Patient personality dimensions, transference patterns and therapeutic alliance: An empirical investigation

Introduction: Transference (meant in this context, as the patient relational patterns expressed towards the clinician) and therapeutic alliance play a crucial role in the treatment of personality pathology. To date, no empirical study examined the association between these two dimensions of the clinical relationship and patients' personality maladaptive traits in psychotherapy. Methods: A national sample of therapists (N = 100) of different theoretical orientations assessed dysfunctional personality features of a patient in their care using a comprehensive and empirically grounded dimensional diagnostic approach from the Shedler–Westen Assessment Procedure-200 (SWAP-200). Moreover, they filled in the Psychotherapy Relationship Questionnaire (PRQ) to identify interpersonal patterns expressed early in treatment by the patients and the Working Alliance Inventory (WAI-T) to evaluate quality of therapeutic alliance. Results: Overall, the most severe and maladaptive dimensions of patients' personality were associated with more negative clinician–patient dynamics and poorer levels of therapeutic alliance in statistically significant and clinically relevant ways. Notably, the hostile transference was predicted by both SWAP Hostility and Psychopathy, whereas the SWAP Narcissism was the strongest predictor of the special/entitled transference. The latter was also predicted by SWAP Emotional Dysregulation; conversely, the SWAP Dysphoria was the most robust predictor of anxious/preoccupied pattern. The SWAP Schizoid Orientation and Psychopathy predicted avoidant/dismissing attachment pattern; moreover, they were strongly and negatively related to the SWAP Psychological Health that was the best predictor of positive transference and alliance. Conclusions: Findings support that therapists' careful understanding of patients' interpersonal ways during early treatment stages may meaningfully inform diagnostic and therapeutic processes.