A Direct-Access Care Model
Firefighters typically receive care for occupational-related MSIs through a workers' compensation system. However, this injury management model results in excessive lost duty time and large financial payouts for fire departments and related municipalities. A model to promote healthcare efficiency and effectiveness is to provide direct access to healthcare providers (e.g., ATs, PTs, OTs, EPs). Early and direct access to HCPs allows for immediate assessment, treatment, and frequent rehabilitation of MSI that are currently delayed and constrained within the traditional workers' compensation system. Expedited treatment can potentially speed up the recovery process and reduce perceptions of pain post-injury—decreasing the time it takes to return to full-duty status.
The direct access care model has been successfully implemented in physically active and tactical populations and is emerging in occupational settings. Early access to physical rehabilitation from an AT in occupational settings can reduce days lost by over 50%(23). The National Athletic Trainers' Association reports a favorable return on investment (ROI) with early direct access to AT in occupational settings(24). Over 80% of companies with an integrated AT on staff reported an ROI of more than $3/employee per $1 invested, and 30% reported more than $7 per employee/$1(24). Simultaneously, 96% of companies experienced a significant decline in workers' compensation claims costs within one year of offering direct access to AT(24). The San Antonio Fire Department documented $593,682 in cost savings with the presence of AT for early assessment and intervention in the first nine months of the initiative(22).
Aim 1
Aim 1 of the project will focus on gathering input from career firefighters and HCPs who work with career firefighters to improve the limited injury surveillance data available on MSIs in career firefighters. Using a qualitative approach, our team will conduct interviews with career firefighters who have sustained injuries during occupational activities or physical training and HCPs, including ATs and PTs, who are currently engaged in treating MSI in career firefighters. The qualitative information gathered from these interviews will be compiled using a mixed-methods analysis to identify categories and themes across participants related to injury mechanisms and environmental circumstances. Specifically, we will identify intrapersonal (e.g., movement pattern during injury, occupational and physical training experience, situational awareness), interpersonal (e.g., peer pressure to exercise beyond capacity), departmental (e.g., operational policies and procedures), and environmental factors (e.g., heat exposure, slippery environment) that may be associated with injury occurrence. This will inform the understanding of the mechanisms of MSI sustained by career firefighters.
Qualitative semi-structured interviews will be conducted with a purposive sample of 20 career firefighters and 10 HCPs (10 HCP interviews were completed in preliminary work prior to the start of this study). Criteria for career firefighter survey participant recruitment and selection include career firefighters who have sustained an occupational MSI within the last two years. These participants will be recruited from a representative sample of fire departments across the United States. Criteria for HCP survey participant recruitment and selection include those who have at least three years of experience in treating and rehabilitating career firefighter injuries, including experience within the past year and are working as a licensed HCP (e.g., AT, PT) with at least five years of experience. These participants will be recruited from across the United States via our partnering fire departments and established firefighter HCP groups.
Aim 1 will follow a phenomenological qualitative design to explore the experiences with injuries incurred by firefighters during physical training and occupational tasks from the perspective of career fighters and HCPs who have worked with career firefighters. Previous studies have used a phenomenological approach when interviewing firefighters(25, 26). The phenomenological approach encompasses understanding a group of individual's experiences and identifying shared phenomenon. Specifically, a semi-structured interview protocol will be drafted by investigators. The interview protocol will include four sections. First, the introductory section will outline the purpose of the study, gain consent, address issues of confidentiality, and establish trust with each participant. Second, transition questions will explore firefighters' or HCPs' experience with occupational injuries. The third section will focus on open-ended questions related to injury mechanisms (e.g., ergonomic, anatomical, physiological factors), environmental factors (e.g., ambient environment), intra- and interpersonal factors, and institutional factors (e.g., personnel staffing, influence of shiftwork) associated with occupational tasks and exercise training, including detailed probes for each question. The last section will include closing questions, allowing the participants to add information not covered in the other sections. Sample questions in the interview include: "In your experience, what factors increase the risk of injuries while exercising?" and "While working, how did the fireground or fire station environment contribute to injuries during work-related tasks?"
Interviews will be conducted using video conference technology (Zoom, Version 5.5.2, San Jose, CA) and recorded for analysis. Each interview will last approximately 60 minutes. Field notes will be gathered during the interview to capture non-verbal interactions (e.g., facial expressions). Video interview data and field notes from each interview will be immediately transcribed and uploaded into Dedoose (SocioCultural Research Consultants, LLC). This cross-platform software program allows for qualitative analysis of text, audio, and video data. Dedoose was created specifically for mixed-methods research and allows teams to collaborate on data analysis. Inductive axial coding will be conducted to identify categories and themes across participants related to injury mechanisms and environmental circumstances. Three methods will be used to ensure the trustworthiness and credibility (i.e., reliability and validity) of the data analysis and findings. First, a limited number of researchers will code the data to help ensure consistency. Second, inter-rater reliability of codes and categories across coders will be conducted via Dedoose. Cohen's Kappa will be used to measure the level of agreement among coders with a goal of 90% agreement(27). Third, member checking (participant validation) will be used to explore the credibility of the findings to ensure they resonate with their experiences.
Aim 2
Using a prospective cohort study design, Aim 2 of the study will determine if implementing a direct access model of HCPs will reduce time loss from MSI and workers' compensation outcomes compared with career fire departments that do not have direct access to HCPs. This will support evidence for an implementation model in career fire departments to maximize the safety and performance of career firefighters while lowering the associated financial burden. For Aim 2, career fire department injury surveillance and workers' compensation data will be used to track injury incidence, reported mechanisms, time loss due to injury, and fiscal outcomes. To accomplish this, we have partnered with career fire departments throughout the United States to prospectively collect these data for two years. These fire service partners will be grouped based on their status of HCP integration: 1) direct access to a full-time AT or PT for injury evaluation treatment and/or prevention (DA); or 2) workers compensation only (WC; i.e., no immediate access to AT/PT). Although no direct human subject data collection will occur between the investigators and the firefighter population, an IRB will be required. De-identified data from the partnering fire departments will be collected and retained using a HIPPA-compliant and encrypted database system hosted on University of Kentucky servers. Critical to this aim will be comparing the injury, time-loss, and financial data between departments with full, limited, or no access to HCPs.
Using a prospective design, we will track MSI incidence, activity during injury, traditional injury cause (current categories reported by NFPA), injury type (strain, sprain, fracture, etc.), and suspected injury mechanism. Additionally, we will collect information on time loss, light duty, and workers' compensation claim data associated with the reported MSI. Details regarding activity during injury description, movement pattern type (e.g. lifting, bending, carrying, stepping down), traditional cause injury, injury type, and suspected injury mechanism (e.g. slip, trip, fall, strain, overexertion) will be collected at the time of injury when filing an incident report as per the department's standard practice. At the individual career fire departments, the firefighter's name and date of birth are kept on record to match the initial report and the final workers' compensation record data. To collect time loss data for each injury, departments' workers’ compensation records will be queried by a designated authorized department personnel who will enter de-identified data on the matched original record in the injury surveillance instrument.
The healthcare model group will report summary statistics for injury incidence, type, activity, cause, mechanism, time loss, and time in light/modified duty. Injury rate ratios with 95% confidence intervals for all MSI will be calculated for each group. Injury rates will be considered significantly different if the 95% confidence intervals do not overlap. Additionally, average time loss and time in light or modified duty per injury will be calculated and assessed for normality using a Shapiro-Wilk test. These average time-loss values will be compared across each group using independent t-tests. If any of the healthcare model group’s time-loss data are not normally distributed, a Mann-Whitney U test will be used to make the non-parametric pairwise comparisons. An alpha level will be set a priori at 0.05. All statistical comparisons will be completed using SPSS (version 26, IBM, Armonk, NY).
Aim 3
Aim 3 will compare two models used to expedite care, reduce occupational time loss, and limit the associated fiscal burden. We will conduct a case study analysis with two fire departments and gather information from key informants to create profiles of the models, including policies, organizational structure, budgeting, and facilitators and barriers to implementing the modules. A case study approach will be used with two primary cases. The cases selected will represent two different modes of HCP access: (1) Direct HCP access for “prehab” conditioning and standard rehabilitation services, and (2) A traditional workers’ compensation model. Key stakeholders interviewed include Executive Fire Department Leadership, Risk Management Director, Workers’ Compensation Representative, and Municipality Leadership. The case with the direct access care model was required to have the model in place for a minimum of one year.
Semi-structured interviews will be conducted with key stakeholders from each case. An interview guide will be established prior to conducting the interview, including questions that address the cross-case and case-specific research questions in congruence with the predetermined theoretical framework. An introductory section of the protocol will focus on providing the interviewee with the purpose of the study, gaining consent, addressing issues of confidentiality, and establishing trust with each participant. The second section of the protocol will include three open-ended questions. Question one will ask the interviewee to describe the model they currently use to treat injured firefighters. The interview protocol will include prompts to ensure key areas of interest related to the major components of their treatment model, why they chose the model, and how the model, if funded, will be addressed. Question two will ask the interviewee to describe major outcomes they feel they have achieved as a result of using the model. Question three will ask what they would consider the strengths and challenges of the model. At the end of the interviews, participants will be asked to identify written documents about the model that can be shared. Various open-ended probing questions outside of the structured interview guide were provided based on the responses of the participants. Interviews will be conducted using video conference technology (Zoom, Version 5.5.2, San Jose, CA) and recorded for analysis. Each interview will last approximately 60 minutes. Field notes will be gathered during the interview to capture non-verbal interactions (e.g., facial expressions). Video interview data and field notes from each interview will be immediately transcribed and uploaded into Dedoose (SocioCultural Research Consultants, LLC), a cross-platform software program that allows for qualitative analysis of text, audio, and video data. One research team member will independently analyze interviews by coding initial transcripts and identifying additional codes. An additional two team members will follow up with analysis to ensure an inter-rater reliability of Cohen’s Kappa coefficient of > 0.80.
As part of the larger research project, data collected in a separate aim will be included in the current cross-case analysis. These data consist of career fire department injury surveillance and workers’ compensation data tracking injury incidence, reported mechanisms, time-loss due to injury and fiscal outcomes. De-identified data from the partnering fire departments will be collected and retained using a password-protected University of Kentucky server. Critical to this aim will be the comparison of the injury, time-loss, and financial data between departments that an embedded HCP versus those using traditional workers’ compensation.
The key stakeholders will be asked to share non-confidential documentation about the direct access care model to further investigate the factors influencing implementation. These include official documents (policies or policy directives), implementation documents (training manuals, financial analyses, funding requests), legal documents (MOUs, cooperative agreements), and working documents (budgets, planning). Information from the documents will be extracted to help corroborate and elucidate key characteristics of the models described by the interviewees (e.g., major components, funding). The analysis process will include the systematic appraisal and synthesis of information across the documents to help support a profile of each model being implemented in the case study sites. The case study approach aims to create a profile for each of the models being implemented. To create the profile, interview data, and field notes from each interview will be immediately transcribed and uploaded into Dedoose as a case. Documents gathered by each site will be uploaded into the case file. A deductive coding process based on interview questions will be used to sort data from interviews and documents into major categories of model description, outcomes, and strengths and weaknesses. A deductive coding process will be conducted on the policy documents, linking information to the major categories covered in the interview guide regarding model structure, outcomes, and barriers/facilitators. Data from the document analysis will be linked to interview data and used to support information gathered from the interviews and for illustrative purposes within the profiles. To ensure the accuracy of the profiles, they will be shared with each site, with opportunities to revise and refine the information so it accurately reflects the model being implemented. Once all profiles have been created and reviewed by the sites, a cross-case analysis will be conducted to identify commonalities and differences across the models and provide examples for the model profiles.