A Cross-sectional Analysis of Traumatic Stress and Burnout Symptoms in Search and Rescue Volunteers

Abstract

Introduction

Search and Rescue (SAR) volunteers are regularly exposed to traumatic events in their work, leading to increased risk of adverse mental health outcomes, burnout, and stress injuries. This study was designed to estimate the prevalence of mental health and burnout symptoms among SAR volunteers.

Methods

A cross-sectional study of SAR volunteers in all 64 counties in Colorado was conducted between November 9^th-30^th 2021. The survey included the Maslach Burnout Inventory, Secondary Traumatic Stress Scale, short-form Beck Hopelessness Scale, and Substance Use Behavior Scale. Analysis of data was conducted using SPSS version 28. This study was approved by the Colorado Multiple Institutional Review Board, Protocol #: 22–0246

Results

We had a 23% response rate with 657 survey responses out of 2800 of all volunteers in the state. 47 out of 52 SAR agencies were represented in the survey. Age of participants was 47 ± 21, 26% identified as female, and 97% identified as non-Hispanic White. We found that 24% are experiencing worsening health status, with 26% with some level of burnout, while 54% are at risk for burnout. 25.5% reported intrusive experiences, 10% were at risk for suicide, and 1/3 met criteria for alcohol use disorder. Between 10–33% of SAR volunteers reported severe enough symptoms to qualify for immediate mental health referrals.

Conclusion

The effects of stress injuries are of paramount importance to the SAR community as they can increase risk factors that lead to adverse physical and mental health outcomes. Burnout and stress levels among SAR are higher than the general population and comparable to other first-responder groups, including firefighters, police, emergency medical systems, and emergency and critical care health workers.

INTRODUCTION

Search and rescue (SAR) work is often dangerous, and volunteers usually do not receive any compensation, healthcare, or mental health services, despite the distinct possibility of personal injury and impacts on mental health from witnessing extreme trauma or death. Professionals working in SAR are exposed to traumatic events in their work, leading to increased risk of adverse mental health outcomes, secondary traumatic stress (STS), burnout, and stress injuries.^1–5 Volunteers’ working conditions and demands have also become more difficult in recent years as more people take advantage of outdoor recreation opportunities without corresponding increases in the volunteer-based system of rescue services for people who have an emergency in the back-country.

First responders are at higher risk than the general population for experiencing traumatic events, including life-threatening situations, grave injuries, and deaths of colleagues and civilians.^3,4,6 They experience taxing work demands with routine exposures to stressors that have been linked to the development of new mental health conditions or exacerbation of pre-existing conditions.^1,2,6–10 There is evidence that mental health conditions are significantly more common among first responders than in the general population.^6,7,10–13 There is also evidence of increased burnout among first responders and people who work with trauma survivors.^14,15 Trauma can affect people even when it is not personally experienced, with additional research showing that stress injuries are linked to work-related exposure in professionals working with trauma survivors.^{11,12,15–17}

Most of the literature on burnout and stress among first responders is focused on police officers, firefighters, and emergency medical services. There is limited research on the prevalence of burnout and mental health symptoms among SAR personnel. Most studies exploring this topic among SAR are limited by small sample sizes, therefore preventing the generalizability of reported findings. There is a growing recognition of the impact of trauma exposure on behaviour and physical health outcomes among SAR communities. To address this problem, clinicians and researchers must define and measure burnout and mental health characteristics of those regularly exposed to stress while doing SAR volunteer work. Our study, therefore, was designed to estimate the prevalence of mental health and burnout symptoms across the total population of SAR volunteers in Colorado.

METHODS

This study was approved by the Colorado Multiple Institutional Review Board, Protocol #: 22–0246. We conducted a cross-sectional survey with Colorado SAR volunteers who were part of a SAR agency in Colorado. Self-identified volunteer first responders and members of nonprofit SAR teams were recruited to participate in the survey. We defined SAR to include responders who are coordinated by a sheriff to provide service during emergencies or disasters in forests, deserts, mountains, canyons, caves, waters, parks, plains, and at times, in more populated areas. Responders were members of volunteer teams that work alongside fire, law enforcement, emergency medical personnel, the Colorado National Guard, and other government employees in disasters or emergencies. Anecdotally, some of these volunteers also have paid employment as first responders (e.g., through police or fire departments), but their SAR work was done as volunteers and on their own time.

Eligible participants were invited to participate with the assistance of the Colorado Search and Rescue Association (CSAR) and Responder Alliance, two pre-existing nonprofit groups that serve the SAR community. The list of BSAR agencies was validated by a steering group that also included county sheriffs, members of SAR agencies, and staff at the Colorado Parks and Wildlife Department. Once a comprehensive list of BSAR agencies was identified, an email with the survey was sent to identified leaders at each one, with a request to distribute the survey link to all agency members. The first request to collect information was sent on Tuesday, November 9, 2021, with reminders sent twice over two weeks. The survey closed on Tuesday, November 30. CSAR estimates that there are 50 SAR agencies in the state with up to 2,800 volunteers. Our overall response rate was at least 23% for SAR volunteers (depending on assumptions about the total number of SAR workers in the state), and 94% of agencies were represented. Study data were collected and managed using REDCap electronic data capture tools (Nashville, TN: Vanderbilt University) hosted in the Colorado Clinical and Translational Sciences Institute at the University of Colorado Anschutz Medical Campus.^18,19 REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies providing, 1) an intuitive interface for validated data capture; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for data integration and interoperability with external resources.

The survey consisted of 154 questions drawn from various validated instruments to assess demographics, experiences, and mental health outcomes, including depression, substance use disorder, posttraumatic stress disorder, chronic stress, and burnout. The questionnaires included one item from the Short Form Health Survey (SF-36), two of three subscales from the Maslach Burnout Inventory (MBI), the full Secondary Traumatic Stress Scale (STSS) measuring symptoms of post-traumatic stress disorder, a 4-item short form of the Beck Hopelessness Scale (BHS), the 2-item Patient Health Questionnaire (PHQ-2) depression screener, and the validated 4-item Substance Use Brief Screen (SUBS). These scales have undergone psychometric testing and are reliable and valid tools for collecting data for similar populations.

Using the SF-36 item we examined global perceived health, where scores less than 4 (very good) indicate increased risk of mortality.²⁰ The MBI scoring was based on a sum of items with cutoff scores from a prior study of healthcare professionals, where personal accomplishment was scored as < 33 = low (worse), 34–39 = moderate, > 40 = high and depersonalization was scored as < 5 = low, 6–11 = moderate, > 12 = high (worse).²¹ Participants who scored either low on personal accomplishment or high on depersonalization were considered to be showing signs of burnout, where participants who scored in the moderate category were considered to be at risk of burnout.²² The STSS score was based on averages compared to original item labels where 0 = low (not at all), 1–2 = moderate (mildly or moderately), and 3–4 = high (severely or very severely).²³ Positive screen for depression was based on the PHQ2, calculated as a sum of items with a score of 3 or higher (out of 6) suggesting clinical depression.²⁴ Risk of suicide was estimated based on a validated 4-item version of the BHS, where a “yes” response on any of 2 of the 4 items suggested an increased risk.²⁵ Estimates of binge drinking, tobacco use, opioid use, and other drug use were based on 4 separate items from the SUBS.²⁶ The opioid use item on this scale represents use of prescribed painkiller medication other than the way in which it was prescribed, and therefore does not include heroin or other non-prescribed opioid use.

Descriptive analyses of quantitative data were conducted using SPSS version 28 (IBM Corporation., Armonk, NY) and graphics were prepared using GraphPad Prism v. 8.3.1 (GraphPad Software, Inc., La Jolla, CA). For continuous and normally distributed data, means with standard deviations were calculated, and numeric counts with percentages were calculated for categorical variables. Mean values or frequencies were expressed as means ± SD or 95% CIs for proportions. The percentage of respondents with missing data was very low (< 10% for all subscales) so missing responses were omitted from the analysis. However, participants with partial responses were still included if a subscale could be prorated, and those with incomplete survey responses on one measure were still included in the analysis of data from other measures. This manuscript follows the guidelines to The Strengthening the Reporting of Observational Studies in Epidemiology.²⁷

There was considerable concern among stakeholders that SAR volunteers would not be sincere on measures that asked about their mental health concerns. To get around this challenge, we administered 2 pairs of items in the survey: In each pair, 1 item asked whether the SAR worker had personally experienced a problem, and the other asked if they had seen peer SAR volunteers affected by the same problem. Across the 2 pairs of items, SAR volunteers reported 18% more distress on questions about others than themselves. We considered this difference to represent the degree to which SAR volunteers minimize their mental health concerns; estimating this difference based on discrepancies between questions about others versus oneself is an established method for demonstrating actor-observer bias²⁸, which is a well-established psychological principle.^29–31 Therefore, in the analyses below, we applied this +18% adjustment to respondents reported mental health symptoms to obtain a more accurate estimate of SAR volunteers’ true level of mental health concerns. Both the adjusted and unadjusted results are presented below.

RESULTS

There were 657 individual SAR volunteers from 47 different SAR teams represented in the survey. Participant characteristics are summarized in Table 1. Of note, the average age of participants was 48 ± 21, the majority identified as white males, and most had completed some form of higher education. Most participants were active field SAR members and coordinators, but administrative leadership was also represented.

Table 1.

Participant demographics and characteristics of 657 search and rescue volunteers.

Descriptive characteristics	n (%)
Age (yrs), M (SD)	48 (21.47)
Gender
Man	480 (74.3%)
Woman	159 (24.6%)
Nonbinary	1 (0.2%)
Prefer not to say	6 (0.9%)
Race
White	615 (96.7 %)
Black or African America	2 (0.3 %)
Native American	12 (1.9 %)
Asian	10 (1.6 %)
Pacific Islander	2 (0.3 %)
Other	11 (1.7%)
Ethnicity
Not Hispanic or Latino	617 (97 %)
Hispanic or Latino	18 (3.4 %)
Level of Education, % (n)
Master’s or other graduate	229 (30.5 %)
Bachelor’s degree / 4-year college	295 (45.5 %)
Associate degree or certificate	74 (11.4 %)
High school diploma or equivalent	78 (12 %)
Less than high school	4 (0.6 %)
Role within Organization
Field active SAR member	421 (64.3 %)
SAR mission leader	119 (18 %)
Administrative member	11 (1.7 %)
Program director/manager/supervisor	16 (2.4 %)
Board member/officer	68 (10.4 %)
Other role	21 (3.2 %)

General Health

SAR volunteers are generally healthy, with 464 (76%, 95% CI: 72.4–79.2) participants reporting either “very good” or “excellent” health. However, 148 (24% 95% CI: 20.8–27.6) of SAR volunteers are at risk for chronic health problems based on the SF-36 perceived health item.²⁰

Burnout

An early warning sign of burnout is feeling less energized or excited about work. This type of experience is captured by the “personal accomplishment” scale of the MBI, in which lower scores suggest a higher risk for burnout.³² As shown in Figure 1. Panel A, we estimate that 322 (54%, 95% CI: 50–58) of SAR volunteers are currently at risk for burnout based on a lack of personal accomplishment at work. Our adjusted values show that up to 470 (78.9%, 95% CI:75.6–82.2) may be at risk for burnout.

Figure 1.

Self-reported experiences at work based on the Maslach Burnout Inventory (MBI). In panel A, scores of < 33 reflect low levels of personal accomplishment and are a sign of burnout, while scores of 34–39 are in the moderate range and indicate a risk for burnout. In panel B, depersonalization scores > 12 are high and suggest current burnout, while scores of 6–11 are moderate and indicate a risk for burnout. Data is reported as proportion of participants in each category with its associated 95% confidence interval. Adjusted scores are based on participants’ actual self-reported STS results (the original scores) plus an 18% correction factor for under-reporting. Out of 647 total respondents, 598 provided sufficient data to score the MBI, a 92.4% response rate.

A second burnout scale, depersonalization, suggests a more severe burnout reaction in which people start to see their teammates or the people they rescue as objects rather than as people. In our sample of SAR volunteers, 157 (26.3%) are experiencing moderate to high levels of depersonalization. Our adjusted values show that 187 (31.3%) may be experiencing moderate to high levels of depersonalization after correction for under-reporting (Figure 1, Panel B).

Traumatic Stress

Figure 2 shows SAR volunteers’ scores on 3 symptom scales related to traumatic stress. Participants’ STSS responses showed that 161 (26.4%) avoid some people or responsibilities related to their SAR work because of stress (Panel A). Our adjusted values show that 228 (37.3%) are experiencing levels of avoidance that probably interferes with their work or everyday life.²³ There were 193 (31.6%) participants experiencing moderate to high levels of stress-related physiological arousal (Panel B), or up to 242 (39.6%) of participants after adjustment for under-reporting. There were 156 (25.5%) participants who experience moderate level of intrusion, compared to our adjusted value of 223 (36.3%) (Panel C).

Figure 2.

Self-reported traumatic symptoms based on the Secondary Traumatic Stress Scale (STSS). The STSS is scored based on the three main traumatic symptoms including arousal (Panel A), avoidance (Panel B), and intrusion (Panel C). STSS level is based on a cumulative number, scored as 0 = low (not at all), 1–2 = moderate (mildly or moderately) and 3–4 = high (severely or very severely). Data is reported as proportion of participants in each category with its associated 95% confidence interval. Adjusted scores are based on participants’ actual self-reported STS results (the original scores) plus an 18% correction factor for under-reporting. Only 48 out of 657 participants did not provide sufficient data to calculate STS scores, resulting in a 92.7% response rate for this instrument.

Depression and Suicide Risk

In an unexpectedly low result, 27 (4.5% 95% CI: 2.9–6.1) SAR respondents reported clinically significant depressive symptoms, or an adjusted value of 32 (5.2% 95% CI: 3.4–7). We also asked a less-stigmatizing question related to depression: due to the nature of my SAR work, I am more irritable with loved ones. We found that 153 (24.4%) of SAR volunteers agreed with this statement, or an adjusted value of 167 (26.7%). Despite the low reported rate of depression, we nevertheless found that 59 (9.6% 95% CI: 7.3–11.9) of SAR volunteers are at significant risk for suicide. After adjusting, we estimate that as many as 69 (11.3% 95% CI: 8.8–13.8) are at risk for suicide.

Substance Use

139 (21.1%, 95% CI: 17.9–24.3) of respondents endorsed binge drinking (defined as 4 or more drinks at a single sitting) at least once or twice during the past year, and an additional 200 (30.3%, 95% CI: 26.7–33.9) said they had engaged in binge drinking 3 or more times in the previous year (Figure 3, Panel A). The total binge drinking rate was estimated to be 51%. Only 13 (2.1%) said that they used any prescription medication (e.g., opioids) other than how it was prescribed (Figure 3, Panel B). Tobacco products were used by 90 (14.8%) of the survey respondents (Figure 3, Panel C). Drugs, including marijuana, were used by 147 (22.2%) of the survey respondents (Figure 3, Panel D).

Figure 3.

Self-reported substance uses on four separate items from the Substance Use Brief Screen (SUBS), showing the number of times using substances in the past 12 months among SAR volunteers. For alcohol (Panel A), this represents the number of episodes of binge drinking (4+ drinks at one sitting). For opioid use (Panel B), represents use of prescribed painkiller medication other than the way in which it was prescribed, and therefore does not include heroin or other non-prescribed opioid use. Tobacco use (Panel C) represents use of any tobacco products. Drug use (Panel D) represents any other drug use including marijuana. Data are reported as proportion of participants in each category with its associated 95% confidence interval. Adjusted scores are based on participants’ actual self-reported STS results (the original scores) plus an 18% correction factor for under-reporting. The number of BSAR workers with missing data ranged from 48 (alcohol item) to 52 (drug use item) out of 657 total respondents, for a 92.1% overall response rate on the SUBS.

Overlap between Problem Areas

As shown in Table 2, only 31.1% of BSAR workers had no problems in any of the three areas of physical health, mental health, or substance use. The other two-thirds had problems that were likely to be clinically important in at least one of the three areas. For 30.6% the issue was substance use only, making this the single most important problem area among BSAR workers. For 5.5% the problem involved physical health only, and for 4.3% the problem involved mental health only. The greatest area of comorbidity was between mental health and substance use, with 11.6% BSAR workers having both of these problems, and another 6.2% having problems in these two areas and also with physical health. Physical health and substance use problems co-occurred without any mental health symptoms in 8.2% of BSAR workers, and the least frequent combination was mental health and physical health symptoms with no reported substance use.

Table 2.

Co-Occurrence of Physical, Mental Health, and Substance Use Problems

		Poor or Moderate Health N (%)	Good or Excellent Health N (%)
Mental Health Problem	With High Risk Substance Use	41 (6.2%)	76 (11.5%)
	Lower Risk or No Substance Use	17 (2.6%)	28 (4.2%)
No Mental Health Problem	With High Risk Substance Use	54 (8.2%)	201 (30.5%)
	Lower Risk or No Substance Use	36 (5.5%)	207 (31.4%)

Note. A significant mental health problem indicates a score in the clinically meaningful range on any of the 3 subscales of the Secondary Traumatic Stress scale, the PHQ-2 depression screener, and/or the short-form Beck Hopelessness Scale suicide screener. “High risk substance use” reflects a score of 3 (binge drinking 3+ times per year) on the SUBS alcohol question, or scores of 2–3 (any use) on the questions about non-prescribed opioid use or other drug use. All counts presented in this table use the more conservative, unadjusted results for each of the survey tools.

DISCUSSION

We found a level of stress and burnout among survey respondents that is likely to have meaningful consequences for SAR volunteers’ performance, relationships, and health. We found considerable evidence for psychological risk factors that are likely to interfere with functioning and can also predispose SAR volunteers to stress injuries and adverse health outcomes. Based on the observed level of risk for physical health problems alone, about 20% of volunteers may need to be replaced within the next 5 years due to medical disability or death.³³ That level of turnover – about 4% a year on average – is only manageable if there is a steady supply of new SAR volunteers willing to do the work and if other factors such as mental health concerns or substance use do not further accelerate the departure of current SAR volunteers from the workforce.

The level of alcohol use reported by SAR volunteers creates additional risk for mental health conditions such as depression and physical health conditions like liver disease, high blood pressure, and diabetes. The prevalence of binge drinking in the general Colorado population has been estimated around 29.7% in previous years.³⁴ Our data suggest that binge drinking rates are higher among our SAR population. Binge drinking on 3 or more days in a year suggests that 30% of SAR volunteers are likely to meet the criteria for alcohol use disorder. A high level of alcohol use is also consistent with our general interpretation that SAR volunteers were under-reporting their mental health concerns because people frequently use alcohol as a form of self-medication to reduce their anxiety, depression, or other mental health symptoms.

The overlap of burnout and secondary trauma with depression and PTSD has been well documented among health care professionals, where burnout affects nearly 10–70% of nurses and 30–50% of physicians, with higher prevalence rates among emergency and critical clinical settings.^6,14,35 Among the general population, in adults aged 18 years or older, the prevalence of major depression has been reported at 7.8%.³⁶ A cross-sectional study of Austrian physicians found that 10% were affected by major depression, and about 51% were affected by burnout. The OR of experiencing burnout in physicians without depression was 3 compared to 10, 47, and 93 of physicians experiencing mild, moderate, and severe signs of depression, respectively.³⁷ By contrast, our findings did not show substantial overlap between burnout and depression among SAR volunteers, given that only a small percentage of reported experiencing depression compared to many more who were experiencing burnout.

Because our findings did not show a substantial overlap between burnout and depression, the use of depression and burnout inventories may not help to identify depression among SAR volunteers; however, these tools still might be used to identify risk factors to prevent the development of depression. It may be more useful to administer tools such as the STSS that identify post-traumatic stress, physical health screening tools like the SF-36, and tools like the SUBS that identify substance use problems. Screening tools have public health implications because, without a sensitive way to screen for stress injuries, such injuries will go largely unnoticed, limiting the resources needed to rehabilitate volunteers effectively.

In a prior cross-sectional survey of 51 rescue volunteers, 82% reported being in excellent or very good health, consistent with our observations of general health status among our population.³ In various cross-sectional studies, rescue volunteers and emergency nurses reported an average prevalence of 10–30% for traumatic stress symptoms, which is lower than our population. However, our numbers are consistent with what is found among other first-responder populations.^11–13,38 Among the general population, rates of PTSD have been reported at 3.6%.³⁹ Our findings suggest that rates of PTSD are significantly higher among our population, consistent with what has been reported previously in the literature for other first-responder groups.

In a systematic review by Haugen et al., 9% of first responders endorsed barriers to care and the need for mental health services⁴, which is much lower than the actual prevalence of mental health problems that we found. The data suggest that SAR volunteers may have a greater need for mental health services than what has been reported previously and that SAR volunteers have a comparable prevalence of traumatic stress to emergency health workers. The level of burnout among SAR volunteers was slightly lower but similar to that among healthcare workers.^{14,16,35,37,40,41} Similarly, our results support the literature showing that more than half of our respondents were considered at risk of developing burnout, indicating that they may develop comparable levels of burnout to health care workers.

The high prevalence of traumatic stress and burnout among SAR volunteers is of clinical importance, as a significant proportion of volunteers may be experiencing adverse effects of trauma, and these symptoms may contribute to emotional exhaustion and increased volunteer turnover.^15,16,42 Our results suggest that medical problems will mainly drive SAR volunteer turnover, but that mental health and substance use problems create additional risk. We also found evidence that SAR volunteers under-reported their problems due to the stigma associated with mental health concerns. The literature shows that 33% of first responders endorse stigma regarding mental health care due to fears of confidentiality and negative career impact, which can potentially lead to underreporting and delayed presentation to care, therefore increasing the risk or chronicity of mental or physical health conditions.⁴

Given that first responders are at greater risk of experiencing stress, they are likely to engage in substance use as a form of coping and therefore have an increased risk for developing substance use disorders. Consistent with our results, in national surveys, it has been found that about 29% of firefighters are at risk for an alcohol use disorder, 25% of the police force report hazardous alcohol consumption, and around 30% of emergency responders are at risk for any substance use disorder.^43–46 In the general population, among adults aged 26 or older, 6.7% had a substance use disorder in 2019.³⁶ Our numbers suggest that SAR volunteers’ substance use is significantly higher than the general population and slightly higher than that of other first responder populations, which might reflect barriers to access for mental health services in the SAR population. Our interpretation is that the stress and trauma that SAR volunteers experience daily drives these professionals towards substance use to cope with the severe psychological harms.

LIMITATIONS

The main strength of this study is that it involved a large cohort comparable to those in similar published studies of other first-responder groups. Furthermore, we had a representation of SAR volunteers across the state and strong external validity. Second, we used validated tools to assess mental and physical health outcomes. The main limitation of our study is that it is cross-sectional, and we were not able to observe any longitudinal changes. Another limitation is that we believe there was under-reporting of problems in our sample; however, we created a correction factor to help estimate the impact of stigma on participants’ willingness to report problems. Finally, comparisons to other literature are limited because our survey was conducted in November 2021, after approximately 2 years of chronic stress created by the COVID-19 pandemic and right at the start of the Omicron wave of SARS-CoV-2 infections. Another limitation we considered is that the quick turnaround for the survey may have served as a barrier and led to selection bias. The extensive survey may have led to a response bias where participants may have filled out the survey for completion rather than answering each question truthfully. Although these limitations are present, we received a response rate between 26%−41% of all SAR volunteers in the state (depending on differences in the size of the denominator as estimated by various SAR experts), which is a respectable response rate even under the most conservative assumptions. Therefore, the survey length and short sampling window are not critical flaws in our design, and our results are still likely to be valid despite these potential limitations.

CONCLUSION

Increasing public participation in outdoor recreation and evolving challenges tied to the pandemic and resource limitations are creating unique mental health challenges for SAR volunteers across the country. This study explored the prevalence of mental health risk factors, stress levels, and burnout among SAR volunteers. The effects of stress injuries are of paramount importance to the SAR community as they can increase risk factors that lead to adverse physical and mental health outcomes and result in high turnover rates. Burnout and stress levels among SAR volunteers are higher than in the general population and are comparable to other first-responder groups, including firefighters, police, emergency medical responders, and emergency and critical care health workers. Results may be helpful to experts in other states or countries who oversee health and wellness initiatives among SAR volunteers in other back-country environments.

DATA AVAILABILITY STATEMENT

Data are available upon reasonable request to the corresponding author.