Occupational post-traumatic stress disorder: an updated systematic review

A literature search was conducted such as PubMed and Google scholar to determine all published reports on occupational PTSD according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines [4]. We used the following search terms: (PTSD OR “post traumatic stress disorder” OR “posttraumatic stress disorder” OR “post-traumatic stress disorder” OR “acute stress disorder” OR “acute stress reaction”) AND (work* OR occupation* OR industry* OR employ*). A literature search was performed to identify relevant articles published in databases from August 1, 2010, to August 31, 2018. Previously, only 140 occupational PTSD-related articles published from January 1, 1950, to July 31, 2010, were reviewed [3].

We included studies that satisfied the following criteria: studies (1) based on the working population, (2) with individuals whose PTSD was diagnosed by a medical professional, (3) categorized as original articles, and (4) that were published in English in a peer-reviewed journal. Exclusion criteria were as follows: articles (1) categorized as reviews, notes, commentaries, or editorials related to occupational PTSD; (2) based on retirees or adolescents; (3) in which PTSD was not the variable of interest; or (4) that described only the guidelines for clinical treatment or the study design.

After searching the indicated databases as described above, the duplicate articles were removed. All articles were initially screened to select the relevant studies using the inclusion and exclusion criteria based on titles and abstracts. After the initial screening, the remaining articles were checked to determine eligibility. Consequently, quality assessments were conducted for the eligible studies (Supplementary Table). The full-text screening was performed by all authors to determine which of the articles fulfilled the inclusion criteria and achieved the purpose of the current study based on unanimous agreement. A flow chart of the process of article selection is displayed in Fig. 1.

We gathered the articles that addressed the following issues: (1) the status (prevalence or incidence) of PTSD among workers, (2) the risk factors of PTSD among workers, and (3) the effect PTSD on the health of the workers.

Nine studies reported on the prevalence of occupational PTSD among workers on duty after a disaster, employees working in emergency services, and workers who experienced work-related trauma (Table 1). The prevalence of occupational PTSD varies greatly among studies (8.4–41.1%) due to variation in the definition of PTSD, the type of traumatic event, the period after exposure, and differences in occupation.

Workers on duty after a disaster had a high prevalence of PTSD. After the terror attacks on the World Trade Center on September 11, 2001, the prevalence of clinician-measured PTSD 4 years post-trauma was 8.4% [5]. After 11–13 years, 9.7% had current, 7.9% had remitted, and 5.9% had partial PTSD [6]; the definition of PTSD for this study was taken from The Diagnostic and Statistical Manual of Mental Disorders (DSM) 4th Edition. Previous study examined the prevalence of PTSD among 16,488 rescue and recovery workers 8–9 years after the 9/11 attack. Five groups had similar score trajectories in the PTSD Checklist (PCL): low stable (53.3%), moderate stable (28.7%), moderate increasing (6.4%), high decreasing (7.7%), and high stable (4.0%) [7]. Fourteen months after the Great East Japan Earthquake, the prevalence of PTSD (PCL-specific version score ≥ 44) was 9.0% among 610 municipalities, 9.3% among 357 medical workers, and 2.6% among 327 firefighters [8]. The prevalence of PTSD after 2 years was 19.7% among 274 nurses who worked in war zones in Gaza [9].

The prevalence of PTSD was 41.1% (Impact of Event Scale revised Korean version; IES-K score ≥ 26) among 3817 police officers in Korea who experienced traumatic events over a 1-year period [10], 28.0% (PCL-Civilian Version score ≥ 50) among 2706 healthcare workers from Chinese hospitals who experienced physical violence [11], 15.7% (DSM-V PTSD diagnosis using Trauma and Loss Spectrum Self-Report; TALS-SR) among the 83 emergency staff in an Italian hospital [12], and 13% (IES score > 34) among 383 bank employees who were victims of robbery in Italy [13].

Occupational PTSD can be associated with negative working conditions such as long working hours, layoffs, workplace stress, the severity of the injury, frequency of exposure, marital status, history of mental disorder or occurrence of psychiatric symptoms at the time of the event, personality, negative interpersonal relationship, etc.

Several previous studies suggested the risk factors of PTSD for workers experiencing industrial accidents (Table 2). Among injured workers who survived after a major factory collapse in Bangladesh, those employees who worked 70 h a week (adjusted odds ratio [aOR] = 2.4; 1.1–5.3), had a concussion injury (aOR = 3.7; 1.4–9.8), and were married (aOR = 3.2; 1.3–8.0) [14] had an increased risk of developing PTSD. After an industrial explosion, the risk factors of PTSD among the nearby workers included trauma, history of mental disorder or occurrence of psychiatric symptoms at the time of the event [15, 16], proximity to explosion site, non-managerial occupation, age over 50 years in both sex, layoffs (aOR = 2.6; 1.5–4.5), and unusable workplace after the explosion (aOR = 1.8; 1.1–2.8) in men [17]. A Canadian study evaluating the health of urban public transit employees after experiencing traumatic events at the workplace indicated difference in PTSD severity according to the severity of depression, gender, ethnicity, and workplace stress [18].

The risk factors of PTSD among workers on duty after earthquake were as follows: injury, experience with water shortage, disconnection from family and friends during the response, having passive coping styles, exhibiting neurotic personalities in case of Chinese medical rescue workers [19], lower education status, and resilience in case of Japanese rescue workers [20]. The risk factors of PTSD among workers on duty during the European refugee crisis were as follows: marital status (single/divorced/widower), age, long operation period (> 14 days, aOR = 2.3;1.4–3.2), long shift hours (> 4 h/day, aOR = 3.9; 3.1–4.7), and handling dead refugees (> 6 refugees, aOR = 3.4; 2.3–4.5) and dead children (> 1 children, aOR = 3.2; 1.9–4.4) [21, 22]. One study conducted on first responders suggested that the variety of exposure and the frequency of exposure were associated with PTSD [23]. The police officers who worked after Hurricane Katrina showed significantly decreased risk of PTSD, increased level of resilience, satisfaction with life, and gratitude [24].

Prison workers, who were exposed to critical incidents, had ambiguity in their job role, had negative relationships with supervisors and co-workers, and had a type D (distressed) personality [25, 26], were at an increased risk of developing PTSD. Workplace bullying was a significant risk factor for PTSD among hospital nurses [27].

Workers with PTSD are more likely to experience worsening of health status, problems in social and occupational functioning, early retirement, and job loss than those with no PTSD (Table 3). Among rescue and recovery workers with asthma, PTSD comorbidity showed worse asthma control, higher rates of inpatient healthcare utilization, and poorer quality of life than those with sub-threshold or no PTSD [28]. In a cohort of workers exposed to the 9/11 attack, PTSD was significantly associated with bronchodilator response at baseline (aOR = 1.43; 1.19–1.72) and predicted incident asthma (aOR = 2.41; 1.85–3.13) [7].

Among male utility workers on duty after the 9/11 attack, PTSD was significantly associated with sleep disturbance (β = 0.52, p-value < 0.01) [29]. Among rescue and recovery workers on duty after the 9/11 attack, workers with PTSD were more likely to experience early retirement and job loss [30] than those with no PTSD, and PTSD severity was positively associated with subjective distress and deficits in social and occupational functioning, over time [31]. Among policemen who responded to the 9/11 attacks and non-traditional responders, a significant relationship was observed between PTSD and lower respiratory symptoms [32, 33]. Meanwhile, no significant association was found between PTSD and work performance among emergency medical service personnel in Pakistan [34].

Most studies on work-related PTSD were primarily based on survey data and have a cross-sectional study design, which precludes inferences of causal explanations. Longitudinal studies are warranted to confirm the causal nature of the relationship. Meanwhile, errors can occur in retrospective studies due to selective bias. Particularly, having a negative mood strongly disturbed the retrospective self-reported recollection of the life of a worker (ref: selective bias in retrospective self-reported of negative mood). Disease status can affect the recollection of memories by an individual as unhealthy people tend to underestimate the quality of their past life (ref: Breast cancer survivors’ recollection of their quality of life: identifying determinants of recall bias in a longitudinal population-based trial).

Another methodological issue is the definition of the traumatic event and the timing of the PTSD assessment in relation to this event, which makes it difficult to draw a definitive conclusion from the findings of various studies. In most of the studies included in our review, which evaluated the prevalence of occupational PTSD, assessment of PTSD symptoms was completed using constructed questionnaires such as Clinician-Administered PTSD Scale, Impact of Event Scale (IES), or PCL with cut-off values indicating a diagnosis of PTSD. However, a strict application of the diagnostic criteria is essential in research on PTSD. There could be confusion between psychopathology and normal reactions to psychosocial stress or other psychiatric problems when applying checklists for subjective symptoms. Participants whose symptom scores are above these thresholds for caseness cannot be successfully diagnosed with PTSD. Therefore, there is a possibility of overdiagnosis of PTSD when the diagnosis is based on the symptom checklist rather than the criteria described in DSM-5 or diagnosis by a physician. Although a symptom checklist is a simple tool for assessing PTSD, it cannot be used as a substitute for full clinical diagnostic criteria [40]. Moreover, there is a risk of self-report bias in these studies, which could lead to an “artifactual covariance between the predictor and criterion variable” since the same person is assessing both measures [41]. In addition, a variety of different questionnaires have been applied across researches making a comparison of the results difficult.

The timing of PTSD assessment varied widely across the studies, although it is critical in understanding the reported prevalence of PTSD. According to DSM-5 criteria, it is necessary to stipulate both the onset of the PTSD symptoms and their duration. Acute PTSD can be defined after a symptom duration of at least 1 month after the traumatic event; if symptom duration exceeds more than 3 months, it is referred to as chronic PTSD. Meanwhile, if the PTSD symptoms occur at least 6 months after the traumatic event, it is referred to as delayed-onset PTSD. However, the timing of the traumatic event remains unclear and depends on an individual experiencing workplace trauma “to remember” when the PTSD symptoms first occurred.