$101 ‐ $500
$501 ‐ $1,000
$1,001 ‐ $5,000
$5,001 ‐ $10,000
$10,001 ‐ $50,000
$50,001 ‐ $100,000
$100,001 ‐ $500,000
$500,001 ‐ $1,000,000
Over $1,000,000
Total
2010
No Loss Reported
$0
$100 and under
$101 ‐ $500
$501 ‐ $1,000
$1,001 ‐ $5,000
$5,001 ‐ $10,000
$10,001 ‐ $50,000
$50,001 ‐ $100,000
$100,001 ‐ $500,000
$500,001 ‐ $1,000,000
Over $1,000,000
Total
All Fires
Nonbuilding
Structure
Fires
Percent
Buildings and
Mobile Property
Structure
Fires
Percent
Vehicle
Fires
Percent
Outside
Fires
Percent
Fires
Percent
150,222
212,895
13,207
22,530
20,650
58,063
23,833
39,070
10,560
7,451
479
272
559,232
26.9
38.1
2.4
4.0
3.7
10.4
4.3
7.0
1.9
1.3
0.1
0.0
100.0.0
2,720
2,289
261
627
420
734
344
555
187
109
5
3
8,254
33.0 30,209
27.7 33,791
3.2
3,777
7.6
9,374
5.1
8,620
8.9 23,317
4.2 12,772
6.7 28,659
2.3
9,339
1.3
6,726
0.1
431
0.0
234
100.0.0 167,249
18.1 77,822
20.2 141,220
2.3
6,284
5.6
3,474
5.2
1,759
13.9
2,341
7.6
554
17.1
567
5.6
111
4.0
93
0.3
17
0.1
16
100.0.0 234,258
33.2 28,475
60.3 35,595
2.7
2,885
1.5
9,055
0.8
9,851
1.0 31,671
0.2 10,163
0.2
9,289
0.0
923
0.0
523
0.0
26
0.0
19
100.0.0 138,475
20.6
25.7
2.1
6.5
7.1
22.9
7.3
6.7
0.7
0.4
0.0
0.0
100.0.0
169,177
233,837
12,776
23,353
21,971
63,488
25,562
40,729
11,003
7,688
507
219
610,310
27.7
38.3
2.1
3.8
3.6
10.4
4.2
6.7
1.8
1.3
0.1
0.0
100.0.0
2,992
2,728
315
731
508
905
330
586
174
103
4
5
9,381
31.9 37,092
29.1 37,750
3.4
4,352
7.8 10,403
5.4
9,906
9.6 26,333
3.5 13,973
6.2 30,604
1.9
9,765
1.1
6,895
0.0
461
0.1
177
100.0.0 187,711
19.8 86,943
20.1 156,731
2.3
5,408
5.5
3,490
5.3
1,665
14.0
2,303
7.4
607
16.3
589
5.2
95
3.7
95
0.2
15
0.1
10
100.0.0 257,951
33.7 42,150
60.8 36,628
2.1
2,701
1.4
8,729
0.6
9,892
0.9 33,947
0.2 10,652
0.2
8,950
0.0
969
0.0
595
0.0
27
0.0
27
100.0.0 155,267
27.1
23.6
1.7
5.6
6.4
21.9
6.9
5.8
0.6
0.4
0.0
0.0
100.0.0
27 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value).
54
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Reported Property
Dollar Loss
2011
No Loss Reported
$0
$100 and under
$101 ‐ $500
$501 ‐ $1,000
$1,001 ‐ $5,000
$5,001 ‐ $10,000
$10,001 ‐ $50,000
$50,001 ‐ $100,000
$100,001 ‐ $500,000
$500,001 ‐ $1,000,000
Over $1,000,000
Total
All Fires
Fires
Percent
173,062
244,170
12,635
22,846
21,427
63,062
25,573
41,009
10,744
7,344
563
249
622,684
27.8
39.2
2.0
3.7
3.4
10.1
4.1
6.6
1.7
1.2
0.1
0.0
100.0.0
September 2014
Nonbuilding
Structure
Fires
3,135
2,910
359
848
534
931
362
594
209
104
7
3
9,996
Percent
Buildings and
Mobile Property
Structure
Fires
Percent
31.4 37,331
29.1 38,763
3.6
4,324
8.5 10,531
5.3
9,876
9.3 26,463
3.6 14,087
5.9 30,863
2.1
9,486
1.0
6,546
0.1
498
0.0
208
100.0.0 188,976
Vehicle
Fires
Outside
Percent
19.8 92,929
20.5 166,270
2.3
5,497
5.6
3,568
5.2
1,909
14.0
2,575
7.5
693
16.3
642
5.0
140
3.5
120
0.3
17
0.1
19
100.0.0 274,379
Fires
Percent
33.9 39,667
60.6 36,227
2.0
2,455
1.3
7,899
0.7
9,108
0.9 33,093
0.3 10,431
0.2
8,910
0.1
909
0.0
574
0.0
41
0.0
19
100.0.0 149,333
Source: NFIRS.
Table 22. Data Quality for Common NFIRS Data Elements Used in USFA Analyses
Distribution of Reported Contents Dollar Loss in Buildings and Mobile Property Structures Fires, 2009‐2011
Reported Contents
Dollar Loss
No Loss Reported
$0
$100 and under
$101 ‐ $500
$501 ‐ $1,000
$1,001 ‐ $5,000
$5,001 ‐ $10,000
$10,001 ‐ $50,000
$50,001 ‐ $100,000
$100,001 ‐ $500,000
$500,001 ‐ $1,000,000
Over $1,000,000
Total
Source: NFIRS.
2010
2009
Fires
30,209
33,791
3,777
9,374
8,620
23,317
12,772
28,659
9,339
6,726
431
234
167,249
Percent
Fires
21.2
30.8
4.1
6.5
5.7
12.6
6.2
9.9
1.9
1.0
0.1
0.0
100.0
Percent
43,804
56,606
7,422
12,076
10,771
23,262
11,151
17,365
3,295
1,703
178
78
187,711
2011
23.3
30.2
4.0
6.4
5.7
12.4
5.9
9.3
1.8
0.9
0.1
0.0
100.0
Fires
Percent
43,576
58,251
7,348
12,300
10,952
23,496
10,972
16,918
3,174
1,737
159
93
188,976
23.1
30.8
3.9
6.5
5.8
12.4
5.8
9.0
1.7
0.9
0.1
0.0
100.0
55
26.6
24.3
1.6
5.3
6.1
22.2
7.0
6.0
0.6
0.4
0.0
0.0
100.0.0
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 23. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Deaths, 2009‐2011
Data Element
2009
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
2010
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to Injury
Percent
Factors Contributing to Injury
Percent
Number and Percent of Reported Deaths
Number of
Null or
Reported Valid
Valid
Invalid
No
Deaths
Known
Unknowna
Entry
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,831
100.0
1,788
97.7
1,110
60.6
786
42.9
1,831
100.0
1,090
59.5
994
54.3
849
46.4
601
32.8
837
45.7
918
50.1
0
0
0
0
73
4
0
0
0
0
234
12.8
0
0
2
0.1
395
21.6
169
9.2
0
0
0
0
42
2.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0.1
648
35.4
1,045
57.1
0
0
507
27.7
837
45.7
980
53.5
835
45.6
825
45.1
913
49.9
1,978
100.0
1,978
100.0
1,978
100.0
1,978
100.0
1,978
100.0
1,978
100.0
1,978
100.0
1,978
100.0
1,978
100.0
1,949
98.5
1,156
58.4
779
39.4
1,978
100.0
1,133
57.3
1,096
55.4
949
48
0
0
0
0
77
3.9
0
0
0
0
253
12.8
0
0
0
0
0
0
29
1.5
1
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
744
37.6
1,199
60.6
0
0
592
29.9
882
44.6
1,029
52
Quality
Index
Usability
Index
100
100
97.7
97.7
64.6
60.6
42.9
42.9
100
100
72.3
59.5
54.3
54.3
46.5
46.4
54.4
32.8
54.9
45.7
50.1
50.1
100
100
98.5
98.5
62.3
58.4
39.4
39.4
100
100
70.1
57.3
55.4
55.4
48
48
56
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number and Percent of Reported Deaths
Number of
Null or Quality
Data Element
Reported Valid
Valid
Index
Invalid
No
Deaths
Known
Unknowna
Entry
Activity When Injured
1,978
616
453
0
909
54
Percent
100.0
31.1
22.9
0
46
Primary Apparent Symptom
1,978
904
162
0
912
53.9
Percent
100.0
45.7
8.2
0
46.1
Primary Area of Body Injured
1,978
975
0
0
1,003
49.3
Percent
100.0
49.3
0
0
50.7
2011
Gender
2,007
2,007
0
0
0
100
Percent
100.0
100.0
0
0
0
Age or Date of Birthb
2,007
1,966
0
41
0
98
Percent
100.0
98
0
2
0
Race
2,007
1,182
83
1
741
63
Percent
100.0
58.9
4.1
0
36.9
Ethnicity
2,007
830
0
0
1,177
41.4
Percent
100.0
41.4
0
0
58.6
Severity
2,007
2,007
0
0
0
100
Percent
100.0
100.0
0
0
0
Cause of Injury
2,007
1,209
232
0
566
71.8
Percent
100.0
60.2
11.6
0
28.2
Human Factors Contributing to Injury
2,007
1,072
0
0
935
53.4
Percent
100.0
53.4
0
0
46.6
Factors Contributing to Injury
2,007
865
0
0
1,142
43.1
Percent
100.0
43.1
0
0
56.9
Activity When Injured
2,007
658
420
0
929
53.7
Percent
100.0
32.8
20.9
0
46.3
Primary Apparent Symptom
2,007
903
166
0
938
53.3
Percent
100.0
45
8.3
0
46.7
Primary Area of Body Injured
2,007
959
0
0
1,048
47.8
Percent
100.0
47.8
0
0
52.2
Usability
Index
31.1
45.7
49.3
100
98
58.9
41.4
100
60.2
53.4
43.1
32.8
45
47.8
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are distinctly
different from blanks and no entries as the lack of information only implies that the information was not given and does not imply that
the element is truly “unknown”.
b. In the NFIRS paper forms and the NFIRS CRG, the full data element name is Age or Date of Birth; in the civilian fire casualty file in the
PDR, the data element is Age.
57
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 24. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Deaths in Buildings28 and Mobile Property Structures, 2009‐2011
Data Element
Number and Percent of Reported Deaths
Number of
Null or
Reported Valid
Valid
Invalid
No
Deaths
Known
Unknowna
Entry
2009
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
2010
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,476
100.0
1,452
98.4
936
63.4
657
44.5
1,476
100.0
896
60.7
822
55.7
682
46.2
508
34.4
710
48.1
763
51.7
0
0
0
0
46
3.1
0
0
0
0
183
12.4
0
0
2
0.1
322
21.8
135
9.1
0
0
0
0
23
1.6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0.1
494
33.5
819
55.5
0
0
397
26.9
654
44.3
792
53.7
646
43.8
631
42.8
713
48.3
1,605
100.0
1,605
100.0
1,605
100.0
1,605
100.0
1,605
100.0
1,605
100.0
1,605
100.0
1,586
98.8
992
61.8
658
41
1,605
100.0
929
57.9
0
0
0
0
48
3
0
0
0
0
200
12.5
0
0
19
1.2
1
0.1
0
0
0
0
0
0
0
0
0
0
564
35.1
947
59
0
0
476
29.7
Quality
Index
Usability
Index
100
100
98.4
98.4
66.5
63.4
44.5
44.5
100
100
73.1
60.7
55.7
55.7
46.3
46.2
56.2
34.4
57.2
48.1
51.7
51.7
100
100
98.8
98.8
64.8
61.8
41
41
100
100
70.4
57.9
28 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value) as for these fires, only the most basic information about the incident is required.
58
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Number and Percent of Reported Deaths
Number of
Null or Quality
Data Element
Reported Valid
Valid
Index
Invalid
No
Deaths
Known
Unknowna
Entry
Human Factors Contributing to Injury
1,605
924
0
0
681
57.6
Percent
100.0
57.6
0
0
42.4
Factors Contributing to Injury
1,605
765
0
0
840
47.7
Percent
100.0
47.7
0
0
52.3
Activity When Injured
1,605
508
374
0
723
55
Percent
100.0
31.7
23.3
0
45
Primary Apparent Symptom
1,605
753
132
0
720
55.1
Percent
100.0
46.9
8.2
0
44.9
Primary Area of Body Injured
1,605
789
0
0
816
49.2
Percent
100.0
49.2
0
0
50.8
2011
Gender
1,593
1,593
0
0
0
100
Percent
100.0
100.0
0
0
0
Age or Date of Birthb
1,593
1,569
0
24
0
98.5
Percent
100.0
98.5
0
1.5
0
Race
1,593
994
47
1
551
65.4
Percent
100.0
62.4
3
0.1
34.6
Ethnicity
1,593
673
0
0
920
42.2
Percent
100.0
42.2
0
0
57.8
Severity
1,593
1,593
0
0
0
100
Percent
100.0
100.0
0
0
0
Cause of Injury
1,593
970
189
0
434
72.8
Percent
100.0
60.9
11.9
0
27.2
Human Factors Contributing to Injury
1,593
850
0
0
743
53.4
Percent
100.0
53.4
0
0
46.6
Factors Contributing to Injury
1,593
656
0
0
937
41.2
Percent
100.0
41.2
0
0
58.8
Activity When Injured
1,593
548
330
0
715
55.1
Percent
100.0
34.4
20.7
0
44.9
Primary Apparent Symptom
1,593
738
125
0
730
54.1
Percent
100.0
46.3
7.8
0
45.8
Primary Area of Body Injured
1,593
754
0
0
839
47.3
Percent
100.0
47.3
0
0
52.7
Usability
Index
57.6
47.7
31.7
46.9
49.2
100
98.5
62.4
42.2
100
60.9
53.4
41.2
34.4
46.3
47.3
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are distinctly
different from blanks and no entries as the lack of information only implies that the information was not given and does not imply that
the element is truly “unknown”.
b. In the NFIRS paper forms and the NFIRS CRG, the full data element name is Age or Date of Birth; in the civilian fire casualty file in the
PDR, the data element is Age.
59
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 25. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Injuries, 2009‐2011
Data Element
2009
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
2010
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Number and Percent of Reported Injuries
Number of
Reported Valid
Valid
Null or
Invalid
Injuries
Known
Unknowna
No Entry
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,582
100.0
9,357
97.7
5,241
54.7
3,792
39.6
9,204
96.1
6,955
72.6
5,824
60.8
5,003
52.2
5,696
59.4
6,101
63.7
5,568
58.1
0
0
0
0
329
3.4
0
0
378
3.9
224
2.3
0
0
5
0.1
679
7.1
104
1.1
2
0
0
0
215
2.2
13
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10
0.1
3,999
41.7
5,790
60.4
0
0
2,403
25.1
3,758
39.2
4,574
47.7
3,207
33.5
3,377
35.2
4,012
41.9
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,376
100.0
10,193
98.2
5,852
56.4
4,355
42
9,992
96.3
7,624
73.5
6,220
59.9
5,204
50.2
6,219
59.9
0
0
0
0
373
3.6
0
0
384
3.7
227
2.2
0
0
2
0
783
7.5
0
0
183
1.8
7
0.1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
4,144
39.9
6,021
58
0
0
2,525
24.3
4,156
40.1
5,169
49.8
3,374
32.5
Quality
Index
Usability
Index
100
100
97.7
97.7
58.1
54.7
39.6
39.6
100
96.1
74.9
72.6
60.8
60.8
52.3
52.2
66.5
59.4
64.8
63.7
58.1
58.1
100
100
98.2
98.2
60
56.4
42
42
100
96.3
75.7
73.5
59.9
59.9
50.2
50.2
67.4
59.9
60
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
2011
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
September 2014
Number and Percent of Reported Injuries
Number of
Quality
Reported Valid
Valid
Null or
Index
Invalid
Injuries
Known
Unknowna
No Entry
10,376
7,018
124
0
3,234
68.8
100.0
67.6
1.2
0
31.2
10,376
6,289
5
0
4,082
60.6
100.0
60.6
0
0
39.3
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,428
100.0
10,203
97.8
5,602
53.7
4,189
40.2
10,004
95.9
7,478
71.7
6,206
59.5
5,154
49.4
6,059
58.1
6,811
65.3
6,195
59.4
0
0
0
0
369
3.5
0
0
424
4.1
285
2.7
0
0
1
0
804
7.7
153
1.5
5
0
0
0
224
2.1
6
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
4,451
42.7
6,239
59.8
0
0
2,665
25.6
4,222
40.5
5,273
50.6
3,565
34.2
3,464
33.2
4,228
40.5
Usability
Index
67.6
60.6
100
100
97.8
97.8
57.2
53.7
40.2
40.2
100
95.9
74.4
71.7
59.5
59.5
49.4
49.4
65.8
58.1
66.8
65.3
59.4
59.4
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are distinctly
different from blanks and no entries as the lack of information only implies that the information was not given and does not imply that
the element is truly “unknown”.
b. In the NFIRS paper forms and the NFIRS CRG, the full data element name is Age or Date of Birth; in the civilian fire casualty file in the
PDR, the data element is Age.
61
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Table 26. Data Element Quality and Usability for Common NFIRS Data Elements Used in USFA Analyses
Reported Injuries in Buildings29 and Mobile Property Structures, 2009‐2011
Data Element
Number and Percent of Reported Injuries
Number of
Reported Valid
Valid
Null or
Invalid
Injuries
Known
Unknowna
No Entry
2009
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to
Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
2010
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
6,988
100.0
6,988
100.0
6,988
100.0
6,988
100.0
6,988
100.0
6,988
100.0
6,988
100.0
6,818
97.6
3,779
54.1
2,760
39.5
6,705
96
5,027
71.9
0
0
0
0
257
3.7
0
0
283
4
189
2.7
0
0
165
2.4
12
0.2
0
0
0
0
0
0
0
0
5
0.1
2,940
42.1
4,228
60.5
0
0
1,772
25.4
6,988
4,205
0
0
2,783
100.0
6,988
100.0
6,988
100.0
6,988
100.0
6,988
100.0
60.2
3,591
51.4
4,138
59.2
4,451
63.7
4,046
57.9
0
5
0.1
506
7.2
84
1.2
2
0
0
0
0
0
0
0
0
0
0
39.8
3,392
48.5
2,344
33.5
2,453
35.1
2,940
42.1
7,639
100.0
7,639
100.0
7,639
100.0
7,639
100.0
7,639
100.0
7,639
100.0
7,639
100.0
7,488
98
4,311
56.4
3,216
42.1
7,359
96.3
5,605
73.4
0
0
0
0
274
3.6
0
0
280
3.7
181
2.4
0
0
151
2
4
0.1
0
0
0
0
0
0
0
0
0
0
3,050
39.9
4,423
57.9
0
0
1,853
24.3
Quality
Index
Usability
Index
100
100
97.6
97.6
57.8
54.1
39.5
39.5
100
96
74.6
71.9
60.2
60.2
51.5
51.4
66.4
59.2
64.9
63.7
57.9
57.9
100
100
98
98
60
56.4
42.1
42.1
100
96.3
75.8
73.4
29 Does not include confined building fires (small fires confined to the object of origin) or outside rubbish fires (fires with no
value) as for these fires, only the most basic information about the incident is required.
62
USFA Response to the OMB
2013 NFIRS Terms of Clearance
Data Element
Human Factors Contributing to
Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
2011
Gender
Percent
Age or Date of Birthb
Percent
Race
Percent
Ethnicity
Percent
Severity
Percent
Cause of Injury
Percent
Human Factors Contributing to
Injury
Percent
Factors Contributing to Injury
Percent
Activity When Injured
Percent
Primary Apparent Symptom
Percent
Primary Area of Body Injured
Percent
September 2014
Number and Percent of Reported Injuries
Number of
Reported Valid
Valid
Null or
Invalid
Injuries
Known
Unknowna
No Entry
7,639
4,570
0
0
3,069
100.0
7,639
100.0
7,639
100.0
7,639
100.0
7,639
100.0
59.8
3,834
50.2
4,623
60.5
5,206
68.2
4,628
60.6
0
2
0
555
7.3
104
1.4
5
0.1
0
0
0
0
0
0
0
0
0
40.2
3,803
49.8
2,461
32.2
2,329
30.5
3,006
39.4
7,853
100.0
7,853
100.0
7,853
100.0
7,853
100.0
7,853
100.0
7,853
100.0
7,853
100.0
7,682
97.8
4,167
53.1
3,116
39.7
7,514
95.7
5,561
70.8
0
0
0
0
280
3.6
0
0
339
4.3
217
2.8
0
0
170
2.2
5
0.1
0
0
0
0
0
0
0
0
1
0
3,401
43.3
4,737
60.3
0
0
2,075
26.4
7,853
4,597
0
0
3,256
100.0
7,853
100.0
7,853
100.0
7,853
100.0
7,853
100.0
58.5
3,815
48.6
4,524
57.6
5,061
64.4
4,571
58.2
0
1
0
582
7.4
125
1.6
5
0.1
0
0
0
0
0
0
0
0
0
41.5
4,037
51.4
2,747
35
2,667
34
3,277
41.7
Quality
Index
Usability
Index
59.8
59.8
50.2
50.2
67.8
60.5
69.6
68.2
60.7
60.6
100
100
97.8
97.8
56.7
53.1
39.7
39.7
100
95.7
73.6
70.8
58.5
58.5
48.6
48.6
65
57.6
66
64.4
58.3
58.2
Source: NFIRS.
Notes: a. Valid Unknown entries are entries coded as “U”, “UU”, or “UUU” (depending on the field length). These coded values are distinctly
different from blanks and no entries as the lack of information only implies that the information was not given and does not imply that
the element is truly “unknown”.
b. In the NFIRS paper forms and the NFIRS CRG, the full data element name is Age or Date of Birth; in the civilian fire casualty file in the
PDR, the data element is Age.
63
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
NFPA SURVEY
The NFPA is a non‐governmental organization whose mission is “to reduce the worldwide
burden of fire and other hazards on the quality of life by providing and advocating consensus
codes and standards, research, training, and education”. In support of their mission, NFPA
administers the annual National Fire Experience Survey, from which they calculate high‐level
national estimates of total fires, civilian deaths and injuries, and dollar loss. As described earlier
in this report, estimates from the NFPA survey and percentages from USFA’s NFIRS are
combined to portray a richer, more detailed picture of the fire situation in the U.S. This section
documents the survey methodology and quality control measures taken by the NFPA to ensure
these estimates are as accurate as practicable.
The remainder of this section is drawn from NFPA’s Fire Loss in the United States During 2012,
published in September, 2013.
Each year, based on a sample survey of fire departments across the country, the NFPA
estimates the national fire problem as measured by the number of fires that public fire
departments attend, and the resulting deaths, injuries and property losses that occur. This
section explains the major steps in conducting the 2012 survey.
Sample Selection
The NFPA currently has 30,145 public fire departments listed in the U.S. in its Fire Service
Inventory (FSI) file. Based on desired levels of statistical precision for the survey results and the
staff available to process, edit, and follow up on the individual questionnaires the NFPA
determined that 3,000 fire departments were a reasonable number for the 2011 sample.
Because of the variation in fire loss results by community size, fire departments were placed in
one of the following 10 strata by size of community protected:
1,000,000 and up
500,000 to 999,999
250, 0000 to 499,999
100,000 to 249,999
50,000 to 99,999
25,000 to 49,999
10,000 to 24,999
5,000 to 9,999
2,500 to 4,999
Under 2,500
64
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Sample sizes for the individual strata were chosen to ensure the best estimate of civilian deaths
in one‐and two‐family dwellings, the statistic that most aptly reflects the overall severity of the
fire problem. All departments that protect 50,000 people or more were included. These 831
departments in the five highest strata protect 153,760,000 people.
For the remaining five population strata, assuming response rates similar to the past two years
for the five highest strata, a total sample of 2,592 was indicated. Sample sizes for individual
strata were calculated using a methodology that assured optimum sample allocations.30 Based
on the average variation in civilian deaths in one‐ and two‐family dwellings by stratum for the
last two years and on the estimated number of fire departments, appropriate relative sample
weights were determined. Then the corresponding sample sizes by stratum were calculated.
The sample size by stratum was then adjusted based on the response rates from the last two
years’ returns. A sample size of 20,385 was found to be necessary to obtain the desired total
response of 3,000 fire departments. For all strata, where a sample was necessary, departments
were randomly selected.
Data Collection
The fire departments selected for the survey were sent the 2012 NFPA Fire Experience
Questionnaire during the 2nd week of January 2013. A second mailing was sent in mid‐March to
fire departments that had not responded to the first mailing. A total of 2,795 departments
responded to the questionnaire 2,097 to the first mailing and 698 to the second. There were
676 departments (24 percent) that responded by using the online version of the survey form.
The overall response rate was 14 percent, although response rates were considerably higher for
departments protecting larger communities than they were for departments protecting smaller
communities. The response rate was 50 percent for departments protecting communities of
50,000 population or more, 23 percent for departments protecting communities of 10,000 to
24,999, and 10 percent for departments protecting communities less than 10,000 population,
which are comprised of mostly volunteers. The 2,795 departments that did respond protect
120,282,500 people or 39 percent of the total U.S. population.
After the NFPA received the surveys, technical staff members of the Fire Analysis and Research
Division reviewed them for completeness and consistency. When appropriate, they followed up
on questions with a telephone call. After the review procedures were completed, the survey
data were keyed to a computer file, where additional checks were made. The file was then
ready for data analysis and estimation procedures.
30 Steve K. Thompson, Sampling, John Wiley, New York, NY, 1992, pp. 107‐111.
65
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
Estimation Methodology
The estimation method used for the survey was ratio estimation with stratification by
community size. 31 For each fire statistic a sample loss rate was computed for each stratum.
This rate consisted of the total for that particular statistic from all fire departments reporting it,
divided by the total population protected by the departments reporting the statistic. Note that
this means that the departments used in calculating each statistic could be different, reflecting
differences in unreported statistics. The sample fire loss rates by stratum were then multiplied
by population weighting factors. The estimates were then combined to provide the overall
national estimate.
If this method of estimation is to be effective, estimates of the total number of fire
departments and the total population protected in each stratum must be accurate. The NFPA
makes every effort to ensure that this is the case. The population weights used for the national
estimates were developed using the NFPA Fire Service Inventory (FSI) file and U.S. Census
Bureau population estimates.
For each estimate, a corresponding standard error was also calculated. The standard error is a
measure of the error caused by the fact that estimates are based on a sampling of fire losses
rather than on a complete census of the fire problem. Due to the fact that the survey is based
on a random sample, we can be very confident that the actual value falls within the percentage
noted in parentheses for the overall national fire loss statistics: number of fires (2.0 percent),
number of civilian deaths (11.3 percent), number of civilian injuries (5.5 percent), and property
loss (3.0 percent).
The standard error helps in determining whether year‐to‐year differences are statistically
significant. Differences that were found to be statistically significant were so noted in the
tables. Property loss estimates are particularly prone to large standard errors because they are
sensitive to unusually high losses, and, as a result, large percentage differences from year to
year may not always be statistically significant. In 2012, for instance, property damage in
educational properties was estimated to be $64,000,000. This represented an increase of 35.0
percent from the year before, but was found not to be statistically significant.
In addition to sampling errors, there are nonsampling errors. These include biases of the survey
methodology, incomplete or inaccurate reporting of data to the NFPA, differences in data
collection methods by the fire departments responding. As an example of a nonsampling error,
most of the fires included in the survey took place in highly populated residential areas,
because the fire departments selected for the surveys are primarily public fire departments
31 William G. Cochran, Sampling Techniques, John Wiley, New York, NY, 1977, pp. 150‐161.
66
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
that protect sizable residential populations. Fires that occur in sparsely populated areas
protected primarily by State and Federal Departments of Forestry are not likely to be included
in the survey results.
The NFPA Fire Incident Data Organization (FIDO) data base was also used in conjunction with
the annual survey to help identify any large loss fires or deaths that the survey might have
missed.
The editors of the survey data attempted to verify all reported civilian deaths in vehicle fires.
They contacted most of the fire departments that reported fire‐related deaths in vehicles and
found that many of the deaths were indeed the results of fire. In some instances, however,
impact was found to have been the cause of death. This effort can have a considerable impact
on the estimates.
The results presented in this report are based on fire incidents attended by public fire
departments. No adjustments were made for unreported fires and losses (e.g., fires
extinguished by the occupant). Also, no adjustments were made for fires attended solely by
private fire brigades (e.g., industry and military installations), or for fires extinguished by fixed
suppression systems with no fire department response.
Fire Experience of Nonrespondents
A telephone follow‐up was made to a sample of nonrespondents to determine whether fire
departments that did not respond to the survey experienced fire loss rates similar to those that
did respond. This would help the NFPA determine whether we received questionnaires only
from departments that had experienced unusually high or low fire losses.
The sample of nonrespondents selected was proportional by state and population of
community to the original sample selected for the survey. As a result of these efforts, 155 fire
departments were successfully contacted and answered some of the questions about their fire
experience.
For communities of 100,000 to 249,999, the rates for respondents and nonrespondents were
similar for deaths, the respondent rate was 55 percent higher for civilian deaths, and 102
percent higher for property loss, while the nonrespondent rate was 12 percent higher for fires.
The result for property loss was statistically significant.
For communities of 50,000 to 99,999, the respondent rate was 61 percent higher for civilian
deaths and 45 percent higher for property loss, while the nonrespondent rate was 18 percent
higher for fires. The result for property loss was statistically significant.
67
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
For communities of 25,000 to 49,999, the respondent rate was 112 percent higher for civilian
deaths, and 10 percent higher for property loss, while the nonrespondent rate was 48 percent
higher for fires. Results for fires and deaths were statistically significant.
For communities of 10,000 to 24,999, the respondent rate was 20 percent higher for civilian
deaths and 11 percent higher for property loss, while the nonrespondent rate was 12 percent
higher for fires. None of these results were statistically significant.
For communities of 5,000 to 9,999, the nonrespondent rate was 50 percent higher for fires,
while the respondent rate was 21 percent higher for property loss. Neither of these results was
statistically significant.
RESOURCES
Several resources are available that provide more detailed documentation on the NFIRS system
and using the NFIRS data. The National Fire Incident Reporting System Complete Reference
Guide provides both instructions for reporting data to NFIRS and an understanding of the data
elements collected by the system. It also serves as a reference for the coding of the data.
The National Fire Incident Reporting System Version 5.0 Fire Data Analysis Guidelines and Issues
discusses analytic considerations and methods of analyzing fire incident data using the NFIRS
data. The topics include the NFIRS 5.0 data structure, general quality assurance (QA) issues, and
definitions and parameters of common fire analyses (e.g., residential structure fires or fires by a
specific cause). The methods, techniques, and considerations discussed are those used by USFA
analysts and do not necessarily reflect methods, techniques, and considerations used by fire
data analysts from other agencies and organizations. NFIRS data partners may (and do) employ
their own methods for analyzing the data and may make differing assumptions when
encountering data issues.
The National Estimates Approach to U.S. Fire Statistics is the original methodology for creating
estimates of the U.S. fire problem using the NFPA annual survey of fire departments and the
NFIRS data. The authors present a detailed consensus procedure for such calculations and the
supporting rationale. National Estimates Methodology for Building Fires and Losses is the
USFA’s application of the national estimates approach to building fires and fire losses. It details
USFA’s current fire data estimation methodology for all building (i.e., residential and
nonresidential) fires and associated losses.
The online USFA Structure Fire Cause Methodology and the links on this USFA web page provide
both a detailed description of the cause hierarchy methodology and the technical hierarchy
itself. The USFA Data Sources and Methodology Documentation provides an in‐depth discussion
68
USFA Response to the OMB
2013 NFIRS Terms of Clearance
September 2014
of the data sources and the methodologies used to incorporate this data into fire analyses. The
Fire Data Analysis Handbook is a resource for those unfamiliar with basic data analysis
techniques and their applicability to fire data based analyses.
Lastly, NFIRSGrams, short bulletins that provide coding help to fire department personnel using
NFIRS, address frequently asked questions and common mistakes made when completing
incident forms. NFIRSGrams also helps NFIRS users to better understand their impact on the
quality of the information from NFIRS at the local, state, and national levels.
These resources are listed below:
1) National Fire Incident Reporting System Complete Reference Guide, USFA, January 2013,
http://www.nfirs.fema.gov/documentation/reference/.
2) National Fire Incident Reporting System Version 5.0 Fire Data Analysis Guidelines and
Issues, USFA, July 2011:
http://www.usfa.fema.gov/downloads/pdf/publications/nfirs_data_analysis_guidelines
_issues.pdf.
3) Hall, J. and Harwood, B., “The National Estimates Approach to U.S. Fire Statistics”, Fire
Technology, Vol 25, No. 2 (1989). 99‐113.
http://www.nfpa.org/~/media/6906FADB2CE149488FB5103F4A750A05.ashx.
4) National Estimates Methodology for Building Fires and Losses, USFA, August 2012:
http://www.usfa.fema.gov/downloads/pdf/statistics/national_estimate_methodology.p
df.
5) “USFA Structure Fire Cause Methodology”:
http://www.usfa.fema.gov/fireservice/nfirs/tools/fire_cause_category_matrix.shtm.
6) “USFA Data Sources and Methodology Documentation”, March 2014:
http://www.usfa.fema.gov/downloads/pdf/statistics/data_sources_methodology.pdf.
7) Fire Data Analysis Handbook, Second Edition, USFA, January 2004:
http://www.usfa.fema.gov/downloads/pdf/publications/fa‐266.pdf.
8) NFIRSGrams, USFA 2014: http://www.usfa.fema.gov/fireservice/nfirs/.
69
| File Type | application/pdf |
| File Title | Microsoft Word - OMB Response FINAL with cover and title page.docx |
| Author | Tricia |
| File Modified | 2014-09-04 |
| File Created | 2014-09-04 |