MARCH 19TH | 4:00 PM - 5:30 PM (ET)

Session Abstract:In the past decade, the psychopathology research literature, including that on psychosis, has emphasized deconstructing disorders into dimensional components. The first presentation by Mihura provides an overview of the key psychosis components and how R-PAS scores map onto those components, summarizing the literature to date. The 2nd presentation by Mihura et al. reports results from (a) 3 datasets towards developing a streamlined version of the R-PAS Cognitive Scores specifically targeting the DSM-5 Disorganized Thinking criterion and (b) a literature review differentiating the Cognitive Scores associated with schizophrenia versus mania. The 3rd presentation by O’Gorman et al. uses a sample of 186 psychiatric patients to cross-validate the Scales of Problematic Communication and Thinking (SPCT), a dimensional method of coding disordered thought processes in Rorschach responses. Based on findings of a previous study using the measure, it is anticipated that the SPCT will provide incremental validity over the standard R-PAS Cognitive Codes for predicting diagnostic severity. The 4th presentation by Buckingham et al. uses a sample of 70 VA patients with schizophrenia, schizoaffective, and major depressive disorder to assess the magnitude of relationship between measures of disorganized thinking, negative symptoms, hallucinations, and delusions on the BPRS and the SCID-P with R-PAS measures targeting those constructs. Analyses were conducted to determine if SPCT dimensions provided incremental validity of predicting disorganized thinking over the traditional R-PAS measures. Using data from a maximum-security inpatient forensic hospital (N = 88), the final presentation by Crittenden et al. evaluates whether automated coding of discourse coherence (using Coh-Metrix) predicts clinician ratings of disorganized thinking on the PANSS and, in addition, provides incremental validity over R-PAS Cognitive Codes and the SPCT.

Chair Information: Joni Mihura, PhD, ABAP | University of Toledo

Discussant Information: James H Kleiger, PsyD, ABPP, ABAP | Private Practice, Bethesda, MD

Presentation 1 Title: Deconstructing Psychosis: The Role of Method in Assessing Psychosis Dimensions

Presentation Abstract: The first presentation by Mihura sets the stage for the subsequent presentations describing the empirical support and nonsupport for using the Rorschach to assess psychosis. This is a particularly timely topic, because the psychoticism dimension is the third most supported dimension in the Hierarchical Taxonomy of Psychopathology (HiTOP) following the internalizing and externalizing dimensions. Assessing the psychotic phenomena in this psychoticism spectrum require multiple methods of assessment. For example, although self-report is the most common assessment methods taught in doctoral programs and can save clinicians’ time, relying on self-reported information alone significantly overestimates psychosis prevalence (Fusar-Poli et al., 2017; Schultze-Lutter et al., 2014). In fact, particularly for disorganized thinking, studies find that associations between self‐reported and clinician‐rated symptoms hover around zero (Becker et al., 2012; Chan et al., 2019; Cochrane et al., 2010; Moreno-Izco et al., 2015; Sumner et al., 2020). This is not to say that the self-report method of assessment is not important in assessing psychotic phenomena, it is essential to assessing the patient’s phenomenological experiences that include the auditory, visual, and other sensory experiences of hallucinations as well as to know what thoughts a person is thinking and how strongly they hold beliefs, and the degree to which the psychotic experiences cause the patient distress. In general, assessing psychotic phenomena requires behavioral observations and the clinical judgment of an experienced assessor as to whether the symptoms meet criteria for psychosis—whether beliefs are in fact delusional; perceptual experiences qualify as true hallucinations; speech and motor behavior are disorganized; and the patient’s motor, vocal, verbal, facial, and other modalities of expression meet criteria for negative symptoms.
This presentation describes how specific methods of assessment map onto the psychotic phenomena with a particular focus on using the Rorschach as a measure to assess visual misperceptions and disorganized thinking that has meta-analytic support for interrater reliability and validity as well as age-based norms for children, adolescents, and adults, as well as cross-cultural replications. Mihura describes research on the Rorschach and visual misperceptions and how the response process fits the interpretation of this psychosis phenomenon. She then describes the R-PAS disordered thinking scores compares them to other standard measures of disordered thinking in the psychosis research literature such as the TLC and TDI. Research published since the Mihura et al. (2013) meta-analyses is also described, followed by the individual paper presentations on this topic.

Presenters: 

Joni Mihura, PhD, ABAP | University of Toledo

Presentation 2 Title: Deconstructing Psychosis and Cross-Validating the R-PAS Variables Targeting its Constructs

Presenters: 

Kirsten Buckingham | University of Toledo, Ohio

Gregory J Meyer, PhD | University of Toledo

Emily T. O'Gorman, MS | University of Toledo

Jeffrey S Kline, PhD | Menlo Park Palo Alto Veterans Administration Medical Center, CA

Joni Mihura, PhD, ABAP | University of Toledo

Presentation 3 Title: Cross-Validation of a Dimensional Approach to Assessing Disordered Thought Processes in Rorschach Responses

Presenters:

Emily T. O'Gorman, MS | University of Toledo

Larson E. Sholander, MA | University of Toledo

Gregory J Meyer, PhD | University of Toledo

Presentation 4 Title: Streamlining Rorschach Cognitive Scores to Target DSM-5 Disorganized Thinking: Results from 3 Datasets

Presentation Abstract: This presentation by Mihura et al. reports the results from 3 archival datasets (a forensic inpatients, a mixed nonpatient/patient sample, and a VA sample) towards developing a streamlined version of the R-PAS Cognitive Scores that specifically targets the DSM-5 Disorganized Thinking criterion. As part of this presentation, Mihura distinguishes between the different types of R-PAS Cognitive Codes (called Special Scores in the Comprehensive System) in the WSumCog and SevCog variables: Codes that target visual versus linguistic disordered thinking—the former (also called ‘combinatory’ scores) more common in mania—and the latter that map onto positive thought disorder or disorganized thinking. The end goal of this presentation is to identify the Cognitive Codes that target disorganized thinking in order to use R-PAS results to specifically target the DSM-5 Disorganizing Thinking criterion using standardized administration to gather the speech sample and age-based international norms to compare the patient’s results.

Presenters: 

Joni Mihura, PhD, ABAP | University of Toledo

Gregory J Meyer, PhD | University of Toledo

Angie Keene | University of Toledo

Joshua Eblin, PhD | Center for Forensic Psychiatry, Saline, MI

Jeffrey S Kline, PhD | Menlo Park Palo Alto Veterans Administration Medical Center, CA

Nicole Kletzka, PhD | Center for Forensic Psychiatry, Saline, MI

Presentation 5 Title: Does Adding Automated Analysis of Speech Cohesion to R-PAS Cognitive Codes Improve Our Ability to Assess Disorganized Thinking?

Presentation Abstract: Background- Automated speech analysis has shown to be highly beneficial for predicting transition to first-episode psychosis (FEP) and diagnosing psychotic disorders (Argolo et al., 2020; MacKinley et al., 2020; Gupta et al., 2017; Rosenberg, 1979). Simultaneously, little research has compared its utility to other standardized measures of disordered thinking (Minor et al., 2019). The Rorschach Performance Assessment System (R-PAS) provides a structured procedure for gathering a speech sample to Rorschach inkblots and meta-analytically supported reliable (Schneider et al., 2020) and valid (Mihura et al., 2013) measure of disordered thinking. In addition, a more recently developed dimensional measure of disordered thinking –the Scales of Problematic Communication and Thinking (SPCT)—is also used to code speech samples collected using the Rorschach. Although various Rorschach methods have been suggested to assess negative symptoms, no formal scores currently exist. Therefore, first, we plan to investigate whether automated speech analysis provides incremental validity over R-PAS and SPCT scores of disordered thinking in predicting disorganized thinking ratings on the Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Second, we will investigate whether an automated coding can target clinician ratings of negative symptoms.

Methods
Patients (N = 90) were recruited from a maximum-security inpatient forensic hospital, ranging in age from 19 to 80 years-old (M = 40) and predominantly male (89%). R-PAS protocols were coded for thought disturbance using the R-PAS language and reasoning cognitive codes and the SPCT. The patient’s primary clinician and psychiatrist provided PANSS ratings. The automated speech analysis software, the Natural Language Processing (NLP) Tools for the Social Sciences, will be used to code the protocols. The NLP tools target various measures of speech, including text cohesion, connectives, lexical diversity, and syntactic complexity. Hierarchical regression analyses will be used to assess whether the computer indices provide incremental validity over the R-PAS scores of disordered thinking and the SPCT ratings in predicating PANSS ratings of Conceptual Disorganization. Correlational analyses will be used to assess which relevant computer indices target PANSS negative symptoms.

Results
Rater agreement was excellent for the SPCT ratings was excellent (ICC = 0.97) and good for the PANSS ratings (ICC = 0.71). R-PAS Severe Cognitive Codes (SevCog) and the SPCT ratings were significantly correlated with PANSS Conceptual Disorganization ratings (r = 0.36, r = 0.38). SPCT ratings provided incremental validity over the R-PAS WSumCog score but not the hypothesized SevCog score. The last incremental validity step is to test whether Rorschach protocols coded by computer indices provide incremental validity over the SevCog score and SPCT ratings in predicting PANSS Conceptual Disorganization. In addition, we will test the correlations between the relevant computer indices and PANSS negative symptoms. Our final goal is to uncover which computer indices provided by the NLP tools specifically target the R-PAS cognitive codes.

Discussion
Our findings hope to provide clinicians with a more efficient and objective alternative to hand-scoring Rorschach protocols for disordered thinking as well as discover an automatic coding that targets negative symptoms. In all, we hope to contribute to the creation of an overarching framework for the use of automated speech analysis in the clinical assessment of emerging psychosis.

Presenters: 

Gregory J Meyer, PhD | University of Toledo

Emily T. O'Gorman, MS | University of Toledo

Joshua Eblin, PhD | Center for Forensic Psychiatry, Saline, MI

Laura K. Allenc

Isra Dar | University of Toledo

Angela Keene | University of Toledo

Omar R. Assaly | University of Toledo

Joni Mihura, PhD, ABAP | University of Toledo

One thought on “Deconstructing Psychosis and its Assessment with Relevant R-PAS Measures (1.5 CE Credits)

Leave a Reply