Clinical Appraisals of Individual Differences in Treatment Responsivity Among Patients With Psychopathy: A Consensual Qualitative Research Study

Abstract

This study addressed which factors expert clinicians consider crucial in successful completion versus dropout in the mandatory forensic psychiatric treatment of psychopathic patients in the Netherlands. Eleven clinicians were interviewed about patient characteristics, treatment (provider) characteristics, and other factors they deemed associated with failure (transfer to another facility) or completion. The interviews were coded using the guidelines of Consensual Qualitative Research (CQR). Overall, extremely high scores on Psychopathy Checklist–Revised (PCL-R) Facets 1 (Deceitful Interpersonal Style) and 2 (Defective Affective Experience) were thought to impede treatment retention, particularly by its negative impact on motivation and therapeutic relationship. Older patients, those with a prosocial network, and/or patients with comorbid borderline traits appeared to fare better. Treatment success was deemed more likely when treatment goals and expectations are stipulated in a concrete fashion, when an extended and gradual resocialization trajectory is offered, and the treatment team is expert, cohesive, and stable.

Keywords

psychopathy forensic psychiatric treatment qualitative analysis qualitative methods responsivity

Both clinical experience and empirical research into inpatient forensic psychiatric treatment for psychopathic patients to date suggest that these patients are particularly difficult to treat. In early treatment studies, justice-involved individuals with psychopathy showed less motivation, less compliance with treatment, and conversely, more often involvement in institutional misconduct, and more drop out as compared to non-psychopathic individuals (Ogloff et al., 1990; Wong & Hare, 2005). However, a handful of more recent reviews is more encouraging (e.g., Hecht et al., 2018; Polaschek, 2014; Polaschek & Skeem, 2018; Reidy et al., 2013). Moreover, these early findings do not rule out that for some subgroups of people with psychopathy, treatment may be effective in reducing criminal behavior. For example, Olver and Wong (2009) reported on the therapeutic response of males convicted of a sex offense in a high-risk program. Psychopathy was found to be a strong predictor of dropout. However, those psychopathic individuals who remained in treatment and made progress on risk-related treatment targets were less likely to recidivate violently than the non-completers. In similar vein, a Dutch study (Klein Haneveld et al., 2018) targeted individual differences in treatability by seeking empirically based subtypes of male violent patients in a forensic psychiatric hospital. Based on latent profile analysis of scores on the Psychopathy Checklist–Revised (PCL-R; Hare, 2003), three psychopathic profiles emerged: one prototypical group, with high PCL-R scores (and thus displaying virtually all defining features of this composite construct); and two groups with moderate psychopathic scores on the PCL-R. Dropout was (again) very high in the prototypical psychopathic group (nearly 50%), but those who remained in treatment showed recidivism rates commensurate with the less severe psychopathic group. Hence, one hurdle to overcome in the search for effective treatments for psychopathic justice-involved people is the issue of attrition.

To our knowledge, only a handful of studies have examined treatment attrition in individuals with psychopathy. These studies (Cullen et al., 2011; Jeandarme et al., 2017; Olver & Wong, 2011; Sewall & Olver, 2019) reported on the association between non-completion and the Two-Factor or Four-Facet models of the PCL-R (Hare, 1991, 2003) or the Two-Factor model of the Screening Version (PCL: SV; Hart et al., 1995). In a group of 154 males convicted of a sex offense receiving a high-intensity treatment program, Olver and Wong (2011) found that all Facets were related to non-completion, but that only the Affective Facet (Facet 2) made a significant unique contribution to the prediction of dropout. Olver and Wong suggested that shallow affect, callousness, and lack of empathy interfere with the ability to “connect” with the patient and to form a working alliance in treatment. In a British study (Cullen et al., 2011) in a group of 84 justice-involved individuals with a mental disorder sampled from six medium security forensic hospitals, dropout was studied as part of a randomized controlled trial of a cognitive skills program. In this group, Factor 2 (impulsive and antisocial behavior) was found to predict treatment attrition. A more recent Belgian study (Jeandarme et al., 2017) also assessed outcome in mentally disordered justice-involved people receiving medium security treatment (N = 224). Contrary to the British study, Factor 1 predicted dropout after controlling for offense-related and clinical variables (Jeandarme et al., 2017). Factor 1 encompasses Facet 1 (Interpersonal) and 2 (Affective). Finally, in a study by Sewall and Olver (2019) in 302 males convicted of sex offenses, only Facet 3 (Lifestyle) made a significant unique contribution to the prediction of dropout, and a trend was noted for the Affective Facet 2 (p = .081). In sum, although the Affective Facet emerges as the only replicated predictor, no clear picture emerges from these studies about which aspects of psychopathy, as measured with the PCL-instruments, are associated with treatment attrition among justice-involved people. Moreover, all studies focused exclusively on patient characteristics associated with dropout. Clinical experience (Kröger et al., 2014) strongly suggests that contextual factors may also play a role, such as characteristics of the treatment team, or institutional policies unrelated to either the patient or the treatment program. We are not aware of any studies reporting on these possible determinants of treatment dropout versus completion in justice-involved patients with psychopathy.

The Present Study

Our goal was to harness clinical experience and to generate more inclusive hypotheses on differences in treatability among patients with psychopathy. In view of the current state of empirical research, we decided to use a qualitative research design. This approach made it relatively straightforward to explore a wider range of possible factors related to dropout beyond patient characteristics only. In our analytic approach, we adhered to the guidelines provided by the Consensual Qualitative Research paradigm (CQR; Hill, 2012; Hill et al., 1997, 2005). CQR is characterized by (a) the use of semi-structured interviews; (b) a method of data analysis in which a primary team of several judges bring multiple perspectives to the data and then work toward consensus about the meaning of the data; (c) a coding system that consists of domains, core ideas, and categories that are used in the cross-analysis; and (d) at least one auditor who checks the work of the primary team. We focused on the following questions:

Which patient characteristics are associated with failure or successful completion of treatment?

Which factors related to the treatment providers are associated with failure or successful completion of treatment?

Which other factors influence failure or successful completion of treatment?

Method

Setting

In the Netherlands, justice-involved people who were deemed psychopathic have been committed for forensic psychiatric treatment since 1928. In that year the so-called “Psychopath Laws” came into existence, which made it possible to impose a TBS order for people with a mental disorder. TBS (“ter beschikking stelling”) is a measure of mandatory intensive treatment, usually of indefinite length, that can be ordered by the Dutch courts, together with a sentence for violent or sexual offenses. The primary goal of treatment is to minimize risk of reoffending while working toward gradual rehabilitation. Notably, assessment and treatment methods have varied widely over the decades. The translation of the PCL-R into Dutch in 2001 (PCL-R; Vertommen et al., 2002) marked the beginning of systematic assessment of psychopathy. Since 2005, the Dutch Ministry of Justice requires a PCL-R assessment for every TBS patient in the Netherlands (if sufficient information is available). The result of this regulation is that the concept of psychopathy as defined by the PCL-R has become embedded in clinical practice.

Participants

This study was conducted at two of 11 high-security hospitals in the Netherlands providing treatment for patients with a TBS order. Hill et al. (1997) suggested that expert interviewees be selected from a homogeneous population that is very knowledgeable about the subject under investigation. For this reason, all participants were licensed psychologists with at least a Master’s degree and a minimum of 5 years of post-license clinical experience in a specialized TBS hospital. We were confident that these requirements would result in a sufficiently knowledgeable and homogeneous sample. The first 10 interviewees (six of whom were female) were employed in the same clinic. During the analysis of the data and the development of the coding system, a licensed psychologist (male) from a second clinic was asked to participate, to determine whether the coding system was applicable to his interview as well, and to ascertain whether new themes emerged (which was not the case).

Treatment

All patients for whom treatment progress was evaluated in the interviews were admitted to a high-secure forensic psychiatric hospital under a TBS order. Patients are admitted immediately after completing a prison sentence. TBS treatment starts with a high-security inpatient phase. When an independent board of professionals assesses that sufficient progress has been made on risk-related targets of treatment, permission for leave outside the hospital can be obtained from the Ministry of Justice; first supervised, then unsupervised leave. In the final phase, patients leave the hospital, but remain under supervision until their eventual release. All decisions about leave are based on an extensive review of the treatment gains, including structured risk assessment. TBS treatment currently takes 7 to 8 years on average, including several years outside the TBS clinic (Tbs inzichtelijk, 2019). The court evaluates progress in treatment every 2 years and decides whether or not to extend or terminate the order. In this study, treatment was considered successful when a patient proceeded through all the phases of treatment, and the court subsequently ended the TBS order. Because it is impossible to drop out of the (mandatory) TBS treatment, a patient was considered a dropout (or treatment failure) when the forensic psychiatric treatment failed or stalled and a transfer to another hospital for alternative treatment was deemed necessary.

Research Team

The research team consisted of four researchers with varying backgrounds and prior experience: one senior researcher specialized in forensic psychiatry, a psychology student working on her Master’s degree, a licensed clinical psychologist/psychotherapist with extensive experience in working with patients with psychopathy (first author), and a full-time university professor/clinical psychologist with substantial forensic research experience. Three were female, and all were native Dutch. The first three formed the primary team, and the university professor served as the auditor. As advised by Sim et al. (2012), the team spent considerable time discussing their expectations and potential biases during the first meeting. On occasion, these biases would surface during the primary research team data analysis efforts (e.g., a mildly cynical reaction to an example of patient behavior while formulating a core idea). When this happened, these biases were immediately discussed to limit their influence. Nevertheless, it is important to acknowledge that the members of the research team were co-constructors of the data and the results. A different team may have come to slightly different results; see, for example, a comparison of cross-analyses by two teams with the same data by Ladany and colleagues (2012).

Procedures

Recruiting Participants

This research project was approved by the University of Amsterdam faculty of Social and Behavioral Sciences ethics review board. Expert psychologists were recruited per email by the first author. The first author knew all participating psychologists as peer colleagues. Of the 14 experienced psychologists who were invited for an interview, 10 consented to participation. All participants gave written informed consent for the recording of the interview and the use of the transcript.

Ideally, in qualitative research, data collection continues until no new themes emerge. In an update of CQR, Hill and her colleagues (2005) argued that it has proven difficult to establish with certainty whether this has been achieved. Based on 25 CQR-based studies, they recommend a sample of 8 to 15 participants, depending on the homogeneity of the group (i.e., with a heterogeneous sample requiring more participants). Given our selection criteria (see under “Participants”), we were confident that our sample of 11 experienced forensic practitioners was sufficiently homogeneous and expert to produce consistent results.

Interviewing

The first author conducted the interviews between November, 2017, and April, 2018. All interviews were conducted by telephone and were recorded using the “TapeACall” app. In two cases, part of the respective interview did not record well. Because the interviewer also took extensive interview notes, the missing information could be recovered from these notes (the same day). In line with recommendations by Burkard et al. (2012), the instructions and the interview questions were sent to the participants prior to the interview (see Table 1). Accordingly, participants can more fully appraise to what they may give their informed consent (or not). Considering the fact that all participants knew the interviewer as a colleague, this also enabled them to consider whether they felt free to discuss their thoughts on this particular subject with her specifically. In addition, having the protocol in advance gave participants time to reflect on their complex experiences with these difficult patients. We asked participants to select two male patients whom they had treated in the past 10 years, both scoring 28 or higher on the PCL-R, with one of the patients having completed treatment as intended (i.e., completion), and the other having been transferred to another clinic due to a failed treatment attempt (i.e., dropout). We felt that discussing two specific cases might serve to elicit more specific and concrete answers, rather than vague generalities. Finally, several steps were taken to guarantee the anonymity of the cases that were discussed by the participants during the interview. Participants were only asked for the PCL-R total score and the year of birth of the patients involved. When accidentally a name was mentioned, we deleted these from the transcripts. Potentially identifying behavioral examples were not used to illustrate the findings in this article.

Table 1:

Semi-Structured Interview

Instructions for the preparation of the interview
- Please select two male patients whom you have treated in the past 10 years, both scoring 28 or higher on the PCL-R. It is important that one of the patients completed treatment as intended, while the other was transferred to another clinic due to a failed treatment attempt. - See the interview questions below. For both cases, the same questions will be asked. To prepare for the interview, please take one of the cases in mind, read the questions, and reflect upon your answers. Then repeat the same procedure for the second case.
Interview schedule
1. Which specific patient characteristics of the case you have chosen are associated with the success/failure of treatment?
2. For each characteristic, can you describe a specific situation or event that illustrates how this characteristic has influenced the course of treatment?
3. Which factors related to the treatment team are associated with the course of treatment?
4. Can you think of a specific example that illustrates how each factor influences treatment?
5. According to you, are there other factors not yet discussed on the basis of these two cases, which may be essential to take into account when treating psychopathic patients? Can you illustrate these factors with specific examples?

Note. PCL-R = Psychopathy Checklist–Revised.

Transcribing and Coding

All interviews were transcribed by the psychology student in a Word document. The mock interview was used for practice. All transcribed interviews were then proof read by the first author (and interviewer). All interviews were uploaded into MAXQDA (2018), a qualitative data analysis software package (VERBI software, Germany) that we used for coding and further analysis.

PCL-R Records

For all cases discussed by the interviewees, a PCL-R score had been determined at the beginning of treatment, as part of the regular assessment procedure in the clinics involved in this study. Available file information from the PCL-R assessment included criminal records, police records, reports from previous institutes and from prison, and information acquired from relatives, former employers, and schools. Each patient was scored independently by two raters, who then determined a final consensus score together. The raters were licensed psychologists who had completed a 3-day training in scoring the PCL-R.

In his 1991 manual, Hare recommended a cut score of 30 to classify a patient as psychopathic, based on an American sample. However, it has become evident that in European samples a lower cut score may be more appropriate (Mokros et al., 2013). Based on a meta-analysis of samples from German-speaking countries, it has been suggested that a score of 25 reflects the same level of psychopathy as a score of 30 in North American samples (Mokros et al., 2013). We instructed the participants to include cases with a score of 28 or higher, to ensure that all cases could be considered highly psychopathic (i.e., well beyond the European cut-off). The standard error of measurement for standard PCL-R assessments is approximately three points (Hare, 2003). See Table 2 for the total scores of the patients discussed by the participants. All patients were male. Successful cases had a mean total score of 33.33 (SD = 2.58); failed cases had a mean total score of 33.56 (SD = 2.85).

Table 2:

PCL-R Scores and Age of Patients Discussed in the Interviews

Year of birth success	PCL-R success	Year of birth failure	PCL-R failure
1968	31.6	1985	32
1979	28.9	1978	31
1974	34.7	1966	30
1979	31.6	1975	34
1983	33.7	1967	34.4
1982	35.6	1984	38.9
1974	34.7	1961	31
1963	37	1986	35.3
1959	34.4	1968	30.6
1964	29.5	1959	36
1967	35	1971	36

Note. The fractions in the total scores (e.g., 31.6) result from the prorating of scores when items have been omitted, in accordance with the instructions in the PCL-R manual (Hare, 2003). PCL-R = Psychopathy Checklist–Revised.

Instruments

Interview

The semi-structured interview was developed by the first author and revised after feedback from two of the senior researchers. Subsequently, a mock interview was held with a licensed psychologist of one of the hosting clinics. This led to a final revision (see Table 1). Each of the questions could be followed up with probes for further clarification. As can be seen in Table 1, the interview first focused on patient characteristics, then switched to factors related to the treatment team that may have influenced the course of treatment, and finally provided the opportunity to describe further pertinent factors not yet discussed. Participants were asked to illustrate their answers with specific examples from clinical practice, as illustrative examples often yield more elaborate material than purely theoretical discussion.

PCL-R

The PCL-R (Hare, 1991, 2003) consists of 20 items, which can be scored 0 (definitely does not apply), 1 (may apply or partly applies), or 2 (definitely applies), leading to a possible maximum score of 40. This maximum score is considered to represent a “prototypical psychopath.” Extensive psychometric properties have been documented in the manual (Hare, 2003). Inter-rater reliability for a sample from the primary hospital involved in this study has been estimated previously, largely based on the same pairs of raters (Hildebrand et al., 2002). The single-measure intraclass correlation coefficient (ICC) was .88 for the PCL-R total score.

In the first edition of the PCL-R (Hare, 1991), exploratory factor analysis led to the identification of two underlying factors, a model that was replicated several times (Hare & Neumann, 2008). Factor 1 contains the personality traits typically associated with psychopathy: shallow affect, lack of empathy, and a manipulative, arrogant interpersonal style. Factor 2 reflects a chronically impulsive, aggressive, and antisocial lifestyle. In the second edition of the PCL-R (Hare, 2003), Hare revised the original model into the more fine-grained Four-Facet model, which was also replicated repeatedly (Neumann et al., 2015). Facet 1 (labeled “Interpersonal”) refers to a glib, arrogant, and deceptive interpersonal style; Facet 2 (“Affective”) contains items describing the callous and unemotional traits; Facet 3 (“Lifestyle”) describes an impulsive and irresponsible lifestyle; and Facet 4 (“Antisocial”) refers to aggressive and antisocial behavior.

Analytic Strategy

Consensus Coding and Audit

The guidelines explicated in the edited CQR handbook (Hill, 2012) were followed in our data analysis. CQR relies on a consensus process, in which the members of the primary research team first examine the data independently and then meet to discuss their ideas until the entire team agrees on the coding. This process is repeated several times. In this study, it was used first to identify the domains, then to assign all relevant text segments into one or more domains, subsequently to develop categories, and finally to assign all core ideas (summaries of text segments) to the categories. After each step in the consensus process, the auditor (final author) was consulted for feedback. This led to a relabeling and a subtle reorganization of the domains and to a collapsing of several categories into a less elaborate but more straightforward system (see Table 3 for the final result).

Table 3:

Results of Cross Analysis

Domains/categories	Frequency
Domain 1: Patient characteristics
• Not measured by PCL-R	General
• Related to PCL-R Facet 1 (Interpersonal)	General
• Related to PCL-R Facet 2 (Affective)	General
• Related to PCL-R Facet 3 (Lifestyle)	Typical
• Related to PCL-R Facet 4 (Antisocial)	Variant
Domain 2: Team characteristics
• Level of expertise	Typical
• Sufficient emotional distance yes/no	Typical
• Consensus about diagnosis and treatment yes/no	Variant
• Continuity in treatment team yes/no	Variant
Domain 3: Treatment strategies
• Underlying treatment philosophy	General
• Reaction to treatment interfering behavior	Typical
• Specific interventions	Typical
Domain 4: Team–patient interactions
• Working relationship	Variant
• Development of trust	Variant
• Splitting	Variant
• Fear or aversion	Variant
• Punishment	Variant
Domain 5: Social Network
• Prosocial network (yes/no)	Typical
• Network support for treatment	Typical

Note. A frequency of General was coded for appearance of the category in 10 or 11 (all) interviews, Typical for appearance in six to nine interviews, Variant for two to five interviews. PCL-R = Psychopathy Checklist–Revised.

Domains and Core Ideas

There are two methods for constructing a domain list (Thompson et al., 2012). The first is to start with a list based on themes that emerge from a review of the literature and to modify this list during coding to ensure that it accurately reflects the data under investigation. Considering the lack of literature in our area of research, this was not a viable option. We therefore used the second method, which involves an inductive approach. We used three randomly chosen interviews to see what topics emerged and constructed an initial list that was modified throughout the coding and the audit discussions. The next step in the data analysis was to capture the content of each coded text fragment into a so-called “core idea,” a succinct version of what the participant said that is both more clear and concise (Thompson et al., 2012). In this study, the developing of core ideas was not prepared independently by the members of the team, but was done by the primary team immediately upon the coding of the domains.

Cross-Analysis and Categories

Once all relevant text segments of each interview were coded into one or more domains, we started the cross-analysis, which aims to capture common themes across the interviews. Each participant of the primary team independently examined all core ideas per domain and developed a list of categories. In a consensus meeting a joint list was created, which was revised again during the process of coding the categories, and subsequently as a result of the audit discussion (e.g., where possible using the same categories for cases that failed and cases that successfully ended treatment, instead of two separate lists of categories). This resulted in a final product.

Results

Over 11 interviews, 369 text fragments were coded; 356 in five domains, 13 (3.5%) in the domain “other.” See Table 3 for the results. The five main domains were labeled as: (a) patient characteristics, (b) team characteristics, (c) treatment strategies, (d) team-patient interactions, and (e) social network. Within the domains, several categories were found. With three exceptions, all categories were applicable to both successful and failed cases. For example, having a social network in a successful case versus lacking such in a failed case were considered two sides of the same coin. Both were coded in the category “Social network.” Therefore, all categories, except “Splitting,” “Fear or Aversion,” and “Punishment,” contain both failed and successful cases. Following CQR methodology (Hill et al., 2005), the frequency labels of “General,” “Typical,” and “Variant” were assigned to all categories. General applied to categories that were found in all interviews, or all but one (10 or 11 interviews). Typical applied to categories found in more than half of the interviews (six to nine), and Variant to those found in two to five interviews. Below, we describe all categories per domain, starting out with the most representative category. To optimally convey its content, we provided illustrative interview quotations of most categories.

Domain 1: Patient Characteristics

As can be seen in Table 3, patient characteristics relevant to success or failure in treatment could be divided into characteristics associated with Facets of the PCL-R and characteristics unrelated to the PCL-R Facets. Facet 1 and 2, the Interpersonal and Affective Facets of the PCL-R, and certain characteristics outside the PCL-R were all mentioned quite frequently (i.e., considered General). Facet 3 (Impulsive) was mentioned in more than half of the interviews (Typical), while Lifestyle Facet 4 was only found in a handful of interviews (Variant).

Not Measured by the PCL-R (General)

This category referred to age, motivation, and comorbidity. Older psychopathic patients were thought to be more successful relative to younger ones, for two reasons: physical deterioration and having “calmed down” over the course of life. Admittedly, “older” is an ambiguous term, but when clinicians mentioned a specific number, it was over 40 or even 50 years of age. Second, some form of motivation was described as vital to treatment, even if merely extrinsic or exclusively based on a wish to be free from interference by the criminal justice system. See the following excerpt:

It appears that he came to a point in his life that he thought, “wait . . . I am now forty, I am stuck in the TBS, maybe I should try to make something of my life in a different way.” Not at all from any intrinsic motivation for treatment, but just simply based on the confrontation with a complete lack of perspective for the future otherwise, and the fact that he had somehow learnt to reflect a little bit on his life. (Treatment success, participant K)

Comorbidity was generally thought to be a drawback, except in the case of features of borderline personality disorder, as defined by the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013). Several participants felt that having some borderline traits in the emotional domain somehow “softened” the psychopathic manifestation and made it more susceptible to supportive interventions.

Related to Facet 1: Superficial Charm, Unreliability, Narcissism (General)

This is the Interpersonal Facet of the PCL-R, relating to superficial charm, unreliability, and narcissism. The degree to which patients were able to accept the authority of their treatment providers despite their grandiose sense of self was described as an important factor in success or failure. Second, the degree to which patients behaved in reliable ways was thought to be crucial. Some lack of reliability was considered normal in this population, but excessive deceit and fickleness during treatment impeded treatment, according to many informants. See the following excerpts:

The first thing that comes to mind is just the constant deceit, and conning, and externalizing, and saying yes while meaning no. Voicing good intentions and then letting them fail. You can’t tell anymore whether it is impulsivity or lack of frustration tolerance (which I do think are important) or just plain deception, agreeing to do something but secretly making his own plans. So, I think conning and manipulation are the most important factors. (Treatment failure, participant G)

Deception is the most important factor in failure. The TBS-system does not have an answer to that. Many other factors can be treated or managed, like impulsivity or lack of empathy, but we have not been able to find a creative way to deal with deception. (Treatment failure, participant K)

Related to Facet 2: Shallow Affect, Lack of Empathy, Failure to Take Responsibility (General)

Facet 2 is the Affective Facet of the PCL-R and is related to shallow affect, lack of empathy, and failure to take responsibility. Participants felt that the degree of callousness was an influential factor in treatment. Treatment providers reported that being able to work with a patient required some minimal emotional connection. If, however, the patient was deemed entirely instrumental or even sadistic, treatment often failed. The same counted for a complete failure to take responsibility for the offenses that led to the TBS. See the following excerpts:

He was someone who was not just indifferent with respect to other people’s interests, he actually enjoyed crossing boundaries and physically hurting people. (Treatment failure, participant K)

He could be humorous, entertaining. And now and then he would show a glimpse of empathy, for example admitting that he was making life very difficult for us. (Treatment success, participant J)

Related to Facet 3: Impulsivity, Thrill Seeking, Irresponsible Lifestyle (Typical)

The items of Facet 3 are related to impulsivity, thrill seeking, and an irresponsible, parasitic lifestyle. Again, the degree to which these issues were characteristic for the patient was reported as critical to treatment outcome. Reportedly, some degree of impulsivity can be managed, but within limits. See the following excerpts:

He did have some impulsivity, but it was just within limits. He didn’t overdo it. [. . .] Otherwise it would have been a very unfavorable factor. (Treatment success, participant E)

It was only temporarily that he managed to stick to the program. As long as he was kept inside the clinic, for a while it went okay. But as soon as he was given more freedom, his behavior tended to deteriorate. (Treatment failure, participant A)

Related to Facet 4: Aggressive and Antisocial Behavior (Variant)

Facet 4 is related to aggressive and antisocial behavior. Apparently, this is considered less influential with regard to the course of treatment, as the representativeness of this category is Variant. Early onset of problematic behavior was named as a risk factor for failure of treatment, as well as extreme aggression while admitted to the TBS hospital:

In my view, what was important is that he was really a very aggressive man. He was also very aggressive inside the clinic. Not all patients have this, or not to that degree. (Treatment failure, participant H)

Domain 2: Team Characteristics

Within this domain, four categories were coded. The first two were labeled Typical, while the third and the fourth were Variant.

Level of Expertise (Typical)

A majority of informants emphasized that teams working with psychopathic patients need to have specialized training. They need to have sufficient knowledge about psychopathy, and the requisite skills, patience, and experience to deal with difficult situations. See the following excerpt:

A team needs to be prepared for unexpected situations. With these patients, you just know that they will occur, and you need to be prepared. They shouldn’t surprise you. So, it is essential that a team is knowledgeable about psychopathy. (Treatment success, participant K)

Sufficient Emotional Distance (Typical)

Likewise, a majority of informants indicated that the ability of team members to see the behavior of the patient as the patient’s problem or handicap, and not to take it personally, was important. See the following excerpt:

[It is important that] you don’t have too many expectations, and that you can really think, “OK, this is part of the job. [. . .] We knew he was unreliable, but he doesn’t do it to spite us, it’s just how he is with everyone. Don’t take it personally.” You try not to mix your own moral values with your work, so to speak. It’s quite difficult, and you need your team to help you with this. (Treatment success, participant J)

Consensus About Diagnosis and Treatment (Variant)

Participants reported that disagreement in a team about the goals of treatment and how to interpret the patient’s behavior often led to failure, while consensus on the treatment strategy appeared to be a requirement for success. The following excerpt illustrates this:

The team did not really agree on limit setting. [. . .] For example, when he had used drugs, part of the team felt that his permission for leave outside the clinic needed to be restricted for a long time, while others were of the opinion that this wasn’t really necessary. Of course, in such situations a decision was always reached, but you could sense that underneath, the team was divided. (Treatment failure, participant B)

Continuity in Treatment Team (Variant)

Participants reported that too much staff turnaround in teams leads to loss of salient experience in the team and to disruption of the treatment process for the patient. See the following excerpt:

New team members generally have a need to work with clear rules. On the other hand, with psychopathic patients, you need to be able to play a little bit, to know when to be flexible and when to set clear limits to certain behavior. In my opinion, you just have to be very experienced to be able to do this well. The constant turnover of young and inexperienced men and women in our clinic makes this impossible. (Treatment failure, participant D)

Domain 3: Treatment Strategies

Underlying Treatment Philosophy (General)

Together with three of the five categories in the domain patient characteristics, this was the only other General (as in almost universally reported) category. Informants reported several elements that they felt were crucial basic treatment principles. The first was creating a clear perspective or goal for the patient. See the following excerpts:

[At first] it was completely unclear what the route towards rehabilitation should be. And he had no idea what his own goal was. So he was sort of at the mercy of his fellow patients in the group. [. . .] But then at one point, it became clear in which residence he could live if only he would commit himself to treatment and gain more liberties. At that point he realized, “I would really like that.” (Treatment success, participant D)

The thing is, we didn’t really have a good plan [. . .] for the route towards rehabilitation. We just tried some things. And when that didn’t go too badly, we said, “OK, let’s try some more,” but we never really had a clear goal for this patient. [. . .] We didn’t discuss this openly with him. [. . .] We gradually and very carefully extended his possibilities for leave outside the clinic, but he didn’t understand why we went so slowly. This was an explosive mix. (Treatment failure, participant K)

Informants also felt that a relatively limited inpatient phase worked best, whereas there should be an extended phase outside the clinic, with gradual, stepwise exposure to situations associated with risk for recidivism. In addition, participants reported that encouraging patients to take responsibility for their own actions, being clear about expectations but allowing some room for mistakes often worked well. See the following excerpt:

He did have some relapses into drug abuse, but the deal was that if he reported these voluntarily, he would not be readmitted to the clinic, unless it happened very often. Beforehand, he had also been asked to think about what kind of arrangement would help him to stay clean permanently. So, he was emphatically invited to take his own responsibility. And he was also given a lot of autonomy. I think that was good for him, that we didn’t take all responsibility away from him, but on the contrary, stimulated him to think for himself. (Treatment success, participant A)

Reaction to Treatment Interfering Behavior (Typical)

Treatment interfering behavior was considered part and parcel to the treatment of patients with psychopathy. When this category was referenced, participants were quite unanimous in how best to deal with it: by immediate and predictable limit setting, and if necessary (for security reasons) by imposing a short period of restriction of freedom, followed by resumption of the treatment program. See the following excerpts:

So, the moment [. . .] he does not stick to the rules, a direct consequence should follow. A short intervention. And then treatment continues. And I think in this case, with this patient, this happened a lot. At some point, you get a kind of behavioral training. He started to understand the right thing to do, or at least how to stay on course and not get in trouble all the time. (Treatment success, participant I)

[The new team] worked with short and relatively predictable interventions. [. . .] And also with a proportionate response to what the patient did. [. . .] For example, when he had done something wrong on his mobile phone. In that case he wasn’t forced to hand in the entire phone, but he had to remove the app involved. While before, with the other team, they retained his mobile phone for an indefinite period of time. (Treatment success, participant C)

Specific Interventions (Typical)

Participants typically reported various kinds of interventions that had been helpful in specific cases, but no clear picture emerged. This was due to the relatively wide range of interventions and due to the fact that some interventions appeared to have equivocal results. For example, Eye Movement Desensitization and Reprocessing (EMDR) was named as a factor in a successful as well as an unsuccessful case.

Domain 4: Team–Patient Interactions

All five categories in this domain were of Variant frequency.

Working Relationship (Variant)

This category refers to the situation when, irrespective of the shallow affect and lack of commitment of the patient, to some modest extent, a working relationship still developed with the staff. Interviewees described that some psychopathic patients inspire a certain amount of sympathy; for example, because of a traumatic life history. Possibly patients sensed this sympathy and felt less defensive; for treatment providers, it can help in enduring treatment interfering or other problematic behavior. When all sympathy is lacking, treatment is more likely to fail, according to the participants. See the following excerpt:

I felt there was some degree of connection between us. Of course, with such a high score on the PCL-R you know the connection probably wasn’t very strong for him. He did deceive us a lot. But I think he also genuinely felt our empathy for him. (Treatment success, participant G)

Development of Trust (Variant)

Trustworthiness is uncommon in psychopathic patients. Vice versa, people with psychopathy are not very likely to trust their therapists. Participants reported that nevertheless, with some patients, a certain amount of mutual trust developed in the course of treatment, or at least a kind of “reciprocal predictability.” On the contrary, in some of the described failed cases, mutual distrust did not abate at all, and the behavior of the patient remained unpredictable for the team. See the following excerpt:

This was a patient with a history of extremely violent behavior, both outside and inside the clinic. Also, he had managed to have a secret intimate relationship with a staff member. [. . .] At the same time, he had periods where he seemed calm and in control. But we never felt that we could predict his behavior, there was always a feeling of tension and distrust. Ultimately, this made it impossible to make progress in treatment. (Treatment failure, participant H)

Splitting (Variant)

This category refers to the situation in which a patient, for example, through charm or manipulation, created a division in the staff. Members of the team disagreed, or even argued about the patient and what would be the optimal treatment planning, and could not overcome their differences of opinion in an effective manner. This was a precursor to several failures of treatment. See the following excerpt:

He allowed some members of the team to be closer to him, while he kept other team members more at a distance. The result was that those whom he appeared to trust more, were more positive about the treatment effects and more willing to make progress, while others remained more suspicious of him. We did not manage to solve this. (Treatment failure, participant B)

Fear or Aversion (Variant)

Sometimes patients were so hostile or dangerous even within the clinic (i.e., causing aggressive incidents), or had committed such horrific crimes, that team members experienced fear and/or aversion. See the following excerpt with an example of aversion:

This patient was part of a group that met in some foreign country to participate in extreme sexual abuse of children. The more you paid, the more you were allowed to do. It was really very horrible to read his file. And also, to listen to him minimizing the impact of his behavior. [. . .] It was very difficult to see this offender as a patient, and to contain the negative countertransference. (Treatment failure, participant G)

Punishment (Variant)

In some cases, teams did not find a way to tolerate and manage problem behavior (such as breaking institutional rules or violating conditions for leave outside the hospital; for example, using drugs, smuggling a smart phone into the facility), leading to ineffective punishment of the patient instead of effective limit setting (e.g., restrictions in the freedom of movement in or outside the clinic, which the respondents felt were disproportionate). See the following excerpt:

When he broke the rules, I felt the staff reacted with anger, punitively, as if they felt betrayed. And they tried to appeal to his conscience, I think. But it didn’t work out that way. What happened was that he was more and more convinced that he could never do it right. (Treatment failure, participant B)

Domain 5: Social Network

Both categories in this domain had a frequency of Typical.

Prosocial Network (Typical)

A majority of participants reported that simply having a social network that was not overly antisocial was a positive factor, for example, through offering support and improving motivation. Conversely, a complete absence of social contacts, or mainly antisocial connections, did not bode well. See the following excerpt:

He reconnected with his old antisocial friends, but they now all had families of their own. They had become more normal. They were still involved in, for example, small scale, illegal trade, but that was it. No more violence. (Treatment success, participant H)

Network Support for Treatment (Typical)

Both active support of and opposition to the TBS order and the treatment were important factors. See the following excerpts:

[The family felt that] the TBS order was completely unjustified. Nothing wrong with him. They thought we were ridiculous. What did we think we were doing? (Treatment failure, participant F)

His social network was very important. His mother, for example, told us he was a different man when he used alcohol. She made sure that during family visits, no alcohol was available to anyone. The patient himself thought he could manage a beer now and then, but because of his mother’s attitude he never tried. This was very important in his treatment. (Treatment success, participant B)

Discussion

In this study, we used a qualitative research design to explore with expert clinicians factors that in their view influenced successful completion versus dropout in the forensic psychiatric treatment of highly psychopathic patients. Individual patient characteristics as well as characteristics of the treatment or team were sought. A variety of factors emerged. First, a number of general forensic treatment-impeding factors were mentioned as equally relevant for patients with severe psychopathic traits: age, motivation, and patients’ network. While it may seem unsurprising that factors relevant for forensic treatment in general also appear relevant in the treatment of psychopathic patients, some of these observations are still notably at odds with general beliefs about psychopathy. A good example concerns the age of the justice-involved individual. In general, it is believed that men high on psychopathic traits start their offending behavior at a young age and, in contrast to “ordinary” antisocial individuals, do not easily quiet down. However, according to our informants, there still appears to be an age at which they do quiet down. In similar vein, the issue of a minimal amount of some kind of motivation is remarkable. It was repeatedly emphasized that it did not matter how “thin” or self-centered the motivation might be; any and all forms of motivation were deemed helpful and some minimum essential. In other words, patients with psychopathy don’t have to do it for the right reasons, as long as they can find any reason. Another interesting finding is that the network, in the eyes of the interviewees, does not have to be exemplary prosocial, just not too antisocial. A little rule violation is not a problem per se, as long as there are clear limits within the network as to the unacceptability of violence. We know of no studies that examine the role of involving family and peers in the treatment of adult psychopathic patients specifically.

Several factors were related to the treatment and the team. Relevant treatment characteristics for patients with psychopathy mostly referred to an underlying philosophy or framework. This framework included (a) matching clear goals and perspectives for the patient with clear and concrete expectations from the patient and (b) keeping the inpatient treatment relatively short, followed by a prolonged and gradual outpatient trajectory. Other important suggestions our interviewees made were tolerating some rule violation, choosing your battles, and applying clear, predictable punishment of short duration, where necessary. This seems applicable to the treatment of all forensic patients, but likely needs more emphasis for psychopathic patients, who are of course among the hardest to treat within this group.

Team characteristics were regularly mentioned as relevant. However, most of the characteristics are relevant factors for any treatment, such as consensus on diagnosis and treatment trajectory. A team characteristic related to psychopathy was expertise on psychopathy in general and, more specific, the ability to keep sufficient emotional distance. It appears that a team that is well informed on the nature and characteristics of psychopathy is therefore better able to view psychopathy as a deficiency and not feel personally betrayed by the failures of their patients. Continuity of treatment providers was also deemed important. This finding seems somewhat counterintuitive. Why would this matter for people with psychopathy who are believed to be notoriously uninterested in and incapable of attachment? The overall findings of this study seem to indicate that the prospects of successful treatment for patients with psychopathy are better if any relationship, no matter how unstable, pragmatic/opportunistic, or superficial, can still grow. However, it seems that these relationships with highly psychopathic patients (a) take long to develop and (b) easily get damaged. Discontinuity in a team is thought to be one of the disturbing factors.

A number of factors more specific to psychopathy emerged from the study as relevant. Patient characteristics associated with Facets 1 and 2 of the PCL-R were found in (nearly) all interviews; those associated with Facet 3 in more than half. Facet 4 was considered less influential. This is in line with three of the four studies cited in the introduction (Jeandarme et al., 2017; Olver & Wong, 2011; Sewall & Olver, 2019). Of note, for all Facets, the informants reported that it was the degree to which a certain Facet was present that counted. That is, some degree of deception, arrogance, callousness, impulsivity, and aggression were all to be expected behaviors of these patients during treatment, but only very high levels were thought to be seriously treatment interfering.

The problems arising from overly pronounced Facet 1 or Facet 2 traits were described in two variations: either no relationship or form of cooperation could develop at all (emphasis Facet 2), or a relationship appears to develop, but is destroyed by dishonest and untrustworthy behavior (emphasis Facet 1). It remains unclear whether this dishonest behavior should be seen as planned manipulation or that psychopathic patients are fundamentally unable to acknowledge their faults and simply declare their own desired truth, which exists separate from reality. Very high traits on Facet 3, and very high traits versus more moderate traits on Facet 4 were mentioned to a lesser extent and mostly in the form of excessive impulsivity or aggression. Most of the Facet 4 items (Poor Behavioral Controls, Early Behavioral Problems, Juvenile Delinquency, Revocation of Conditional Release, Criminal Versatility) refer to historical facts and behavior and to a far lesser extent to personality traits. Therefore, it is not surprising that Facet 4 is less often nominated as a treatment-interfering patient characteristic. Nevertheless, Facet 4 was mentioned as relevant in a few cases, specifically regarding patients that caused extremely serious violent incidents within the walls of the facility. Serious violent incidents within the facility caused by psychopathic patients are relatively rare in our experience, but cause strong backlash, and are often the prelude to transfer, according to our informants.

The recurring suggestion that comorbid borderline personality traits may be favorable for treatability of patients with psychopathy is intriguing. How may comorbid borderline traits make psychopathic patients more susceptible to treatment? Stipulating recommendations for further research on psychopathy, Lilienfeld (2018) argued that it is important to incorporate measures of other disorders, especially antisocial, histrionic, borderline, and narcissistic personality disorders, also known as the cluster B personality disorders. He noted that although comorbidity generally is considered to increase impairment, there may be notable exceptions. He cited two studies (Short et al., 2016; Walker et al., 1991) that found that in children and adolescents with conduct disorder, comorbid symptoms of depression or anxiety tended to be associated with better outcome. Lilienfeld hypothesized that the presence of an internalizing disorder may attenuate risk. A similar mechanism may hold for borderline traits comorbid with psychopathy. One might conjecture that the presence of emotional borderline features is incompatible with extreme emotional callousness—a factor that was frequently mentioned as a risk factor for treatment failure. We know of no studies that examined the effect of borderline traits on the treatment of patients with psychopathy, and further research seems warranted.

This study has several limitations. First, we asked clinicians to choose two specific cases, in the expectation that illustrative examples would allow for in-depth discussion. This requirement also ensured that truly psychopathic patients were chosen. However, it is possible that this approach led to clinicians choosing particularly memorable (instead of representative) cases. Another strategy would have been to frame the questions in a more general way, focusing on psychopathic patients in general. At the end of the interview, we did provide the clinicians the opportunity to add general comments (see Table 1; “According to you, are there other factors not yet discussed on the basis of these two cases, which may be essential to take into account when treating psychopathic patients?”). Second, we did not specify in great detail how to prepare for the case interviews. It became clear during the interviews that some participants had taken the time to study the files, while others appeared to rely predominantly on their memory. It is not immediately clear how this variation may have affected the findings. It bears mentioning that we consistently requested informants to bolster their comments by providing concrete, specific examples to best illustrate their answers. Also, the argument could be made that the most salient factors would have remained in long-term memory anyway. Third, this study focused exclusively on factors related to treatment completion in highly psychopathic patients. While limiting dropout is an important goal in developing effective treatment for these patients, this study did not focus on the effects of treatment on reducing recidivism. In addition, our research design precludes determination of the extent to which these factors of treatment success (or failure) are specific to patients with psychopathy. When possible however, we provided some informed conjectures on this issue. Fourth, no specific interventions for reducing the responsivity problems in psychopathy were consistently identified. Respondents sometimes mentioned a certain form of treatment that was effective (or not) in a specific case, but no general pattern emerged. At this stage, it seems advisable to try various interventions and to not give up too easily. Even if no specialized interventions have yet been developed for patients with psychopathy in general, some may still be effective in particular cases. Much more research is needed in this domain. Fifth, for the purpose of homogeneity, all but one of the interviewees were working in the same hospital, albeit in different units. Therefore, the results of this study may be vulnerable to embedded culture effects of this particular setting. Future studies should widen the scope, either by replicating the current study in another hospital or by including more than one institute in the research design. A final limitation is that this study is limited to the perspective of the treatment providers. As of yet, there are no data on how patients reflect on their experiences in forensic treatment. We argue this remains a worthwhile empirical question. Eliciting the patients’ perspective may well be a profitable avenue for the further development of specialized treatment programs.

Notwithstanding these limitations, tentative conclusions regarding the ideas of experienced treatment providers about the treatability of forensic patients with psychopathy can be drawn. Overall, the results imply that patients with psychopathy may be retained in treatment successfully lest they are not exceedingly psychopathic, especially with regard to arrogant and deceitful interpersonal style (Facet 1) and defective affective experience (Facet 2). However, not only specific psychopathic traits are thought to be important, other patient characteristics also appear influential. Older patients and those with a somewhat prosocial and supportive network may fare better. If some form of a working relationship can develop, no matter how shallow, and some form of motivation can be found, no matter how self-centered, treatment of psychopathic patients can proceed successfully. Future research into the treatability of patients with psychopathy should therefore take into account other patient characteristics than psychopathic traits only.

Of note, our study also indicates that certain characteristics of the treatment framework were deemed highly relevant for treatment success or failure. A treatment program that stipulates clear and concrete goals and expectations; provides a long and gradual resocialization trajectory; and is offered by a knowledgeable, cohesive, and stable team appears to be more successful. This means that forensic treatment as usual may need to be adapted for patients with psychopathy. For example, not all forensic patients need the extensive outpatient phase of treatment that is recommended in our study. In addition, teams working with psychopathic patients probably need extra training, not only to limit the possible detrimental effects of working with these patients, but also to enhance their ability to retain these patients in treatment in the first place. All these suggestions of course need to be validated in future (quantitative) studies. At this point, developing a treatment framework using some of these suggestions appears to be a priority.

In this respect, it is intriguing to recall that 25 years ago, after many years of clinical pessimism about the prognosis, dialectical behavior therapy (DBT; Linehan, 1993) was the first evidence-based psychotherapy for borderline personality disorder. One of the explicit goals of DBT was limiting attrition, and the first clinical trial showed significantly less dropout when compared to treatment as usual, 16.7% versus 50% (Linehan et al., 1991). In her book, Linehan first devoted no less than 200 pages to describing the theoretical framework and the specification of the parameters for treatment, including integration of supervision and consultation for therapists, before introducing any concrete therapeutic techniques for patients. As the treatment of psychopathic patients is without a doubt just as challenging for treatment providers, developing such a manual would be of great value for this clinical condition as well. Some of the topics emerging from this study may eventually be essential reading in the first chapters.

Footnotes

Authors’ Note:

We have no conflicts of interests to disclose.

ORCID iD

Evelyn Klein Haneveld

Evelyn Klein Haneveld, MSc, is a licensed clinical psychologist and psychotherapist. She is director of treatment at FPC Oostvaarderskliniek, a high secure forensic psychiatric hospital in the Netherlands, and also serves as counsel in the Penitentiary Chamber of the Court of Appeal of Arnhem-Leeuwarden. She is working on a thesis which focuses on individual differences in the treatability of patients with psychopathy.

Wineke Smid, PhD, is head of the research department at forensic psychiatric hospital Van der Hoevenkliniek in Utrecht, the Netherlands. She divides her time between conducting and supervising empirical research projects, carrying out individual (risk) assessments of residential high-risk forensic psychiatric patients, and providing feedback of state-of-the-art knowledge to various stakeholders in Dutch society, such as police officers, prosecutors, judges, treatment providers, and probation officers.

Kelsey Timmer, MSc, studied clinical psychology at the University of Amsterdam (UVA) in the Netherlands. She worked as a sociotherapist in forensic psychiatry, while completing her Master’s degree. She currently works as a psychologist at the outpatient clinic of GGZ inGeest, a mental health facility.

Jan H. Kamphuis, PhD, is a full professor at the University of Amsterdam (UvA), serving as the director of its Clinical Psychology program, and also has a research/clinical appointment at the Viersprong, a specialized clinic for the assessment and treatment of personality disorders in the Netherlands. His research focuses on promoting the treatment utility of clinical personality assessment, with special interests in personality disorders and forensic applications.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). American Psychiatric Publishing.

Burkard

A. W.

Knox

Hill

C. E.

(2012). Data collection. In Hill

C. E.

(Ed.), Consensual qualitative research: A practical resource for investigating social science phenomena (pp. 83–101). American Psychological Association.

Cullen

A. E.

Soria

Clarke

A. Y.

Dean

Fahy

(2011). Factors predicting dropout from the reasoning and rehabilitation program with mentally disordered offenders. Criminal Justice and Behavior, 38(3), 217–230. https://doi.org/10.1177/0093854810393659

Hare

R. D.

(1991). Hare Psychopathy Checklist-Revised. Multi-Health Systems.

Hare

R. D.

(2003). Hare Psychopathy Checklist-Revised (PCL-R) (2nd ed.). Multi-Health Systems.

Hare

R. D.

Neumann

C. S.

(2008). Psychopathy as a clinical and empirical construct. Annual Review of Clinical Psychology, 4, 217–246. https://doi.org/10.1146/annurev.clinpsy.3.022806.091452

Hart

S. D.

Cox

D. N.

Hare

R. D.

(1995). Manual for the Hare Psychopathy Checklist: Screening Version (PCL: SV). Multi-Health Systems.

Hecht

L. K.

Latzman

R. D.

Lilienfeld

S. O.

(2018). The psychological treatment of psychopathy: Theory and research. In David

Lynn

S. J.

Montgomery

G. H.

(Eds.). The state of the science and practice (pp. 271–298). Wiley-Blackwell.

Hildebrand

de Ruiter

de Vogel

van der Wolf

(2002). Reliability and factor structure of the Dutch language version of Hare’s Psychopathy Checklist-Revised. International Journal of Forensic Mental Health, 1(2), 139–154. https://doi.org/10.1080/14999013.2002.10471169

10.

Hill

C. E.

(Ed.). (2012). Consensual Qualitative Research: A practical resource for investigating social science phenomena. American Psychological Association.

11.

Hill

C. E.

Knox

Thompson

B. J.

Williams

E. N.

Hess

S. A.

Ladany

(2005). Consensual Qualitative Research: An update. Journal of Counseling Psychology, 52(2), 196–205. https://doi.org/10.1037/0022-0167.52.2.196

12.

Hill

C. E.

Thompson

B. J.

Williams

E. N.

(1997). A guide to conducting Consensual Qualitative Research. The Counseling Psychologist, 25(4), 517–572. https://doi.org/10.1177/0011000097254001

13.

Jeandarme

Pouls

Oei

T. I.

Bogaerts

(2017). Forensic psychiatric patients with comorbid psychopathy: Double trouble? International Journal of Forensic Mental Health, 16(2), 149–160. https://doi.org/10.1080/14999013.2017.1286414

14.

Klein Haneveld

Neumann

C. S.

Smid

Wever

Kamphuis

J. H

. (2018). Treatment responsiveness of replicated psychopathy profiles. Law and Human Behavior, 42(5), 484–495. https://doi.org/10.1037/lhb0000305

15.

Kröger

van Beek

van der Wolf

Klein Haneveld

van Geest

Geraerts

(2014). Treatment of psychopathy: A mission impossible? Forum Educatief.

16.

Ladany

Thompson

B. J.

Hill

C. E.

(2012). Cross-analysis. In Hill

C. E.

(Ed.), Consensual Qualitative Research: A practical resource for investigating social science phenomena (pp. 117–134). American Psychological Association.

17.

Lilienfeld

S. O.

(2018). The multidimensional nature of psychopathy: Five recommendations for research. Journal of Psychopathology and Behavioral Assessment, 40(1), 79–85. https://doi.org/10/1007/s10862-018-9657-7

18.

Linehan

M. M.

(1993). Cognitive-behavioral treatment of borderline personality disorder. The Guilford Press.

19.

Linehan

M. M.

Armstrong

H. E.

Suarez

Allmon

Heard

H. L.

(1991). Cognitive-behavioral treatment of chronically parasuicidal borderline patients. Archives of General Psychiatry, 48, 1060–1064.

20.

MAXQDA. (2018). VERBI Software GmbH [Version 2018]. https://www.maxqda.com

21.

Mokros

Hollerbach

Vohs

Nitschke

Eher

Habermeyer

(2013). Normative data for the Psychopathy Checklist–Revised in German-speaking countries: A meta-analysis. Criminal Justice and Behavior, 40(12), 1397–1412. https://doi.org/10.1177/0093854813492519

22.

Neumann

C. S.

Hare

R. D.

Pardini

D. A.

(2015). Antisociality and the construct of psychopathy: Data from across the globe. Journal of Personality, 83(6), 678–692. https://doi.org/10.1111/jopy.12127

23.

Ogloff

J. R.

Wong

Greenwood

(1990). Treating criminal psychopaths in a therapeutic community program. Behavioral Sciences & The Law, 8(2), 181–190. https://doi.org/10.1002/bsl.2370080210

24.

Olver

M. E.

Wong

S. C.

(2009). Therapeutic responses of psychopathic sexual offenders: Treatment attrition, therapeutic change, and long-term recidivism. Journal of Consulting and Clinical Psychology, 77(2), 328–336. https://doi.org/10.1037/a0015001

25.

Olver

M. E.

Wong

S. C.

(2011). Predictors of sex offender treatment dropout: Psychopathy, sex offender risk, and responsivity implications. Psychology, Crime & Law, 17, 457–471. https://doi.org/10.1080/10683160903318876

26.

Polaschek

D. L. L.

(2014). Adult criminals with psychopathy: Common beliefs about treatability and change have little empirical support. Current Directions in Psychological Science, 23(4), 296–301. https://doi.org/10.1177/0963721414535211

27.

Polaschek

D. L. L.

Skeem

J. L.

(2018). Treatment of adults and juveniles with psychopathy. In Patrick

(Ed.), Handbook of psychopathy (2nd ed., pp. 710–731). Guilford Press.

28.

Reidy

D. E.

Kearns

M. C.

DeGue

(2013). Reducing psychopathic violence: A review of the treatment literature. Aggression and Violent Behavior, 18(5), 527–538. https://doi.org/10.1016/j.avb.2013.07.008

29.

Sewall

L. A.

Olver

M. E.

(2019). Psychopathy and treatment outcome: Results from a sexual violence reduction program. Personality Disorders: Theory, Research, and Treatment, 10(1), 59–69. https://doi.org/10.1037/per0000297

30.

Short

R. M.

Sonuga-Barke

E. J.

Adams

W. J.

Fairchild

(2016). Does comorbid anxiety counteract emotion recognition deficits in conduct disorder? Journal of Child Psychology and Psychiatry, 57(8), 917–926. https://doi.org/10.1111/jcpp.12544

31.

Sim

Huang

T. C.

Hill

C. E.

(2012). Biases and expectations. In Hill

C. E.

(Ed.), Consensual Qualitative Research: A practical resource for investigating social science phenomena (pp. 59–69). American Psychological Association.

32.

Tbs inzichtelijk [Understanding Tbs]. (2019, April). Forensische zorg. https://www.forensischezorg.nl/files/infographic_tbs_april_2019.pdf

33.

Thompson

B. J.

Vivino

B. L.

Hill

C. E.

(2012). Coding the data: Domains and core ideas. In Hill

C. E.

(Ed.), Consensual Qualitative Research: A practical resource for investigating social science phenomena (pp. 103–116). American Psychological Association.

34.

Vertommen

Verheul

de Ruiter

Hildebrand

(2002). Hare’s Psychopathie Checklist: Handleiding [Hare Psychopathy Checklist-Revised]. Swets and Zeitlinger.

35.

Walker

J. L.

Lahey

B. B.

Russo

M. F.

Frick

P. J.

Christ

M. A. G.

McBurnett

Loeber

Stouthamer-Loeber

Green

S. M.

(1991). Anxiety, inhibition, and conduct disorder in children: Relations to social impairment. Journal of the American Academy of Child and Adolescent Psychiatry, 30(2), 187–191.

36.

Wong

S. C. P.

Hare

R. D.

(2005). Guidelines for a psychopathy treatment program. Multi-Health Systems.