Abstract
We explored the relationship between severity of personality pathology, cluster type and therapeutic interventions (psychodynamic–interpersonal [PI] and cognitive–behavioural [CB]) in 76 outpatients across two early sessions (3rd and 9th) of psychodynamic psychotherapy, while accounting for patients’ baseline global symptom severity. Pretreatment personality pathology severity was assessed using the Personality Disorder Index (PDI), where DSM‐IV Axis II PD was assigned a value of 2, subclinical traits and features were assigned a 1 and absence of Axis II psychopathology was assigned a 0. Interrater reliability of personality pathology severity was excellent (ICC [1, 1]: 0.85). Interrater agreement for Cluster A (κ = 0.75), Cluster B (κ = 0.92) and Cluster C (κ = 0.70) was high. Interventions were coded with Comparative Psychotherapy Process Scale (CPPS) from videotapes, and reliability was excellent (CPPS‐PI = 0.86; CPPS‐CB = 0.78). Stepwise linear regressions indicated that therapists’ focus on mood shift/topic avoidance (B = 0.29, p = .009) and future events (B = −0.26, p = .020) predicted Axis II severity. Overall use of PI techniques and Cluster A personality disorder (CLA) were positively correlated (r = .312, p = .006). Stepwise binomial logistic regressions indicated that therapists’ focus on uncomfortable feelings (B = 1.915, p = .008) and explaining rationale behind approach (B = 1.276, p =. 038) predicted CLA. All results remained significant when controlling for patients’ baseline general symptomatology (Brief Symptom Inventory‐Global Severity Index [BSI‐GSI]), except for the relation between explaining rationale and CLA. Discussion highlights how using psychodynamic treatment model, therapists’ focus on patient’s in‐session affect expression and explaining rationale behind approach are highly relevant when working with CLA patients.