A Match Made in Heaven?: A Pilot Study of Patient–Therapist Match

Abstract

The authors report on a study of patient– therapist match in 50 psychodynamic psychotherapy dyads. Sixty-six percent of patients and therapists agreed about the quality of the match, with 58% of patients and 56% of therapists reporting that the match was positive. Positive match correlated with positive patient and therapist assessments about the progress and process of therapy, but not with perceived similarity of personal characteristics. Patients' and therapists' perceptions about their similarities and differences from one another did not correlate. This study suggests it is both possible and important to gather data from both patient and therapist when studying match.

The recognition that psychotherapy is a two-person system naturally created interest in the complex and multifaceted relationship between patient and therapist. The term patient–therapist match attempts to capture an important dimension of that relationship. The concept of match originated in the developmental perspective of “goodness of fit” between mother and child. Baby watchers believe that there is a strong correlation between goodness of fit and successful progression through the phases of separation and individuation and the process of identity formation. Conversely, problems in psychological development can arise when the fit is discordant. The concept of fit assumes that a child's intrapsychic development results from a highly interactive interpersonal relationship, as opposed to the child's passively absorbing the mother's influence.

As the psychotherapeutic relationship was reconceptualized as a two- rather than a one-person system, the concept of “goodness of fit” was adapted to the psychotherapeutic relationship and renamed patient–therapist match. As Kantrowitz¹ notes, “The similarities and differences between analyst and patient have an impact on whether the analyst can or cannot resonate with the patient, just as the similarities and differences between parent and child affect the synchrony between them” (p. 295).

Match is considered to be a “reality” that creates a context that significantly influences the subsequent development of all other dimensions of the therapeutic process. Often an idea of a fantasied match serves as the basis for the referral of a patient to a particular therapist. Thus, match is an autonomous reality that stands apart from the therapeutic process; however, it influences the development of other patient–therapist interactions considered critical to the therapeutic process (therapeutic alliance, transference, countertransference). If match is seminal in the development of key elements of the therapeutic process, this naturally leads to the hypothesis that “good” patient–therapist match correlates with therapeutic benefit. This hypothesis parallels the developmentalists' belief that goodness of fit between mother and child is necessary for healthy psychological development.

Although psychotherapists consider match to be distinct from such entities as therapeutic alliance, transference, and countertransference, the boundaries between these concepts are far from clear. In fact, although match is a concept that intuitively resonates for most psychodynamic clinicians, like many fundamental psychodynamic terms it has no consensual definition. If a core concept of psychodynamic metapsychology has no definition, clinical discussions that use the term are compromised.

While clinicians have been discussing patient–therapist match in conceptual terms, psychotherapy researchers have focused on gathering descriptive information. However, these studies have focused exclusively on match as it relates to the initial process of choosing a therapist rather than examining the evolution of match longitudinally. In a study by Alexander and colleagues,² patients were given the opportunity to choose between two therapists and selected the one they felt would be more helpful and whom they “liked” more. However, the study was not able to explicate the reasons that a patient liked a particular therapist, and it was unclear how liking the therapist related to the concept of being well matched.

Hollander-Goldfein and co-workers³ studied 97 patients who were given the option to choose among three therapists. They chose the therapist whom they believed to be the most competent and the most understanding and who demonstrated qualities that they wished to emulate. Neither demographic variables nor the patient's perception of similar personality characteristics significantly influenced the choice of therapist. An intriguing finding in this study was a significant correlation between the therapist rating the patient as likable and the patient subsequently choosing that therapist.

Grunebaum⁴ interviewed 23 psychotherapists to determine what they considered most important when choosing their own therapists. Most of the subjects had had more than one experience with psychotherapy. Characteristics such as warmth, supportiveness, and a nonjudgmental attitude were frequently cited, but the most important characteristic was the level of therapist activity in the session; therapists wanted therapists who would be active, who were “talkers,” and there was a clear aversion to the inactive, “blank screen” approach.

In a study looking for predictors of outcome, a cognitive match between patient and therapist (as assessed by the Interpersonal Discrimination Test) was associated with decreased dropout in the first 2 weeks of therapy.⁵ Furthermore, patient–therapist pairs that were cognitively matched appeared to have a faster rate of improvement in the first 12 weeks of therapy as assessed by the Global Assessment Scale. However, at 24 weeks cognitive match scores did not predict outcome.

Nelson and Stake⁶ studied the correlation between patient and therapist perceptions of the therapy relationship and personal qualities. Although some match between patient and therapist scores in “thinking-feeling” and “judging-perceiving” dimensions correlated with the patient's perception of a positive therapeutic relationship, once again the most powerful predictor of both patient and therapist mutual satisfaction was a high degree of therapist activity.

In a study of patient–therapist match in a sample of psychoanalytic training cases, Kantrowitz and colleagues^7,8 reported that patient–therapist match significantly influenced outcome in 13 of 21 cases. However, as Kantrowitz herself notes, this study is limited by several methodological problems, including small sample size, retrospective evaluation of match and outcome, and the concept of match being assessed only from the analyst's perspective, not the patient's. Perhaps the most important methodological problem was that Kantrowitz did not offer a definition of match.

In fact, a frequent problem with studies of patient–therapist match has been the lack of an a priori definition of what match is and—just as important—what it is not. For example, Kantrowitz says:

Match . . . covers a broader field of phenomena in which countertransference is included as one of many types of match. The individual history, characteristics, attitudes, and values of each analyst and patient predispose them respectively to certain countertransference and transference reactions. Match, however, also can refer to observable styles, attitudes and personal characteristics which are rooted in residual and unanalyzed conflicts, shared or triggered in any patient–analyst pair.⁷ (p. 895)

Such a broad definition fails to distinguish match from other phenomena such as transference, countertransference, and character armor. Thus, when two clinicians or researchers are talking about match they may use the same word, but they are unlikely to mean the same thing.

In this article we report a study in which patient–therapist dyads who were engaged in twice-weekly psychodynamic psychotherapy for an average duration of 1 year simultaneously reported perceptions of each other, the psychotherapeutic process, and the state of the match. The goal of this study was to assess the following questions:

1. Do patient and therapist agree about the quality of match?

2. Does positive match correlate with other variables?

3. Do patient and therapist respond similarly to questions about perceived similarities and differences?

Methods

All psychiatric residents at the New York State Psychiatric Institute/Columbia Presbyterian Medical Center were asked to participate in this study, which involved completing a self-report questionnaire. If the resident agreed to participate, he or she presented the study to psychotherapy patients in open-ended, twice-weekly psychodynamic psychotherapy and asked whether they would agree to be contacted by one of the investigators (A.D.). Patients and therapists were both informed that neither party would see the other's answers to the questionnaire. If the patient agreed to participate, patient and therapist simultaneously completed the questionnaire. Patient and therapist questionnaires were answered anonymously, but patient–therapist pairs were identified by a code for the purposes of data analysis.

Patients participating in the study are predominantly Columbia University undergraduate and graduate students referred by the student health service for long-term psychodynamic psychotherapy. Treatment at this clinic is free of charge. Patients meet twice weekly with therapists who are supervised by the faculty of the Columbia Center for Psychoanalytic Training and Research.

After the study was explained in detail, all participants signed written informed consent. This study was approved by the New York State Psychiatric Institute Internal Review Board.

The questionnaire elicited information in three areas:

1. Demographic variables.

2. Attitudes toward the therapy and the therapist (or patient).

3. Perceptions of similarities and differences between therapist and patient in terms of characterological traits such as humor and cognitive style.

Statistical Methods

Patient–therapist match was evaluated in terms of three questions:

1. Do patient and therapist agree about the quality of match?

2. Does positive match correlate with other variables?

3. Do patient and therapist respond similarly to questions about perceived similarities and differences?

Patient and therapist responses were compared by using Pearson's chi-square test of independence and Spearman's rho correlation. Logistic regression was applied when the predictive variable was continuous (e.g., age and duration of therapy). The spirit of the data analysis was exploratory. As such, no corrections for multiple comparisons were applied, and all patient and therapist responses were assumed to be independent, ignoring the fact that the same therapist might be treating up to 3 different patients. Significant relationships are reported by using an alpha level of P <0.05.

Results

Ninety-one percent (50/55) of patients and 94% (31/33) of therapists responded to the survey. Seventeen therapists were treating 1 patient, 9 therapists were treating 2 patients, and 5 therapists were treating 3 patients. The mean duration of therapy at the time of the study (± SD) was 11.6 ± 8 months.

Demographics

Mean patient age was 27.4 ± 5.6 years (range 19–41 years). Mean therapist age was 33.2 ± 3.1 years (range 27–40 years). Sixty percent (30/50) of patients and 61% (19/31) of therapists were female. The gender matches of therapist and patient were F/F in 21 cases, M/M in 10, F/M in 10, and M/F in 9. More than 85% of both patients and therapists were Caucasian. Eighty-one percent of patients had a college degree. Both patient and therapist reports of religious background were evenly distributed among Protestant, Catholic, Jewish, and none.

Do Patient and Therapist Agree About the Quality of Match?

Match was rated on a five-point scale ranging from “bad” to “excellent.” No patient or therapist rated the match as “bad.” In order to create cells of sufficient size for the purposes of data analysis, the categories were collapsed into positive (“good” or “excellent”) or negative (“problematic” or “good enough”). Patients and therapists concurred about the quality of the match 66% of the time (Pearson χ² = 7.2, P = 0.007; Table 1). In the 33 patient–therapist dyads who agreed on the rating of match, 20 pairs agreed that there was a positive match and 13 pairs reported a negative match. It is of interest that 58% of patients and 56% of therapists considered the match to be positive.

TABLE 1.

Does Positive Match Correlate With Other Variables?

Patient and therapist report of a positive match correlated positively with affirmative responses to the questions “Do you feel the therapy is progressing?” and “Do you share a sense of how to proceed?”, and with a rating of a high level of therapist activity during sessions. The robustness of these correlations was stronger for the patient group than for the therapist group.

Interestingly, female therapist gender correlated with patient, but not therapist, rating of positive match (P = 0.025). This finding is accounted for by the fact that 5 of 6 patients who rated the match as “problematic” had male therapists, while 6 of 6 patients who rated the match as “excellent” had female therapists. The 6 patients who rated the match as “problematic” included 3 males and 3 females, and the 6 who rated the match as “excellent” included 4 females and 2 males (Table 2).

TABLE 2.

The rating of positive match by either patient or therapist was not significantly correlated with a perceived similarity of personal characteristics such as cognitive style, sense of humor, political values, personality style, or hierarchy of personal values.

Do Patient and Therapist Respond Similarly to Questions About Perceived Similarities and Differences?

Patients and therapists responded similarly and positively more than 90% of the time to five questions:

1. Does the therapist like the patient?

2. Does the patient like the therapist?

3. Does the patient feel accepted and respected by the therapist?

4. Do therapist and patient agree on the kinds of changes the patient would like to make in therapy?

5. Do patient and therapist feel they are working together?

Responses to these questions were so strongly correlated with each other that these questions do not appear to assess distinct domains.

In seven questions, patients and therapists were asked to rate whether they perceived themselves as similar or different in personal characteristics, including:

1. Sense of humor.

2. Political values.

3. Personal values.

4. Cognitive style.

5. Personal style.

6. Sense of discipline.

7. Level of activity versus passivity.

In one question, “political values,” the number of responses was too small to allow for meaningful analysis. In the other six characteristics, there was no significant correlation between patient and therapist rating; that is, patient and therapist did not agree as to whether they are similar or different.

Discussion

Perhaps the most important result of this study was to demonstrate that it is feasible to collect data from both members of the patient–therapist dyad. Indeed, the results of this study indicate that it is not only possible but necessary to do so. In 33% of patient–therapist pairs, the rating of sense of match was discordant; that is, there was disagreement regarding whether the match was positive or negative. In addition, in the entire sample, patients and therapists did not give similar answers to the majority of questions other than match. Not surprisingly, patients and therapists do not perceive things the same way. To talk about match in the psychotherapy dyad with input from only half of the pair is like attempting to do couples therapy with only one person.

A rating of positive match correlated with positive responses to questions that assess therapeutic alliance, therapeutic process, and outcome. In addition, positive match was strongly correlated with a high level of therapist activity. This finding replicates the work of Grunebaum⁴ and Nelson and Stake⁶ and suggests that therapist activity is important in establishing a sense of match and of therapeutic alliance.

In contrast, match was not associated with a perception of shared personal characteristics. If a patient is engaged in treatment, the match is positive regardless of whether the patient and therapist perceive each other as similar. A “therapeutic” match, which is most likely really the therapeutic alliance, is not dependent on patient and therapist personality concordance.

Our findings must be interpreted in light of the fact that mean duration of treatment at the time of assessment of match was almost 1 year. Because we do not have longitudinal data on the evolution of match, we cannot exclude the possibility that in the early stages of treatment, perceived similarity of personal characteristics is correlated with a perception of match, or perceived differences with a bad or problematic match. In the latter circumstance, patients may well have dropped out early and therefore not been included in this study. However, it is striking that therapy continued among the 58% of patient–therapist pairs in which at least one participant rated the match as negative—and despite the fact that 43% of patients rated the match as negative. Although longitudinal studies of match are clearly needed, our data seem to suggest that the concept of match does not add to our understanding of the therapeutic relationship or process, at least in the midphase of long-term treatment.

If the concept of match is ever to be useful, the term needs to be defined, its assessment operationalized, and the phenomenon studied longitudinally and with data collected from both members of the patient–therapist dyad. Only in this way can we answer the question of whether the patient–therapist match is indeed a match made in heaven or a partnership created by the therapeutic process itself.