Does Where You Work and What You Do Matter? Testing the Role of Organizational Context and Job Type for Future Study of Occupation-Based Secondary Trauma Intervention Development

Abstract

Organizational context (e.g., criminal justice, community-based, and healthcare) and job type (e.g., police, social workers, and healthcare providers) may impact the extent of occupation-based secondary trauma (OBST). Survey data collected from a multiphase community-based participatory research project were analyzed from a variety of professionals, who were likely to “encounter the consequences of traumatic events as part of their professional responsibilities” (n = 391, women = 55%, White = 92%). Results document high trauma exposure (adverse childhood experiences [ACEs] and workplace) and OBST-related outcomes (Maslach Burnout Inventory, Secondary Traumatic Stress Scale, post-traumatic stress disorder symptom checklist for DSM-5) for the entire sample with important differences across organizational context and job type. Using multivariate regression, the strongest determinants of suffering, however, were not related to a provider’s specific profession but to their number of years on the job and their ACEs (e.g., adjusted R² = 0.23, b = 2.01, p < .001). Likewise, the most protective factors were not profession specific but rather the provider’s age and perceived effectiveness of OBST-related training (e.g., b = 2.26, p < .001). These findings inform intervention development and have implications for rural and other often under-resourced areas, where the same OBST-related intervention could potentially serve many different types of providers and organizations.

To what extent do differences in organizational context (e.g., criminal justice, community based, and healthcare) and job type (e.g., police officers, social workers, and healthcare providers) impact occupation-based secondary trauma (OBST) and related correlates? The general literature on secondary trauma has theorized the process through which working with traumatized victims leads to secondary trauma (e.g., Ellis & Knight, 2021; Figley, 1999; McCann & Pearlman, 1990), documented the subsequent individual and organizational toll (Baum, 2016; Cieslak et al., 2014; Hensel et al., 2015; Knight et al., 2018; Wood et al., 2019), and begun to offer solutions (Bober & Regehr, 2006; Clements et al., 2018; Sansbury et al., 2015; Vilardaga et al., 2011). For instance, Ellis and Knight (2021) developed a theoretical model of secondary trauma that shows the links among (a) primary trauma, (b) victim service provision, (c) repeated empathetic engagement with victims, and (d) damage to the perception of self in the form of unreliable self-agency, untrust-worthy coherence of others, desensitized self-affectivity, and fractured self-history. In terms of individual and organizational toll, Knight et al. (2018) found in their review of the literature (which includes several other systematic literature reviews) that being a woman, having a history of trauma, and the number of hours worked were some of the most predominant risk factors. Following extensive research, Clements et al. (2018) offered solutions to begin reducing secondary trauma, which is organized into three domains: (a) building awareness by learning to better understand the different forms of trauma, recognizing its consequences, and weighing risk factors against resilience; (b) practicing self-care by focusing on ongoing assessment of symptoms and cultivating body-based mindfulness practices; and (c) creating an organizational response by implementing trauma-informed practices in workplaces and improving social support among workers.

Overall, however, this body of research tends to focus on only one kind of professional (e.g., police officers or social workers), thus not accounting for how OBST and related correlates vary across different types of professions and organizations. Attending to this gap in research has important implications for the study and prevention of OBST (Bercier & Maynard, 2015), which is defined here simply as a form of trauma that results from repeated empathetic engagement with populations who experience primary trauma (Knight et al., 2018). If there are important differences in OBST across organizational contexts and job types, then careful attention to their role in the development, study, and implementation of interventions is needed. On the other hand, if few differences exist, then interventions can be used to widely serve different types of professionals working within a number of contexts and positions. This will be especially helpful to workers in less populated or under-resourced areas. Data from a multiphase community-based participatory research project (CBPR; Israel et al., 2013; Wallerstein & Duran, 2006) are used to answer this and related questions. Specifically, this study is descriptive in identifying risk and protective factors that may have interesting clinical implications warranting future study. Ultimately, our research will help providers—a term used here to inclusively refer to all criminal justice, community and social service, and healthcare workers who work interpersonally with traumatized others or in traumatizing contexts—and the organizations they work for more effectively address OBST.

The focus of the current study was to design and implement a survey to quantitatively assess OBST-related outcomes among providers at a community-wide level. We assessed OBST-related outcomes including symptoms of secondary traumatic stress (STS; Bride et al., 2004), post-traumatic stress disorder symptoms (PTSD; Weathers et al., 2013), and burnout-related exhaustion and cynicism as indicators of burnout (Maslach et al., 2016). Typically, these OBST-related outcomes are high in victim service professions, including nurses (Adriaenssens et al., 2015; Gomez-Urquiza et al., 2016), school personnel (Borntrager et al., 2012), social workers (Choi, 2011), child-protection workers (Conrad & Kellar-Guenther, 2006), police officers (Schaible & Gecas, 2010), and physicians (Shanafelt et al., 2012).

Researchers, however, debate the extent to which these OBST-related outcomes are overlapping or distinct constructs. Summarizing four extensive literature reviews, Elwood et al. (2011) argue that OBST constructs are subtly distinct, Molnar et al. (2017) see them as largely overlapping, as does Newell et al. (2016), whereas Sabin-Farrell and Turpin (2003) note both. This debate has been made more complex by changes to the diagnostic criteria for PTSD in the fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association [APA], 2013; Pai et al., 2017). The diagnostic criteria now include a new exposure type, “repeated or extreme exposure to aversive details of a traumatic event,” which applies specifically to “workers who encounter the consequences of traumatic events as part of their professional responsibilities” (p. 3). Few studies have compared differences in PTSD and other OBST-related outcomes across occupations (for exception, see Rhineberger-Dunn & Mack’s comparison of STS among juvenile detention and probation officers, 2020). We attend to this debate in the research by measuring four outcomes (STS, PTSD, and burnout-related exhaustion and cynicism) across three different organizational contexts (criminal justice, community and social services, and healthcare) and for several different job types (law enforcement, social workers, and healthcare providers among others).

A subsequent goal of the project involved assessing determinants of OBST. Prior research demonstrates that sociodemographic factors are important. Women (Baum, 2016) and younger workers (Baird & Jenkins, 2003; Bonach & Heckert, 2012; Lerias & Byrne, 2003; Sprang et al., 2007), for example, are at greater risk. Likewise, years on the job (Beaton et al., 1995; Cornille & Meyers, 1999) and workplace trauma exposure (Elwood et al., 2011; Lerias & Byrne, 2003) are important predictors. However, disaggregating workplace trauma from the trauma derived from ongoing adversities (e.g., having a low household income, a disability, or veteran status) and adverse experiences in childhood (e.g., sexual, physical, or emotional abuse; Centers for Disease Control & Prevention [CDC], 2019) is paramount to debates about the accuracy of data documenting prevalence and determinates, as well as debates about hiring practices and organizational accountability (Ellis & Knight, 2021). Research suggests that individuals exposed to early or ongoing trauma may select victim service professions at higher rates, which can have both advantages and disadvantages (Elliott & Guy, 1993; Farber et al., 2005; Nikčević et al., 2007). Few studies, nevertheless, assess the extent to which the effects of these risk factors can be mitigated by training in OBST (Molnar et al., 2017). Our study responds to this gap in research by measuring the amount of training providers have received and their perceptions of its effectiveness, as well as their interest in receiving more training. Ultimately, the goals of this study (e.g., examining how OBST and related correlates, including trauma history, vary across different types of professions and organizations) were informed by a theoretical model that we developed based on our qualitative research (Ellis & Knight, 2021). Our theoretical model of secondary trauma is grounded in the symbolic interaction perspective, specifically Stern’s (1985) concept of the core self. The current study will help empirically support this model which then can be used to guide training and inform future intervention work.

The Current Study

A better understanding of the different forms of trauma exposure (e.g., childhood versus workplace), the prevalence of OBST-related outcomes (e.g., STS, PTSD, and burnout-related exhaustion and cynicism), and their determinants (e.g., sociodemographic and training) by organizational context and job type will further the effectiveness of our larger CBPR study and the community’s response. This understanding will help the development of an evidence-based intervention while also contributing to the larger body of literature on OBST. To that end, and in close collaboration with our community advisory board (CAB) members, we developed, piloted, and implemented an online survey asking providers working in a variety of organizations to answer the following four research questions: (1) Are providers exposed to trauma? (2) Are they suffering from OBST-related outcomes? (3) Are there differences across organizational contexts and job types? and (4) What other factors are associated with increases and decreases in OBST-related outcomes?

Method

Project Background

After findings from a statewide qualitative victim need assessment documented OBST as a growing concern (Ellis & Knight, 2021), we launched a multiphase CBPR project. CBPR is a theoretically driven methodology focused on finding manageable, effective, sustainable, and ethical ways to respond to public health problems (Holkup et al., 2004). The first phase of the project involved organizing a cross-sector CAB representing seven different victim service professions (e.g., law enforcement, advocacy, healthcare), coauthoring a toolkit (booklet) with them for distribution to local victim service providers (Clements et al., 2018), and hosting a daylong training for providers to encourage organizational response to the issue. While not the focus of this study, findings showed that execution of the first phase of the project was overwhelmingly positive (Knight et al., 2022). Results from evaluation questionnaires representing a wide range of occupations document that the implementation of the project was successful, toolkits created for the project were useful to both individual participants and organizations, and training outcomes improved significantly.

Data

Recruitment and data collection unfolded in the following stages. First, given their employment with or knowledge of victim service agencies in the local community, CAB members nominated 11 agencies for possible recruitment into the study: five law enforcement agencies (one county sheriff’s office, two city police departments, one university police department, and one airport authority), one detention center, one child protective service (CPS) agency, one hospital (targeting emergency-room personnel and sexual assault nurse examiners), and three community-based organizations, each providing a variety of programs and services (e.g., responding to domestic violence, sexual assault, suicide, and housing and transportation crises). Second, emails soliciting participation in the survey were sent to these agencies’ leaders. Participants had the option to receive all three of the following incentives: (a) individual participants would receive a $20 e-gift card for completing the survey, (b) individual participants would be eligible for selection into a pilot randomized controlled trial testing a multiday training intervention for secondary trauma, and (c) each participating agency would receive an individualized report summarizing the anonymized results of their employees’ surveys. Third, the principal investigators met with all but two agencies to further explain the study. A CAB member, serving as a trusted gatekeeper, was also present at seven of these meetings. All 11 agencies agreed to participate (yielding a 100% participation rate for all organizations contacted). Fourth, agency leaders were emailed a link to an online survey that they could then forward to their employees. In addition, agency leaders were given flyers and stacks of small cards that advertised the survey. Fifth, weekly reminder emails were sent to encourage participation.

The survey was designed in Qualtrics and included one validity item asking participants to “please select ‘very often’ for this response” to help assess whether responses were valid. The survey took approximately 30 to 45 minutes to complete. At the end of the survey, if participants elected to receive the $20 e-gift card, they were redirected to a separate web page to provide contact information. The study was conducted during the 2018 to 2019 academic year and all procedures were approved by the Montana State University Institutional Review Board.

Sample

The recruitment and data collection procedures yielded a community-wide sample of n = 431 providers. However, participants who worked as volunteers (n = 8; 2%), participants who failed to complete more than 35% of the survey (n = 19; 4%), and those who failed to answer the validity item correctly (n = 13; 3%) were dropped from the analysis. Altogether, this resulted in a final analytic sample for this study of n = 391. To calculate participation rates for all potential participants within each agency, agency leaders were asked to provide the number of employees working at the time the survey was administered. Only the hospital was not able to respond to this request. At the agency level, the participation rate for this online survey was 50, 50, 54, 71, and 78% for the law enforcement agencies; 70% for the detention center; 77% for CPS; and 49%, 84%, and 95% for the community-based agencies. In terms of organizational context, the mean participation rate across agencies was 61% (standard deviation [SD] = 12%) for criminal justice, 62% (SD = 12%) for community and social services, and the data were not available for healthcare.

Measures

For the current study, the following types of measures are included: (a) job type and organizational context, (b) trauma exposure, (c) OBST-related, (d) training, and (e) demographic variables.

Job Type and Organizational Context Variables.

Participants were asked to provide their job title (by completing a text-entry response), occupation (by selecting from 22 U.S. Department of Labor categories; U.S. Bureau of Labor Statistics, 2020), employer (by choosing from a drop-down menu of recruited agencies or typing in an “other” response), and years on the job. A total of 193 different but often similar job titles were given (e.g., police, police officer, and police detective). These were compared to the two questions asking participants about their occupation and employer and then coded into 10 job types: administrators, detention officers, educators and trainers, healthcare technicians, law enforcement officers, nurses, office and administrative support, physicians, social workers, and others. In addition, 18 different employers were provided and then coded into three organizational contexts: criminal justice (e.g., law enforcement agencies and the detention center), community and social services (e.g., community-based organizations and CPS), and healthcare (e.g., the hospital and emergency medical services). Results are presented in Table 1.

Table 1.

Organizational Contexts and Job Types (n = 391).

	Sample Size
	n	%
Organizational context:
Criminal justice	188	48.3
Community and social services	127	32.6
Healthcare	74	19.0
Job type:
Law enforcement officers	127	32.5
Social workers	80	20.5
Detention officers	38	9.7
Healthcare technicians	38	9.7
Office and administrative support	32	8.2
Nurses	31	7.9
Educators and trainers	16	4.1
Other	15	3.8
Administrators	8	2.0
Physicians	6	1.5

Note. Two cases missing data for organizational context.

Trauma Exposure Variables.

Participants were asked about potential trauma exposure in childhood and in the workplace. Adverse childhood experiences (ACEs) were assessed using the Behavioral Risk Factor Surveillance System Adverse Childhood Experiences Module (CDC, 2019). The questionnaire consists of 11 items asking about the participant’s exposure to eight different types of negative experiences before the age of 18: emotional abuse, physical abuse, sexual abuse, intimate partner violence, household substance abuse, household mental illness, incarcerated household member, and parental separation or divorce. Following Merrick et al. (2018), a prevalence indicator was created for each of the eight negative experiences. These were summed to generate the total number of ACEs. Participants’ level of workplace trauma exposure was assessed by asking, “Overall, how much of your work involves (directly or indirectly) engaging with traumatized individuals or traumatic material?” Responses were on a 5-point scale ranging from 0 (none at all) to 4 (a great deal). Although research shows that retrospective measures like the ACEs questionnaire have methodological limitations, they nevertheless have practical utility and are predictive of important outcomes (Baldwin et al., 2019).

OBST-Related Variables.

Participants were asked about four outcomes related to secondary trauma. The Maslach Burnout Inventory General Survey (MBI-GS; Maslach et al., 2016) is a fee-based self-report instrument designed to assess professional burnout. Participants were asked to respond to 16 statements describing job-related feelings, such as, “I feel emotionally drained from my work” and “I have become more cynical about whether my work contributes anything.” Responses are on a 7-point scale denoting how often the participant feels that way about their job, ranging from 0 (never) to 6 (every day). The MBI-GS includes two subscales: burnout-related exhaustion (5 items; α = .92), or the sense of being overextended and drained by work; and burnout-related cynicism (5 items; α = .87), or the level of indifference felt toward work. Scores were calculated in two ways, each using all 16 items. First, items are summed, and a higher score indicates a higher risk for burnout. Second, we calculated means, ranging from 0 to 6, because Maslach et al. (2016) and the literature more broadly do not provide cut-scores. Participants with scores less than or equal to 2.99 were categorized as low, 3 to 4.99 as medium, and greater than or equal to 5 as high. Note that the second method is only used for comparative purposes in our descriptive analyses.

The Secondary Traumatic Stress Scale (STSS; Bride et al., 2004) is a self-report instrument designed to assess STS resulting from the workplace. Participants were asked to read 17 statements, such as, “My heart started pounding when I thought about my work with clients,” “It seemed as if I was reliving the traumas experienced by my clients,” and “I avoided people, places, or things that reminded me of my work with clients.” Participants responded by indicating how frequently each statement was true for them in the past 7 days using a 5-point scale ranging from 1 (never) to 5 (very often). The STSS contains three subscales: intrusion (5 items), avoidance (7 items), and arousal (5 items). Following Bride et al. (2007), all items are summed to create a total STS symptoms score (α = .93). Respondents with a score lower than 27 are considered to have little or no STS, a score of 28 to 37 indicates mild STS, a score of 38 to 43 qualifies as moderate, a score of 44 to 48 is considered high, and a score of 49 and greater is severe. Moderate or higher STS prevalence was calculated for respondents with a score of at least 38 (n = 172).

The PTSD checklist for the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) (PCL-5; Weathers et al., 2013) is a self-report questionnaire used to screen for symptoms of PTSD, which can occur from work-related repeated or extreme exposure to aversive details of traumatic events (APA, 2013). The PCL-5 includes a list of 20 symptoms that a person might have in response to a very stressful experience, such as having disturbing dreams, losing interest in previously enjoyable activities, feeling cut off from other people, struggling to fall asleep, and acting uncharacteristically irritable or aggressive (Weathers et al., 2013). The participant is asked to rate how often they were bothered by each symptom in the past month. The rating scale includes five response options ranging from 0 (not at all) to 4 (extremely). The PCL-5 contains items that correspond to the DSM-5: criterion B, symptoms of intrusion (5 questions, α = .90); criterion C, trauma-related stimuli (2 questions, α = .86); criterion D, negative thoughts and feelings (7 questions, α = .80); and criterion E, arousal symptoms (6 questions, α = .90). First, the total PTSD symptoms score (α = .95) was obtained by summing all 20 questions and the score can range from 0 to 80. Then, a prevalence variable was created to indicate a score of 33 or higher which is an optimal score for diagnosing PTSD (Bovin et al., 2016).

Last, a prevalence variable was created to indicate which providers were suffering from at least one of the OBST-related outcomes at the moderate level (i.e., participants who scored medium or higher on exhaustion or cynicism, who scored moderate or higher on STS, or who may be suffering from PTSD given that they had a score of 33 or higher).

Training Variables.

Four questions asked participants about their secondary trauma training experiences. To assess the amount of training received, one question asked, “How much training have you received on secondary trauma (including topics like vicarious trauma, secondary traumatic stress, burnout, and moral distress)?” Responses to training amount were on a 5-point scale ranging from 0 (none at all) to 4 (a great deal). For participants who had at least a little training, they were asked, “Altogether, how effective were your training on secondary trauma (including topics like …?” Responses to training effectiveness were on a 5-point scale ranging from 0 (not effective at all) to 4 (extremely effective); participants with 0 training amount were coded as 0 on training effectiveness. Participants were also asked, “Do you want more training on secondary trauma (including topics like …?” Responses to training interest were on a 5-point scale ranging from 0 (definitely not) to 4 (definitely yes).

Demographic Variables.

Participants were asked demographic questions including age and gender (female, male, or other which allowed a text entry). They could select all applicable racial categories listed (American Indian, Alaskan Native, Asian, Black or African American, Hispanic or Latino, Native Hawaiian or Other Pacific Islander, White, or Other which allowed a text entry). Participants were asked to indicate the highest level of education that they had achieved (high school or equivalent, some college or post-high school training, an associate’s or 2-year degree, bachelor’s degree, graduate or professional degree, or did not complete high school). They were also asked about marital status (married, living with a partner but not married, divorced or separated, single, widow, or other which allowed a text entry box). Last, participants were asked about household income (using 14 categories ranging from under $9,999 to $130,000), disability status (attention deficit hyperactivity disorder/attention deficit disorder, learning, psychological, traumatic brain injury, or other which allowed a text entry), and veteran status (yes or no).

Analytic Strategy

Data were analyzed using SPSS, Version 25. To answer the first and second research questions, descriptive statistics assessed the extent of trauma exposure and the prevalence of OBST-related outcomes. To answer the third research question, OLS regression models tested differences in four OBST-related outcomes by both organizational context and job type. Note that intraclass correlations were calculated, and findings suggested little clustering on outcomes at the agency level (p < .05). As such, multilevel regression modeling was not warranted. To answer the fourth research question, four multivariate OLS regression models tested the extent to which social demographic, trauma exposure, work, and training factors were associated with OBST-related outcomes. Missing data are minimal (2% on average, reported in Table 2) except for four variables that were added to the survey approximately 1 month after the survey was launched: disability status (28% missing), veteran status (28% missing), and 1-item assessment of workplace trauma exposure (19% missing) and of training amount (7% missing), separately. To help retain the overall sample size, these missing cases were controlled for by creating dummy variables, including a missing indicator, and categorizing the two variables, workplace trauma exposure and training amount, into low, moderate, and high levels (whereby low served as the reference category). Otherwise, case-wise deletion was used. For each model, multicollinearity was tested, standardized beta coefficients were used to compare the relative strength of predictors, and adjusted R² was reported to assess explained variance.

Table 2.

Descriptive Statistics: Social Demographic, Work, and Training Factors (n = 391).

	Variable	Prev	Mean	SD	Range	Missing	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.	16.	17.
1.	Age	—	37.43	11.14	19–72	9	—
2.	Men	45.00	—	—	0–1	0	0.06	—
3.	Persons of color	7.70	—	—	0–1	2	−0.03	−0.13**	—
4.	Bachelor’s degree or higher	66.80	—	—	0–1	0	−0.03	−0.23***	0.00	—
5.	Married or living together	72.20	—	—	0–1	3	0.18***	0.18***	0.02	0.04	—
6.	Low household income	19.40	—	—	0–1	5	−0.27***	−0.30***	0.03	−0.01	−0.23***	—
7.	Disability	13.50	—	—	0–1	110	−0.03	0.07	0.08	−0.15*	0.00	0.05	—
8.	Veteran	16.80	—	—	0–1	111	0.15*	0.31***	0.02	−0.11	0.05	−0.16**	0.41***	—
9.	Years in job	—	5.79	6.18	0.08–25	2	0.47***	0.22***	−0.40	−0.13**	0.16	−0.29***	0.01	0.03	—
10.	Adverse childhood experiences	19.70a	2.02	1.96	0–8	11	0.02	−0.23***	0.21***	−0.01	−0.14**	0.07	0.07	0.06	−0.05	—
11.	Workplace trauma exposure	87.4b	2.97	1.11	0–4	74	−0.15**	0.01	0.05	0.06	0.02	0.04	−0.02	−0.04	−0.01	0.09	—
12.	Burnout-exhaustion	49.2c	14.01	7.60	0–30	3	−0.09	−0.11*	0.03	0.13**	−0.05	0.08	0.15**	0.05	0.08	0.13**	0.14*	—
13.	Burnout-cynicism	28.5c	10.59	7.46	0–30	4	−0.02	0.12*	0.00	0.01	0.06	−0.12*	0.06	0.04	0.25***	0.00	0.07	0.68***	—
14.	STS symptoms	46.1d	36.89	12.82	17–75	18	−0.12*	−0.09	0.07	0.06	−0.12*	0.04	0.12*	0.07	0.06	0.24***	0.23***	0.66***	0.59***	—
15.	PTSD symptoms	15.1e	16.54	14.63	0–65	19	−0.17**	−0.12*	0.11*	−0.03	−0.18***	0.15**	0.18**	0.10	−0.01	0.32***	0.15**	0.53***	0.39***	0.74***	—
16.	Training amount	—	1.71	1.12	0–4	27	−0.01	0.01	0.12*	0.01	0.04	−0.10*	0.13*	0.11	0.06	0.05	0.24***	−0.10	0.13*	−0.04	−0.01	—
17.	Training effectiveness	—	1.81	1.02	0–4	16	0.07	−0.11*	0.10	0.07	0.01	0.18	−0.04	0.00	−0.07	0.00	0.01	−0.29***	−0.31***	−0.23***	−0.18***	0.59***	—
18.	Training interest	—	2.74	1.13	0–4	17	−0.16***	−0.30***	0.09	0.14**	−0.09	0.18***	0.00	−0.09	−0.15**	0.16**	0.16**	0.12*	−0.10*	0.17**	0.25***	−0.02	0.06

Note. Variables 7, 8, 11, and 16 were added late to the survey. Prev = prevalence; SD = standard deviation; MBI = Maslach burnout inventory; STS = secondary traumatic stress; PTSD = post-traumatic stress disorder; PCL-5 = post-traumatic stress disorder symptom checklist for DSM-5.

^aPercent of cases that scored four or more ACEs.

^bPercent of cases that scored moderate or higher amounts of workplace trauma exposure.

^cPercent of cases that scored medium or higher on MBI.

^dPercent of cases with a score of 38 or higher (moderate STS; Bride et al., 2007).

^ePercent of cases with a score of 33 or higher using PCL-5 items (optimal score for diagnosing PTSD; Bovin et al., 2016).

^***p ≤ .001.

^**p ≤ .01.

^*p ≤ .05, t ≤ .10 (two-tailed).

Results

Participants reported demographic characteristics including age (M = 37, SD = 11.14) and whether they identified as male (45%), identified as a person of color (8%), had earned a bachelor’s degree or higher (67%), were married or living together (72%), had a low household income (19%; defined as less than $29,999, the third lowest of 14 categories ranging from under $9,999 to $130,000), had a documented disability (14%), and had veteran status (17%, see also Table 2). These demographic variables are often correlated with trauma and other related mental health issues more broadly (Baum, 2016; Elwood et al., 2011).

Research Question 1: Are Providers Exposed to Trauma (in Childhood and Later in the Workplace)?

First, we assessed providers’ exposure to trauma, both in their childhoods and in the workplace. For the entire sample, 71.8% had experienced at least one ACE. Emotional abuse (54.0%), household mental illness (37.9%), and household substance abuse (29.2%) were the most frequent categories. When counting ACEs (range = 0–8, M = 2.02, SD = 1.96), 19.7% reported four or more, which is higher than the national average of 15.81% (Merrick et al., 2018). In terms of workplace trauma exposure, 1.9% of respondents reported none, 10.7% reported a little, and 87.4% reported at least a moderate amount or more (range = 0–4, M = 2.97, SD = 1.11). The correlation between ACEs and workplace trauma exposure was not statistically significant. Nonetheless, the majority of providers (63.4%) reported exposure to both adversity during their childhoods (at least one ACE) and trauma in their workplaces as adults (moderate or higher). Altogether, these results suggest that employees have a high probability of trauma exposure, either in childhood or adulthood, and are at risk for OBST-related outcomes.

Research Question 2: Are Providers Suffering From OBST-Related Outcomes (Exhaustion, Cynicism, STS, and PTSD)?

Second, four measures helped assess the extent to which providers were suffering from OBST-related outcomes: exhaustion, cynicism, STS, and PTSD. For exhaustion (range = 1–30, M = 14.01, SD = 7.60), 40.4% of respondents were in the medium range and 8.7% were in the high range of the MBI-GS. Cynicism was somewhat lower (range = 1–30, M = 10.59, SD = 7.46) with 23.3% in the medium range and 5.1% in the high range of the MBI. STS was prevalent (range = 17–75, M = 36.89, SD = 12.82) with 28.2% reporting little to none, 25.7% mild, 15.5% moderate, 11.0% high, and 19.6% reporting severe symptoms on the STSS. Likewise, providers appeared to be suffering from STS with 46.1% reporting moderate STS symptoms or higher. For PTSD symptoms (range = 0–65, M = 16.54, SD = 14.63), the rate was lower than STS, with 15.1% reporting symptoms indicative of a provisional diagnosis of PTSD using the PCL-5. Altogether, these results suggest that almost two-thirds (62.7%) of providers were suffering from at least one of the four outcomes (at a moderate level or higher) and that outcomes were highly correlated (see Table 2).

Research Question 3: Are There Differences in OBST-Related Outcomes Across Organizational Contexts and Job Types?

Third, significant differences in exhaustion, cynicism, STS, and PTSD were explored by organizational context and job type. The most frequent organizational context, criminal justice, and most frequent job type, law enforcement officer, served as the reference categories in the OLS models used to test differences. Results are reported in Table 3, and significant findings are summarized below.

Table 3.

Differences in OBST-Related Outcomes Across Organizational Contexts and Job Types (n = 391).

	Burnout-Exhaustion				Difference			Burnout-Cynicism				Difference			STS Symptoms				Difference			PTSD Symptoms				Difference
	n	Mean	SD	Range	b	SE	p	n	Mean	SD	Range	b	SE	p	n	Mean	SD	Range	b	SE	p	n	Mean	p	Range	b	SE	p
Organizational context
Criminal justice (reference)	187	12.94	8.10	0–30	—	—	—	186	11.44	8.04	0–30	—	—	—	176	34.80	13.17	17–72	—	—	—	178	14.36	14.72	0–65	—	—	—
Community and social services	126	14.54	6.86	1–29	1.60	0.87	0.06t	126	9.01	6.61	0–28	−2.43	0.85	0.01**	127	38.14	12.61	17–74	3.33	1.47	0.02*	124	19.12	14.99	0–63	4.76	1.70	0.01**
Healthcare	74	16.01	6.95	1–29	3.08	1.03	0.00***	74	11.30	6.88	1–27	−0.14	1.02	0.89	69	40.20	11.29	17–75	5.40	5.40	0.00***	69	17.78	12.92	0–57	3.42	2.06	0.10t
Job type
Law enforcement officers (reference)	126	13.19	8.28	0–30	—	—	—	125	12.23	7.97	0–30	—	—	—	119	36.13	13.35	17–72	—	—	—	120	14.19	13.48	0–56	—	—	—
Social workers	79	14.65	7.09	1–29	1.46	1.09	0.18	79	8.37	6.35	0–28	−3.87	1.06	0.00***	80	38.60	12.57	17–74	2.47	1.83	0.18	77	19.65	16.42	0–63	5.46	2.12	0.01**
Detention officers	38	12.45	7.86	0–30	−0.74	1.40	0.60	38	9.87	8.57	0–29	−2.36	1.37	0.09t	34	31.24	12.94	17–69	−4.90	2.46	0.05*	36	13.47	17.66	0–65	−0.72	2.76	0.79
Healthcare technicians	38	15.39	7.27	1–29	2.20	1.40	0.12	38	11.68	6.21	2–27	−0.55	1.37	0.69	38	41.26	12.93	17–75	5.13	2.36	0.03*	37	19.62	14.20	0–48	5.43	2.73	0.05*
Office and administrative support	31	12.26	7.80	0–27	−0.93	1.52	0.54	31	9.97	7.84	0–26	−2.26	1.48	0.13	29	33.59	12.68	17–63	−2.55	2.62	0.33	30	18.13	14.87	0–57	3.94	2.96	0.18
Nurses	31	15.39	7.42	2–28	2.20	1.52	0.15	31	10.51	7.41	1–25	−1.72	1.48	0.25	29	36.31	8.33	18–51	0.18	2.62	0.95	29	12.83	8.83	0–39	−1.36	3.00	0.65
Educators and trainers	16	17.56	6.66	5–28	4.37	2.01	0.03*	16	10.56	7.60	20–64	−1.67	1.96	0.40	16	40.56	11.87	20–64	4.43	3.37	0.19	15	19.60	12.36	0–39	5.41	3.97	0.17
Other	15	14.47	6.28	2–23	1.33	1.56	0.40	15	10.93	6.10	0–22	−2.40	1.52	0.12	15	40.47	14.19	17–61	1.90	2.66	0.48	15	21.93	15.31	0–50	4.27	3.04	0.16
Administrators	8	13.13	5.22	9–25	—	—	—	8	7.75	5.70	1–19	—	—	—	8	35.13	13.64	20–61	—	—	—	8	17.00	15.24	2–49	—	—	—
Physicians	6	16.50	3.50	13–23	—	—	—	6	9.83	8.13	13–23	—	—	—	6	35.40	9.94	26–48	—	—	—	6	10.40	4.93	6–18	—	—	—

Note. OBST = Occupation-Based Secondary Trauma; SD = standard deviation; b = regression coefficient; SE = standard error; p = p-value.

^***p ≤ .001.

^**p ≤ .01.

^*p ≤ .05, t ≤ .10 (two-tailed).

Organizational Contexts (Criminal Justice, Healthcare, and Community Based).

Compared to participants working in criminal justice, only participants working in healthcare reported more exhaustion (b = 3.08, p < .01) and only participants working in community-based organizations reported less cynicism (b = −2.43, p < .01). In addition, compared to criminal justice, participants working in both community-based (b = 3.33, p < .05) and healthcare (b = 5.40, p < .01) organizations reported more STS symptoms. Lastly, compared to criminal justice, only participants working in community-based organizations reported more PTSD symptoms (b = 4.76, p < .01). Stated somewhat differently, participants in healthcare reported the most exhaustion, participants in community-based organizations reported the least cynicism, participants in community-based and healthcare reported the most STS symptoms and participants in community-based organizations reported the most PTSD symptoms. Altogether, these findings suggest important differences in outcomes that warrant further exploration.

Job Types.

To further explore differences in outcomes, eight different job types were compared (job types with fewer than 15 cases were recoded into the “other” category). Compared to law enforcement officers, only educators and trainers reported more exhaustion (b = 4.37, p < .05) and only social workers reported less cynicism (b = −3.87, p < .001). In addition, compared to law enforcement officers, detention officers reported fewer STS symptoms (b = −4.90, p < .05), whereas healthcare technicians reported more (b = 5.13, p < .05). Last, compared to law enforcement, only social workers (b = 5.46, p < .01) and healthcare technicians (b = 5.43, p < .05) reported more PTSD symptoms. As stated differently, educators and trainers reported the most exhaustion, social workers reported the least cynicism, detention officers reported the least STS symptoms, whereas healthcare technicians reported the most, and social workers and healthcare technicians reported the most PTSD symptoms. When taken together, these findings suggest unique differences in outcomes by both organizational context and job type.

Research Question 4: What Factors Are Associated With Increases and Decreases in OBST-Related Outcomes?

Fourth, the extent to which organizational context and job type were associated with outcomes was assessed, net of other factors. Specifically, we tested social demographic, work, and training factors in four multivariate OLS regression models where the dependent variables were either exhaustion, cynicism, STS, or PTSD symptoms. Results for Models 1 through 4 are reported in Table 4 and are organized below by factors associated with increases in outcomes and by factors associated with decreases in outcomes.

Table 4.

Multivariate OLS Regression Models Testing Social Demographic, Work, and Training Factors.

	Model 1				Model 2				Model 3				Model 4
	DV = Exhaustion				DV = Cynicism				DV = Secondary Traumatic Stress Symptoms				DV = Post-Traumatic Stress Symptoms
	b	SE	b *	p	b	SE	b *	p	B	SE	b *	p	b	SE	b *	p
Constant	13.34	2.71	—	.00**	15.32	2.62	—	.00**	33.07	4.34	—	.00**	13.67	4.94	—	.01**
Age	−0.09	0.04	−0.13	.05**	−0.12	0.04	−0.18	.01**	−0.19	0.07	−0.17	.01**	−0.28	0.08	−0.21	.00**
Men	−1.57	1.19	−0.10	.19	−0.63	1.16	−0.04	.59	−0.68	1.93	−0.03	.73	1.15	2.18	0.04	.60
Persons of color	0.17	1.53	0.01	.91	0.66	1.48	0.02	.66	0.46	2.44	0.01	.85	2.19	2.83	0.04	.44
Bachelor’s degree or higher	2.27	0.94	0.14	.02*	1.40	0.91	0.09	.13	1.05	1.52	0.04	.49	−0.94	1.72	−0.03	.59
Married or living together	−0.54	0.91	−0.03	.56	0.47	0.88	0.03	.59	−2.08	1.47	−0.08	.16	−2.91	1.69	−0.09	.09t
Low household income	0.30	1.20	0.02	.81	−2.19	1.16	−0.12	.06t	−3.21	1.94	−0.10	.10t	0.97	2.20	0.03	.66
Disability	4.07	1.45	0.16	.01**	1.14	1.40	0.05	.42	4.04	2.33	0.10	.08t	4.54	2.64	0.09	.09t
Veteran	0.54	1.45	0.02	.71	−0.15	1.40	−0.01	.91	1.80	2.38	0.05	.45	3.67	2.62	0.08	.16
Adverse childhood experiences	0.41	0.21	0.11	.05*	0.20	0.20	0.05	.34	1.39	0.34	0.22	.00**	2.01	0.39	0.27	0.00***
Community-based organization	−0.55	2.36	−0.03	.82	0.09	2.28	0.01	.97	−2.13	3.76	−0.08	.57	−6.15	4.41	−0.20	.16
Healthcare organization	2.50	2.28	0.13	.27	−0.20	2.20	−0.01	.93	4.97	3.81	0.15	.19	0.89	4.29	0.02	.84
Social workers	0.21	2.47	0.01	.93	−2.77	2.39	−0.15	.25	1.05	3.96	0.03	.79	6.73	4.65	0.19	.15
Detention officers	1.09	1.49	0.04	.46	−0.11	1.44	0.00	.94	−3.57	2.46	−0.08	.15	−0.94	2.72	−0.02	.73
Healthcare technicians	−1.40	2.83	−0.06	.62	1.38	2.73	0.06	.61	0.54	4.69	0.01	.91	0.57	5.26	0.01	.91
Office and administrative support	−1.72	2.19	−0.06	.43	−0.97	2.12	−0.04	.65	−1.07	3.53	−0.02	.76	6.32	4.02	0.12	.12
Nurses	−2.59	2.67	−0.09	.33	−3.27	2.58	−0.12	.21	−5.68	4.38	−0.12	.20	−3.10	4.92	−0.06	.53
Other	0.47	2.56	0.02	.85	−1.71	2.48	−0.08	.49	1.14	4.10	0.03	.78	6.20	4.81	0.14	.20
Workplace trauma exposure—Moderate	1.05	1.59	0.05	.51	0.55	1.54	0.03	.72	2.94	2.55	0.09	.25	−1.88	2.88	−0.05	.51
Workplace trauma exposure—High	2.52	1.37	0.16	.07t	1.76	1.33	0.12	.19	7.60	2.23	0.30	.00**	4.04	2.48	0.14	.11
Years in job	0.22	0.08	0.18	.01**	0.33	0.08	0.27	.00***	0.36	0.12	0.18	.00**	0.34	0.14	0.14	.02*
Training amount—Moderate	0.32	1.05	0.02	.76	0.07	1.02	0.00	.95	0.48	1.71	0.02	.78	−0.54	1.93	−0.02	.78
Training amount—High	−0.62	1.29	−0.03	.63	0.47	1.25	0.03	.71	−0.11	2.08	0.00	.96	1.05	2.35	0.03	.66
Training effectiveness	−1.37	0.45	−0.20	.00***	−1.76	0.44	−0.26	.00***	−1.70	0.74	−0.15	.02*	−1.60	0.83	−0.12	.05*
Training interest	0.23	0.38	0.03	.55	−0.55	0.37	−0.08	.14	0.77	0.61	0.07	.21	2.26	0.69	0.17	.00***
Analytic n			355				355					347				345
Adjusted R2			0.10				0.11					0.17				0.23

Note. All models control for missing cases due to items being added late to the survey (i.e., disability, veteran, workplace trauma exposure, and training amount). Criminal justice, law enforcement officers, and low workplace trauma exposure served as reference categories. OLS = ordinary least squares; DV = dependent variable; b = regression coefficient; SE = standard error; b* = standardized regression coefficient; p = p-value.

^***p ≤ .001.

^**p ≤ .01.

^*p ≤ .05, t ≤ .10 (two-tailed).

Factors Associated With Increases in Outcomes.

In terms of social demographic factors, Model 1 shows that having a bachelor’s degree or higher (b = 2.27, p = .02) and having a disability (b = 4.07, p = .01) were significantly and positively associated with exhaustion. In addition, Models 1, 3, and 4 show that the number of ACEs was significantly and positively associated with three outcomes: exhaustion (b = 0.41, p = .05), STS (b = 1.39, p < .00), and PTSD (b = 2.01, p < .00) symptoms.

In terms of work factors, Model 1 shows that high workplace trauma exposure is significantly and positively associated with one outcome: STS symptoms (b = 7.60, p < .00). Moreover, Models 1 through 4 show that number of years in the job was significantly and positively associated with all four outcomes: exhaustion (b = 0.22, p = .01), cynicism (b = 0.33, p < .00), STS (b = 0.36, p < .00), and PTSD (b = 0.34, p = .02) symptoms. Last, training interest was significantly and positively associated with PTSD symptoms (b = 2.26, p < .00).

Comparing the relative strength of standardized beta coefficients within models, a number of years on the job was the strongest factor associated with exhaustion (Model 1), cynicism (Model 2), and STS symptoms (Model 3), whereas the number of ACEs was the strongest factor associated with PTSD symptoms (Model 4).

Factors Associated With Decreases in Outcomes.

In terms of social demographic factors, Models 1 through 4 show that being older in age was significantly and negatively associated with all four outcomes: exhaustion (b = − .09, p = .05), cynicism (b = − .12, p = .01), STS (b = − .19, p = .01), and PTSD symptoms (b = −0.28, p = .00). Likewise, in terms of work factors, training effectiveness was also significantly and negatively associated with all four outcomes: exhaustion (b = −1.37, p < .00), cynicism (b = −1.76, p < .00), STS (b = −1.70, p = .02), and PTSD symptoms (b = −1.60, p = .05).

Comparing the relative strength of standardized beta coefficients within models, training effectiveness was the strongest factor associated with decreases in exhaustion (Model 1), cynicism (Model 2), and STS symptoms (Model 3), whereas being older in age was the strongest factor associated with decreases in PTSD symptoms (Model 4).

In sum, these findings suggest that the most frequent and strongest predictor of suffering was a number of years on the job and, to a somewhat lesser extent, the number of ACEs and workplace trauma exposure. Conversely, both being older in age and training effectiveness were the most protective predictors associated with less suffering. Organizational contexts and job types were not significant in any of the models controlling for other factors. Altogether, these findings have important implications for victim service organizations, the people they employ, and intervention research that is needed to protect them.

Discussion

This study builds on prior work on OBST (for reviews, see: Adams et al., 2017; Bercier & Maynard, 2015; Bride et al, 2007; Baum, 2016; Elwood et al., 2011; Gomez-Urquiza et al., 2016; Knight et al., 2018; Maslach et al., 2016; Molnar et al., 2017; Newell et al., 2016; Sabin-Farrell & Turpin, 2003) by analyzing survey data collected from a variety of victim service providers and organizations as part of a multiphase CBPR project. We examined the role of organizational context and job type on OBST-related outcomes while controlling for important confounds. In doing so, findings move forward our understanding of interpersonal trauma on several fronts.

We began by assessing the extent to which providers had been exposed to trauma in their childhood and their workplaces. Then, four different measures helped evaluate the degree to which providers were suffering from OBST-related outcomes: exhaustion, cynicism, STS, and PTSD. Next, significant differences in these outcomes were explored by organizational context and job type. Last, the extent to which organizational context and job type were associated with outcomes, net of other factors, was tested. Together, these strategies can help inform the development of much-needed scientifically rigorous interventions for OBST (Bercier & Maynard, 2015; Molnar et al., 2017).

Six core findings emerged from the analysis. First, the majority of providers (63.4%) reported exposure to both interpersonal trauma during their childhoods (at least one ACE) and in their workplaces (moderate or higher). Second, not surprisingly, almost two-thirds (66.2%) of providers were suffering from at least one of the four OBST-related outcomes (at the moderate level or higher) and outcomes were highly correlated, ranging from (r = 0.39–0.74, p < .001). Both of these findings support past research (Knight et al., 2018) and future research should explore how to best support providers and their organizations, given the likelihood that workers will be contenting with trauma symptoms activated by both past and ongoing trauma exposures. We recommend educating workers on the different types of trauma exposures and the potential conflation of symptoms.

Third, findings suggest unique differences in outcomes by both organizational context and job type. Looking at just organizational context, for example, participants in healthcare reported the most exhaustion, participants in community-based organizations reported the least cynicism, participants in community-based and healthcare organizations reported the most STS symptoms, and participants in community-based organizations reported the most PTSD symptoms. Future studies should explore how the culture of organizations can explain and reduce these findings. Research has shown that the current culture of healthcare is increasing rates of burnout, given shortages in providers and the overwhelming need for services (Montgomery et al., 2019). Likewise, research on the culture of law enforcement has shown that cynicism is high among law enforcement officers (Alves et al., 2023). Lastly, the culture of community-based organizations is that women tend to be over-represented in this work (Knight et al., 2021; McPhail, 2004) and, in general, women have higher rates of PTSD compared to men (Valentine et al., 2019). Ultimately, we recommend educating organizations about these differences and encouraging them to use evidence-based organizational improvement processes (Hallinan et al., 2021).

Fourth, results show that the most frequent and strongest predictor of suffering was a greater number of years on the job and a greater number of ACEs. Conversely, both being older in age and training effectiveness were the most protective predictors associated with less suffering. Interestingly, organizational contexts and job types were not significant in any of the multivariate models controlling for these and other factors. These findings support prior research on younger employees (Bonach & Heckert, 2012; Lerias & Byrne, 2003; Sprang et al., 2007) and those who have been in the field for some time (Beaton et al., 1995; Cornille & Meyers, 1999). We recommend research that tests whether varying the responsibilities of employees who have a high number of years on the job reduces their workplace trauma exposure and OBST-related symptoms. We also suggest that organizations work to ensure that providers, especially those who are new to the work and who are younger, receive effective training in trauma (both ACEs and ST).

Fifth, our findings align with the theoretical perspectives that informed this study. This includes our prior research (Knight et al., 2018, 2021, 2022) and the conceptual model of secondary trauma (Ellis & Knight, 2021) that is based on Stern’s (1985) theory of the core self. Applying this theoretical framework of interpersonal trauma, findings from this study can be used to support policies and practices that enhance employees’ trust in themselves (i.e., self-agency), and trust in others (i.e., coherence of others), presence and mindfulness (i.e., self-affectivity), as well as professional growth over time (i.e., self-history).

Last, the differential impacts of diversity on outcomes when comparing bivariate and multivariate analyses warrant an important discussion. At the bivariate level, older participants scored lower on workplace trauma exposure, STSS, PTSD, and training interest. Likewise, compared to participants who identified as a person of color, participants who were White scored lower on ACEs, PTSD symptoms, and training amount. Furthermore, compared to women, men scored lower on ACEs, exhaustion, and PTSD symptoms, as well as training effectiveness and interest. Married participants also reported fewer ACEs, STS, and PTSD symptoms. In sum, at the correlational level, the dominant groups in our society (e.g., older, White, married, men) appear to be more protected from OBST-related outcomes. Other than age, however, these correlational relationships were not significant in the multivariate models. Future research should interrogate the differences in these findings. It is possible that the characteristics of this sample (e.g., small number of providers who identify as a person of color and large rural setting) and analyses (e.g., moderating effects were not tested) are hinting at but ultimately hiding important demographic disparities that warrant future research.

Before concluding, it is necessary to recognize this study’s limitations. To start, a convenience sample was obtained. Employees were not required to participate in the survey, and it is possible that the results do not represent providers at the tail ends of the spectrum (i.e., those suffering the least and most from the symptoms of OBST). In addition, we used a CBPR framework where it is paramount to build a trusting and equitable relationship with community partners (Israel et al., 2013; Wallerstein & Duran, 2006). As such, a simple cross-sectional research design was used in this early phase of the project, which limits causal interpretations. Longitudinal data collection is a goal of future research. Last, we did not collect comparison data from individuals working in less trauma-exposed organizations and job types. However, we statistically controlled for workplace trauma exposure. Future research could also explore how controlling for job types with lower trauma exposure might impact findings (e.g., using samples with more administrative staff and a more detailed measure, such as the life events checklist for the DSM-5; Weathers et al., 2013). Despite these limitations, we successfully surveyed a variety of providers working in various organizational contexts and achieved a high response rate (between 50% and 95% depending on the agency). The CBPR approach is an important strength of this study, allowing us to survey a hard-to-access population and strengthen ecological validity. In particular, the CAB and incentives provided are unique strengths. We believe that the study will make a substantial contribution to the existing literature.

In conclusion, these findings have implications for research and practice. Interpersonal trauma exposure and OBST were high among providers, suggesting that interventions for OBST are warranted. Second, these interventions will need to attend to the possible overlap in different forms of trauma exposure (e.g., childhood and workplace) experienced by participants. Third, results suggest that interventions should be implemented, especially for both younger employees who may have received little training and for those who have been in the field for a long time whose exposure to trauma may be higher. Four, findings signaled that a single intervention for OBST might be appropriate for different kinds of victim service professionals, given that there were no differences in outcomes for the different job types and organizational contexts using the multivariate models. This may help rural and other often under-resourced communities, where one intervention may need to serve many different types of providers and organizations.

Biographies

Kelly E. Knight, PhD, is an associate professor in the Department of Sociology and Anthropology at Montana State University. Her research examines developmental, life course, and intergenerational trajectories of victimization, trauma, substance use, and offending.

Colter Ellis, PhD, is an associate professor in the Department of Sociology and Anthropology at Montana State University. His research examines rural communities, trauma, and secondary trauma.

Tristan Miller graduated Montana State University with a BA is Sociology and worked as a research assistant with the Center for American Indian and Rural Health Equity at Montana State University.

Joshua Neu graduated Montana State University with a BA is Sociology and worked as a research assistant with the Center for American Indian and Rural Health Equity at Montana State University.

Leah Helfrich graduated Montana State University with a BA is Sociology and worked as a research assistant with the Center for American Indian and Rural Health Equity at Montana State University.