Risk Management with Clients Who Stalk, Threaten, and Harass Mental Health Professionals

Most mental health professionals will be victims of stalking, threatening, or harassing behavior (STHB) by a client at some point in their career (1–6). Despite the predictability of it, most clinicians feel unprepared to manage these situations (4). This finding is consistent with McIvor and Petch’s (7) description of STHB as a significant yet underrecognized and poorly managed problem. Compounding the problem is the lack of empirically derived risk management strategies or guidelines for responding to these clients. Although two sets of general guidelines for managing risk with clients who engage in STHB have been published based on the respective authors’ clinical experiences (8, 9), neither has been empirically tested.

In the only known examination of clinicians’ risk management responses to clients engaging in STHB that incorporates the perceived effectiveness of these interventions, 17 clinicians from various professional backgrounds were interviewed regarding the frequency and effectiveness of 15 risk management strategies across 28 cases of STHB (6). All 15 risk management responses were considered by clinicians to have made things better most of the time; however, clinicians reported a relatively high risk of adverse consequences for some responses (6). In fact, some of the most commonly used responses were among those most likely to be perceived to make the situation worse. For example, of the 15 risk management responses considered, direct confrontation was the third most commonly used response but was judged to have made things better less frequently than any other response except for one. Additionally, several of the risk management responses included by Sandberg et al. (6) appear relatively specific to clinicians working in hospital settings (e.g., notified security).

Purcell et al.’s (4) large-scale survey of 830 Australian psychologists examined clinicians’ use of 13 risk management responses to client stalking. These authors found a 20% (N=162) lifetime prevalence rate of stalking victimization among psychologists. Resentment was the most commonly perceived motivation (N=68, 42%) for stalking, followed by infatuation (N=31, 19%). Clinicians responded in a variety of ways to client stalking; most (N=115, 71%) modified some aspect of their professional and/or personal lives in response to being stalked, and almost all (N=153, 95%) stalked psychologists sought assistance from others (e.g., colleagues, family, friends). Purcell et al.’s risk management items (4) appear more generalizable to clinicians working in diverse settings but do not include clinical risk management responses in which the psychologist and patient interact to reduce risk.

As Sandberg et al.’s (6) data highlight, the frequency with which clinicians use various risk management responses may not necessarily match the perceived effectiveness of these interventions. Purcell et al.’s (4) study did not include data on the perceived effectiveness of the risk management responses, and mental health professionals remain without an empirically informed approach to managing risk with clients who engage in STHB. Attention to the perceived effectiveness of differing risk management responses in adequate samples may allow for better understanding of when, with whom, and why differing risk management strategies are effective with some clients—essential information for clinicians seeking empirically informed risk management practices with such clients.

Characteristics of Clients Who Engage in STHB

Research considering the characteristics of clients who stalk mental health professionals has consistently focused on the twin pillars of motivation and personality pathology. Several general typologies of stalkers have been described that emphasize the stalker’s motivation and the nature of the relationship between stalkers and their targets (10, 11). For example, Mullen et al.’s (10) five-part typology distinguished between rejected, intimacy-seeking, resentful, predatory, and incompetent stalkers. This type of multi-axial integration of motivational and relational characteristics (12) is, by definition, limited in research with those who engage in STHB toward mental health professionals, because the nature of the relationship is predefined. As a result, research specific to those who stalk mental health professionals has tended to emphasize the motivational axis in distinguishing clients who engage in these behaviors. Summarizing the literature on the motives of those who engage in STHB toward clinicians, Pathé and Meloy (13) note that research has consistently identified two predominant motives—anger/resentment and infatuation. These two motives have been found to be useful in differentiating among clients who engage in STHB in research examining the targeting of this behavior toward physicians (14), psychiatrists (15, 16), and psychologists (4).

In the general population, epidemiological research has found prevalence rates of 14.8% for personality disorders (17) and 31.4%–52% among treatment-seeking outpatient psychiatric clients (18, 19). In Keown et al.’s (19) British psychiatric sample, the cluster A and cluster C diagnoses of schizoid and anxious/avoidant personality disorders emerged as the most common personality disorder diagnoses. Researchers have found generally similar rates of psychopathology across stalkers and general offender populations, although stalkers differ in the nature and extent of their personality pathology. For example, stalkers appear more likely than other offenders to have a personality disorder but are less likely to have antisocial (20) or psychopathic (21) personality disorders. Mullen et al. (10) found that 51% of their sample of stalkers had been diagnosed as having a personality disorder, with most in cluster B. Galeazzi et al. (15) similarly found that 35% of their sample of individuals who had stalked mental health professionals met criteria for a personality disorder and that 93% of these personality disorders fell in cluster B. Meloy (22) suggested that this distinct pattern of personality pathology among stalkers is the “product of an attachment disorder that is preoccupied rather than dismissive.”

In this study, we examined overall and relative perceived effectiveness of 18 risk management responses to STHB in a sample of 112 board certified psychologists who endorsed having at least one experience of STHB. We examined client characteristics that might further our understanding of which risk management strategies were perceived as most, and least, effective. Specifically, we examined the effect of various personality disorders, higher or lower levels of personality organization, and client motivation (resentment or infatuation) for engaging in STHB. We then examined frequency of use for 18 risk management responses to determine whether clinicians tended to use differing risk management strategies with clients with differing levels of personality organization and motivation. Next, we examined the clinicians’ perceived effectiveness of these 18 responses overall and then specifically with clients diagnosed as having varying personality disorders, levels of personality organization, and motivation. Our results provide a first step toward advancing empirically informed risk management strategies matched to patients diagnosed as having specific disorders, levels of organization, and motivation for engaging in STHB.

Methods

Participants

A sample of 112 board-certified psychologists (50% male) who endorsed experiencing at least one episode of STHB was identified from a larger sample (N=157) of psychologists certified by the American Board of Professional Psychology (ABPP). These participants had been licensed to practice psychology for 22.02±12.09 years (mean±SD). Fewer than half (46%) were ages 55 or younger, and nearly one-quarter (22%) of the sample was older than age 65. Specialty boards with the most respondents included clinical (32%; N=36), clinical neuropsychology (26%, N=29), and forensic (15%, N=17). More than one-third of the participants worked in public organizations (36%) and private practice (37%), and the remaining participants worked in private organizations (17%), home-based private practices (7%), or other (4%) office environments. Most of these psychologists endorsed being the recipient of unwanted phone calls (62%) or unsolicited letters, faxes, or e-mails (50%). Other common types of reported harassment included experiencing unwanted approaches (38%); having malicious gossip spread or false reports filed with professional boards (32%); having clients loitering around work, home, or other places frequented by the psychologist (32%); being spied upon or kept under surveillance (23%); or followed on foot or in a car by a client (15%). Most of the psychologists in this sample had experienced multiple types of harassment; 75% experienced two or more and 48% experienced at least three forms of harassment. Four (4%) of the psychologists surveyed had been physically assaulted.

Procedures

All psychologists holding board certification status with the ABPP were invited to participate. Participant e-mail addresses were obtained through the ABPP’s online directory. Potential participants were e-mailed an invitation to participate that included a link to an online questionnaire through an encrypted Web site. The invitation described the questionnaire as examining mental health professionals’ experiences with intrusive or harassing behavior by their clients and the risk management responses they had used. To minimize potential self-selection bias and to avoid misconceptions or preconceptions, the term stalking was not used (4). Research has shown that distinct rates of reported stalking vary depend on whether behavioral definitions of stalking (such as those used in this study) versus self-identification as a stalking victim are used (23). Participants were sent a follow-up e-mail an average of seven days after the initial study invitation was sent. All participants who completed the survey were eligible to win one of five $20 Amazon gift cards. Only psychologists endorsing at least one experience with client STHB directed toward them were included in the study.

We used two strategies to test for nonresponse bias. First, using established extrapolation methods to estimate systematic differences between responders and nonresponders, we compared early responders to late responders, the latter of which are thought to be similar to nonresponders (24). Using Lindner et al.’s (24) guidelines, we identified 30 late responders as the final group to respond to our survey, all of whom responded approximately 14 days after the initial invitation (approximately seven days after the follow-up invitation). No significant differences emerged between early versus late responders on key study variables, including rates of reported stalking victimization, ratings of perceived effectiveness across the 18 strategies, or nine of 10 Shedler-Westen Assessment Procedure Prototype (SWAP-P) ratings (the mean score for late responders, 3.42±1.30, was significantly lower than that for early responders [mean=4.07±1.13] for one category). Second, in using Miller and Smith’s (25) method of comparing respondents to the population to control for nonresponse error, we found significant similarities on key study variables in prevalence research conducted in the United States (26) and Australia (4), which further supported the validity of our findings. Therefore, the low response rate does not appear to reflect a systematic selection bias but appears to be the result of the relative length of the survey coupled with the general absence of compensation.

This study was approved by the University of Indianapolis Institutional Review Board. All participants provided informed consent prior to participation.

Measures

Demographic features and professional background of participants.

Demographic information was collected on participants’ gender, age, years licensed to practice independently, primary area of practice, theoretical orientation, and board certification status.

Stalking, threatening, harassing, and attacking behavior questionnaire.

The Stalking Victimization Survey Project developed by Purcell et al. (4) was adapted for this study to measure psychologists’ self-reported experiences with unwanted intrusions by clients in order to allow for direct comparisons to past research. Consistent with Purcell et al. (4), respondents who had experienced unwanted intrusions by multiple clients were instructed to describe only the most recent client to ensure responses captured a discrete stalking episode instead of an aggregation of experiences. Intrusion items included whether a client had ever “(a) followed them; (b) kept them under surveillance; (c) loitered around their workplace, home, or other places they frequent; (d) made unwanted approaches; (e) made unsolicited telephone calls; (f) sent unwanted letters, faxes, or e-mail; (g) sent offensive materials; or (h) interfered with their property” (4). Respondents endorsing any of these items were presented with follow-up questions regarding the frequency and persistence of the unwanted intrusions, whether the behavior caused them fear, whether they were ever threatened or physically attacked, given or inferred reasons for the STHB, and the risk management responses and perceived effectiveness of each response endorsed. For this study, we expanded the risk management section from Purcell et al. (4) in two ways. First, we added five risk management items from Sandberg et al. (6), which are specific to clinical risk management, to Purcell et al.’s 13 items, for a total of 18 risk management options. Second, consistent with Sandberg et al.’s methods, we added a measure of perceived effectiveness for each risk management strategy used by each clinician. The perceived effectiveness of each strategy was measured on a Likert scale, ranging from 1, “it made things much worse,” to 5, “it made things much better,” with 3 reflecting “no impact.” Effectiveness was then recoded into ineffective (“made things much worse,” “made things somewhat worse,” or “no difference”) or effective (“made things somewhat better” or “made things much better”).

SWAP-P scoring (27).

Respondents completed the SWAP-P for the most recent client to engage in STHB toward them. The SWAP-P is derived from the Shedler-Westen Assessment Procedure–II (SWAP, cf. 28), a 200-item clinician-scored Q-sort. The SWAP-P provides empirically derived prototypical descriptors for 10 personality disorders and asks respondents to rate the extent to which their patient matches this prototype on a scale of 1 “no match (description does not apply” to 5 “very good match (patient exemplifies this disorder; prototypical case).” Scores of four and five both indicate that the patient meets diagnostic criteria for this disorder, allowing for dimensional as well as categorical data for each personality disorder. Personality disorders are organized under four hierarchical categories, including those on the internalizing (depressive, anxious-avoidant, dependent-victimized, and schizoid-schizotypal personalities), externalizing (antisocial-psychopathic, paranoid, and narcissistic personalities), borderline dysregulated (borderline dysregulated personality), and neurotic (obsessional and hysteric-histrionic personalities) spectrums. Field research suggests that clinicians can readily, reliably, and accurately assess patients with the SWAP-P (27).

We also created a dichotomous higher- versus lower-level personality organization variable based upon Kernberg’s (cf. 29) object relations theory. Higher-level personality organization was coded for those meeting criteria for (a) no personality diagnoses, (b) for internalizing personality disorders, and/or (c) for neurotic personality disorders only (i.e., no externalizing or borderline dysregulated diagnoses given). Lower-level personality organization was coded for clients with externalizing and/or borderline and/or dysregulated personality disorders (whether alone or in tandem with other higher-level personality diagnoses). This distinction is consistent with Clarkin et al.’s (29) model of personality organization, which included the following descriptive diagnoses within higher-order and neurotic personality organization: obsessive-compulsive, depressive-masochistic, hysterical, avoidant, dependent, histrionic, and nonmalignant narcissism. Lower-order personality organization included paranoid, schizotypal, schizoid, borderline, malignant narcissism, and antisocial personality disorders (29).

Results

Descriptive Statistics

More than seven in 10 (N=112, 71%) psychologists from the larger sample reported experiencing some form of STHB. Nearly one in seven (N=22, 14%) endorsed experiences that met research criteria for having been stalked (defined as 10 or more intrusions over a minimum of two weeks that cause the clinician fear), more than one in five (N=33, 21%) reported having been threatened, and 4% (N=4) endorsed having been physically attacked by a client at some point in their career. Except for physical assault, which was experienced only by male psychologists, no gender differences emerged in rates of STHB victimization. Clients who engaged in STHB were approximately even across gender (male=55%, N=60). Male and female clients engaged in STHB toward male and female psychologists at similar rates (χ²=0.05, df=1, N=108, p=0.83).

Clients who engaged in STHB were, on average, ages 39±12 years. Most of the clients who engaged in some form of STHB had a psychiatric diagnosis (N=72, 67%). The primary psychiatric diagnoses for clients who engaged in STHB are provided (Table 1). After personality disorders, psychotic and mood disorders were most common. Most clients who harassed psychologists met SWAP-P criteria for at least one personality disorder (65%; N=68) and a substantial minority met criteria for at least two (31%; N=33). Nearly one in six (15%; N=16) met criteria for three or more personality disorders. Among clients who engaged in STHB, externalizing spectrum personality disorders were the most common, followed by borderline/dysregulated spectrum, internalizing spectrum, and neurotic spectrum disorders (Table 2).

Data Analytic Strategy

Of the 112 psychologists who experienced STHB, 105 reported the interventions they had used (Table 3). The sample size on which effectiveness ratings were calculated varied considerably, because the number of psychologists who responded yes to using each strategy varied for each strategy (Table 3). Also, some respondents who indicated use of a strategy did not rate its effectiveness. Sample size was further constricted because not all respondents completed ratings of client personality organization (N=97) or motivation (N=105) (Table 3). Further, of the 105 psychologists who responded to the motivation query, 67 endorsed infatuation or resentment, 22 endorsed other, and 16 indicated no discernible motivation. Because only infatuation and resentment were considered for this analysis, these responses also reduced the sample size for analyses of motivation.

Risk Management Strategies: Overview of Use and Perceived Effectiveness

Changes to personal or professional life.

Of the 112 psychologists responding, 33% (N=35) reported increasing workplace security (Table 3). Approximately twice as many respondents rated this strategy as effective (66%, N=21) than ineffective (34%, N=11). Of the respondents, 18% (N=19) reported increasing home security, and this was rated about equally effective (53% effective, N=9; 47% ineffective, N=8). The respondents rarely reported changing their office phone number (1%, N=1) or location (1%; N=1) or their home phone number (4%, N=4), and none reported relocating their home. Only 4% (N=4) reported reducing their social outings.

Clinical risk management responses.

Of the psychologists responding, 55% (N=58) confronted the client directly. Of these, most (57%, N=27) rated this direct discussion of the behavior as effective (Table 3). Overall, 36% (N=38) of the respondents reported referring the individual who engaged in STHB for treatment or evaluation elsewhere, and less than half (47%, N=15) reported that referral was effective. Eleven percent (N=12) of psychologists reported having hospitalized the individual engaging in STHB. Only 25% (N=2) of respondents rated hospitalization as effective overall in reducing or stopping the STHB. Very few of the psychologists involved law enforcement to manage STHB. About 5% (N=5) reported having the individual arrested, and 7% (N=7) reported obtaining a restraining order against the individual. However, among this small group, the respondents were three times more likely to rate arrest as effective than ineffective and five times more likely to rate a restraining order as effective than ineffective.

Seeking advice or assistance.

Of the 35% (N=37) of respondents reporting seeking advice or assistance from family members or friends, 80% (N=24) rated this strategy as effective. Of the 74% (N=78) of psychologists who reported seeking advice or assistance from colleagues or supervisors, 78% (N=47) reported that this was an effective strategy. Overall, 22% (N=23) of respondents reported seeking advice or assistance from police. Of these, most (68%, N=15) reported that this strategy was effective. Of the 28% (N=29) of respondents who reported seeking advice or assistance from a lawyer, 77% (N=17) reported that it was effective. About 11% (N=11) reported seeking the assistance of a psychotherapist, and 56% (N=5) reported that this was effective. Of the 18% (N=19) of respondents who reported seeking advice or assistance from a professional indemnity provider, 67% (N=8) reported that this was effective.

Risk Management Strategies: Use and Perceived Effectiveness by Client Motivation and Personality Organization

Clients motivated by resentment.

For STHB by clients motivated by resentment, the respondents were most likely to report seeking assistance from colleagues and/or supervisors (77%, N=39), family and/or friends (43%, N=22), and legal counsel (37%, N=19). Among these popular strategies, assistance from colleagues and/or supervisors was perceived as effective 74% (N=20) of the time, and family and/or friend assistance was perceived as effective 81% (N=13) of the time. Lawyer assistance was rated as the most effective (92%, N=12) strategy for resentful clients. Other common strategies included confronting the client directly (37%, N=19) and increasing workplace security (33%, N=17) (Table 3). Increasing workplace security was rated as effective 57% (N=8) of the time, whereas directly confronting clients perceived to be motivated by resentment was rated as effective only 36% (N=5) of the time (Table 3).

Less common strategies used with resentful clients included seeking assistance from a professional indemnity provider (26%, N=13), increasing home security (24%, N=12), referring the client elsewhere (24%, N=12), and seeking police assistance (22%, N=11). The respondents rated professional indemnity provider assistance as effective 57.1% (N=4) of the time, increasing home security as effective 50% (N=5) of the time, and referring the client as effective 44% (N=4) of the time; whereas police assistance was rated as effective 80% (N=8) of the time. Rarely used strategies included assistance from a psychotherapist (12%, N=6), client hospitalization (10%, N=5), client arrest (6%, N=3), reduction of the psychologist’s social outings (4%, N=2), placing a restraining order against client (4%, N=2), and changing office or home phone numbers (both 2%, N=1). Perceived effectiveness of these rarely used strategies is presented (Table 3).

Clients motivated by infatuation.

For STHB by clients motivated by infatuation, the respondents were most likely to report seeking assistance from colleagues and/or supervisors (75%, N=12), confronting the client directly (69%, N=11), referring the client elsewhere (69%, N=11), and increasing workplace security (50%, N=8). Among these popular strategies, assistance from colleagues and/or supervisors was rated as the second most effective overall (91%, N=10) against STHB by infatuated clients. Psychologists perceived confronting the client as effective 50% (N=5) of the time, referring the client as effective 36% (N=4) of the time, and increasing workplace security as effective 63% (N=5) of the time.

Less common strategies used with infatuated clients included seeking assistance from family and/or friends (38%, N=8), legal counsel (31%, N=5), and police (25%, N=4). Seeking assistance from family and/or friends was rated as effective 100% (N=6) of the time, seeking legal counsel was rated as effective 40% (N=2) of the time, and police assistance was rated as effective 50% (N=2) of the time. Rarely used strategies included increasing home security (19%, N=3), seeking assistance from a professional indemnity provider (13%, N=2), obtaining a restraining order against the client (13%, N=2), hospitalizing the client (6%, N=1), and seeking psychotherapy (6%, N=1). Perceived effectiveness of these rarely used strategies is presented (Table 3).

Clients with higher-level personality organization.

For clients with higher-level personality organization, the most commonly endorsed strategies included seeking assistance from colleagues and/or supervisors (69%, N=29) and confronting the client directly (50%, N=21). Of these most popular strategies, colleague and/or supervisor assistance was rated as effective 78% (N=18) of the time, and confronting the client was rated as effective 67% (N=12) of the time. Less commonly used strategies included referring the client elsewhere (36%, N=15), increasing workplace security (33%, N=14), seeking assistance from family and/or friends (29%, N=12), seeking police assistance (29%, N=12), and seeking assistance from legal counsel (21%, N=9). Seeking assistance from family and/or friends was rated as the most effective strategy (80%, N=8) for clients with higher-level personality organization. Perceived effectiveness was greater than 50% for the remaining less-common strategies: referring the client was rated as effective 58% (N=7) of the time, increasing workplace security was rated as effective 54% (N=7) of the time, seeking police assistance was rated as effective 64% (N=7) of the time, and seeking legal assistance was rated as effective 57% (N=4) of the time. Rarely used strategies included increasing home security (14%, N=6), seeking psychotherapy (12%, N=5), seeking assistance from a professional indemnity provider (12%, N=5), obtaining a restraining order (10%, N=4), hospitalizing the client (5%, N=2), having the client arrested (5%, N=2), reducing the psychologist’s social outings (5%, N=2), and changing the psychologist’s home phone number (2%, N=1). Perceived effectiveness of these rarely used strategies is presented (Table 3).

Clients with lower-level personality organization.

For clients with lower-level personality organization, seeking assistance from colleagues and/or supervisors was the most commonly used strategy (78%, N=42), followed by confronting the client directly (56%, N=31). Colleague and/or supervisor assistance was rated as effective 75% (N=24) of the time, and confronting the client was rated as effective 54% (N=13) of the time. Less commonly used strategies for clients with lower-level personality organization included seeking assistance from family and/or friends (40%, N=22), referring the client elsewhere (35%, N=19), increasing workplace security (31%, N=17), seeking legal counsel (31%, N=17), and seeking assistance from a professional indemnity provider (24%, N=13). Of these less commonly used strategies for clients with lower-level personality organization, seeking assistance from family and/or friends (77%, N=13) was rated as the most effective strategy. Increasing workplace security was rated as effective 75% (N=12) of the time, referring the client was rated effective 44% (N=7) of the time, seeking legal assistance was rated as effective 85% (N=11) of the time, and seeking assistance from a professional indemnity provider was rated effective 56% (N=5) of the time. Rarely used strategies included increasing home security (18%, N=10), hospitalizing the client (17%, N=9), seeking police assistance (17%, N=9), seeking psychotherapy (9%, N=5), obtaining a restraining order against the client (6%, N=3), changing the psychologist’s home phone number (6%, N=3), having the client arrested (4%, N=2), changing the psychologist’s office phone or address (2%, N=1), and reducing the psychologist’s social outings (2%, N=1). Perceived effectiveness of these rarely used strategies is presented (Table 3).

Comparison of Strategy Use and Perceived Effectiveness Across Client Characteristics

Across client characteristics, few statistically significant differences in strategy use or perceived effectiveness were observed. Nevertheless, differences emerged across client characteristics with the use of client confrontation, client referral, client hospitalization, and legal counsel—all among the more common risk management strategies used. Specifically, respondents were almost twice as likely to confront infatuated clients versus resentful clients (χ²=4.36, df=1, N=58, p<0.05) (Table 3). Psychologists were three times more likely to refer infatuated clients than resentful clients (χ²=10.33, df=1, N=65, p<0.001); however, this strategy had the lowest ratings for perceived effectiveness in the infatuated client group compared with the all the other client groups offered referrals (Table 3). Respondents were 3.5 times more likely to hospitalize clients with lower-level personality organization (χ²=3.30, df=1, N=96, p=0.06). Finally, the psychologists rated seeking assistance from a lawyer as more effective for resentful clients than for infatuated clients, (χ²=5.72, df=1, N=18, p<0.05).

Discussion

This study builds upon previous research into mental health professionals’ responses to STHB. Published global guidelines for interventions have been based on clinical experience and general effectiveness data. However, empirically informed guidelines and specificity for different client populations are lacking. Our study describes characteristics of clients who engage in STHB, the prevalence of psychologists’ interventions in response to STHB, and the perceived effectiveness of each of these risk management strategies as a function of both intervention type and client characteristics.

Client Characteristics

Clients engaging in STHB in this sample were, on average, in their late 30s and equally likely to be male or female. Most had psychiatric diagnoses, with personality disorders most common, followed by mood and psychotic disorders. In contrast to previous research, substance use disorders were relatively rare.

Risk Management Strategies by Prevalence and Perceived Effectiveness

In this study, psychologists who had experienced STHB reported on their use and perceived effectiveness of risk management strategies from three broad categories: seeking advice or assistance within personal and professional relationships, making changes to personal or professional life, and using clinical risk management strategies. Strategies involving seeking assistance from important relations or other mental health professionals were among the most popular, and they were generally rated as effective.

The most common strategy, endorsed by more than three-quarters of the respondents, was to seek support from colleagues or supervisors. Seeking colleague or supervisor support was generally rated as effective regardless of client characteristics. Although fewer psychologists endorsed seeking support from family and/or friends—about one-third of the sample—this strategy was rated about as effective as seeking support from colleagues. Thus, one conclusion is that a combination of professional and social-emotional support was perceived as a vital component in managing STHB of a client. This finding supports Meloy’s (9) guideline of using a team approach to STHB. Moreover, because seeking social-emotional support was less popular than conferring with colleagues, although both were considered effective, it may be important to understand barriers and develop resources for psychologists to use this strategy. For example, training and continuing education programs would be ideal settings in which to help psychologists navigate how to seek social support for STHB while maintaining client privacy and confidentiality.

The most common change to personal or professional life was to increase workplace security, with one-third of respondents endorsing this strategy. About two-thirds of respondents who increased workplace security reported that this strategy was effective. The second most common strategy was to increase home security, although this was comparatively uncommon; only 18% of respondents reported making this change. Overall, the respondents rated increased home security as effective about half the time; however, this rate was significantly higher for the respondents in private practice or who had a home-based office. The use and effectiveness of changes to workplace and home security also varied by client characteristics, discussed below. Other changes to personal or professional life, such as changing home or office telephone number or address and reducing social outings, were rare. Overall, the respondents who reported making changes in their personal or professional lives in response to STHB were adhering to Meloy’s (9) guideline of taking personal responsibility for safety.

The most commonly used clinical risk management response, and second most common response overall, was to confront clients directly. Slightly over half of the respondents who confronted their clients reported that this strategy was effective. The overall effectiveness of confronting clients was almost identical to that of a previous study (6). Thus, consistent with previous research, confronting clients about STHB was popular, although only modestly effective. Similarly, referring clients elsewhere for services was the third most common response overall, yet it too was perceived as only modestly effective, by less than half. However, the use and effectiveness of confronting and referring clients varied based on client characteristics. This finding indicates the importance of approaching STHB within the context of individual client characteristics, which is discussed below.

Effective Risk Management Strategies by Client Characteristics

For clients with higher-level personality organization, the most popular strategies were generally the most effective. Specifically, most forms of support or assistance were more effective than not and confronting the client directly and referring the client elsewhere for services were perceived as effective for a modest majority of these higher-functioning clients. Furthermore, direct confrontation had the highest rate of perceived effectiveness for clients with higher-level personality organization across all types of clients in our study, and, on average, referral elsewhere was rated as effective slightly more than half of the time for higher-functioning clients. Of the rarely used intensive clinical risk management approaches—hospitalization, arrest, and restraining order—only obtaining a restraining order was rated as an effective strategy for higher-functioning clients. Although comparatively few psychologists consulted with their professional indemnity provider regarding clients with higher-level personality organization (compared with the overall sample), all who did so reported that this strategy was effective. Finally, increasing workplace and home security were both rated as effective about half of the time for STHB in higher-functioning clients. Taken together, responses that were perceived as effective for clients with higher-level personality organization tended to be the least invasive and most popular. The perceived effectiveness of these strategies may reflect these higher-functioning clients’ ego strength, which allowed them to tolerate direct conversations about the psychologist-client relationship and modify their behavior in response to minimally intrusive limit setting.

For clients diagnosed as having lower-level personality organization, effective management of STHB appeared to involve respondents taking a complex self-protective stance. Increasing workplace security and seeking assistance from a lawyer were rated as especially effective for clients diagnosed as having lower-level personality organization and seeking assistance from police also was generally perceived as effective with this group. Changing office phone number or address, having the client arrested, and obtaining a restraining order were each rare but also generally rated as effective with clients diagnosed as having lower-level personality organization. Support from family and/or friends, professional colleagues, and psychotherapy were effective at rates similar to the overall sample. Rates of hospitalization were highest for clients with lower-level personality organization, although this response often was rated as ineffective for curbing STHB of these clients. Confronting clients with lower-level personality organization was common but was only slightly better than equivocal in its perceived effectiveness. These findings highlight that the most effective pattern of risk management for clinicians dealing with STHB in a client with a severe personality disorder occurred through professional protection (e.g., increased workplace security and legal assistance). More popular, less intensive approaches, such as referring the client elsewhere, were more often rated as ineffective. This result is likely a reflection of these clients’ difficulties with cognitive, behavioral, and personality dysregulation and disorganization, which require more concrete forms of limit setting within the professional relationship. Given that this is a clinically difficult and heterogeneous client group, training and continuing education would likely be useful for helping clinicians navigate the experience of preventing and responding to STHB when treating and assessing clients who have severe psychopathology. Finally, although client hospitalization was often ineffective against STHB, hospitalization may be clinically indicated for clients with severe psychopathology, regardless of its effect on STHB.

For clients motivated by infatuation, psychologists’ support seeking was perceived as the most effective constellation of strategies. Specifically, seeking support within close relationships, from a therapist, and from professional colleagues and/or supervisors, was always or almost always rated effective by the respondents with infatuated clients engaging in STHB. Increasing home and workplace security was rated as more effective than not, although not dramatically so. Confronting the infatuated client directly was effective only half the time. Referring infatuated clients to other providers was a popular strategy but was rated as generally ineffective. However, given the small number of clients described as motivated by infatuation, future research is needed to discern whether there are additional strategies that may be effective with this group. For example, approaches may vary based on individual clients’ overall functioning and perceived dangerousness. Further, the social support required to help a client work through a strong, yet nonpsychotic transference reaction, will and should differ from the response needed to cope with and contain STHB in response to therapist-directed erotomania.

For clients motivated by resentment, approaches designed to contain the STHB with external assistance were perceived as the most effective. Almost all psychologists who sought assistance from police or attorneys for resentful clients reported that this strategy was effective, and police and attorney support had the highest effectiveness rates for resentful clients compared with all other types of clients assessed. Similarly, although rare, restraining orders, arrests, and changing office phone number were unanimously rated as effective in addressing STHB of resentful clients. Effectiveness of increasing home and workplace security in response to STHB among resentful clients was slightly lower than overall rates, whereas the effectiveness of social and mental health professional support was similar to that for the overall sample. Alternatively, despite their popularity, confronting and referring resentful clients were generally ineffective strategies, as was hospitalizing the client. These findings emphasize that, although less forceful approaches may sometimes work to reduce STHB of resentful clients, strategies perceived as effective tended to involve setting higher-intensity limits. As such, a different kind of team approach is likely necessary for these clients. In addition to a strong professional and social-emotional support network, effective coping with STHB from resentful clients may require a team of law enforcement and legal advisors. Further, resentful clients are more likely to require clinicians to adhere to Meloy’s (9) guidelines of “protection orders, law enforcement, and prosecution.”

Strengths, Limitations, and Future Directions

This study strengthened our knowledge of strategies perceived by clinicians to be effective for dealing with clients engaging in STHB, using responses from a sample of experienced clinicians. One limitation of these findings is that it is not clear how the risk management strategies were applied. Confrontation and referral with a psychotherapy client and a forensic assessment client are necessarily different tasks, and both differ from confronting a client hiding in the bushes outside one’s home. Thus, future research should continue to investigate how psychologists faced with STHB implement risk management strategies. Also, because clinicians tend to use multiple strategies and implement them at different points over the course of STHB, it will be important for future researchers to examine issues related to the timing and sequencing of risk management responses.

Further, while the use of the SWAP-P allowed for the first standardized assessment of the personality characteristics of clients who engage in STHB, this study was limited by having respondents rate previous clients, given that SWAP validation studies have tended to rely on ratings of current clients (30, 31). Because STHB is a far less common experience for clinicians than working with clients with personality difficulties, a large number of clinicians would likely be required to identify an adequate sample to allow researchers to focus exclusively on those currently experiencing STHB. Additional research supporting the reliability and validity of retrospective applications of the SWAP-P over varying lengths of time would further strengthen this study’s findings. It is also possible that clinicians’ SWAP-P ratings were affected by their experience with the client who engaged in STHB. This possible limitation is not unique to our study and likely reflects a challenge common to clinicians working with patients with severe personality pathology that evokes strong countertransference reactions.

As with other surveys examining STHB, self-selection sampling bias may have systematically influenced the results. However, because the primary focus of this study was to evaluate the effectiveness of a range of risk management responses to STHB in a nonrandom sample of psychologists who had endorsed experiences of STHB—as opposed to establishing the prevalence of STHB or different risk management responses—it is expected that our findings are relatively robust to these sampling issues. Although systematic error may have been introduced if study participation was directly related to perceptions of risk management effectiveness with specific clients, we consider this outcome unlikely. Although the possibility of systematic bias cannot be definitively ruled out, further support for a lack of systematic bias was provided by extrapolation analyses and comparisons to population findings, both of which further support the cautious generalizability of our findings.

Conclusions

Most clients who engaged in STHB had moderate to severe psychopathology and were viewed as having these conditions at the time they engaged in STHB. Whereas several risk management responses to STHB were perceived as effective for all clients, others evidenced distinct patterns of perceived effectiveness depending on client personality and motivation. With clients who had more severe personality pathology and who were motivated by resentment, clinicians appeared to benefit most from interventions that integrated external supports into the clinical dyad, whereas clinicians working with higher-functioning clients who were motivated by infatuation also reported some benefit from various clinical interventions within the psychologist-client relationship. Overall, our results indicate that the STHB itself and the perceived effectiveness of risk management strategies may vary depending on client characteristics. Further, these findings point to a disconnect between the prevalence of some risk management responses and their perceived effectiveness, highlighting the need for empirically based risk management guidelines.