Abstract
The issue of gender is largely ignored in studies of secondary traumatization (STS). This article addresses the question of gender differences in susceptibility to STS among clinicians who treat traumatized clients. It does so by systematically reviewing the very limited body of published findings on this subject to date. These are 10 published studies that measure STS by post-traumatic stress disorder (PTSD) symptomatology and 4 studies that measure it using Stamm’s Professionals Quality of Life Survey (ProQOL), which queries PTSD symptomatology along with other difficulties that may arise in helping traumatized clients. Almost all the studies based on PTSD symptomatology show greater female susceptibility. Although the pattern is less clear in the ProQOL studies, the article argues that the research to date does not really show mixed findings, as is repeatedly claimed, but greater susceptibility among female clinicians. It also points out that the findings do not mean that male clinicians are unaffected by their traumatized clients and notes the various manifestations of their distress reported in the reviewed studies. The article offers a variety of explanations for the heightened female susceptibility.
It is by now well established that therapists who work with traumatized clients are emotionally affected by the disturbing stories, shocking images of horror, and suffering that are characteristic of serious traumas (Bride, 2007; Connally, 2012; Elwood, Mott, Lohr, & Galovski, 2011; McCann & Pearlman, 1990). When clinicians listen to, discuss, and help such clients work through their feelings and empathetically engage with them, they may experience the client’s contents as if they were their own internally generated ones (Canfield, 2005). Repeatedly hearing their clients’ painful and often graphic accounts of sexual and/or physical abuse, torture, humiliation, betrayal, or violent death may cause considerable stress and result in any of a variety of overlapping stress responses. Responses that have been named are burnout, compassion fatigue, vicarious traumatization, and secondary traumatic stress (STS). This article focuses on STS. More specifically, it addresses the neglected issue of gender differences in susceptibility.
STS
The term secondary traumatic stress is used in the literature both in a broad and in a narrow sense. In the broad sense, it refers to all the above named manifestations of distress following indirect exposure to a traumatic experience. There is a huge amount of clinical, theoretical, and empirical literature about STS in this sense of the term (for a comprehensive review see Sabin-Farrell & Turpin, 2003).
In its narrow sense, the term has been defined by Figley, as “a syndrome of symptoms nearly identical to PTSD” that arises from “exposure to knowledge about a traumatizing event experienced by a significant other” (Figley, 1995, p. 8). There are three parts to this definition. One, secondary traumatization is PTSD, that is, post-traumatic stress disorder, without direct exposure to a traumatic event. Two, it is PTSD “caught” from a person who had experienced the traumatic event directly and somehow transmitted his or her firsthand knowledge and experience of the event to the secondary victim. Three, the trauma victim from whom it is caught is not just anyone, but a person in close contact with the secondary victim and meaningful in his or her life, presumably a family member, a close friend, or a therapy client. Figley thus referred to it as “the cost of caring” (Figley, 1995, p. 12).
STS in the narrow sense of the term is distinct from the other manifestations of distress that might follow upon engaging with clients’ traumatic experiences. Burnout is an outcome of a large variety of occupational stressors, not necessarily people related, and marked by emotional exhaustion, erosion of idealism, and reduced sense of accomplishment and achievement (Maslach, 1982). Compassion fatigue, which Figley (1995) sometimes used interchangeably with the term “secondary traumatization,” is marked by PTSD symptoms. But unlike STS, it can result from caring for persons with any type of infirmity or disability and encompasses, in addition to PTSD symptoms, manifestations of burden and/or burnout (e.g., Bride, Radey, & Figley, 2007). Vicarious traumatization, like STS, stems specifically from exposure to trauma survivors and often shows subclinical levels of PTSD symptoms, but unlike STS is marked largely by cognitive changes, in meanings, beliefs, schemas, and adaptations (Sabin-Farrell & Turpin, 2003). STS in the narrow sense of the term is the only response that both stems solely from exposure to traumatized individuals and the only one that is measured solely by symptoms virtually identical to those of PTSD. It is also the only reaction in which the trauma and its attendant symptoms—the intrusion, avoidance, and hyperarousal that characterize PTSD—are transmitted from the survivor to a person who did not experience the traumatic event at first hand.
This article focuses on STS in the narrow sense of the term, as the very specific syndrome defined by Figley (1995) and characterized by PTSD symptoms, for two related reasons. One is that, of all the painful emotional effects of indirect exposure to traumatic events, PTSD symptoms are the only ones that can be directly linked to the knowledge of the event. Without knowledge of the event, it is impossible to meet two key Diagnostic and statistical manual of mental disorders (DSM) criteria for PTSD: intrusive recollections of the event and avoidance of such recollections (Diagnostic and statistical manual of mental disorders [4th ed., Text Revision (DSM-IVTR)]; American Psychiatric Association, 2000). 1
In contrast, adverse effects such as depression and anxiety, which may also follow from indirect exposure, can have a multitude of sources or triggers. So can the cognitive changes in meanings, beliefs, and schemas, and adaptations that mark vicarious traumatization (Sabin-Farrell & Turpin, 2003). Similarly, symptoms of burnout (Maslach, 1982) and compassion fatigue (Figley, 1995) can readily be attributed to the difficulties of taking care of a traumatized individual rather than to knowledge of the trauma and traumatic event. The other reason is that PTSD symptoms may be found in reactions that are sometimes studied under other names, for example, vicarious traumatization (e.g., McCann & Pearlman, 1990) and compassion fatigue (e.g., Stamm, 2005). This enables us to study STS in the narrow sense irrespective of how it is labeled.
Implications of STS for Mental Health Professionals
Quantitative and qualitative studies and clinical self-reports on mental health professionals who treat traumatized clients show the same patterns of intrusion, avoidance, and hyperarousal that characterize their clients’ PTSD (e.g., Arnold, Calhoun, Tedeschi, & Cann, 2005; Knight, 2013). They experience intrusive thoughts and images of clients’ traumatic events and distress or physiological reactivity in response to reminders of clients’ traumatic experiences. Some report dreaming their clients’ dreams (e.g., Cerney, 1995; Nelson-Gardell & Harris, 2003). Like their traumatized clients, they may make active efforts to avoid thoughts, feelings, activities, and situations reminiscent of the traumatic event. Some show diminished affect and/or diminished interest in previously pleasurable activities relief of stress. Hyperarousal symptoms such as sleep disturbances, difficulty concentrating, startle responses, feelings of agitation or irritability, and hypervigilance have all been reported (e.g., Palm, Polusny, & Follette, 2004). Gastrointestinal problems, headaches, and heart palpitations have also been observed (e.g., Figley, 1995; L. Miller, 2003; Pistorius, Feinauer, Harper, Stahmann, & Miller, 2008).
Therapists’ STS can affect their families and close friends (e.g., Cerney, 1995; Goldblatt, Buchbinder, Eisikovits, & Arizon-Mesinger, 2009). Several researchers have observed that therapists suffering from STS may withdraw emotionally from their families and be chronically unavailable to them, in much the same way that direct trauma victims are (e.g., Cerney, 1995; Dutton & Rubinstein, 1995).
Various scholars suggest that the emotional and physical manifestation of their STS may prevent mental health professionals from working effectively (e.g., Clemans, 2004; Harris, 1995). They note the detrimental impact the increased fatigue or illness, emotional numbing, social withdrawal, and feelings of hopelessness and despair may have on their work (Etherington, 2007; Saakvitne, Gamble, Pearlman, & Tabor, 2000). Some scholars have also noted that therapists who have internalized the client’s traumatic imagery and experiences cease to be able to listen to the client’s traumatic material (e.g., Clark & Gioro, 1998; Clemans, 2004; Figley, 1995).
To limit the consequences of STS among helping professionals, extensive effort has been made to identify risk and protective factors. For the most part, the endeavor has focused on work-related factors such as the level of exposure to traumatized clients (e.g., Pinsley, 2000; Sabin-Farrell & Turpin, 2003), whether the traumatized clients are children or adults (e.g., Good, 1996; Pistorius et al., 2008), the source of the clients’ trauma (e.g., sexual abuse, act of terror; K. Baird & Kracen, 2006; Cunningham, 2003), workplace support (e.g., supervision, training to work with trauma survivors (e.g., Baird & Jenkins, 2003; Lybeck-Brown, 2001; Naturale, 2007; Perron & Hiltz, 2006; Slattery, 2003), and practice experience (e.g., Badger, Royse, & Craig, 2008; Birck, 2001). Individual factors have also been examined, though to a lesser extent. These include the therapists’ personal history of trauma (e.g., Kassam-Adams, 1999; Nelson-Gardell & Harris, 2003), the personal support they receive from family and friends (e.g., Harrison & Westwood, 2009; Hunter & Schofield, 2006; Saakvinte & Pearlman, 1996), and demographic features such as the practitioner’s age (e.g., Ghahramanlou & Brodbeck, 2000; Good, 1996; VanDeusen & Way, 2006) and marital status (e.g., Byrne, 2006). In all these literature, however, the effect of gender on STS is hardly mentioned.
The scant attention to gender in STS contrasts sharply with the considerable attention to gender differences in post-traumatic symptomatology after direct exposure to a wide range of traumatic events. Gender differences in PTSD symptoms have been a central issue of studies of responses to direct exposure to both man-made collective disasters (e.g., Galea et al., 2002), such as wars (e.g., Johnson & Thompson, 2008), and natural disasters (e.g., Norris, Perilla, Ibañez, & Murphy, 2001). Gender differences in PTSD symptomatology have also been studied in the wake of personal violence such as domestic violence (e.g., Breslau, Chilcoat, Kessler, Peterson, & Lucia, 1999; Tolin & Foa, 2006), sexual abuse (e.g., Banyard, Williams, & Siegel, 2004), and car accidents (e.g., Fullerton et al., 2001). The lack of attention also contrasts sharply with the ample attention to gender differences in studies of burnout and post-traumatic growth (e.g., Purvanova & Muros, 2010; Vishnevsky, Cann, Calhoun, Tedeschi, & Demakis, 2010).
Gender in STS Studies
The pattern of ignoring gender in studies of STS among mental health professionals can be traced back to two early studies: Follette, Polusny, and Milbeck (1994) and Pearlman and Mac Ian (1995). Both these studies sought to ascertain the contribution of the therapists’ personal trauma history and caseload of traumatized clients to their STS. The findings, which showed that both variables made substantial contributions, shed light on the personal and professional antecedents of STS among therapists. However, although both studies reported the gender distribution in the antecedent variables, neither reported the gender distribution in the therapists’ STS.
The bulk of the studies on secondary traumatization among professionals, as among other groups, is either single-sex studies or studies that include subjects of both genders without comparing them (Pearlman & Mac Ian, 1995). Where the population studied included a small number of male professionals, they were generally (though not always) excluded from the studies (e.g., Schauben & Frazier, 1995) to maintain the sample’s homogeneity.
A repeated explanation for the non-examination of gender in STS among therapists is that there are not enough male therapists to conduct a comparison (e.g., Creamer & Liddle, 2005). While this explanation may hold for some studies (e.g., Schauben & Frazier, 1995), it certainly does not apply to all of them. There are a good number of studies in which practitioners of both genders participated. Most had between 15% and 30% males in their samples (e.g., Buchanan, Anderson, Uhlemann, & Horwitz, 2006; Devilly, Wright, & Varker, 2009), which are enough for comparison.
A small number of studies of professionals in the fields of sexual abuse and domestic violence do provide data on the two genders. However, the authors of these studies note that a valid comparison is impossible because the male professionals usually treat the perpetrators, while the female professionals usually treat the victims (e.g., Ennis & Horne, 2003). Nonetheless, there are studies that do investigate both genders and provide data that enable comparison. These include studies of mental health professionals who treat traumatized children following sexual or physical abuse, victims of family violence, and/or survivors of terror attacks and natural disasters.
Some of these studies report gender differences and state that they are important. Most of them, however, warn against generalizing from their findings. The cautions go back to Kassam-Adams (1999) who found that female professionals treating traumatized clients experienced higher STS than their male counterparts even after controlling their age and years of experience. Despite this finding, she warned that “… we should not conclude that secondary trauma is primarily a risk for women” (p. 45). These words continue to echo in the literature. For example, Creamer and Liddle (2005), who found that female professionals treating traumatized patients after September 11 experienced slightly higher STS than their male counterparts, maintained that looking at gender as a risk factor for STS without considering other factors may be simplistic. Choi (2011), who similarly found that female social workers working with victims of abuse were more likely to sustain STS than their male counterparts, cautioned that “it is preliminary to conclude that females…have a greater risk of experiencing STS than males” (p. 237).
The reluctance to generalize from single, sometimes small, studies is appropriate. But it also underscores the need for a systematic examination of gender vulnerability to determine—in the words of Figley (2002, p. 6)—“… who is most vulnerable, in what type of work setting and under what type of conditions.” We may also ask whether, in fact, one gender is more vulnerable than the other. Answering these questions has theoretical and practical importance. On a theoretical level, the answers may contribute to a better understanding of gender differences and similarities. Among other things, they will permit review of the generally accepted, but not thoroughly examined, claim that findings on gender differences in STS are mixed. On a practical level, the importance inheres in the fact that clinicians of both genders may engage with traumatized individuals in the course of their work. The knowledge gained may be used in designing and applying gender-sensitive interventions with clinicians, whether preventive or after the fact.
Rationale and Method of the Review
To address these questions, the article systematically reviews findings on gender in STS reported in published studies on various groups of mental health professionals. In principle, the association between gender and STS among these professionals can be examined by comparing the STS of males and females in a specific sample. Such a study, however, would of necessity provide findings only on a single group of professionals (e.g., psychologists, social workers, psychiatrists), working in a particular kind of setting (e.g., child protection services, centers for family violence and sexual assault), usually with a particular group of clients (e.g., children, adults), traumatized in a single type of traumatic event (e.g., war, terror, sexual or physical abuse, etc.). The generalizability of the findings to other groups of therapists would be open to question.
To avoid this limitation, I, along with two colleagues, embarked on a systematic survey of the English-language literature on STS on mental health professionals. We started by looking for English-language papers on STS published in professional journals between 1990 and December 2013. We searched in Med-line, Psych-net, Social Work Abstract, and Google Scholar data bases under the key words secondary traumatization, secondary trauma, STS, STS disorder, secondary survivor, and the initials STS and STSD. We also conducted a manual search of reference lists in recent articles that reviewed the literature on secondary traumatization. The initial search generated thousands of references. These encompassed articles on STS not only among mental health professionals, but in a variety of populations including policemen, firemen, journalists, and others.
For the purpose of the present survey, we then sought those papers that dealt with STS mental health professionals and that provide or enable gender comparison.
Inclusion and Exclusion Criteria
These papers were then reviewed by all three experts in accord with the following criteria. (1) The study participants were mental health professionals working with someone who had been traumatized. (2) The study had to be conducted using a structured questionnaire based on DSM criteria of PTSD, which, as noted above, has the same symptoms as STS. The study instrument did not have to include all the DSM symptoms but did have to include the symptoms of intrusive thoughts and avoidance, the key symptoms of PTSD and STS. The most common instrument used in the examined papers was Horowitz, Wilner, and Alvarez’s 30-item Impact of Events Scale (IES, 1979). Also examined were papers using Figley’s (1995) measure of Compassion Fatigue. This consists of three separate scales. We used only the 5-item scale that queried PTSD symptoms, which is comparable to the IES in that it, too, measures the key symptoms of PTSD: intrusion and avoidance symptoms, and allows the investigators to distinguish between low and high levels of STS. (3) Gender comparison of secondary traumatization was made or could be made on the basis of the data. This restricted inclusion to studies that reported findings or presented data on each gender separately rather than on the sample as a whole. (4) The study participants had not been directly exposed to the traumatic event. Studies that did not meet this criterion were excluded so as to avoid confounding the results of direct and indirect, or primary and secondary, traumatization. (5) The study did not examine therapists of perpetrators (e.g., Ennis & Horne, 2003; Hatcher & Noakes, 2010) who by definition participated in causing trauma. (6) Also removed were studies of secondary traumatization of mental health professionals working with clients suffering from a severe illness, because such illness may be better understood as a severe stressor for the sufferer than as a traumatic event. Each prospective study was examined simultaneously for all six criteria. In the end, the process of elimination was left with only 10 papers.
Gender Differences in Varied Work Settings
The surveyed studies examine professionals working in three frameworks: child protection services, centers for family violence and sexual assault, and care of disaster survivors. Across these frameworks, the studies vary in the kind of data they report. Some papers report both the percentage of men and women who suffered STS and the means and standard deviations on the overall score and various subscales. Others report only the overall score, yet others report the scores on intrusion or avoidance and still others report both. The findings are presented separately for each group of study participants. For ease of interpretation, where possible, the effect of gender on STS in each study is presented in Cohen’s (1988, p. 40) effect size (standardized mean difference), followed by the exact significance level of the effect (p). In 4 out of 14 studies reviewed below (Adams & Riggs, 2008; Creamer and Liddle, 2005; Nelson-Gardell & Harris, 2003; Wee & Myers, 2002), the authors did not provide the information required for computing a standardized effect-size measure (the author tried to obtain this information but failed). The raw data that were originally reported in each study and the standardized, computed measures of association between gender and STS are summarized in Table 1.
Research on STS in Working Therapeutically With Trauma Clients.
Note. NYC = New York City; ST = secondary traumatization; TSI = Trauma Symptom Inventory; STSS = secondary traumatic stress scale; ProQOL = Professionals Quality of Life Survey; IES = Impact of Events Scale; STS = secondary traumatic stress; M = mean; SD = standard deviation; PSTD = Posttraumatic stress disorder; CF = Compassion Fatigue; RR = Relative Risk; IES-R = Impact of Events Scale - Revised; SCL = Symptom Checklist.
Child Protection Service Workers
Three studies were conducted on child protection workers, whose work brought them into close contact with children who had been traumatized by the maltreatment they suffered. Two studies found statistically significant gender differences. The first was conducted by Cornille and Meyers (1999) on 183 workers who constituted 57% of the 360 child protection services (CPS) workers who received the questionnaire. Eighty-two percent (n = 150) were female and 18% (n = 33) male. PTSD symptom severity was rated on the IES-R (Weiss & Marmar, 1997). Their findings showed that gender was strongly associated with symptoms of IES-R (d = 0.87, p < .001), with females more likely to report symptoms than men. Their hyperarousal scores were especially high in comparison to those of men.
The second study, by Meyers and Cornille (2002), seems to have been another data analysis on an overlapping but somewhat larger sample, consisting of 34 males and 169 females. This analysis showed that the female professionals scored higher than their male counterparts on all three PTSD subscales. Females reported more severe avoidance (d = 0.47, p = .013), intrusion (d = 0.38, p = .043), and hyperarousal (d = 0.34, p = .069) than male professionals.
The third study, by Nelson-Gardell and Harris (2003), collected data from 166 social workers working in child welfare. STS was measured on Figley’s 23-item Compassion Fatigue Risk subscale (Figley, 1995). The authors state that t-tests showed no significant gender differences in mean STS scores without providing any related statistics that can be converted into effect-size measure.
Professionals Working With Adult Victims of Sexual Abuse and Spousal Violence
Three studies comparing the secondary traumatization of male and female mental health professionals who treated adult victims of sexual abuse and spousal violence have been published to date. Two of them report that female practitioners showed higher levels of STS than their male counterparts.
The first of these, by Kassam-Adams (1999), examined STS in a sample of experienced psychotherapists working in outpatient clinics in West Virginia and central Maryland. The aim of the study was to explore the association between STS and the level of exposure to victims of sexual trauma. PTSD symptom severity was measured on the IES (Horowitz, Wilner, & Alvarez, 1979). Without providing the supporting data, the author wrote that female therapists reported more symptoms of secondary traumatization than their male counterparts, regardless of their age or years of experience. The study also found that gender was correlated with primary traumatization, r(271) = .40, p < .001, with females reporting more than males. Regression analysis showed that gender and primary traumatization together predicted STS (Adjusted R 2 = .12; F = 7.27, p < .001).
The second study, by Choi (2011), examined predictors of STS in a sample of 154 social workers who treated victims of sexual abuse and family violence, of whom 78.6% were women and 21.4% men. Study participants were recruited from the National Association of Social Workers (NASW) mailing list. The response rate was 29% symptoms were measured on the 17-item Secondary Traumatic Stress Scale constructed by Bride, Robinson, Yegidis, and Figley (2004) to assess STS in helping professionals. The findings revealed somewhat small and marginally significant (d = 0.36, p = .064) gender differences, with women scoring 3.7 points (some 10%) higher than men on STS.
The third study, by Adams and Riggs (2008), showed no significant gender differences. This study examined the predictors of STS (and vicarious trauma) among 129 trainees doing their internship in clinical psychology and counseling among trauma victims. Of the trainees, 83.7% were women. STS was tested on the five subscales of the Trauma Symptom Inventory (Briere, 1995). The researchers provide no data on gender differences, but note in the limitations section that no gender differences emerged in any of the study variables. Number of previous semesters experience with trauma work was M = 4.86 (SD = 6.09), with 40% of the sample reporting two or fewer semesters working with trauma clients.
Professionals Treating Disaster Victims
Four studies comparing the secondary traumatization of male and female professionals who treated traumatized victims of terror attacks have been published to date. All of them show greater female tendency to STS.
Wee and Myers (2002) examined mental health workers who provided crisis counseling to persons affected by the Federal Building Bombing in Oklahoma City. The Compassion Fatigue Self-Test for Helpers (Figley, 1995) was used to identify the severity of their STS. The questionnaires were mailed to 74 workers 9 months after the bombing. Thirty-four questionnaires were returned (28 by females and 6 by males), yielding a return rate of 45.9%. The female professionals (M = 37.07) experienced a considerably higher level of STS than male (M = 34.50) professionals, yet the sample sizes in this study are too small for reaching any valid conclusions.
Creamer and Liddle (2005) examined 80 mental health professionals who served as disaster mental health workers in response to the September 11th attack in 2001. Questionnaires were mailed to 100 disaster mental health workers in 28 states who responded to announcements posted on disaster and other mental health–related listservs or by APA Disaster Response Network coordinators. To avoid confounding with primary PTSD, the researchers excluded persons who were within 15 miles of the attack when it occurred or who had a family member or close friends in the vicinity during the attack. Participants were 50 female (62%) and 30 male (38%) professionals. At the time of the study, most of them (70%) were treating survivors of other traumas (e.g., incest survivors, combat veterans) as well. PTSD symptom severity was rated on the IES. The authors state that the female therapists reported higher levels of STS than the male therapists, though the difference was not statistically significant. The data are not reported, however.
Another study of professionals post-September 11 was carried out by Adams, Figley, and Boscarino (2008). This study examined 47 male and 182 female social workers who engaged in direct intervention with trauma victims in New York. Twenty months after the attack, a short version (5 items) of Figley’s (1995) STS scale was mailed to 600 members of the National Association of Social Work (NASW) living in New York City. Effective response rate was 39%. All the respondents had provided counseling to victims of the attack, and some were also counseling survivors of other forms of physical or sexual violence. The findings reveal small gender differences though not significant. While 15.9% of the females reported high secondary traumatization, 10.6% of the males did so.
McLean, Wade, and Encel (2003) conducted a mail survey to examine the predictors of various manifestations of distress among Australian psychologists, social workers, and other mental health professionals who treated traumatized clients. One hundred and sixteen professionals, of whom 85 were women and 31 men, returned the questionnaire. According to their reports, 47% worked mainly with child clients, 35% mainly with adult clients, and 18% with a fairly even mixture of children and adults. The findings show that females report a significantly higher level of IES. (d = 0.45, p = .041). Gender differences were not found in the other manifestations of distress, namely, vicarious traumatization and burnout that were examined.
To the knowledge of the author, the 10 studies reviewed above are the only studies that examine STS in mental health professionals using validated measures of PTSD based on DSM criteria, which provide findings on both genders separately. In view of the very small number, I decided to look at studies of STS in a somewhat broader sense of the term. To this end, I looked at studies based on the Compassion Fatigue subscale of Stamm’s (2005) Professionals Quality of Life Survey (ProQOL). This 10-item subscale contains 3 items tapping the three key symptoms of PTSD (intrusion, avoidance, and hypervigilance) and 7 items querying thoughts and feelings deriving from the respondent’s role as a helper of traumatized individuals. It is a well-known, commonly used scale, which Stamm (2005) describes as an instrument that measures STS. Its reported α reliability is .80 (Adams, Boscarino, & Figley, 2006). The present survey relates to the 10 items in the Compassion Fatigue subscale.
Studies of STS Using Stamm’s ProQOL
I found four studies of mental health professionals that have used Stamm’s (2005) ProQOL scale and also provide gender data. One study was conducted by Van Hook and Rothenberg (2009) and the other three were conducted by Sprang and colleagues (Craig & Sprang, 2010; Sprang, Clark, & Whitt-Woosley, 2007; Sprang, Craig, & Clark, 2011). These studies present a somewhat more complicated picture than studies that are using solely PTSD criteria.
Van Hook and Rothenberg (2009) examined STS among 175 child welfare workers in a variety of assignments. Of these, 136 (82.9%) were males and 28 (17.1%) females. Findings showed that females reported more STS symptomatology than males (d = 0.40), yet the effect is marginally significant (p = .06).
The studies by Sprang and colleagues were large-scale surveys, each conducted on a different sample of professionals. Each showed a different pattern of results. The first, by Sprang, Clark, and Whitt-Woosley (2007), examined a sample of 1,121 mental health professionals working in community mental health settings, other public agencies, private practice, or another setting. Findings showed that female professionals had higher mean scores than males (d = 0.19, p = .003). They also showed, however, that the psychiatrists, who, the researchers state, were predominantly male, experienced higher levels of STS than the other professionals.
The second study, by Craig and Sprang (2010), was conducted among a total of 532 experienced clinical social workers and psychologists who identified themselves as having some expertise in trauma treatment. Sixty-five percent of the study participants were women. The study reports that gender made no significant contribution to STS (d = 0.06, p = .54).
The third study, by Sprang, Craig, and Clark (2011), examined 669 professionals from across the United States. More than half were either outpatient mental health workers or child welfare workers. Among males, nearly a third were child welfare workers. The study finds that males experience a significantly higher level of CF, compared with females (d = −0.46, p < .001).
Table 1 summarizes the results reported in all studies reviewed above. For each study, it also shows the type of the STS instrument on which the comparison is based, the sample size of women and men and a brief description of the findings as originally reported by the authors. Also reported are the computed effect-size measure and its related test statistic (Z-score) and exact p-value. Nine studies provided complete data that could be converted into effect size using Comprehensive Meta-Analysis program (Version 2.2.064; Biostat, Englewood, NJ). Among these studies, six found greater female susceptibility to STS under a conventional (Cornille & Meyers, 1999; McLean, Wade, & Encel, 2003; Meyers & Cornille, 2002; Sprang et al., 2007) or more lenient significance threshold (Choi, 2011; Van Hook & Rothenberg, 2009). Gender’s weighted effect on STS was calculated by the random effects method (Borenstein, Hedges, Higgins, & Rothstein, 2009, pp. 69–75). The analysis shows that across the nine effect sizes combined, women experience higher level of STS, yet the mean effect of gender is relatively small (d = 0.261), it is only marginally significant (z = 1.90, p = .057) and varies considerably from one study to the next, Q(8) = 78.38, p < .001.
The 14 papers surveyed above constitute a small percentage of the more than 140 papers on STS among clinicians working with traumatized clients published in English-language professional journals in the last two decades or so. This small number brings home just how very little methodical research attention has been paid to gender differences in susceptibility to STS among clinicians working with traumatized individuals.
Consistent Findings or Mixed Findings?
Six of the 10 studies that examine STS using solely PTSD criteria report greater susceptibility to STS among female mental health professionals. Of the four studies using the ProQOL, two report greater female susceptibility, one reports no significant gender difference, and one reports greater male susceptibility. On the face of it, these findings seem to support the repeated claims that studies of STS show mixed gender results (e.g., Creamer & Liddle, 2005). But do they really? Let us reconsider the studies examining the two different conceptions of STS.
The 10 studies that examine STS on the basis of PTSD symptoms show a fairly consistent gender pattern. None of them shows greater male susceptibility to STS. Eight of them report greater female susceptibility. Only two studies, one among child protection workers (Nelson-Gardell & Harris, 2003), the other among professionals treating adult trauma victims (Adams & Riggs, 2008), report no significant differences.
These exceptions do not really show a mixed pattern of gender susceptibility, however. Nelson-Gardell and Harris (2003) do not report the trauma caseloads of the child welfare workers they examined. The professionals in Adams and Riggs’ (2008) study were student trainees who had worked with trauma victims for relatively short periods of time—40% of them no more, and some less, than two semesters. The lack of gender differences in both these studies may be due to the clinicians’ low exposure to traumatized clients. Unfortunately, this cannot be said for certain, as differences in the exposure data provided by the various studies make levels of exposure difficult, if not impossible, to compare. Some of the studies provide information on the number of years treating traumatized clients (e.g., Choi, 2011), others on the number of hours per week (e.g., Creamer & Liddle, 2005). Yet others report the percentage of traumatized clients in their caseloads (Adams et al., 2008). The two studies that did not find significant gender differences may suggest that gender differences emerge only after a certain level of exposure.
When they do emerge, the gender differences show themselves in all the work settings examined (e.g., child protection services, domestic violence centers) and irrespective of the populations treated and the source of their traumatization: child abuse, adult spousal abuse, or a natural or man-made disaster. The findings show the same pattern whether the primary victims were children or adults and whether the event that precipitated the primary trauma was a collective trauma, such as a terror attack, or an individual trauma, such as sexual abuse or spousal abuse.
The frequently repeated statement in the literature that the gender findings are mixed (e.g., Creamer & Liddle, 2005) apparently stems (at least in part) from a misreading of Wee and Myers’ (2002) findings. These findings actually show greater susceptibility to STS (in the narrow sense of the term) among female therapists than males. The authors clearly write, “The mean CFS score for males was 34.5, for females it was 37.07 …” (p. 73 lines 8–9). They further state that “males had more signs and symptoms of distress than females on all measurement scales except compassion fatigue and burnout” (p. 74 lines 10–11). In other words, the authors indicate that the female professionals scored higher on compassion fatigue (measured by PTSD symptomatology), while the male professionals scored higher on other expressions of distress following engagement with their clients’ traumas. What Wee and Myers’ (2002) findings do indicate is that greater female susceptibility to STS does not necessarily mean that male therapists suffer less as a result of their engagement with their clients’ traumas.
The more diverse picture provided by the studies using Stamm’s ProQOL may have two sources. One is the studies’ samples. These are very large but include many clinicians who work little with trauma clients. As suggested above, a certain load of trauma clients seems to be necessary for gender differences in STS to emerge. The other source lies in the contents of the questionnaire and its underlying conceptualization. In addition to the three PTSD symptoms, the questionnaire taps a variety of other reactions, including edginess, depression, and preoccupation and enmeshment with and dissociation from their traumatized clients. Moreover, 8 of the 10 statements explicitly ascribe the troublesome response to helping their traumatized clients. Rather than simply query responses, many of the statements on the questionnaire require the respondent to assess and make attributions for his or her situation (e.g., “I think that I might have been affected by the traumatic stress of those I help”). The demand for self-assessment invites a certain intellectualization and emotional distancing, which men are more prone to than women are (e.g., Petraglia, Thygesen, Lecours, & Drapeau, 2009).
Explanations of the Gender Differences
Three types of explanations for the greater female propensity to STS may be offered. One is a certain confounding in the studies between primary and secondary traumatization. As is well known, direct exposure to a traumatic event is a major risk factor for STS. For this reason, all but one of the described studies (Choi, 2011) inquired about such exposure, whether in childhood or adulthood and whether in the subjects’ personal or professional lives. Thus, we cannot know to what extent the symptoms reported derive from the therapists’ engagement with traumatized clients and to what extent they derive from their own direct exposure to a traumatic event. The question arises in the two studies of therapists who treated terror victims (Adams et al., 2008; Wee & Myers, 2002), as some of the therapists in these studies had been directly exposed to the attack. The confounding has implications for the gender findings. Repeated studies show that females are at greater risk for PTSD than males are (e.g., Norris et al., 2001). If the therapists’ STS is affected by previous or concurrent PTSD, then the females’ greater vulnerability may reflect females’ greater susceptibility to PTSD, not to STS.
A second set of explanations may be drawn from the literature on PTSD. The most common of these are that the differences in vulnerability are anchored in (a) physiological differences between the genders (e.g., Norris et al., 2001; Saxe & Wolfe, 1999) and (b) females’ greater tendency to report emotional distress (e.g., Gavranidou & Rosner, 2003). Both these explanations have been set forth at length and their applicability to STS is obvious. A third explanation that can be adapted from the literature on PTSD is that STS is a product of fear. The role of fear in PTSD is well known. It is widely understood that PTSD develops when the individual loses his/her sense of safety and feels threatened in life or limb (e.g., Foa & Kozak, 1986). According to Tolin and Foa (2002) variations in susceptibility to PTSD following exposure to traumatic events may be explained by individual differences in their “fear structures.” So too can gender differences in STS, where fear may be aroused by the stories told and emotions conveyed by the PTSD casualty. According to Stamm, STS among helpers is “a negative feeling driven by fear and work-related trauma” (2010, p. 8). Substantial evidence indicates that women report greater fear than men (McLean & Anderson, 2009) and that they are more likely to perceive situations as threatening (Olff, Langeland, Draijer, & Gersons, 2007). These differences may be intensified by the greater propensity of women than men to ruminate on troubling matters (Nolen-Hoeksema, Larson, & Grayson, 1999; Nolen-Hoeksema, Morrow, & Fredrickson, 1993).
A third set of explanations is anchored in the view of the STS literature that the direct casualty’s PTSD symptoms are communicable (e.g., Weinberg, 2011). Three nonexclusive and somewhat overlapping dynamics may be particularly applicable to the found gender differences: empathy, which is a conscious response, and contagion and receptivity to other persons’ projections, which are unconscious (Sabin-Farrell & Turpin, 2003). Empathy is the tendency to feel and understand the other person’s experiences and emotions (e.g., Hoffman, 1977). Writers on secondary traumatization suggest that empathy can lead persons in close contact with trauma survivors to experience their trauma symptoms as their own (Maloney, 1988; Rosenheck & Nathan, 1985). An early review of the literature (Eisenberg & Lennon, 1983) shows that in self-report and laboratory studies, women evidenced greater empathy than men. These findings are reinforced by subsequent studies, showing that women are better at decoding facial expressions than men (Ickes, Gesn, & Graham, 2000; McClure, 2000), are more reactive to others’ emotions (Thunberg & Dimberg, 2000), and score higher on measures of empathy and empathic concern (Britton & Fuendeling, 2005).
Emotional contagion occurs when “an individual observing another person experiences emotional responses parallel to that person’s actual or anticipated emotions” (Miller, Stiff, & Ellis, 1988, p. 254). Writers on secondary traumatization argue that emotional closeness to a traumatized person can lead to the transmission of trauma symptoms through “contagion,” “infection” (Catherall, 1992a; Figley, 1995), or “mimicry” (Coughlan & Parkin, 1987). Empirical support for this view is provided by repeated findings showing that women in a variety of professions score higher than men on a measure of emotional contagion (Crumpei & Dafinoiu, 2012a, 2012b; Doherty, Orimoto, Singelis, Hatfield, & Hebb, 1995) and that women evidence more emotional contagion than men in response to photographs of faces with different emotional expressions (e.g., Wild, Erb, & Bartels, 2001).
Receptivity to others’ projections is discussed in the literature on the transmission of trauma under the rubric of projective identification (Catherall, 1992b; Weingarten, 2004). Various scholars suggest that traumatized survivors project the distress stemming from their traumatic experience onto their therapists who then experiences it as their own (Danieli, 1982; Meyers & Cornille, 2002). Although there is no hard evidence that females are more prone than males to accept the projections of their close others, this may be the case when the person doing the projection is in need, as traumatized persons invariably are. Most scholars who have tried to understand the development of STS view these dynamics as working in conjunction with one another (e.g., Goff & Smith, 2005; Sabin-Farrell & Turpin, 2003).
Male Professionals’ Manifestations of Distress
The findings of greater female susceptibility to STS do not mean that male professionals are immune to emotional consequences of exposure to their clients’ PSTD. Findings of three of the studies using solely PTSD criteria suggest that male therapists manifest their distress differently from their female colleagues. Cornille and Meyers (1999), who examined child protection service workers, found that the males among them scored higher on several measures of the Brief Symptom Inventory (BSI) and the Global Distress Index (GSI) than the females. Namely, males reported higher levels of distress in interpersonal relations and more depression, phobic anxiety, paranoid ideation, and psychoticism. A similar pattern was reported by Kassam-Adams (1999) in her study of professionals treating adult survivors of sexual abuse. Although the gender differences did not reach statistical significance, she found that the male professionals scored higher on hostility, phobic anxiety, and paranoid ideation as measured by the BSI and the short version of Symptom Checklist-90-R. Wee and Myers (2002), who examined mental health workers who provided crisis counseling to persons affected by the Federal Building Bombing in Oklahoma City, found that male professionals had a higher mean score on the GSI than their female counterparts. These findings reinforce the suggestion of Schauben and Frazier (1995), suggesting that “male counselors … may respond very differently to working with survivors” than females (Schauben & Frazier, 1995, p. 62).
Conclusions
The gender picture presented above provides a partial and preliminary response to Figley’s (2002, p. 6) call to examine “… who is most vulnerable [to STS], in what type of work setting and under what type of conditions.” Taken together, the studies on which the picture is based point to heightened female susceptibility to STS when it is measured by PTSD symptom, but not necessarily when it is measured by the ProQOL. As noted above, the mixed findings on the ProQOL may be attributed both to the fact that this instrument measures the difficulties of helping traumatized clients and to the small trauma loads of many of the queried clinicians.
Having said that, it must be kept in mind that all the studies reviewed were observational studies, which varied in the quality of their sample and the data they provided (see Table 1). Of the 14 studies only 7 were based on random samples, and their response rates ranged from a very low 15% to a better 57%. The other seven studies used convenience samples of volunteers, most from a single agency. Almost all the studies had many more females than males—between 2 and 5 times as many. It is impossible to know how the discrepancies in the numbers of males and females studied or other sampling biases may have affected the findings. At the same time, 7 of the 14 studies that showed greater female susceptibility reported gender differences in symptomatology at 0.05 or higher.
Much more study is required to either confirm or refute the pattern of female susceptibility that emerges from the very small number of studies of STS that provide gender data or enable gender comparison. Although a recently published survey of studies of gender receptivity to STS in the family showed a similar pattern (Baum, 2014), more studies are needed to enable better understanding and interpretation of the findings. These should be studies that give gender a more central place than the marginal position it has in the studies cited above. They should also provide information that is missing from many of the cited studies. This includes fuller information on the clinicians’ exposure to traumatized clients, such as the number of traumatized clients and their proportion of the caseload and the number of weekly hours and duration of the work with traumatized clients. It also includes information about the clinicians’ direct traumatization, both prior and concurrent, in both their personal and professional lives. Most studies of STS to date provide data on direct traumatization but do not connect the data to the gender of the subject. This association must be made. Study is also needed to test the explanations offered for the apparently greater female susceptibility to STS. Much of this can be done using existent reliable and validated instruments measuring empathy, emotional contagion, coping strategies, and fear. The field would also be enriched by studies that investigate matters that have not been examined, such as possible differences in gender susceptibility under different circumstances. For example, does it vary with the gender of the traumatized clients and with whether or not the clients are adults of children? Moreover, further exploration of the various differences in the way that men and women manifest their distress is urged as a means of obtaining a fuller and more reliable picture of gender similarities and differences.
Although a great deal remains to be learned, there are enough indications that female clinicians are more susceptible to STS than males, while male clinicians manifest their distress in other ways. Ample thought has already gone into ways of preventing and alleviating STS in clinicians, whether through supervision, collegial support, or self-care. It is urged that gender be incorporated into the discussion.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.