Evaluating negative process: a comparison of working alliance, interpersonal behavior, and narrative coherency among...

The aim of this study was to investigate the interrelationships of three measures of the therapeutic relationship and their validity in predicting treatment outcome, including the early identification of two treatment-failure conditions. Forty-eight patient-therapist dyads, in 30-session therapies for personality-disordered patients, were classified as premature dropout (DO), poor outcome (PO), or good outcome (GO) cases. Poor and Good Outcomes were determined by a reliable change score. Dropout cases were terminated during the first third of treatment, and patients cited dissatisfaction with the therapy or therapist. Assessment of working alliance, interpersonal behavior and a new measure of narrative coherency in the first third of treatment revealed that DO dyads had significantly poorer alliances and less coherent narratives in early sessions, while PO dyads, who ultimately completed the 30-session treatment protocol, unexpectedly demonstrated the highest degree of hostile complementarity. Clinical implications of the results are discussed.