Early Transference Interventions With Male Patients in Psychotherapy

Abstract

Transcripts of early sessions for 7 personality-disordered male subjects participating in an ongoing naturalistic long-term dynamic psychotherapy project were rated for therapist interventions and alliance. Early transference interpretations were followed by increased defensiveness even when there was a solid alliance. Omitting transference interpretations in the face of an early negative transference was equally problematic. However, the rapid sequence of early transference and defense interpretations, or early defense interpretations alone, enhanced therapeutic work without increasing defensiveness. Caretaking of the alliance after early interpretive work was also investigated. Two different styles of handling affect emerged from the sample.

In a previous study looking at the impact of interpretations on the alliance in the psychotherapy of female patients with borderline personality disorder,¹ we found that early transference interpretations were risky and more often led to further deterioration of an already low alliance. However, if the early alliance was solid and if supportive interventions and defense interpretations were used to pave the way, early transference interpretations were generally well tolerated and were followed by enhanced therapeutic work. Early defense interpretations did not carry the same risk and were generally followed by emotional elaboration in this sample of female borderline patients.

In our search to further elucidate complexities of the psychotherapy process, we turned to a sample of male patients to investigate once again the effect of interventions on the alliance. As before, we were interested in how males receive transference interpretations early in their treatment and also in the role of other interpretive and noninterpretive interventions in maintaining the early alliance.

The literature on interpretations in psychotherapy can be organized in two parts: one that examines the relationship of interpretations to outcome and another that looks at immediate in-session patient responses and impact on the alliance. Psychoanalytic theory proposes that interpretations are mutative and produce change through the acquisition of insight.^2,3 Transference interpretations are advocated as having the greatest potential to produce such change.^2,4 However, empirical studies on the effectiveness of transference interpretations have offered mixed results at best. Although several studies have found a positive correlation with outcome for interpretations in general,^5–7 earlier studies showing a positive relationship between transference interpretations and outcome have been followed by more recent research yielding minimal evidence for such a relationship.^8–10

Studies that correlate interventions with treatment outcome have been criticized by several authors.^11–13 More specifically, Hill¹⁴ notes that “the correlational method is not sensitive to issues such as quality and timing of the intervention…. To rectify these methodological problems, many researchers have begun to study the immediate effects of therapist interventions.” From a psychoanalytic viewpoint with regard to timing and quality of transference interpretations, Freud¹⁵ cautioned that the transference should not be interpreted early on, unless it becomes a resistance and threatens the viability of the treatment (p.139). He differentiated between negative “hostile” transference, which provided resistance, and positive transference, which he further subdivided into erotic and affectionate. He termed the latter subtype of positive transference the “unobjectionable positive regard for the analyst”; he proposed that this was the “vehicle of success” for the treatment and thus needed no interpretation¹⁶ (p.105). Similarly, Meissner,¹⁷ referring to more disturbed patients, advised that early transference interpretations could be experienced as critical attacks. Buie and Adler¹⁸ recommended using transference interpretations in the latter half of therapy after having created a soothing and holding environment with the use of support in the earlier stages.

Studies that examine the therapeutic interaction or dialogue^19–21 have added to our understanding of the impact of transference interpretations. These studies have focused on patient responses to interventions as predictors of outcome. Transference interpretations that are followed by affective disclosure rather than increased defensiveness have been consistently associated with positive outcome.¹⁹ However, this relationship also holds for interpretations that do not address the transference. Furthermore, transference interpretations were found to be twice as likely to elicit a defensive response when compared with other interpretations.

We propose to study interactive sequences between patient and therapist in order to shed further light on the impact of transference interventions. Specifically, we attempt to answer the following questions. How are transference interpretations in the early phase of psychotherapy received by male patients as compared with female patients? Does the negative or positive “valence” of early transference interpretations (i.e., whether they address an ambivalent or hostile versus a warm or idealizing patient–therapist relationship) influence their impact on the therapeutic alliance? And what techniques are used by therapists in order to manage the alliance following early interpretive work?

METHODS

Subjects were 7 males from an ongoing naturalistic long-term dynamic psychotherapy project at the Sir Mortimer B. Davis–Jewish General Hospital and McGill University in Montreal, Canada. The larger sample from which the only 7 males were obtained included 45 patients at the time of the study. Selection criteria for the sample as a whole include presence of one or more personality disorder diagnoses on Axis II and/or presence of a current mood and/or anxiety disorder diagnosis on Axis I. Subjects must be at least 18 years of age, cannot have a diagnosis of organicity or psychosis, and must not be acutely suicidal or abusing substances in a way that would interfere with their treatment. All subjects gave informed consent for their participation in the project.

Treatment consisted of once-weekly dynamic psychotherapy with experienced clinicians who had dynamic or psychoanalytic training. The majority of therapists (5/7) were male. All therapy sessions were audiotaped.

All Axis I to V diagnoses were arrived at by using the Guided Clinical Interview (GCI; J.C. Perry, poster presentation, American Psychiatric Association Annual Meeting, 1999), a semistructured diagnostic interview administered by an experienced researcher independent from the three authors. The Structured Clinical Interview for DSM-III-R²² (SCID II) was administered by a research assistant to confirm diagnostic information. For each subject, three early sessions were transcribed verbatim, usually sessions 3, 4, and 5. Session transcripts were rated for therapist interventions by using the Psychodynamic Interventions Rating Scale (PIRS; S. Cooper, M. Bond, unpublished manuscript, University of California–San Francisco, 1992) which has satisfactory reliability and construct validity.²⁰ Self-report therapeutic alliance scores were obtained by using the California Psychotherapy Alliance Scale (CALPAS).²³

The PIRS was scored by experienced raters trained in the use of this scale. Final ratings were obtained by process of consensus. The scale is outlined in Appendix A. Briefly, it presents two broad categories of interventions, interpretive and noninterpretive, which are outlined in a manual with definitions and examples to help discriminate between near-neighbor interventions. Interpretive interventions consist of transference (TI) and defense interpretations (DI), each of which are further scored for depth of interpretation on a level of 1 to 5. As an example, merely alluding to the relationship with the therapist would qualify as a level 1 transference interpretation. (We acknowledge that some may disagree with such a broad definition of transference interpretation and we note that using this method could potentially result in overestimating the presence of TIs.) Noninterpretive interventions consist of acknowledgments (A), clarifications (Cl), questions (Q), therapist associations (Ass), reflections (R), work-enhancing strategies (WES), support strategies (SS), and contractual arrangements (CA).

In order to rate the immediate effects of interventions on the alliance, the three authors first examined session transcripts independently for “turning points.” These are defined as notable changes in the patient's degree of emotional elaboration²⁴ or level of defensiveness within a session. Only those segments with turning points for which there was complete agreement among all three investigators were selected. Turning points tended to occur around movements toward or away from a more interpretive stance, and therefore the level of agreement was high.

From these selected segments with turning points, the three authors rated, on a qualitative rather than a quantitative basis, the moment-to-moment in-session change in alliance associated with interpretations. The change in alliance was rated independently by assessing immediate patient responses to interpretations. If these included elaboration of material preceding the intervention, or conveyed a sense of collaboration and emotional connectedness or attunement between patient and therapist, the alliance was deemed to have been enhanced or at least maintained by the intervention. If, however, the response included increased defensiveness or resistance, the alliance was deemed deteriorated or in danger of being so if reparative techniques of support did not ensue. The individual rating was followed by the process of reaching consensus. Only segments where full agreement among the three authors took place were included in the study.

It may be argued that alliance is not accurately assessed solely by examining written transcripts for what is said by therapist and patient because the complex nonverbal interaction will be missed.²⁵ This is a potential limitation in our methodology, and addressing it is beyond the scope of this paper. Further, our in-session microassessment of alliance may be criticized for lacking a validated systematic approach despite our use of agreement among three judges.

CLINICAL DATA AND INTERPRETATION OF RESULTS

Impact of Early Transference Interpretationson the Alliance

Negative Transference Interpretation Alone:

Mr. A. has a mixed narcissistic and dependent personality disorder. He presents while in remission from polysubstance abuse with 15 months of sobriety and intake Global Assessment of Functioning (GAF) of 57, indicating “moderate symptoms or moderate difficulty in social, occupational or school functioning.” Early self-report alliance score is relatively high (total CALPAS score=5.75, sample mean=5.50±0.76). The following excerpt illustrates the impact of a TI early in the fifth session addressing a negative transference.

Hearing the patient commiserating about all the sources of pressure in his life, the therapist inquires about negative feelings toward the therapist:

TI-1TH.: So you remember, too, that last Tuesday I had asked you to think about coming twice a week.

PT.: Yeah, it's cool.

TI-1TH.: Do you think that was also another added sense of pressure?

PT.: No, no, 'cause, I mean, my assumption right away up front is that I'm not going to be waking up at six o'clock in the morning to make this appointment. And that's fine, and if it becomes a pressure situation I'll just let you know and we'll just—we'll change it immediately, you know. But that didn't add to it at all.

QTH.: You usually stay up late?

PT.: I mean, I don't give myself a time to go to sleep.

Here an early interpretation of potentially negative transference involving the patient's feeling pressured for more closeness by the therapist is followed by defensive denial and a move toward more superficial content. To preserve the alliance, the therapist moves out of the transference with a nonthreatening question. Later in the session, he turns to a defense interpretation (DI) addressing the same conflict over closeness and fear of rejection. This interpretation is followed by positive emotional elaboration and seems to be integrated by the patient without posing a threat to the alliance:

DI-3H.: But that there's probably some fantasy operating in there that has a bad outcome. That you are imagining that this is not—something about this is not going to turn out well and so you protect yourself, right off the bat, by not pursuing it.

PT.: Interesting. I do think about this girl, though. It's her face, you know. She's very pretty. Very pretty girl. I mean, is it abnormal to think if like a girl she could be your girlfriend right away? You know, I don't know if that is normal or not. Maybe (laughs) you know, there's a song, “she ain't pretty, she just looks that way”.…

DI-2TH.: See, there already is the bitter fantasy that she may look pretty, but she is not going to be a nice person.

PT.: I might just—I think I'm just plain afraid. This is a pattern I established after I failed the grade, how I shut myself in and never admitted to someone that I liked them or never knew how to say: “Hey,” you know, “what are you doing?” But what do you do—you go out for a coffee or something like that?

Follow-up data for Mr. A. reveal that he remained in therapy for 2.5 years until he moved to another city. At termination, he reported being pleased with his progress and had taken part in 125 sessions, yielding a session density of 50 sessions per year. Session density provides information on the “dose” of psychotherapy that cannot be gleaned by looking at duration alone.

An excerpt from the therapy of another patient further illustrates the risk of early transference interpretations alone. Mr. B., a patient with severe narcissistic personality disorder, presents with a current major depressive episode and comorbid polysubstance abuse. Despite a favorable GAF score of 64 obtained at the intake interview, this patient is quite fragile, having entered psychotherapy shortly after his hospitalization for a suicide attempt. Indeed, in session 3, despite his high self-report alliance score (6.7), merely alluding to the therapeutic relationship is felt to be threatening and is followed by defensive denial.

Here the therapist intervenes after the patient's angry complaint directed at his spouse, co-workers, and boss that issues should be discussed right away:

PT.: …And I'm a funny type of person because I tell other people, you know, if you have a complaint on Monday, don't make it an issue on Wednesday 'cause it doesn't have any validity. It happened at ten o'clock in the morning, discuss it at ten o'clock in the morning when it's happening. The context is right, the timing is right, the seriousness of the situation or the frivolity of the situation is then and there, and it can be accepted, dismissed, dissected, whatever at that moment. If you discuss it with me three days later, it's totally out of context and the whole issue might revolve around contextual situation.

WESTH.: Well, I guess what I would say is if you have anything to say on Monday, try to hang onto it as best you can till Thursday, and when you are here on Thursday try to say whatever comes through your mind. I mean the first.

PT.: No, no, I'm not talking about you. I was talking in reference—

ASSTH.: No, I know. I know that you weren't talking for me. I'm just saying that this is what you would say to your friend or employee or whatever, and just what I'm saying to you is if you have anything Monday, try to hang onto it.

Here a work-enhancing strategy is experienced as a level-1 negative transference interpretation. The therapist preserves the alliance by following the patient's cue and moving away from the transference. This strategy is helpful, and the patient continues to rate his early alliance very favorably (6.5–6.75). After 2 years he remains in therapy, with a session density of 46, and has had no recidivism of his suicide attempt. However, given the patient's initial pathology, little structural change has occurred as measured by his defensive functioning at yearly dynamic interviews, where denial, acting out, passive aggression, omnipotence, splitting, devaluation, and idealization remain the prevailing defense mechanisms.

Negative Transference Interpretation Followed by Defense Interpretation:

Mr. C. has mixed narcissistic and obsessive personality disorder and presents with a major depressive episode and comorbid substance abuse. His intake GAF is 69 and his early self-report alliance is good (CALPAS=5.75). In session 4, a low-level transference interpretation (TI) is offered in response to the patient's request at the very beginning of the session that the therapist be more active:

PT.: At this stage I do need more dialogue and more, ugh, not just sitting—not just you sitting there and taking notes, but also talking to me. I don't know if that's a problem. I don't know if that's possible, but because of this very one-sided thing that I have had with you right now, I feel quite uncomfortable with you.

TI-2TH.: What does this mean, this discomfort at the ball being in your court? What is this discomfort at speaking on your own behalf, from yourself, as opposed to my being much more active and directive, which then defines the situation from me, the other person?

PT.: Uh-huh.

DI-1TH.: And further still, last week I connected that to your experience that other people take certain positions in your life which tend to define you. The example that sticks out in my mind was the issue of your wife moving with your kids to H. and your experience of that, that she should do whatever she wants because you wanted her to feel free of the marriage and so on.

Here the therapist interprets the patient's emerging ambivalence toward him and toward the frame with a level-2 TI and rapidly follows with a defense interpretation. The patient responds with elaboration:

PT.: Certainly I recognize from exactly that, that—and I did talk at length about that last week—that I felt often that my life has always been in the control of others and it's one of my—one of the reasons why I'm sort of strongly reacting against my present family and marital situation is maybe because I just can't stand that any more. And I certainly am reacting against that, maybe in order to take control over my life. So in that same vein, the idea that maybe I'm not yet comfortable with even controlling this discussion—I may have the same tendency to give somebody else control.

This negative TI rapidly followed by a mix of DI and support is well tolerated. The patient is able to make connections to important relationships in his life and appears in touch with the affect that accompanies these early insights. This patient remained in therapy for 7 months and received 28 sessions (session density=48).

Missed Negative Transference Interpretation:

Mr. D. has mixed narcissistic, obsessive-compulsive, and passive-aggressive personality disorder. He presents with a major depressive episode and comorbid dysthymia, alcohol abuse, obsessive-compulsive disorder (both obsessions and compulsions) and generalized anxiety disorder. His intake GAF score is 57 and his early self-report alliance score is high (CALPAS=5.75). In session 8, he repeatedly refers to his ambivalence toward the therapist. This is not picked up, and no interpretations addressing the negative transference are given. Increasingly superficial discourse follows, and the alliance deteriorates within the session.

The patient, talking about his boss but likely referring unconsciously to the therapist, warns him of his fear of staying the course and getting close:

PT.: … I have tried a lot of things, but I have never stuck with anything. Most of the time he [the boss] just listens and other times he tries to provide solutions, but the solutions are too (laughs) too simple and not realistic, you know. Don't take into consideration the reality of the situation. But I feel that he is not providing me the support that he should as a boss. If I were in his shoes, I would do something. I would try something else until—try and try until we succeed.

QTH.: I'm wondering if there was more that you wanted to say about your, quotes, fight with your wife this morning—if it is worth looking at the details step by step.

After briefly acceding to the therapist's request to explore the relationship with his wife, the patient seems to withdraw from exploring his internal world and moves to the more emotionally distant topic of medication:

PT.: Well (sighs) I was disappointed that, ugh, that she didn't apologize to me, I guess. Just expressed my disappointment and I saw she wasn't too happy with my disappointment and she was disappointed that I was disappointed. Sometimes I wonder if the Prozac is going to—the effect is going to maintain itself, you know. I had a couple of days this week I didn't feel I was in control, at work, of my emotions at work, you know. Am I getting immune to it? Are the effects dying down? I was coming home more stressed.

QTH.: What were the types of things that stressed you?

PT.: Oh, conversations with my boss, you know. He would say things that would irritate me, you know, I would find ridiculous, you know.

QTH.: Like?

PT.: Ah, jeez. I have a hard time remembering. (sighs)

Here a negative transference is left uninterpreted and deterioration of the alliance follows, as witnessed by feelings of isolation and anger in the patient and movement of the material away from a stance of self-exploration. This patient had high self-report early alliance scores (5.87, 6.20) but a low session density (26 sessions per year) and stopped at 2 years, having received 52 sessions.

Defense Interpretation Alone:

Mr. A. has been described previously. Early in his therapy, in session 4, a DI alone is offered. It is followed by deepening and further elaboration of the material and avoids the anxiety and defensiveness seen in session 5 following a TI (as presented earlier).

The patient talks with sadness about his father's early childhood and the impact of these experiences on himself:

PT.: But also, in that family there is a whole list of illness and mental illness, 'cause his sister went nuts and was institutionalized and her children were taken away from her. This is not a healthy family—completely dysfunctional and this is what he grew up in. I guess he never escaped it because it kept following him. And he had children (laughs) and, you know, so it spills over. Here I am (laughs) telling you about it, you know.

DI-1TH.: You have a big smile on your face, but I'm sure often [unclear].

PT.: Well, yeah, I mean, I don't think much about them, you know. And if I do smile, then maybe that's my own defense mechanism. But that's okay for me, because I'd much rather smile than not smile. It's just it's healthier for me to be this way (laughs). What do you think when you hear that?

DI-2TH.: I think that you've been through a lot of sadness and you're trying to cope with some very tough things and thats— Your way of coping with it is to put on a smile and ride with it.

PT.: Yeah. Well, I mean, the smile doesn't necessarily mean I'm a happy person inside, you know, but it's also—

DI-2TH.: You're trying your hardest to be happy and to cope.

The patient responds well to this DI and expresses some expectable anxiety to which the therapist responds reassuringly with a support strategy:

PT.: Yeah, to let go, too. It's dangerous to get stuck in looking into your life so much and asking questions about the past, because you get out of the now, right now. I want to know the source of this pain in my life, so that I can look at it, heal it and just say okay, and accept it and let it go and give it up.

SSTH.: In fact, that's really what the purpose is, is to allow yourself ultimately to be in the now even more.

PT.: Right.

SSTH.: Not to be so tied—

PT.: …the past…

SSTH.: —whatever there is in the past, and the way to do that is to really understand.

PT.: To look at it.

Following this interchange, the patient is able to continue in the session and elaborate further on his feelings about his relationships.

Managing the Alliance Following Early Interpretive Work

Two distinct styles of managing the alliance after early interpretive work are found in the early transcripts of our 7 male subjects. Of the 7 therapists in our sample, 4 out of 5 male therapists show an intellectualizing style, whereas both female therapists adopt an affect-probing and somewhat more interpretive style. One remaining male therapist does not seem to fit neatly into one or the other category, but rather has a mixed style that alternates between intellectualization and affect probing. Illustrations of these two techniques follow:

Intellectualizing Style:

Mr. C. has been described above. Following interpretive work, in session 4, he begins to manifest intense anxiety and mild disorganization. In response to his anxiety linked to the unstructured frame, his male therapist adopts a reassuring style with intellectual explanations and support:

PT.: I just have a bunch of rambling thoughts right now. Total series of disconnected rambling thoughts that, uh, sort of nothing to do with anything.

WESTH.: But it's exactly those that we would be interested in. It's unstructured, and that's how it works best. But that has made you uncomfortable to start with because you're not—well, it's a new experience for you.

REF

This appears to settle the patient, who then continues in a calmer and more intellectual vein:

PT.: You've probably noticed also that I tend to try to structure my thoughts and I try to sort of have one thing follow from A to B to C to D rather than just bounce around.

Early session material for this obsessional patient is consistent with a fear of emotional intensity and a fear of losing control. Later in the same session, the patient's anxiety is again apparent:

PT.: I know—I have the strong feeling that in time it will be that, the nonrejection of either one or the other. The acceptation of both these things is going to ultimately be very fruitful for me and will be very positive. But right now they are a source of a big problem. And sometimes I get frightened and I say well, maybe I am making a mistake. Maybe it's just one or the other.

Despite the intellectual language, the therapist responds to a buildup of emotion reflected in the patient's allusion to “strong feeling,” “big problem,” and “frightened” and adopts a reassuring intellectual stance:

DI-1TH.: Okay, again I don't have an idea yet about what this is all about because I don't know you well enough. But as I have said before, I'm impressed with how exigent you are with yourself. You're quite self-critical and there's—quite self-critical in the sense of being severe and harsh with yourself. I agree that a certain self-observation, a certain self-criticism is necessary for everyone, but not in a sense that I think that I'm hearing from you in terms of the harshness and severity of your criticisms of yourself. You see, there is a parallel between this process of psychotherapy and the kind of work you do. There is creativity there. In a certain way you have to let your mind go in whatever direction, and then something forms. Well, it's similar here. The way this process works best is that if you try your best to tell me everything that you are aware of that's rushing by.

WES

The therapist offers a defense interpretation couched in intellectual and emotionally distancing terms followed by a return to noninterpretive interventions in the form of a work-enhancing strategy. This appears to calm the patient without threatening his self-esteem and preserves the alliance, as seen by the patient's response that follows:

PT.: Well, that makes me feel good, hearing that, because I was wanting to try to be more structured and more coherent.

Later in session 5, the patient discusses his extramarital affair, with emotional intensity carried in the fractured and disorganized discourse. The therapist senses that Mr. C. tolerates poorly this depressive affect and distracts him away from these painful feelings with a task, in essence an invitation to intellectualization:

PT.: …And of course also had I had—to suffer through a long period of loneliness or whatever, which hasn't happened. So when I say the other day, I think I do feel guilty about, uh, because it's happened, uh, I'm not giving a chance to our marriage. But at the same time I'm not—I don't feel capable whatsoever to, uh, try again.

QTH.: It might help if you told me more about your own personal set of morals and values that you have as a result of your own life experiences and how that fits in with this.

This again settles the patient, who continues:

PT.: I'm a great believer in justice and equality.

In this dyad we see a male therapist managing the alliance, when intense affect emerges early in therapy, by distraction and intellectualization and by normalizing his patient's feelings. The alliance remains favorable.

This particular style is again illustrated in the following excerpt from Mr. E.'s third session. This patient demonstrates dependent personality traits and presents with a major depressive episode and coexisting dysthymia, panic disorder, and generalized anxiety disorder. Pre-treatment GAF is 52, demonstrating moderate impairment in overall functioning. In this session, Mr. E. brings material with themes of rage and revenge:

PT.: He goes—and then he started talking, uh, that I can't remember exactly how he put it, but he said, “Fuck you, man.” And it pissed me off. And I ended up having to meet that doctor one more time to see the results and I said, “That's it, I am never going again after a second time.” He just pissed me off. I really wanted to hit the man.

I find I really want to get revenge on so many people. I saw my family doctor and told him what happened. He goes, “Well, maybe you have a tumor.” I mean, he finds it amusing. I find I'd like to get revenge on a lot of people. I'm very leery of people. I can be very trusting and then I always get shit on.

DI-1TH.: It seems to me that your reactions to what are obviously unfair, pretty well outrageous comments by a professional is, you could say, well, just a reaction to that situation, but it seems to me that you pretty well recognize, maybe you completely recognize, that your wish to kill them is more complicated than reacting to that person.

This male therapist responds to the emotionally charged material with a DI couched in intellectual terms and continues his move away from affect-laden material by bringing in a discussion of medication:

SSTH.: …Essentially, “You hurt me, screw you,” in reaction. What I think you're possibly doing is that as you get to feel somewhat better—whether you're starting medication or it's your psychological belief that it's the medication, or what it is—but the point is, as you are feeling better, you notice you are less vulnerable to that. You don't get drawn in and you feel better and you can stand back and let your daughter play. I think it is too early to know whether the problems you are having are related to the medication or are related to what you already know about the medication. Could well be one of the medications—and any medication, to get licensed, has to go through the test phase and they have to report it. So one of them they reported—and placebos—dizziness coming from the medication, 3%; dizziness from the placebo, 11%.

ASS

PT.: I was wondering, because, you know, if I hear something all of a sudden, I'm not a hypochondriac, I find that mentally I do take it on, and that was a concern.

The patient, who rates his early alliance as good (CALPAS=5.58), appears to be calmed and reassured by the therapist's intervention and moves away from his angry outburst.

A final illustration is seen in Mr. B.'s third therapy session. This patient, as noted earlier, is fragile and had previously been suicidal. In this session, he brings emotionally charged material carrying themes of death and hopelessness, to which his male therapist responds with distraction in the form of a question:

PT.: But I, on the other hand, I try and—I complain to myself a lot. I'll say or I'll feel what I have to feel and it deadens me after a period of time, a long period of time, to the point where I almost just don't want to bother, because ultimately I knew then and there that if I confronted the issue with that particular person on the moment, it would have been a waste of time in any case. I believe that I could have walked out of the hospital on June 14 and been a totally new person on June 15 because of my own special personality because of me being me, and yet I know that there are people, when I walk in here, there are people that will be probably walking up and down the steps for the next 20 years and won't change.

QTH.: Do you know your I.Q.?

PT.: I don't know, I really couldn't tell you. You know, the idiot savant.

The question serves to distract the patient away from his pessimistic and futile mood and appears to calm him. He continues in the session with affect that he can tolerate, despite the superficial content of the ensuing discourse. The early alliance in this dyad is once again preserved.

Affect-Probing Style:

With Mr. F., his female therapist adopts a different approach in the early stages of therapy. In session 3, when painful affect is defended against, the therapist confronts it directly. The patient is asked to elaborate further on his painful emotions as the therapist empathically offers an opportunity to normalize these feelings that are being avoided:

DI-1TH.: Well, you're talking about what this means for the organization. But I imagine what must be very difficult to metabolize is what this means at a personal level.

PT.: Well, obviously I think that—what I think is upsetting to me is the way this whole thing was done when I was not there.

The patient goes on to describe his emotions with more feeling and less intellectualization and displacement. Later in the same session, we see further probing for painful affect:

DI-1TH.: So you're going back again in the organization—as opposed to the enormity of what this must mean on a personal level, given that this isn't the first time that something like this has happened, where you sort of played by the rules, the way it's supposed to be, and then suddenly you discover that other people are on a different planet with respect to the kind of rules that.—but you know this is not the way it is supposed to be when you worked many years for the organization.

PT.: Of course. I mean, I think—I feel very badly for myself. I mean, I'm very depressed today, there's no question. I guess I'm trying to rationalize it in my usual fashion. I'm trying to deflect the—

DI-3TH.: It must stir up a lot of stuff about your father.

Later still, the therapist asks the patient to deepen elaboration of his painful feelings:

DI-1TH.: When you talk about being depressed, how is that—

PT.: Well, I—

DI-1TH.: How is that? What's that?

PT.: Well, what do I feel like now?

ATH.: Uh-huh.

PT.: Well, obviously I—I, uh—I feel like that there's like a—a certain sadness that I feel. I feel a certain aura, like a certain weight on my shoulders.

In this excerpt, direct yet empathic confrontation of dysphoric affect through repeated defense interpretations brings on good emotional elaboration and therapeutic work and takes the patient away from his more typical obsessional defenses of intellectualization, undoing, displacement, and rationalization. However, this subject presents with less severe pathology than others in this sample. At intake, he manifested obsessional personality traits and a tendency to somatize. He initially presented with dysthymia and a GAF of 72. The affect-probing and highly interpretive style that his female therapist adopts may be guided by the patient's pre-treatment characteristics rather than by therapist gender. At 5.5 years, this patient still remains in therapy with a yearly session density of 41.

Another example of this style aimed at uncovering affect is found in Mr. G.'s fifth session with his female therapist. Contrary to the previous patient, Mr. G. has significant pathology. He has mixed passive-aggressive and self-defeating personality disorder as well as narcissistic, antisocial, borderline, and dependent traits. He presented to therapy with dysthymia and generalized anxiety disorder and has moderate difficulty in overall functioning (intake GAF=55).

PT.: …It was really tough.

DI-1TH.: But, you know, here is somebody that, after all, you were looking forward to a very serious relationship. And yet, what do I hear when she tells you that she's staying with her husband? I don't hear anything about how much you cared for her, or how much you loved her, or the fact that you're going to miss her. What I hear is, you know, she led me a merry chase. I, I did all this for her. I was so supportive. I went with her. I did for her. And now, at the eleventh hour, she turns around and says she's staying with her husband. So, uh, you know, it's more, it sounds to me it's not so much that you were in love with N. But that you expected her, you expected some return on your investment.

Here the therapist pursues affect directly. She sympathizes and empathizes with the feeling the patient should appropriately be having. This has the effect of normalizing it and permitting it, all the while delivering a defense interpretation. The patient responds to this intervention with some limited emotional elaboration and rates the early alliance as good (CALPAS=5.9):

PT.: Well, I'm keeping it cold in that way because I'm very upset inside. I'm a real softie, hard shell but soft inside. And I take it very hard.

Mr. G. remained in therapy for 13 months and stopped when he moved to another city. He had received a total of 46 sessions, yielding a yearly session density of 43.

DISCUSSION

Ever since Luborsky and colleagues²⁶ published the paper widely known as “Everybody Has Won and All Must Have Prizes,” there has been a debate about the relevance of specific interventions and specific styles of psychotherapy. Despite conflicting messages in the field of psychodynamic psychotherapy regarding the use of early transference interpretations, some consensus that they are high-risk and possibly high-gain interventions²⁷ exists. There are inevitably times when patients present transference issues early in therapy. They may involve “therapy-interfering behaviors” that both Linehan and Kernberg would agree require some confrontation.^28,29 Thus the therapist may not have any choice but to address issues that are impinging on the patient–therapist relationship. While remembering that Piper et al.³⁰ found an inverse relationship between the proportion of transference interpretations and both therapeutic alliance and favorable therapy outcome for patients with a high quality of object relations, the therapist must also heed Linehan and Kernberg's recommendations. Then the question becomes how best to deal with early transference. This series of case studies offers some guidance.

Early in their therapy, most male patients in our sample were expressing ambivalence toward the therapist, especially through defensive aspects of closeness to and dependency on the therapeutic relationship. Our findings indicate that transference interpretations alone, even at level 2 on the PIRS, stimulate anxiety and threaten the moment-to-moment alliance, as seen by patients' increased defensiveness. These do not have to be interpretations of the negative transference in order to be threatening. This was illustrated by Mr. A.'s denial that he might feel pressured by the therapist's invitation to increase the session frequency to twice per week and by Mr. B.'s vehement disavowal that he was referring to the therapy when expressing his frustration at hanging onto things rather than discussing them “when it's happening” and not “three days later” when “it's totally out of context.” Therapists usually followed up by trying to attenuate the anxiety with questions, explanations, and support strategies or with defense interpretations.

Failure to offer transference interpretations in the face of negative transference also appears problematic for the alliance. It is as if the therapist is avoiding acknowledging an issue within the dyad, and the patient consequently starts distancing himself. This process could be seen with Mr. D.: his discourse manifested themes of abandonment, isolation, and withdrawal after the therapist did not draw parallels between the patient's intense ambivalence toward his ineffectual and unsupportive boss and the possibility that he might be feeling alone and unhelped by the therapist and by the frame in therapy.

In our sample, early transference interpretations, when immediately followed by defense interpretations, preserve the alliance. Here the effect is to outline the patient's ambivalent feelings toward the therapist and then to point out that this pattern seems to exist outside the therapeutic dyad as well, in the patient's relationships with loved ones. Mr. C. manifested a negative transference: namely, his wish to be the passive recipient of the therapist's insight and work on the one hand and his anger about feeling controlled by the therapist on the other hand. The therapist remarked on this conflict within the therapeutic relationship and then broadened it with a defense interpretation that addressed this same relationship pattern between the patient and his wife. This sequence of interventions diffused the anxiety associated with the early transference interpretation, preserved the alliance, and allowed for elaboration and deepening of the conflictual material brought by the patient. Bales's³¹ “equilibration” model suggests that the therapeutic environment must promote disclosure and allow for the reduction of anxiety, with patient and therapist both participating in the equilibration process. Perhaps directing the discussion away from the transference early on helps this equilibration, enhancing the alliance and reducing the tension in the therapeutic relationship.

In these early-session transcripts with personality-disordered male patients, the intervention that seemed best at enhancing alliance in a nonthreatening way was the defense interpretation. As seen with Mr. A., Mr. G., and Mr. F., as long as defense interpretations were offered from a stance of support and empathy, they helped the patient to further express emotions and wishes with which he was struggling, but kept the focus out of the room so that the therapist and patient shared an alliance. In support of this finding, Frances and Perry³² point out that many patients relate better to interpretations regarding third parties and find transference interpretations contrived or irrelevant. Milbrath et al.,²⁰ in a sequential analysis of therapist interventions and patient disclosures using single therapy-hour transcripts from 20 patients, found that therapist use of defense interpretation resulted in patient disclosure of emotional material. Looking at the impact of interventions on outcome, they found a direct association between defense interpretations and patient self-report of symptom reduction 5 months after completion of therapy. Interestingly, there was no significant association between transference interpretation and outcome in their study.

All seven subjects in our sample had a high early alliance as measured by the CALPAS. Still, the alliance deteriorated following both transference interpretations alone and missed transference interpretations. The alliance was, however, preserved by the sequence of early transference interpretation followed by defense interpretation, and it was enhanced by the judicious use of defense interpretations. Regarding valence of transference interpretations, few if any therapists interpreted a truly positive patient–therapist relationship in these early transcripts. However, any mention of this relationship early on, even in a neutral way, elicited resistance in the patient and had a negative impact on the moment-to-moment alliance.

In keeping with our findings, Milbrath et al.²⁰ found that transference interpretations were followed by dyselaborations that distorted the significance of the interpretation or intervention (Z=3.13, effect size=0.55). This component of the sequence probably reflects the anxiety that immediately follows when the therapist broaches an issue in the therapeutic relationship. However, the full sequence was this: patient elaboration, followed by transference interpretation, followed by patient elaborating on significance and then a brief period (1 to 5 thematic units) of distorting significance.

Two distinct styles of how therapists handle the period of anxiety and manage the alliance following early interpretations emerged from our sample. A first group of therapists tended to move into a noninterpretive and somewhat intellectual mode, using questions which at times encouraged topic change and also using work-enhancing strategies and support strategies. Overall, these therapists tended to focus away from affect. A second group tended to use fewer transference interpretations in general yet more defense interpretations, and they “kept the heat on” with continued defense interpretations, empathically probing and confronting further when affect-laden material emerged.

Although both strategies were effective at maintaining the alliance in this personality-disordered male sample, almost all (4 of 5) male therapists endorsed the first style, whereas both of the female therapists (2 of 2) adopted the latter. With such a small sample, we cannot draw conclusions linking style of managing the alliance with therapist gender, especially given the potential confound of severity of patient psychopathology. Indeed, the female therapists' use of this affect-probing style may have been guided by their intuitively sensing that their patient was healthier and could tolerate this work. Similarly, the male therapists in our sample may have colluded with their male patients and avoided affect because their patients seemed fragile. This remains nonetheless an interesting finding at a hypothesis-generating level and merits further study.

Literature on the potential effects of therapist gender on technique, especially with respect to the caretaking of the early therapeutic alliance, is scarce. On the topic of gender and psychotherapy process and outcome, the focus has been mainly on female patients and whether male or female therapists are best suited to treat them.^33–35 Little has been written on male patients and the impact of interpretive work on the alliance. Moreover, in the existing literature few authors seem to agree. In the National Institute of Mental Health Treatment of Depression Collaborative Research Project, Zlotnick and co-workers³⁶ found that neither therapist nor patient gender was significantly related to measures of treatment process or outcome for subjects receiving either interpersonal therapy, cognitive-behavioral therapy, imipramine plus clinical management, or placebo plus clinical management. Their findings were unchanged after controlling for modality of treatment and for pre-treatment depressive symptom severity. Similarly, Mogul³⁷ found that “no specific conclusions as to optimal patient–therapist matches on the basis of therapist sex appear warranted.” Jones et al.,³⁸ however, found that women therapists rated themselves as more “successful,” especially with female patients and that patients of both genders rated their alliance with female therapists as stronger than with male therapists. They also found that despite male therapists' descriptions of their patients “in less socially desirable terms” than those used by female therapists, their patients, both male and female, reported significant improvement with therapy.

The main limitations in this study include our small sample size of 7 males and the variability in patients' levels of pathology and in diagnoses, especially on Axis II. It may be difficult to generalize our findings to a particular patient profile or diagnosis, but the findings probably reflect more accurately the patient population that comes for psychotherapy: a heterogeneous mix of mostly Cluster B and C personality-disordered patients with likely several comorbid Axis I disorders, most often mood and/or anxiety disorders. Another limitation is that we did not formally assess accuracy of interpretations in our study. One could argue that inaccurate interpretations, or those with poor timing, would likely have a negative impact on the alliance, precluding any useful conclusion about the role of interpretations. Furthermore, the PIRS does not allow for rating valence of transference interpretations as either positive or negative. Finally, our method of assessing moment-to-moment change in the alliance within the session in addition to using a recognized standard instrument, the CALPAS, may be seen as lacking a systematic approach. Our use of consensus among three judges was an attempt to address this limitation.

In summary, we conclude that like female patients, male patients with personality disorders are highly sensitive to early transference interpretations, whereas defense interpretations are a low-risk, high-gain intervention. Both positively and negatively valenced transference interpretations seem disruptive. However, when a patient manifests early ambivalence toward the therapist or the frame in a fashion that may threaten the viability of the treatment, this negative transference needs to be interpreted. In this instance it seems that either offering transference interpretations alone or avoiding transference interpretations altogether will likely threaten the alliance. Here, accurate transference followed by defense interpretations within a context of support seems best suited to maintaining the alliance. The context of support is crucial for both male and female patients so that interpretations will enhance the alliance rather than disrupt it. Another finding, to be viewed as very preliminary and interpreted with caution given our small sample size, was that male therapists tend to manage the alliance with a somewhat intellectualizing style while female therapists tend to pursue further the underlying affect with male patients. We acknowledge that these findings will need replication in a larger sample covering a variety of diagnostic categories.

Appendix.