Treating posttraumatic stress disorder in first responders: A systematic review
Highlights
► Of 845 studies identified only 17 focus on PTSD in first responders. ► We identified no medication RCTs and only 2 psychosocial RCTs. ► Effect sizes in psychosocial RCTs are large. ► Based on the first responder literature identified, treatment guidelines are questionable. ► Additional treatment studies of PTSD in first responders are sorely needed.
Introduction
There are a limited number of occupations or professions that repeatedly put those so employed squarely in harm's way; two prominent examples are combat soldiers and first responders. Indeed, the latter group often comprises members of the former. In the context of this article, we use the term first responders to refer to a heterogeneous grouping of both paid professionals and volunteers who provide critical services in emergencies; for many their main occupational task is first response—e.g. fire fighters. Typical first responders have specialized training, sometimes with explicit certification, which both prepare them and entitle them to take action to safeguard the health and safety of those victimized. This action usually occurs on an individual basis and for the public at large, most often in emergencies. Large-scale disasters (e.g., the events of September 11, 2001) have expanded the occupational groupings who engage in first response to include construction and utility workers, laborers, and public sector workers (Benedek et al., 2007, Herbert et al., 2006), but these groups are not primarily first responders. Like many other groupings, first responders can be conceptualized as a fuzzy set where those at the margins are less prototypic than those in the center. Here the margins include occasional disaster workers, and the center includes first responder occupations.
First responder occupations have historically included police officers (e.g., Cardozo et al., 2005, McCaslin, Metzler, et al., 2006, Tak et al., 2007), fire fighters (e.g., Bryant & Harvey, 1995), search and rescue personnel (e.g., Brandt, Fullerton, Saltzgaber, Ursano, & Holloway, 1995), and ambulance personnel (emergency medical technicians and paramedics; e.g., Weiss, Marmar, Metzler, & Ronfeldt, 1995). These positions are characterized by high levels of work demands (Penalba, McGuire, & Leite, 2009) and routine exposure to both physical and psychological stressors (e.g., Galloucis et al., 2000, McCaslin, Rogers, et al., 2006). Examples of the former include overtime and special duty shifts, and arrest quotas under public and press scrutiny. Examples of routine exposure to physical stressors include heavy personal equipment for fire fighters, lifting gurneys for EMTs, and foot pursuit for police officers. Psychological stressors include routine work demands such as labor and management conflicts, harassment, and work demands with poor or outdated equipment (Liberman et al., 2002). The key psychological stressors, however, are those that are typically considered traumatic stressors. These include exposure to incidents that put the first responder or those around him or at risk for death or severe injury (e.g., backdrafts for firefighters, being attacked with a weapon for a police officer), witnessing or participating in incidents where rescue involves preventing death or mitigating serious or severe injury, and various levels of witnessing such incidents. These traumatic stressors are the primary aspect of what distinguish first responders from virtually all other occupations.
Considerable research has shown that these types of exposure increase the likelihood of posttraumatic stress disorder (PTSD), other psychiatric disorders, and burn-out. In the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2000), the symptoms of PTSD are grouped into three clusters: reexperiencing of the traumatic event; avoidance of trauma-relevant stimuli and numbing of general responsiveness; and persistent hyperarousal. The revisions to the criteria for PTSD proposed in the upcoming DSM-5 include specific reference to issues of frequency and severity of exposure to traumatic stressors using the salient example of first responders to illustrate the issue. In addition to PTSD (e.g., Chang et al., 2003, Fullerton et al., 2004, North et al., 2002, Tak et al., 2007), depression (Cardozo et al., 2005, Fullerton et al., 2004, Tak et al., 2007), somatic or psychosomatic complaints (Chang et al., 2003, Morren et al., 2005, Witteveen et al., 2006), chronic fatigue (Morren et al., 2005, Spinhoven and Verschuur, 2006, Witteveen et al., 2006), and difficulties with alcohol (North et al., 2002, Stewart et al., 2004) have all been documented in first responders. Some authors have speculated that the cumulative nature of the stressors may result in a unique symptom profile of PTSD in first responders (e.g., Duckworth, 1986, Tolin and Foa, 1999). There is also a literature suggesting that burn-out may be another consequence of service as a first responder (e.g., Alexander and Klein, 2001, Mitani et al., 2006), with evidence of an impact on HPA axis function, just as in PTSD (e.g., Chida & Steptoe, 2009).
Estimation of lifetime and current rates of full and partial PTSD among first responders has come from small-scale studies, the vast majority using self-report rather than gold standard structured clinical interviews (e.g., Maia et al., 2007). An exception that did employ a structured interview involved 132 Canadian police officers (Martin, Marchand, Boyer, & Martin, 2009) and found that 7.6% of the participants developed full PTSD, whereas 6.8% had partial PTSD following a work-related exposure to a traumatic stressor. We know of no nationally representative, large-scale study of first responders. Consequently, to provide an estimate of the potential need for treatment of symptoms of PTSD among first responders in the United States, we used data from the 2008 National Employment Matrix (Bureau of Labor Statistics, 2008). In 2008, 1,503,100 individuals were employed in these roles (see Table 1). We believe this is a conservative estimate of the number of individuals employed in first responder professions, since it omits volunteers and other non-traditional first responders. Available studies have presented a range of rates of current PTSD from a low of 7%–19% in active duty police officers (Carlier and Gersons, 1997, Gersons, 1989, Maia et al., 2007, Robinson et al., 1997) to 46% in volunteer disaster workers responding to an airline disaster (Mitchell, Griffin, Stewart, & Loba, 2004), with many other estimates falling between these extremes. These rates are far from definitive, in that the measures and methods for indexing exposure and procedures for determining a diagnosis are not at the level required for a high quality epidemiologic estimate (Kulka et al., 1991), quite apart from the representativeness of the samples. We chose to estimate, however, from a recently published meta-regression analysis of the worldwide current prevalence of PTSD in rescue workers (Berger et al., 2011) as this is preferable to selecting any single study. That summary of 28 studies yielded a current prevalence of approximately 10% for full PTSD suggesting that 150,310 first responders may meet criteria for current PTSD, and could benefit from treatment.
It is clear from other evidence (e.g., Stein, McQuaid, Pedrelli, Lenox, & McCahill, 2000) that those with partial PTSD also experience significant impairment (Kassam-Adams, Fleisher, & Winston, 2009). Though there was no estimate made for partial PTSD by Berger et al. (2011), these authors describe the likelihood that the 10% underestimates the scope of impairment for first responders, including for reasons of not meeting full criteria. In order to make an estimate for partial PTSD, we used the relative proportion (77%) between full current PTSD and partial PTSD that was reported in the National Vietnam Veterans' Readjustment Study (NVVRS; Weiss et al., 1992), as this was one of a very few nationally-representative studies that used clinical interviewing to estimate the partial prevalence. If the need for treatment for those with partial PTSD is included and is about ¾ of the rate for full current PTSD in first responders, this suggests that an additional 115,739 first responders likely meet criteria for partial PTSD and could benefit from treatment—bringing the total of U.S. first responders affected to approximately 266,049. Nationwide, there may well be a quarter of a million first responders impaired by symptoms of PTSD for whom effective intervention would be both compassionate and utilitarian. If these are underestimates, which Berger et al. (2011) entertain as a possibility, this only emphasizes the need to know the state of the evidence about treatment for first responders. In addition, treatments for occupation-related traumatic disturbance need to take into account that patients may (a) continue to work in environments in which they may be re-exposed and (b) may return, posttreatment, to the same environment. This has implications for both the provision of services and the nature of the treatment protocol.
Given that the substantial majority of prevalence estimates show that only a minority of those exposed remain chronically symptomatic, conceptualizations have considered a role for either risk factors or differences as a function of type of exposure. There is some evidence that there may be differences across subcategories of first responders as well, with police officers being less affected than other first responders (Ersland et al., 1989, Norris and Alegria, 2005, Perrin et al., 2007) and more similar to nonexposed populations (e.g., Marmar et al., 1999). Other studies, however, have not found such differences (Johnsen et al., 1997, Jones, 1985, Ursano et al., 1995). The Berger et al. meta-regression found that ambulance personnel tended to have a reliability higher rate of current PTSD than police or firefighters. There are a number of possible factors that could help explain why exposure to traumatic stress may have less effect on police than on other groups of first responders: (a) self-screening, (b) pre-employment screening, (c) training, (d) type of exposures, and (e) prior exposure to similar experiences (e.g., in the military and through the course of years of duty), though this last item is conceptually more complicated (see Moran, 1998 for a discussion). The type of critical incident to which first responders are exposed may also play a differential role; for example, Clohessy and Ehlers (1999) reported that incidents involving children were especially problematic for paramedics. More studies are needed to clarify this issue.
A substantial body of literature (Bisson and Andrew, 2007, Bradley et al., 2005, Van Etten and Taylor, 1998) supports the effectiveness of several specific psychological treatments for PTSD in a variety of populations (e.g., veterans, victims of sexual assault, and those who have had motor vehicle accidents). These treatments include cognitive behavioral therapy (CBT), eye movement desensitization reprocessing (EMDR), prolonged exposure (PE), and stress inoculation therapy (SIT). As a consequence, professional societies, and government agencies have issued treatment guidelines (American Psychiatric Association, 2004; International Society of Traumatic Stress Studies in Foa et al., 2009, Australian Centre for Posttraumatic Mental Health, 2007, National Institute for Health and Clinical Excellence, 2005, VA/DoD Clinical Practice Guideline Working Group, 2003) in line with the recent emphasis on the need for an evidence base to justify delivering a specific treatment. Forbes et al. (2010) have summarized and reviewed a number of these guidelines. Although these recommendations are intended to be definitive, there is controversy in the literature regarding this process (e.g., Benish et al., 2008, Ehlers et al., 2010, Wampold et al., 2010). As well, the current guidelines often do not address theoretically important moderating variables such as specific patient groups, including first responders, details of the trauma, including time elapsed since exposure, or whether or not re-exposure is likely.
There are also guidelines for treatment of PTSD with medication included in the larger overall guideline documents (e.g., Friedman, Davidson, & Stein, 2009). The specific psychobiological dysfunctions associated with the disorder (Bonne et al., 2004, Yehuda and McFarlane, 1995) have helped to provide a rationale for the use of medications (Ipser & Stein, 2011) aside from considerations of comorbidity (e.g., depression) and available agents. Several reviews and meta-analyses (e.g.,Friedman et al., 2009, Ipser and Stein, 2011) have concluded that the evidence supports the use of medication for the short-term treatment of PTSD. Selective serotonin reuptake inhibitors (SSRIs) have the largest evidence base in both number of studies and size of trials. The various guidelines differ, however, as to whether the strength of the evidence is sufficient to recommend SSRIs as an alternative to psychosocial therapies for first-line intervention, as in the VA/DoD, International Society for Traumatic Stress Studies and the American Psychiatric Association guidelines or as a second-line intervention when psychosocial therapy is not available, acceptable, or suitable as in the National Institute for Health and Clinical Excellence; National Health and Medical Research Council; American Academy of Child and Adolescent Psychiatry guidelines (see Forbes et al., 2010).
A number of treatment guidelines also suggest that combining pharmacotherapy and psychosocial treatments for PTSD may lead people to recover more effectively than using either treatment alone as has been shown most convincingly for depression (Hollon, Thase, & Markowitz, 2002). In contrast, a recent Cochrane Review of combined pharmacotherapy and psychological therapies for PTSD (Hetrick, Purcell, Garner, & Parslow, 2010) concluded that there are too few studies to be able to draw conclusions about whether combination treatments result in better outcomes. As with the psychosocial treatments, the studies reviewed in guidelines for the pharmacotherapy of PTSD have not included any with first responders as participants.
Our objective was to conduct a thorough literature search of the status of treatment outcome studies for PTSD in first responders. The aims were to investigate (a) the degree to which first responders are acknowledged and discussed as an identified group (similar to veterans or those exposed to intimate partner violence) in treatment guidelines and (b) to what degree the findings from first responders have contributed to these guidelines. We focused on tertiary treatment: psychosocial and pharmacological treatment of first responders with diagnosed PTSD and other posttraumatic psychiatric disorders. We explicitly did not focus on approaches for secondary prevention or early intervention (e.g., Agorastos, Marmar, & Otte, 2011) because these have been adequately reviewed elsewhere (e.g., Deahl, 2000, Larsson et al., 2000); as well, early intervention.
Section snippets
Method
Inclusion criteria for studies were as follows: (a) a psychological or pharmacological intervention was delivered (b) subjects were first responders (c) subjects had a primary diagnosis of PTSD based on DSM or ICD criteria and (d) PTSD diagnosis or symptom status was the chief study outcome. Studies whose subjects had psychiatric disorders comorbid with PTSD were not excluded. To be included, psychosocial treatment studies had to compare two active treatment groups, or one active group to a
Results
The search strategy identified 845 potentially relevant articles. Study titles and abstracts were reviewed by a team of research assistants and the authors to determine whether they met inclusion criteria. Fig. 1 shows the results of the review; the exclusions were not mutually exclusive. Twenty-one studies were not published in English and could not be further reviewed. Of the remaining 824 articles, 672 were excluded because they did not study a specific psychosocial or pharmacological
Acknowledgments
We acknowledge the assistance of Rachel Goldman, Gillian Griffin, Eva Leven, and Aileen McCrillis. This publication was supported by Grant/Cooperative Agreement Number 3U10OH008223-05S2 from the Centers for Disease Control (CDC). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of CDC.
References (105)
- et al.
The relative efficacy of bona fide psychotherapies for treating post-traumatic stress disorder: A meta-analysis of direct comparisons
Clinical Psychology Review
(2008) - et al.
Adaptive and maladaptive psychobiological responses to severe psychological stress: implications for the discovery of novel pharmacotherapy
Neuroscience and Biobehavioral Reviews
(2004) - et al.
Cortisol awakening response and psychosocial factors: A systematic review and meta-analysis
Biological Psychiatry
(2009) - et al.
Do all psychological treatments really work the same in posttraumatic stress disorder?
Clinical Psychology Review
(2010) Psychological treatment of three urban fire fighters with post-traumatic stress disorder using Eye Movement Desensitisation Reprocessing (EMDR) therapy
Complementary Therapies in Nursing & Midwifery
(2004)- et al.
Post-traumatic stress symptoms in an elite unit of Brazilian police officers: Prevalence and impact on psychosocial functioning and on physical and mental health
Journal of Affective Disorders
(2007) - et al.
The diagnostic accuracy of the PTSD Checklist: A critical review
Clinical Psychology Review
(2010) - et al.
Effect of autogenic training on cardiac autonomic nervous activity in high-risk fire service workers for posttraumatic stress disorder
Journal of Psychosomatic Research
(2006) - et al.
Behavioral activation for comorbid PTSD and major depression: A case study
Cognitive and Behavioral Practice
(2004) - et al.
Posttraumatic stress disorder in the primary care medical setting
General Hospital Psychiatry
(2000)