Injury Data Evaluation & Analysis (IDEA)
Firefighter injury data are gathered by multiple organizations, including the NFPA (National Fire Prevention Association) and NFIRS (National Fire Incident Reporting System). The resulting injury estimates are very different from one another and do not allow for analysis of the granular data the fire service needs - data that allow departments and states to compare themselves to others, and thoroughly investigate what is happening to their members. In response to FEMA’s Assistance to Firefighter Grants program calling for proposals that “focus on the design and feasibility of a new database system”, previously non-integrated data was linked (workers’ compensation, safety office, human resources, hospital/ED visits, 911 calls). Two studies were completed to develop model firefighter nonfatal injury surveillance systems through linkage of previously unrelated data systems: The Firefighter Injury Research and Safety Trends (FIRST) project and the Firefighter Injury Research and Safety Trends Reliability Study (FIRST-RS).
FIRST & FIRST-RS were created to research and develop the minimum data elements necessary to conduct public health surveillance of non-fatal firefighter injuries in the United States. [FIRST ultimately grew to become the umbrella under which all research described in this dossier resides, thus it became the name of our Center]. These investigations employed the following strategies:
- learn from existing firefighter data collection;
- explore how non-firefighter data sources (e.g., BRFSS, cancer registries, hospital encounters) can contribute to a comprehensive, national database of firefighter injuries;
- examine relevant federal, state, and local regulations that empower or challenge the functions of non-fatal firefighter injury data collection;
- develop and test data systems in the city of Philadelphia (Pennsylvania), Boston (Massachusetts), and the State of Florida; and
- provide recommendations for a national implementation plan for non-fatal injury data collection to FEMA.
Master injury data systems were created for the Philadelphia Fire Department, the Boston Fire Department, and the State of Florida Fire Marshal’s office. Upon completion of linkage, analysis of the resulting comprehensive non-fatal injury data was conducted. The results of this research recommended core data elements to FEMA for national implementation. The findings also determined that a national firefighter nonfatal injury data system could be developed through 3 paths:
By linking injury reports to relevant existing data sources, a fire department-level system can be developed. This methodology was aptly demonstrated by the Philadelphia Fire Department pilot site. Through standardizing of the resulting master database, the data can be used for comparative purposes. At the department level, it was found that linking dispatch run data, human resources demographic data, risk management data, and disability data were critical to create a robust and longitudinal understanding of injuries to firefighters. Data linkage allowed for: (1) the creation of a new variable - years of experience which illuminated the time during a firefighter’s tenure when injury was most prevalent, (2) a reduction in injury cause misclassification leading to more accurate estimates of cost, and (3) the visualization of injury rates by response type through GIS.
Data at the fire department-level yield the most detailed information, ranging from first aid/treated on-scene injuries, to hospitalizations and lost time injuries, to line of duty deaths. However, capturing this information is the most resource-intensive of the three paths because there are more than 30,000 fire departments in the United States, and linkage would have to be done individually for each of them.
The State of Florida served as a pilot site providing accurate and complete workers’ compensation, and hospital and emergency department discharge data. Data were linked to a statewide registry of firefighters thus creating a state-level system. Results helped build the case for national policy (below) in that, while we could determine if firefighters went to the emergency department or hospital for injuries could be determined, we were unable to determine if visits were work-related because industry and occupation information were not captured. Other limitations of the state-level system were that it did not contain the depth and breadth of injury data available at the department-level. The state model offered a vision for faster national implementation than the fire department-level model because there are only 50 states in which data linkage would need to be conducted.
One of the high risk/high reward activities undertaken through our FEMA funding was to change national policy regarding data standards for occupational surveillance. The ultimate goal was to convince national data standards maintenance organizations (under HIPAA and the US Department of Health and Human Services) to include the data elements of industry and occupation (I/O) in allhospital encounters. We proposed using existing federal standards to facilitate this: the North American Industrial Classification System (NAICS) and standard occupational codes (SOC) developed by the Bureau of Labor Statistics (BLS). A national coalition of stakeholders in support of this national policy change was developed, including resolutions adopted by the Council for State and Territorial Epidemiologists and the Congressional Fire Services Institute. The halfway point was reached in the policy adoption process by obtaining ‘Accredited Standards Committee X12’ approval. To date we have been unsuccessful in getting NUBC approval. Two manuscripts (fast-tracked in review to inform policy) highlight the costs, benefits, and facility with which these missing data elements could benefit all workers:
Policy can be changed to adopt standards for industry and occupation data collection in all hospital-level healthcare encounters in the United States. Achieving NUBC approval would enable all inpatient and emergency department hospital visits by career and volunteer firefighters to be systematically captured. The documentation of industry and occupation information will not only benefit the fire service, but everyone who works. The politically charged reaction to the proposed policy changes hampered the initiative. The national coalition built around this initiative continued its efforts through the FIRST Reliability Study’s pilot study to assess the time it takes to capture industry and occupation during the patient registration process. This pilot study was conducted at two hospitals in Philadelphia, PA:
The models of injury data linkage in the above studies have become essential components to new programs with FIRST. For example, FIRST's latest grant, “Stress and Violence to fire-based EMS Responders”, uses the department-level model to accurately assess verbal and physically violent encounters experienced by first responders. This type of linkage requires very sensitive negotiations among the parties because it uses personal identifiers like employee ID and in some cases, social security number. FIRST works very closely with Drexel’s Office of General Counsel, the Chief Privacy Officer, and the HIPAA legal team to ensure that agreements are protective of study participants, the partner’s privacy concerns, and the university’s liability. The sensitive nature of these agreements can take months to years to effectuate. While this has obvious implications for the speed of our research productivity, we believe the quality and impact of our work is the more important metric.
If you would like more information, please email Dr. Jennifer Taylor, firstname.lastname@example.org.