Psychotherapy 2015, Vol. 52, No. 2, 228 –237

© 2014 American Psychological Association 0033-3204/15/$12.00 http://dx.doi.org/10.1037/a0037919

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Does the Severity of Psychopathological Symptoms Mediate the Relationship Between Patient Personality and Therapist Response? Vittorio Lingiardi and Annalisa Tanzilli

Antonello Colli

Sapienza University of Rome, Italy

“Carlo Bo” University of Urbino, Italy

Countertransference can be viewed as a source of valuable diagnostic and therapeutic information and plays a crucial role in psychotherapy process and outcome. Some empirical researches have showed that patients’ specific personality characteristics tend to evoke distinct patterns of emotional response in clinicians. However, to date there have been no studies examining the impact of patients’ symptomatology on the association between their personality and therapists’ responses. This research aimed to (a) investigate the relationship between patients’ symptom severity and clinicians’ emotional responses; and (b) explore the possible mediated effect of symptom severity on the relationship between patients’ personality pathology and countertransference responses. A sample of psychiatrists and clinical psychologists (N ⫽ 198) of different theoretical orientations completed the Shedler–Westen Assessment Procedure-200 and the Therapist Response Questionnaire on a patient currently in their care, who then completed the Symptom Checklist-90-Revised. The findings showed that patients’ symptomatology partially mediates the relationship between their specific personality disorders (in particular, schizotypal, borderline, histrionic, and avoidant) and therapists’ emotional responses, but in general, the impact of symptom severity is less sizable than one aroused by patients’ personality style. Higher levels of patients’ symptom severity are most associated with an intense feeling of being overwhelmed, disorganization, helplessness, and frustration in clinicians. These countertransference reactions are not accounted for by therapists’ different therapeutic approaches and other variables (as gender, age, profession, and experience). The clinical implications of these results are addressed. Keywords: therapist emotional response, personality disorders, symptom severity, SWAP-200, TRQ

unconscious reaction of the analyst to the patient’s transference. This approach, close to the use of the construct as first introduced by Freud (1910), tends to conceptualize it as a result of the analyst’s neurotic conflicts and an obstacle to the psychotherapy process that must be overcome (Freud, 1912). In the totalistic perspective, countertransference is considered as the totality of what the therapist experiences and feels together with the patient (Heimann, 1950). According to this broader view, that advocates a more active technical use of it in the analytic situation, a large part of the analyst’s reactions (conscious and unconscious, emotional and cognitive, intrapsychic and behavioral) reflects the patient’s interpersonal modes and personality. This conceptualization implies that, although the countertransference should certainly be resolved, it is useful in gaining more understanding of the patient core central dynamics (see also Gabbard, 1995, 2001; Gelso & Hayes, 2007; Muran & Barber, 2010). Some psychodynamic concepts, such as complementary and concordant countertransference (Racker, 1957), role responsiveness (Sandler, 1976), projective identification (Ogden, 1982), as well as cognitive constructs, such as the cognitive interpersonal cycle (Safran, 1984), and dynamic system terms, such as dyadic interaction patterns (Burgoon, Stern, & Dillman, 1995), share the idea that the therapist’s inner experience partially derives from the patient’s relational schemas and, as a consequence, can be used to understand his/her core problems. From this point of view, careful consideration of clinician’s responses to the patient’s personality and interpersonal functioning has critical relevance to tailoring and

A therapist’s emotional response to a patient can be viewed as a source of valuable diagnostic and therapeutic information, as well as plays a crucial role in the psychotherapy process and outcome across different forms of psychopathology and various therapeutic approaches (Hayes, Gelso, & Hummel, 2011). Despite its psychoanalytic origins, countertransference has been discussed in the literature from a wide range of theoretical and clinical perspectives (Fauth, 2006; Gelso & Hayes, 2007). Moreover, empirical research suggests that clinicians of all theoretical orientations should attend to and, when possible, make use of information provided in the context of the therapeutic relationship (Norcross, 2011). As Kernberg observed (1965), two main contrasting approaches in regard to the concept of countertransference could be considered: the classical and the totalistic one. From the classical and “narrow” point of view, countertransference is defined as the

This article was published Online First November 10, 2014. Vittorio Lingiardi and Annalisa Tanzilli, Department of Dynamic and Clinical Psychology, Faculty of Medicine and Psychology, Sapienza University of Rome; Antonello Colli, Department of Human Science, University “Carlo Bo” of Urbino. The authors thank all the clinicians who contributed their data to this research. Correspondence concerning this article should be addressed to Vittorio Lingiardi, Department of Dynamic and Clinical Psychology, Faculty of Medicine and Psychology, Sapienza University of Rome, Via degli Apuli, 1, 00185, Rome, Italy. E-mail: [email protected] 228

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

SYMPTOMS, PATIENT PERSONALITY, THERAPIST RESPONSE

managing the diagnostic and therapeutic process, especially with personality-disordered patients (Bateman & Fonagy, 2006; Clarkin, Yeomans, & Kernberg, 2006; Gabbard, 2009; McWilliams, 2011; OPD Task Force, 2008; PDM Task Force, 2006). Some empirical studies have investigated the relationship between therapists’ emotional experiences and patients’ personality diagnoses at the DSM–IV cluster level (Betan, Heim, Zittel Conklin, & Westen, 2005; Røssberg, Karterud, Pedersen, & Friis, 2007). Limited research has sought to explore clinicians’ responses elicited by specific personality disorders, particularly borderline (Brody & Farber, 1996; Colli, Tanzilli, Dimaggio, & Lingiardi, 2014; McIntyre & Schwartz, 1998; Schwartz, Smith, & Chopko, 2007). Betan et al. (2005) identified eight patterns of emotional response on a sample of 181 clinicians that completed the Therapist Response Questionnaire (TRQ; Zittel Conklin & Westen, 2003) on a randomly selected patient in their care. The countertransference factors were: (a) overwhelmed/disorganized, (b) helpless/inadequate, (c) positive, (d) special/overinvolved, (e) sexualized, (f) disengaged, (g) parental/protective, and (h) criticized/mistreated. These dimensions were clinically and conceptually coherent, as well as independent of clinicians’ theoretical orientation. In fact, to rule out the hypothesis that the underlying factor structure of TRQ reflected the theoretical beliefs of participating clinicians, the authors conducted a second factor analysis, removing from the complete sample all clinicians who reported a psychoanalytic or psychodynamic approach. Using the same rotation and estimation procedures, the factor structure was reproduced, suggesting that this finding was not an artifact of therapists’ preconceptions. The authors assessed patients’ personality pathology asking clinicians to rate the presence or absence of each criterion of all the DSM–IV axis II diagnoses. This procedure furnished both a categorical diagnosis of each disorder (obtained by applying DSM–IV cutoffs) and a dimensional measure of DSM personality disorders (number of criteria met for each disorder) that permitted to run the correlations between therapist responses and all the DSM–IV axis II clusters. The results showed that: Cluster A (paranoid, schizoid, and schizotypal personality disorders) was associated with criticized/mistreated dimension; cluster B (antisocial, borderline, hystrionic, and narcissistic personality disorders) correlated to overwhelmed/disorganized feelings, helplessness, hostility, disengagement, and sexual attraction; and cluster C (avoidant, dependent, and obsessive– compulsive personality disorders) was associated with therapists’ nurturant and warm feelings. Consistent with these findings, Røssberg et al. (2007) compared the clinicians’ emotional reactions (measured with Feeling Word Checklist-58; Røssberg, Hoffart, & Friis, 2003) toward patients with DSM–IV Axis II cluster A, B, and C. The patients were diagnosed first with Structured Clinical Interview for DSM–IV Axis II Personality Disorders (SCID-II; First, Gibbon, Spitzer, Williams, & Benjamin, 1997). The proposed diagnoses were then discussed at a staff conference and compared with other information, such as referral letters, patients’ own written narratives, and several evaluation interviews. Using SCID II diagnoses as categorical variables, the authors examined to what extent patients with cluster A and B personality disorders evoked distinct countertransference reactions than patients with cluster C among psychotherapists. The study revealed that patients with cluster A and B elicited more negative countertransference reactions than those with cluster C. Finally, another study found that the fewer overall DSM–IV per-

229

sonality disorder criteria, the more confident was the countertransference (Dahl, Røssberg, Bøgwald, Gabbard, & Høglend, 2012). Colli et al. (2014) used the TRQ (Zittel Conklin & Westen, 2003) to examine specific clinicians’ emotional responses evoked by patients’ personality and psychological functioning, assessed with the Shedler–Westen Assessment Procedure-200 (SWAP-200; Shedler & Westen, 2007; Westen & Shedler, 1999a, 1999b). In that study, paranoid and antisocial personality disorders were related to criticized/mistreated therapist response, while helpless/ inadequate, overwhelmed/disorganized, and special/overinvolved countertransference was associated with borderline personality disorder. Disengaged therapist response was positively related to schizotypal and narcissistic personality disorders and negatively related to dependent and histrionic personality disorders. Positive countertransference was associated with avoidant personality disorder, which was also related to both parental/protective and special/overinvolved therapist responses. Schizoid personality disorder was related to helpless/inadequate response. Obsessive– compulsive personality disorder was negatively associated with special/overinvolved countertransference. In general, with lowerfunctioning patients, clinicians’ reactions were characterized by stronger negative feelings. The authors evaluated whether the associations between countertransference responses and patients’ personality pathology was dependent on the therapist’s theoretical approach. Analyses using a subsample of cognitive therapists did not differ from those in the full sample (including psychodynamic therapists), suggesting that the results were not affected by clinicians’ theoretical preconceptions. Patient’s symptomatology is another variable that could influence therapists’ responses. Røssberg et al. (2010) examined the extent to which the patients’ subjective psychiatric symptoms, assessed with the Symptom Checklist-90-Revised (SCL-90-R; Derogatis, 1994), evoked specific emotional reactions among therapists in an inpatient setting. Their findings indicated that the amount of symptomatology was positively associated with clinicians’ feelings of being inadequate, on guard, and rejected, and negatively related to confident countertransference response. In summary, studies exploring the relationship between patients’ personality and therapists’ reactions have not addressed the possible impact of patients’ levels of self-reported symptoms (Betan et al., 2005; Røssberg et al., 2007). Similarly, research investigating the association of psychiatric symptoms on therapists’ feelings has neglected the role of patients’ personality in evoking therapists’ responses (Røssberg et al., 2010). The separation between patients’ symptoms and personality in the investigation of their relationship with therapists’ emotional reactions is not justified by clinical and empirical literature. Comorbidity among major psychiatric disorders (ex-axis I of DSM–IV) and personality disorders (ex-axis II) is the norm rather than the exception (Westen, Gabbard, & Blagov, 2006). Moreover, empirical research suggests that symptoms can be better understood inside the context of personality structure (Powers & Westen, 2009; Thompson-Brenner & Westen, 2005). From this point of view, investigating conjointly personality and symptoms severity could provide precious information for a deeper understanding of psychopathological functioning and clinical practice. This research aimed to investigate the following hypotheses: Hypothesis 1: Patients with higher levels of symptom severity tend to evoke stronger degrees of clinicians’ negative emo-

LINGIARDI, TANZILLI, AND COLLI

230

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tional responses that cannot be accounted for by their theoretical orientation and other variables, such as gender, age, profession, and experience. Hypothesis 2: Symptom severity mediates the association between patients’ personality pathology and therapists’ responses. In line with studies mentioned above, we examined if severe symptomatology acts as a mediator on the relationship between (a) odd, eccentric, or socially detached personality disorders and therapists’ disengaged or inadequate responses; (b) dramatic, emotionally dysregulated, and impulsive personality disorders and overwhelmed/disorganized or inadequate patterns of countertransference; (c) anxious and fearful personality disorders and clinicians’ parental or overinvolved reactions.

Method Sampling Procedure We selected a sample of clinicians with at least 3 years’ postpsychotherapy licensure experience, who performed at least 10 hr of direct patient care per week, from the rosters of the two largest Italian associations of psychodynamic and cognitive– behavioral psychotherapy and from centers specializing exclusively in the treatment of personality disorders. We then requested that they select a patient who was at least 18 years old, had no psychotic disorder, and was not on drug therapy for psychotic symptoms; whom the therapist had seen for a minimum of eight sessions and a maximum of 6 months (one session per week); and who agreed to participate in a research protocol on psychological assessment. To minimize selection biases, we directed clinicians to consult their calendar to select the last patient they saw during the previous week who met the study criteria. To minimize rater-dependent biases (i.e., therapist effects), each clinician was allowed to describe only one patient. Clinicians and their patients did not receive any remuneration, and we had a response rate of 81% of therapists (N ⫽ 203) and 79% of their patients (N ⫽ 198). All participants (therapists and patients) provided written informed consent. In this study, we considered only the complete data relative to 198 separate therapist–patient dyads.

Therapists The therapist sample consisted of 198 White therapists, including 109 (55%) women and 89 (45%) men; 65% were psychologists and 35% were psychiatrists. Their mean age was 42.9 years (SD ⫽ 8.6, range ⫽ 34 –52). Two main clinical–theoretical approaches were represented: psychodynamic (N ⫽ 103) and cognitive– behavioral (N ⫽ 95). The average length of clinical experience as a psychotherapist was about 10 years (SD ⫽ 2.8, range ⫽ 3–17), and the average time spent per week practicing psychotherapy was 15.8 hr (SD ⫽ 3.8, range ⫽ 13–25). Seventy percent of the patients described were from independent practice and the remaining 30% were from public mental health institutions.

Patients The patient sample consisted of 198 White patients, including 115 (58%) women and 83 (42%) men. Their mean age was 33.9

years (SD ⫽ 4.4, range ⫽ 29.5–38.5). Sixteen patients were of low educational level (10 years), 80 of middle educational level (⬍13 years), and 102 of high educational level (⬎18 years). The majority (N ⫽ 152) were middle class (income/year between 15,000 and 30,000 euro), 12 working class (income/year ⬍15,000 euro), and 34 professionals (income/year ⬎30,000 euro). Ninety-seven participants were married, 43 were single, and 58 were divorced. Fifty-nine patients had only a DSM–IV axis I diagnosis, 70 had only an axis II diagnosis, 44 had comorbid axis I and axis II diagnoses, and 25 had a double axis II diagnosis. Among patients with axis I diagnoses (alone and comorbid with axis II disorders), 28 had a generalized anxiety disorder, 25 had a panic disorder, 23 had an eating disorder, 15 had a substance (cannabis) use disorder, and 12 had a dysthymic disorder. The mean Global Assessment of Functioning (GAF) score was 55.9 (SD ⫽ 11.7). The length of treatment (one session per week) averaged about 5 months (SD ⫽ 0.9; range ⫽ 2– 6).

Measures Shedler–Westen Assessment Procedure-200. The SWAP200 is an assessment procedure of personality pathology and health to provide clinicians of all the theoretical orientations a standard vocabulary for case descriptions (Westen & Shedler, 1999a, 1999b; Shedler & Westen, 2007; Lingiardi, Shedler, & Gazzillo, 2006). This Q-sort instrument includes 200 statements describing several aspects of personality, each of which may describe a given patient well, somewhat, or not at all. The clinician ranks these statements into eight categories from those that are most descriptive (assigned value of 7) to those that are not descriptive (assigned a value of 0). The SWAP-200 employs an assessment method based on matching prototype. It furnishes: (a) a personality diagnosis expressed as the matching of the patient assessment with 10 personality disorder scales, which are prototypical descriptions of DSM–IV axis II disorders, and (b) a personality diagnosis based on the correlation/matching of the patient’s SWAP description with 11 Q-factors/styles of personality derived empirically. In this study, we used only the personality disorder scales (PD scales). The SWAP-200 also includes a dimensional profile of healthy and adaptive functioning, and its scales or factors/styles can be used to obtain both categorical and dimensional diagnoses. It is important to observe that the validity of SWAP-200 diagnoses relies much more on therapist experience rather than specific instrumental training. In fact, SWAP-200 has shown good interrater reliability and convergent and discriminant validity both with clinicians that have not been trained in the use of the instrument (Blagov, Bi, Shedler, & Westen, 2012; Cogan & Porcerelli, 2004; Shedler & Westen, 2004; Westen & Shedler, 1999a, 1999b) and with clinicians that followed a specific instrumental training (Bradley, Hilsenroth, Guarnaccia, & Westen, 2007). Therapist Response Questionnaire. The TRQ (Betan et al., 2005; Zittel Conklin & Westen, 2003) is a clinician-report instrument designed to provide a psychometrically valid instrument for assessing countertransference patterns in psychotherapy. TRQ’s 79 items measure a wide range of thoughts, feelings, and behaviors expressed by therapists toward their patients that range from relatively specific feelings (e.g., “I feel bored in sessions with him/her”) to complex constructs, such as “projective identifica-

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

SYMPTOMS, PATIENT PERSONALITY, THERAPIST RESPONSE

tion” (e.g., “More than with most patients, I feel like I’ve been pulled into things that I didn’t realize until after the session was over”). Those statements are written in everyday language, without jargon, so that clinicians of all theoretical orientations can use the instrument without bias. The clinicians assess each item on a 5-point Likert scale, ranging from 1 (not true) and 5 (very true). The factor structure of the TRQ comprises eight countertransference dimensions: (a) overwhelmed/disorganized (9 items) indicates a desire to avoid or flee the patient and strong negative feelings, including dread, repulsion, and resentment; (b) helpless/ inadequate (9 items) describes feelings of inadequacy, incompetence, hopelessness, and anxiety; (c) positive (8 items) indicates the experience of a positive working alliance and close connection with the patient; (d) special/overinvolved (5 items) describes a sense of the patient as special relative to other patients, and includes “soft signs” of problems in maintaining boundaries, including self-disclosure, ending sessions on time, and feeling guilty, responsible, or overly concerned about the patient; (e) sexualized (5 items) describes sexual feelings toward the patient or experiences of sexual tension; (f) disengaged (4 items) describes feeling distracted, withdrawn, annoyed, or bored in sessions; (g) parental/ protective (6 items) is marked by a wish to protect and nurture the patient in a parental way, above and beyond normal positive feelings toward the patient; (h) criticized/mistreated (18 items) describes feelings of being unappreciated, dismissed, or devalued by the patient. The scales’ scores are obtained by calculating the average score of the items that make up each countertransference factor. In the present study, the eight factors showed excellent internal consistency (Streiner, 2003). The following Cronbach’s alpha values were obtained: overwhelmed/disorganized (␣ ⫽ .84), helpless/ inadequate (␣ ⫽ .82), positive (␣ ⫽ .76), special/overinvolved (␣ ⫽ .75), sexualized (␣ ⫽ .73), disengaged (␣ ⫽ .80), parental/ protective (␣ ⫽ .72), and criticized/mistreated (␣ ⫽ .80). The Symptom Checklist-90-Revised. The SCL-90-R (Derogatis, 1994) is a widely used self-report psychotherapy change measure assessing 90 psychiatric symptoms on a 5-point Likert scale, ranging from 0 (not at all) to 4 (extremely), indicating how much the client had “been distressed” by the symptom within the past seven days. The Global Severity Index (GSI), which is the mean rating across all 90 items, summarizes the client’s general psychiatric symptom severity. Clinical questionnaire. We constructed an ad hoc questionnaire for clinicians to furnish general information about themselves, patients, and therapies. Clinicians provided basic demographic data, including their profession (psychiatrist or psychologist), years of experience, theoretical orientation, employment address, hours of work, number of patients in treatment, and also gender, as well as patients’ age, gender, race, education level, socioeconomic status, and DSM–IV axis I diagnoses. Clinicians also provided data on the therapies, such as length of treatment and number of sessions.

Procedure After we received the clinicians’ and patients’ consent to participate, we provided them with the material to conduct the study. Clinicians were asked first to evaluate their emotional responses to the selected patient using the TRQ, and then, between one and

231

three weeks later, to evaluate the same patient’s personality using the SWAP-200. We used this interval because the TRQ and SWAP-200 requires different time of commitment. While the TRQ, a faster and user-friendly measure, was completed immediately after a session with the designated patient, the SWAP-200, more time consuming assessment method, was completed later. Another aim of this interval was to reduce any possible effect that clinicians’ rating own emotional response might have on a concurrent evaluation of patients’ personality. We asked therapists to deliver the Symptom Checklist-90-Revised to patients at the end of the session where they rated the TRQ. Clinicians asked their selected patients to return the completed test the next week. Thus, therapists and patients filled out the measures after the same session. Five patients did not return the SCL-90-R.

Statistical Analysis SPSS 20 for Windows (IBM, Armonk, NY) was used to conduct all analyses. To investigate the relationship between patients’ symptom severity and clinicians’ emotional responses, we calculated the bivariate correlations (Pearson’s r, two-tailed) between the GSI of SCL-90-R and TRQ factors. To verify if this relationship accounted for by several therapists’ variables—theoretical orientation, as well as gender, age, profession, and experience—we performed a series of hierarchical (block) multiple regression analyses. All the multiple regressions— one for each TRQ factor as the dependent variable—were estimated in three steps. The first step (i.e., block) contained clinicians’ gender and age. Profession (psychiatrist or psychologist), experience, and theoretical orientation (psychodynamic or cognitive– behavioral) were added in the second step. The GSI of SCL-90-R was entered in the third step. Change in R2 was considered as a measure of significance of each step. The F test, which is referred to as the F-change, was used to test if R2 improvement was statistically significant (with a significance level of p ⱕ .05). To test whether patients’ symptom severity mediated the relationship between their personality pathology (assessed using the SWAP-200 personality disorders scales, PD scales) and patterns of therapists’ emotional response, we performed a mediation analysis. According to Baron and Kenny (1986), mediation is estimated by multiple regression and occurs when the following criteria are satisfied: (1) the predictor is associated with the criterion variable (path c or total effect), (2) the predictor is associated with the mediator (path a), (3) the mediator significantly affects the criterion variable after controlling for the predictor (path b), and (4) the association between the predictor and criterion variable is significantly reduced when controlling for the effect of the mediator (path c’ or direct effect). “Complete” mediation holds when, controlling for the mediator, the predictor does not statistically significantly predict the criterion variable any further; in other worlds, the previously association is no longer significant (path c’ is zero). “Partial” mediation holds when, controlling for the mediator, the association between the predictor and criterion variable decreases, but not to zero (path c’ ⬍ path c). It is possible to find an “inconsistent mediation” when the mediator acts like a “suppressor variable” (MacKinnon, Fairchild, & Fritz, 2007). Typical inconsistent mediation holds when, controlling for the mediator, the association between the predictor and criterion variable is larger than the total effect (path c’ ⬎ path c). To establish the signifi-

LINGIARDI, TANZILLI, AND COLLI

232

cance of the mediation model, we used the Sobel test (Sobel, 1982). Finally, to estimate the magnitude of indirect or mediated effect, we calculated the percentage/proportion of the mediated effect (or the product of path a per b) to the total effect (or the path c): 100(ab/c). This method is based on the proportionate separation of effects into direct and indirect components and allows a simpler comparison of such effects (Alwin & Hauser, 1975; Preacher & Kelley, 2011).

Results

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Therapist Response and Patient Symptom Severity Our first aim was to investigate the direct (bivariate) relationship between patients’ symptom severity and therapists’ emotional responses. We found that the GSI of SCL-90-R was positively associated with criticized/mistreated (r ⫽ .21, p ⱕ .01), helpless/inadequate (r ⫽ .72, p ⱕ .001), overwhelmed/disorganized (r ⫽ .82, p ⱕ .001), and special/overinvolved (r ⫽ .14, p ⱕ .05) countertransference, and negatively with parental/protective (r ⫽ ⫺.41, p ⱕ .001), positive (r ⫽ ⫺.43, p ⱕ .001), and disengaged (r ⫽ ⫺.38, p ⱕ .001) one. No significant correlation between the GSI and sexualized countertransference emerged. To verify if these findings were dependent on clinicians’ theoretical approach (psychodynamic or cognitive– behavioral) and other variables (as gender, age, profession [psychiatrist or psychologist], and experience), we performed a series of hierarchical multiple regression analyses (see Table 1).

Therapist Response, Patient Personality Pathology, and Symptom Severity: A Mediation Analysis Our second aim was to investigate if patients’ symptom severity mediated the relationship between their specific personality disorders and therapists’ responses. Following the Baron and Kenny’s approach (1986), in the mediation model all the three variables (SWAP-200 personality disorder scales, GSI of SCL-90-R, and TRQ factors) must be significantly associated. We excluded the analyses regarding the SWAP-200 paranoid, schizoid, antisocial, narcissistic, dependent, and obsessive personality disorder scales because they were not significantly related to the GSI (p ⬎ .05). We ran 28 total regression analyses: We considered the remaining four SWAP-200 personality disorder scales (schizotypal, borderline, histrionic, and avoidant) used in separate mediational analyses for each of seven different countertransference patterns of TRQ— excluding the sexualized one because the bivariate correlation between this variable and the GSI was not significant. Of 28 total regression analyses, 8 met criteria for partial or inconsistent mediation (see Figure 1). The SWAP-200 schizotypal disorder scale was significantly related to both the disengaged countertransference pattern (␤ ⫽ 0.29, t ⫽ 4.31, p ⱕ .001) and the GSI (␤ ⫽ ⫺0.23, t ⫽ ⫺3.35, p ⱕ .001). In the last regression, the inclusion of the GSI as mediator implied a reduction of the total effect of the schizotypal personality on disengaged therapist response (␤ ⫽ 0.22, t ⫽ 3.28, p ⱕ .001). The Sobel test confirmed the significance of causal chains in this mediation model: Sobel z-value ⫽ 2.65, SE ⫽ 0.002, p ⱕ .01. The percentage/ proportion of the GSI’s mediated effect to the total effect of the schizotypal personality on disengaged countertransference was 26.17%.

The SWAP-200 borderline disorder scale was significantly related to the helpless/inadequate (␤ ⫽ 0.56, t ⫽ 9.63, p ⱕ .001), overwhelmed/disorganized (␤ ⫽ 0.66, t ⫽ 12.57, p ⱕ .001), and special/ overinvolved (␤ ⫽ 0.46, t ⫽ 7.31, p ⱕ .001) countertransference patterns, as well as the GSI (␤ ⫽ 0.60, t ⫽ 10.55, p ⱕ .001). In the last regression, the inclusion of the GSI as mediator implied a different reduction of the effects of the borderline personality on specific therapist responses: helpless/inadequate (␤ ⫽ 0.20, t ⫽ 3.47, p ⱕ .001), overwhelmed/disorganized (␤ ⫽ 0.26, t ⫽ 5.79, p ⱕ .001), and special/overinvolved (␤ ⫽ 0.33, t ⫽ 5.04, p ⱕ .001). The Sobel test confirmed that patients’ symptom severity partially mediated the relationship between the borderline personality and the helpless/inadequate (Sobel z-value ⫽ 7.38, SE ⫽ 0.004, p ⱕ .001), overwhelmed/ disorganized (Sobel z-value ⫽ 8.20, SE ⫽ 0.002, p ⱕ .001), and special/overinvolved (Sobel z-value ⫽ 2.58, SE ⫽ 0.003, p ⱕ .01) countertransference. The percentage/proportions of GSI’s mediated effects to the total effect of the borderline personality on each of therapist response were as follows: 63.45% of helpless/inadequate, 58.38% of one overwhelmed/disorganized, and 27.39% of special/ overinvolved. The SWAP-200 histrionic disorder scale was significantly related to both the disengaged countertransference pattern (␤ ⫽ ⫺0.31, t ⫽ ⫺4.64, p ⱕ .001) and the GSI (␤ ⫽ 0.44, t ⫽ 6.88, p ⱕ .001). In the last regression, the inclusion of the GSI as mediator implied a large reduction of the total effect of the histrionic personality disorder on disengaged therapist response (␤ ⫽ ⫺0.18, t ⫽ ⫺2.51, p ⱕ .05). The Sobel test confirmed the significance of causal chains in this mediation model: Sobel z-value ⫽ ⫺3.46, SE ⫽ 0.003, p ⱕ .001. The percentage/proportion of the GSI’s mediated effect to the total effect of the histrionic personality on disengaged countertransference was 42.58%. The SWAP-200 avoidant disorder scale significantly related to the positive (␤ ⫽ 0.27, t ⫽ 4.02, p ⱕ .001), parental/protective (␤ ⫽ 0.66, t ⫽ 12.41, p ⱕ .001), and special/overinvolved (␤ ⫽ 0.20, t ⫽ 2.95, p ⱕ .01) countertransference patterns, as well as the GSI (␤ ⫽ ⫺0.24, t ⫽ ⫺3.49, p ⱕ .001). In the last regression, the inclusion of the GSI as mediator implied a different reduction of the effect of the avoidant personality on specific therapist responses: positive (␤ ⫽ ⫺0.36, t ⫽ ⫺5.59, p ⱕ .001) and parental/protective (␤ ⫽ ⫺0.29, t ⫽ ⫺5.63, p ⱕ .001). The Sobel test confirmed that patients’ symptom severity partially mediated the relationship between the avoidant personality and positive (Sobel z-value ⫽ 3.01, SE ⫽ 0.003, p ⱕ .01) and parental/protective (Sobel z-value ⫽ 2.91, SE ⫽ 0.002, p ⱕ .01) countertransference. The percentage/proportions of the GSI’s mediated effects to the total effect of the avoidant personality on each of therapists’ responses were as follows: 31.60% of positive and 10.25% of parental/protective. Regarding the effect of GSI on the relationship between the SWAP-200 avoidant disorder scale and special/overinvolved countertransference, we found an inconsistent mediation. The GSI worked as a suppressor variable. In the last regression, the inclusion of the GSI implied an increase of the effect of avoidant personality on special/overinvolved therapist response (␤ ⫽ 0.25, t ⫽ 3.59, p ⱕ .001). The Sobel test confirmed the significance of causal chains in this inconsistent mediation: Sobel z-value ⫽ ⫺2.11, SE ⫽ 0.001, p ⱕ .05. The percentage/ proportion of the GSI’s mediated effect to the total effect of the avoidant personality on special/overinvolved countertransfer-

SYMPTOMS, PATIENT PERSONALITY, THERAPIST RESPONSE

233

Table 1 Hierarchical Multiple Regression Analyses Predicting TRQ Factors From Clinicians’ Variables and Patients’ Symptom Severity (GSI of SCL-90-R) (N ⫽ 198)

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Countertransference, clinicians’ and patients’ variables Criticized/mistreated Step 1: Gender and age Gender (1 ⫽ female; 2 ⫽ male) Age Step 2: Profession, experience, and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R) Helpless/inadequate Step 1: Gender and age Gender (0 ⫽ female; 1 ⫽ male) Age Step 2: Profession, experience, and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R) Positive Step 1: Gender and age Gender (0 ⫽ female; 1 ⫽ male) Age Step 2: Profession, experience and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R) Parental/protective Step 1: Gender and age Gender (0 ⫽ female; 1 ⫽ male) Age Step 2: Profession, experience, and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R) Overwhelmed/disorganized Step 1: Gender and age Gender (0 ⫽ female; 1 ⫽ male) Age Step 2: Profession, experience, and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R) Special/overinvolved Step 1: Gender and age Gender (0 ⫽ female; 1 ⫽ male) Age Step 2: Profession, experience and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R)

R

R2

0.114

0.013

Standardized ␤

F change (model)

p

1.307

.273

0.650

.584

7.523

.007ⴱⴱ

0.826

.439

1.526

.209

0.078 ⫺0.083 0.150

0.010 0.001 ⫺0.075 ⫺0.033

0.242

0.036 0.192

0.091

0.008 0.024 ⫺0.048

0.175

0.023 0.031 ⫺0.074 ⫺0.051

0.727

0.497

206.606

.000ⴱⴱⴱ

0.711 0.119

0.014

0.215

0.032

⫺0.144 0.066

1.433

.241

2.221

.087

0.001 0.118 0.097 0.466

0.171

0.098

0.010

0.169

0.019

0.434

0.160

0.119

0.014

⫺0.417 ⫺0.043 0.054 ⫺0.073 0.080 0.049 ⫺0.403

42.815

.000ⴱⴱⴱ

0.972

.380

1.280

.283

38.567

.000ⴱⴱⴱ

1.426

.243

1.459

.227

0.013 ⫺0.063 0.188

0.021 0.023 ⫺0.071 ⫺0.028

0.825

0.646

396.957

.000ⴱⴱⴱ

0.809 0.119

0.014

1.441

.239

0.042

.998

4.092

.044ⴱ

0.087 0.071 0.122

0.001 0.022 0.034 0.004

0.187

0.020 0.143

(table continues)

LINGIARDI, TANZILLI, AND COLLI

234 Table 1 (continued)

Countertransference, clinicians’ and patients’ variables

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Disengaged Step 1: Gender and age Gender (0 ⫽ female; 1 ⫽ male) Age Step 2: Profession, experience, and theoretical orientation Profession (1 ⫽ psychologist; 2 ⫽ psychiatrist) Experience Theoretical orientation (1 ⫽ cognitive-behavioral; 2 ⫽ psychodynamic) Step 3: Patient symptom severity Global Severity Index (SCL-90-R) ⴱ

p ⱕ .05.

ⴱⴱ

p ⱕ .01.

ⴱⴱⴱ

R

R2

0.118

0.014

Standardized ␤

F change (model)

p

1.401

.249

0.203

.894

0.133 ⫺0.025 0.130

0.003

0.418

0.158

⫺0.010 ⫺0.114 ⫺0.001 ⫺0.404

37.414

.000ⴱⴱⴱ

p ⱕ .001.

ence (taking the absolute values of the direct and indirect effects; see Alwin & Hauser, 1975) was 24%.

Discussion These findings point to several considerations. First, patients with higher symptom severity tend to evoke in clinicians of different

therapeutic approaches stronger degrees of negative emotional responses, such as a feeling of being overwhelmed, disorganization, helplessness, and frustration. Therapists report intense countertransference reactions characterized by apprehension, dread, anxiety, and several difficulties in experiencing positive and nurturing feelings to a patient distressed by severe symptomatology. In particular, in our study the magnitude of the associations between the patient’ symptom

Global Severity Index (SCL-90-Rb) β path a=–0.23*** Schizotypal Personality Disorder (SWAP-200a)

β path b=–0.33*** β path c=0.29*** β path c’=0.22***

Disengaged Countertransference

Global Severity Index (SCL-90-Rb) β path a=0.60*** Borderline Personality Disorder (SWAP-200a)

β path b=0.60*** β path c=0.56*** β path c’=0.20*** β path c=0.66*** β path c’=0.26*** β path c=0.46*** β path c’=0.33***

Helpless/ Inadequate Countertransference

β path b=0.66*** Overwhelmed/ Disorganized Countertransference

β path b=0.21** Special/ Overinvolved Countertransference

Figure 1. The mediated effect of patients’ symptom severity (GSI of SCL-90-R) on the relationship between TRQ factors and SWAP-200 personality disorders (N ⫽ 198). Note. a SWAP-200 ⫽ Shedler–Westen Assessment Procedure–200; b SCL-90-R ⫽ Symptom Checklist-90-Revised. Mediation significant paths are presented by solid lines. An inconsistent mediation path is presented by dashed lines. Baron and Kenny’s path diagrams include completely standardized path coefficients obtained using a series of multiple regressions to construct the mediation models: - Step 1: Regression of the dependent variable (CT) on the independent variable (PD) (path c); - Step 2: Regression of the mediator (GSI) on the independent variable (PD) (path a); - Step 3 and 4: Regression of the dependent variable (CT) on both the mediator (GSI) and independent variable (PD) (path b and c’). ⴱ p ⱕ .05. ⴱⴱ p ⱕ .01. ⴱⴱⴱ p ⱕ .001.

SYMPTOMS, PATIENT PERSONALITY, THERAPIST RESPONSE

235

Global Severity Index (SCL-90-Rb) β path a=0.44*** Histrionic Personality Disorder (SWAP-200a)

β path b=–0.30*** β path c=–0.31*** β path c’=–0.18*

Disengaged Countertransference

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Global Severity Index (SCL-90-Rb) β path a=–0.24*** Avoidant Personality Disorder (SWAP-200a)

β path b=–0.36*** β path c=0.27*** β path c’=0.19** β path c=0.66*** β path c’=0.59*** β path c=0.20*** β path c’=0.25***

Positive Countertransference

β path b = –0.29*** Parental/ Protective Countertransference

β path b=0.20* Special/ Overinvolved Countertransference

Figure 1 (continued).

severity and therapists’ helpless/disorganized reactions is very large. It would seem that patients presenting with serious psychiatric disturbances evoke mostly emotional responses of inadequacy and being overwhelmed in clinicians, provoking several problems to manage the clinical relationship. It is possible that these countertransference patterns associated to patients’ symptom severity may exacerbate and interfere with the construction and maintaining of a good enough therapeutic relationship more than other characteristics of the patients (Hayes et al., 2011). Concerning the positive association between special/overinvolved countertransference and high levels of patients’ symptomatology, we may suppose that clinicians could experience an excessive commitment in reaction to a special urgency and severity of the psychopathological picture of the patient. Our study is consistent with other research that revealed positive correlations between patients’ self-reported symptoms and therapists’ feelings of being on guard, rejected, and inadequate (Røssberg et al., 2010). Furthermore, therapists of different therapeutic approaches (or gender, age, profession, and experience) showed similar emotional reactions in treating patients characterized by high levels of symptom severity, suggesting that any set of techniques and theoretical preferences of clinicians does not affect what they experience in treatment with them. These findings seem to support the previous studies that highlighted the relationship between therapists’ response, patients’ personality (Betan et al., 2005; Røssberg et al., 2007), and their symptomatology (Røssberg et al., 2010). In conclusion, our results also confirm the link between patient personality and symptoms (Powers & Westen, 2009; Thompson-Brenner & Westen, 2005; Westen, Gabbard, & Blagov, 2006).

In this study, it is important to note that a broad range of personality disorders (paranoid, schizoid, antisocial, narcissistic, dependent, and obsessive) was not mediated by symptom severity or global psychopathology. The absence of such a mediated effect seems to suggest that the personality characteristics and interpersonal functioning of patients suffice to elicit distinct emotional responses in clinicians. On the other hand, where the mediated effect of symptomatology was apparent (with schizotypal, borderline, histrionic, and avoidant personality disorders), this average impact was fairly moderate (about 30%). Thus, the patient’s symptoms may partially mediate the relationship between his or her personality pathology and some specific countertransference responses, but in general, their impact is less sizable than one aroused by patients’ personality style. In addition, we can observe the mediated effect of symptom severity is not uniform and may vary considerably depending on the personality disorder and therapists’ emotional responses. For example, the relationships between borderline personality disorder and helpless or disorganized countertransference were greatly affected by psychopathological symptoms (accounting for 63% and 58% of the total effects). This finding seems to support the idea that these patterns of countertransference response, typically related to borderline personality disorder (Clarkin et al., 2006; Gabbard, 1993; Kernberg, 1984), are mostly elicited by higher levels of patients’ symptom severity rather than by their personality structure. Conversely, therapist over involvement or sense of specialness in the treatment of borderline patients seems to be more related to issues of personality in the therapeutic dyad. In addition, clinicians’ emotional responses evoked by other personality disorders appear affected to a lesser

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

236

LINGIARDI, TANZILLI, AND COLLI

degree by patients’ symptomatology. For example, positive and parental/protective countertransference responses were mostly elicited by patients’ avoidant personality characteristics rather than by their psychopathological symptoms. In this study we found that patients’ symptom severity acts also like a suppressor in a model of inconsistent mediation. More specifically, symptomatology increases the magnitude of the relationship between patients’ avoidant personality disorder and special/overinvolved countertransference. This finding seems to suggest that clinicians tend to experience stronger feelings of involvement and concern toward avoidant patients with more severe psychopathology. Perhaps, they could feel a sense of more urgency to develop and maintain a connection and engagement with an individual who is highly avoidant or fearful of affective–relational bonds. This study has some limitations. First, social desirability bias in therapists’ answers might have affected the results: for example, no significant correlation emerged between patients’ symptomatology and the sexualized countertransference pattern. This is not consistent with clinical experience and literature in which erotic and romantic feelings are widely described, especially in the treatment of severe personality disorders (Clarkin et al., 2006; Gabbard, 2009; McWilliams, 2011; PDM Task Force, 2006). Second, clinicians assessed their own countertransference reactions. Although they are wellsuited to describe their reactions toward own patients, the measure of countertransference we used shares the inherent limits of self-report measures, such as defensive biases and failure to recognize processes that an outside observer might identify. Future investigations examining therapists’ responses in psychotherapy by other methods of measurement and perspective (e.g., external observer) would benefit from more articulated research designs. Third, we undertook 28 total regression analyses (four different SWAP-200 scales used in the analysis of mediation) and only about 30% of the results met criteria for partial or inconsistent mediation, whereas the remaining 70% did not meet criteria for full, partial, or inconsistent mediation. These findings are nevertheless consistent with our hypotheses based on the results of previous research (Betan et al., 2005; Colli et al., 2014; Røssberg et al., 2007, 2010). Our study is the first to explore the relationship between patients’ personality, their symptomatology, and therapists’ emotional responses in a mediation model. We hope that future research will be able to examine this relevant topic using mediation models, perhaps also considering the effects of other mediator variables (of both patients and therapists), especially in specific clinical populations. In addition, some clinicians reported that the TRQ was easy to complete and clinically informative, but sometimes too long. In this study, every clinician completed only one questionnaire, but if the aim of a study is to capture the therapist’s inner experience after every session, a short version might be more suitable. In summary, the TRQ is likely to be appealing to investigators interested in exploring countertransference dimensions as a part of a multimodal assessment of the subjective experience and nature of therapists’ reactions (Gazzillo, Lingiardi, & Del Corno, 2011). Finally, in this study we tested a linear model of the relationship, assuming that patient pathology leads to therapist countertransference responses. In future research we intend to investigate a more complex, holistic model of the relationship, assuming the interaction and reciprocal influence of patient and therapist characteristics, seeing therapist responses as part of a relational matrix, and a valuable source of information about countertransference responses.

References Alwin, D. F., & Hauser, R. M. (1975). The decomposition of effects in path analysis. American Sociological Review, 40, 37– 47. doi:10.2307/ 2094445 Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. doi:10.1037/0022-3514.51.6.1173 Bateman, A. W., & Fonagy, P. (2006). Mentalization based treatment for borderline personality disorder: A practical guide. Oxford, UK; New York, NY: Oxford University Press. doi:10.1093/med/9780198570905 .001.0001 Betan, E., Heim, A. K., Zittel Conklin, C., & Westen, D. (2005). Countertransference phenomena and personality pathology in clinical practice: An empirical investigation. The American Journal of Psychiatry, 162, 890 – 898. doi:10.1176/appi.ajp.162.5.890 Blagov, P. S., Bi, W., Shedler, J., & Westen, D. (2012). The ShedlerWesten Assessment Procedure (SWAP): Evaluating psychometric questions about its reliability, validity, and fixed score distribution. Assessment, 19, 370 –382. doi:10.1177/1073191112436667 Bradley, R., Hilsenroth, M., Guarnaccia, C., & Westen, D. (2007). Relationship between clinician assessment and self-assessment of personality disorders using the SWAP-200 and PAI. Psychological Assessment, 19, 225–229. doi:10.1037/1040-3590.19.2.225 Brody, F., & Farber, B. (1996). The effects of therapist experience and patient diagnosis on countertransference. Psychotherapy: Theory, Research, Practice, Training, 33, 372–380. doi:10.1037/0033-3204.33.3 .372 Burgoon, J. K., Stern, L. A., & Dillman, L. (1995). Interpersonal adaptation: Dyadic interaction patterns. New York, NY: Cambridge University Press. doi:10.1017/CBO9780511720314 Clarkin, J. F., Yeomans, F. E., & Kernberg, O. F. (2006). Psychotherapy for borderline personality: Focusing on object relations. Washington, DC: American Psychiatric Publishing. Cogan, R., & Porcerelli, J. (2004). Personality Pathology, Adaptive Functioning, and Strengths at The Beginning and end of Psychoanalysis. Journal of the American Psychoanalytic Association, 52, 1230 –1231. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, MI: Erlbaum. Colli, A., Tanzilli, A., Dimaggio, G., & Lingiardi, V. (2014). Patient personality and therapist response: An empirical investigation. The American Journal of Psychiatry, 171, 102–108. doi:10.1176/appi.ajp .2013.13020224 Dahl, H. S. J., Røssberg, J. I., Bøgwald, K. P., Gabbard, G. O., & Høglend, P. A. (2012). Countertransference feelings in one year of individual therapy: An evaluation of the factor structure in the Feeling Word Checklist–58. Psychotherapy Research, 22, 12–25. doi:10.1080/ 10503307.2011.622312 Derogatis, L. R. (1994). Symptom checklist-90-R: Administration, scoring, and procedures manual (3rd ed.). Minneapolis, MN: National Computer Systems. Fauth, J. (2006). Toward more (and better) countertransference research. Psychotherapy, 43, 16 –31. doi:10.1037/0033-3204.43.1.16 First, M. B., Gibbon, M., Spitzer, R. L., Williams, J. B. W., & Benjamin, L. S. (1997). Structured clinical interview for DSM–IV Axis II personality disorders (SCID-II). Washington, DC: American Psychiatric Press, 1997. Freud, S. (1910). The future prospects of psychoanalytic therapy (Standard Edition, 11, pp. 139 –151). London: Hogarth Press. Freud, S. (1912). Recommendations to physicians practicing psychoanalysis (Standard Edition, 12, pp. 111–120). London: Hogarth Press. Gabbard, G. O. (1993). An overview of countertransference with borderline patients. Journal of Psychotherapy Practice and Research, 2, 7–18. Gabbard, G. O. (1995). Countertransference: The emerging common ground. The International Journal of Psychoanalysis, 76, 475– 485.

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

SYMPTOMS, PATIENT PERSONALITY, THERAPIST RESPONSE Gabbard, G. O. (2001). A contemporary psychoanalytic model of countertransference. Journal of Clinical Psychology, 57, 983–991. doi:10.1002/ jclp.1065 Gabbard, G. O. (Eds.). (2009). Textbook of psychotherapeutic treatments. Washington, DC: American Psychiatric Publishing. doi:10.1176/appi .books.9781585623648 Gazzillo, F., Lingiardi, V., & Del Corno, F. (2011). Toward the validation of three assessment instruments derived from the PDM P Axis: The Psychodynamic Diagnostic Prototypes, the Core Preoccupations Questionnaire and the Pathogenic Beliefs Questionnaire. Bollettino di Psicologia Applicata, 265, 31– 45. Gelso, C. J., & Hayes, J. A. (2007). Countertransference and the therapist’s inner experience: Perils and Possibilities. New Jersey, NJ: Erlbaum. Hayes, J. A., Gelso, C. J., & Hummel, A. M. (2011). Managing countertransference. Psychotherapy, 48, 88 –97. doi:10.1037/a0022182 Heimann, P. (1950). On counter-transference. The International Journal of Psychoanalysis, 31, 81– 84. Kernberg, O. F. (1965). Notes on countertransference. Journal of the American Psychoanalytic Association, 13, 38 –56. doi:10.1177/ 000306516501300102 Kernberg, O. F. (1984). Severe personality disorders: Psychotherapeutic strategies. New Haven, CT: Yale University Press. Lingiardi, V., Shedler, J., & Gazzillo, F. (2006). Assessing personality change in psychotherapy with the SWAP-200: A case study. Journal of Personality Assessment, 86, 23–32. doi:10.1207/s15327752jpa8601_04 MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation analysis. Annual Review of Psychology, 58, 593– 614. doi:10.1146/ annurev.psych.58.110405.085542 McIntyre, S. M., & Schwartz, R. C. (1998). Therapists’ differential countertransference reactions toward clients with major depression or borderline personality disorder. Journal of Clinical Psychology, 54, 923– 931. doi:10.1002/(SICI)1097-4679(199811)54:7⬍923::AID-JCLP6⬎3 .0.CO;2-F McWilliams, N. (2011). Psychoanalytic diagnosis: Understanding personality structure in the clinical process (2nd ed.). New York, NY: Guilford Press. Muran, J. C., & Barber, J. P. (2010). The therapeutic alliance: An evidence-based guide to practice. New York, NY: Guilford Press. Norcross, J. C. (2011). Psychotherapy relationships that work (2nd ed.). New York, NY: Oxford University Press. doi:10.1093/acprof:oso/ 9780199737208.001.0001 Ogden, T. (1982). Projective identification and psychotherapeutic technique. New York, NY: Jason Aronson. OPD Task Force. (Ed.) (2008). Operationalized Psychodynamic Diagnosis OPD-2: Manual of diagnosis and treatment planning. Ashland, OH: Hogrefe & Huber Publishers. PDM Task Force. (Ed.) (2006). Psychodynamic diagnostic manual. Silver Spring, MD: Alliance of Psychoanalytic Organizations. Powers, A., & Westen, D. (2009). Personality subtypes in patients with panic disorder. Comprehensive Psychiatry, 50, 164 –172. doi:10.1016/j .comppsych.2008.07.006 Preacher, K. J., & Kelley, K. (2011). Effect sizes measures for mediation models: Quantitative strategies for communicating indirect effects. Psychological Methods, 16, 93–115. doi:10.1037/a0022658 Racker, H. (1957). The meanings and uses of countertransference. The Psychoanalytic Quarterly, 26, 303–357.

237

Røssberg, J. I., Hoffart, A., & Friis, S. (2003). Psychiatric staff members’ emotional reactions toward patients. A psychometric evaluation of an extended version of the Feeling Word Checklist (FWC-58). Nordic Journal of Psychiatry, 57, 45–53. doi:10.1080/08039480310000257 Røssberg, J. I., Karterud, S., Pedersen, G., & Friis, S. (2007). An empirical study of countertransference reactions toward patients with personality disorders. Comprehensive Psychiatry, 48, 225–230. doi:10.1016/j .comppsych.2007.02.002 Røssberg, J. I., Karterud, S., Pedersen, G., & Friis, S. (2010). Psychiatric symptoms and countertransference feelings: An empirical investigation. Psychiatry Research, 178, 191–195. doi:10.1016/j.psychres.2009.09.019 Safran, J. D. (1984). Assessing the cognitive-interpersonal cycle. Cognitive Therapy and Research, 8, 333–347. doi:10.1007/BF01173309 Sandler, J. (1976). Countertransference and role-responsiveness. International Review of Psycho-Analysis, 3, 43– 47. Schwartz, R. C., Smith, S. D., & Chopko, B. (2007). Psychotherapists’ countertransference reactions toward clients with antisocial personality disorder and schizophrenia: An empirical test of theory. American Journal of Psychotherapy, 61, 375–393. Shedler, J., & Westen, D. (2004). Refining personality disorder diagnosis: Integrating science and practice. The American Journal of Psychiatry, 161, 1350 –1365. doi:10.1176/appi.ajp.161.8.1350 Shedler, J., & Westen, D. (2007). The Shedler-Westen Assessment Procedure (SWAP): Making personality diagnosis clinically meaningful. Journal of Personality Assessment, 89, 41–55. doi:10.1080/ 00223890701357092 Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290 –312. doi:10.2307/270723 Streiner, D. L. (2003). Being inconsistent about consistency: When coefficient alpha does and doesn’t matter. Journal of Personality Assessment, 80, 217–222. doi:10.1207/S15327752JPA8003_01 Thompson-Brenner, H., & Westen, D. (2005). Personality subtypes in eating disorders: Validation of a classification in a naturalistic sample. The British Journal of Psychiatry, 186, 516 –524. doi:10.1192/bjp.186 .6.516 Westen, D., Gabbard, G. O., & Blagov, P. (2006). Back to the future: Personality structure as a context for psychopathology. In R. F. Krueger & J. L. Tackett (Eds.), Personality and psychopathology: Building bridges (pp. 335–384). New York, NY: Guilford Press. Westen, D., & Shedler, J. (1999a). Revising and assessing axis II, part I: Developing a clinically and empirically valid assessment method. The American Journal of Psychiatry, 156, 258 –272. Westen, D., & Shedler, J. (1999b). Revising and assessing axis II, part II: Toward an empirically based and clinically useful classification of personality disorders. The American Journal of Psychiatry, 156, 273–285. Zittel Conklin, C., & Westen, D. (2003). The therapist response questionnaire. Atlanta, Emory University, Departments of Psychology and Psychiatry and Behavioral Sciences. Retrieved from http://www.psychsystems.net/Manuals/ Countertransference/Westen%20countertransference%20questionnaire.pdf

Received February 20, 2014 Revision received August 12, 2014 Accepted August 13, 2014 䡲