M119. DECONSTRUCTING PSYCHOSIS AND CROSS-VALIDATING THE R-PAS VARIABLES TARGETING ITS CONSTRUCTS

BACKGROUND: Consistent with the contemporary literature that psychosis constructs are best represented as continuous syndromes, this study aims to determine if dimensional psychosis measures outperform traditional categorical measures, thereby improving detection of symptom severity. The Rorschach Performance Assessment System (R-PAS) contains meta-analytically supported internationally normed scales for assessing disordered thinking and reality testing that have been replicated in many countries. Given the literature trend of utilizing a dimensional approach when assessing psychosis, a dimensional R-PAS scale for assessing disordered thinking was recently developed. Therefore, it is important to determine if this new measure outperforms the traditional measure. We also attempt to replicate recent research by deconstructing the key components of psychosis (e.g., disorganized thinking, hallucinations, and negative symptoms) and evaluating the validity of the R-PAS measures designed to assess these constructs. METHODS: Our study uses an archival clinical sample of 70 male inpatients with schizophrenia, schizoaffective, and major depressive disorder (Mean age = 41.9, Range 20 to 63) that were collected as part of an IRB-approved research project. Two trained diagnosticians independently interviewed the patients using the Structured Clinical Interview for DSM (SCID) and blindly assigned diagnoses as well as Brief Psychiatric Rating Scale (BPRS) ratings. Interrater reliability of their ratings using ICCs will be computed. The Rorschach was administered and relevant R-PAS variables were scored by trained research assistants. The new R-PAS dimensional measure of disorganized thinking (SPCT) will be coded by the first author, and a subset of protocols will be blindly coded by the third author. Interrater reliability will be computed for all variables. RESULTS: First, we will conduct correlational analyses to test the relationship between clinician ratings of disorganized thinking (on the BPRS Conceptual Disorganization and relevant SCID criteria) and the traditional measure of disorganized thinking on R-PAS (WSumCog). We will then use hierarchical regression analyses to determine whether the new dimensionalized measure of disorganized thinking (SPCT) provides incremental prediction of the clinician ratings of disorganized thinking on the BPRS and SCID-P over the traditional R-PAS measure (WSumCog). To replicate previous research, we will test the relationship between negative symptom ratings (on the BPRS and SCID) and R-PAS measures of behavioral, perceptual, and emotional task engagement (Complexity & FQ-%). Further, we anticipate that clinician ratings of delusions (on the BPRS and the SCID) will correlate with R-PAS measures of inaccurate understanding of human intention and action (M-) and illogical thinking (SPCT Illogical Thinking subscale). Lastly, we use correlational analyses to test the relationship between clinician ratings of hallucinations (on the BPRS and SCID) and an R-PAS measure of visual misperceptions (FQ-%). DISCUSSION: Implications of this research provide additional validation for assessing key components of psychosis with a standardized internationally normed measure. Psychosis components (e.g., poor reality testing) limit the accuracy of patients’ self-reported symptoms and inflate rates of misdiagnosis; these R-PAS measures provide a framework for clinicians to behaviorally assess symptoms on a continuum ranging from nonclinical to severe psychosis-level disturbance. This research will aid in more accurate symptom assessment, thereby improving prognosis and treatment planning.