Personality Assessment Inventory Clinical Scales in Relation to Patient and Therapist Rated Alliance Early in Treatment

Abstract

We examined relationships between the Personality Assessment Inventory (PAI) clinical scales (e.g., Somatic Complaints [SOM]) and subscales (e.g., Conversion [SOM-C]) with patient- and therapist-rated alliance early in treatment (third or fourth session). We also replicated and extended findings from a previous study examining PAI treatment scales (Treatment Rejection, Treatment Process Index) and early session therapist-rated alliance. We used PAI protocols from a clinical outpatient sample ( N = 84). Data were analyzed using stepwise linear regressions. Results suggest that patients who report lower early session alliance also report more antisocial features (β = −.219, p = .050, f2 = 0.05) specifically more antisocial behaviors (β = −.315, p = .004, f2 = 0.11). Additionally, therapists report higher early session alliance with patients who report more anxiety-related disorders (β = .274, p = .012, f2 = 0.08), specifically traumatic stress (β = .325, p = .003, f2 = 0.12). No significant relationships were found between patient- or therapist-rated alliance and Treatment Rejection and Treatment Process Index, consistent with prior findings. Clinical implications are discussed.

Personality assessment inventory items in relation to patient‐ and therapist‐rated alliance

“…In this case, it appears that groups of items emerged around similar levels of significance as well son who pays a lot of attention to detail;" β = À.225)" was the only trending p value at .099. One of the two most significant items in this model is from the ARD-T subscale, which also aligns with our prior finding that ARD-T clinical subscale was significantly related to therapist-rated alliance early in treatment (Cersosimo et al, 2021).…”

“…A subsequent backwards linear regression exploring RXR items in relation to early session patient-rated alliance also produced a nonsignificant model (F(1, 67) = 1.087, p = .301, R 2 = .016. In our earlier work (Cersosimo et al, 2021), analyses from a bivariate Pearson correlation matrix yielded no significant findings between RXR and Therapist-Rated Alliance (WAI-T) (r = À0.008, p = .939) and RXR and patient-rated alliance (CASF-P) (r = À0.036, p = .751). This prior full-scale nonsignificant finding is consistent with the current item-level nonsignificant finding at the third or fourth session in treatment.…”

“…Thus, generally speaking, ANT-A items, "Antisocial Behaviours," were among the most significant, followed by ANT-S, "Stimulus Seeking" and ANT-E, "Egocentricity." In part, this maps onto findings from our prior work, which found the subscale of ANT-A to be most significantly, and negatively, related to patientrated alliance early in treatment (Cersosimo et al, 2021). In the discussion section, we will elaborate further on mix of positive and negative item relationships that emerged.…”

“…• Findings demonstrate that a unique constellation of antisocial (ANT) and anxiety-related disorders (ARD) items is related to patient-and therapist-rated alliance early in treatment In our previous study (Cersosimo et al, 2021), we found that patients who reported lower early session alliance also reported more ANT, specifically more antisocial behaviours. Additionally, therapists reported higher early session alliance with patients who reported more ARD, specifically those with traumatic stress.…”

“…Additionally, no studies have attempted to extend this work or develop a specific PAI index predictive of alliance early in treatment despite alliance being a robust predictor of treatment outcome (Flückiger et al, 2018). The current study aims to fill this gap by expanding on our previous work (Cersosimo et al, 2021) to further…”