Review by:
Faisal Roberts M.A.
WKIPC Psychology Intern
The presence of a personality disorder (PD) can profoundly impact an individual’s quality of life in addition to the management of comorbid mental health issues, therefore screening for PDs should be an integral part of the mental health evaluation process. Although somewhat subjective and imperfect, standardized clinical interviews (SCI) are currently considered to be the most reliable and valid methods available to screen for PDs. However, SCIs can be time consuming. While self-report instruments can be effective regarding efficiency and time conservation, the drawbacks are that a self-report inventory may have relatively poor specificity (bereft of elaboration from a clinician), the patients must possess, at minimum, a fundamental reading level, and the possibility of patient fatigue due to having to read and concentrate during the self-report assessment. The authors of this article suggest a compromise between an SCI and a self-report assessment in the form of a brief structured interview.
For this study, the authors employed the Quick Personality Assessment Schedule (PAS-Q), which is a brief structured interview that takes approximately 15 minutes to complete. The PAS-Q begins with open questions regarding character traits, personality traits, interpersonal relationships, occupational performance, substance use issues, and legal history. The next area, comprised of eight general sections, assesses constructs relevant to PDs: 1) Suspiciousness & Sensitivity (Paranoid PD); 2) Aggression & Callousness (Antisocial PD); 3) Aloofness & Eccentricity (Schizoid PD); 4) Impulsive & Borderline (Borderline PD); 5) Childishness & Lability (Histrionic PD); 6) Conscientiousness & Rigidity (Obsessive Compulsive PD); 7) Anxiousness & Shyness (Avoidant PD); and, 8) Resourcefulness & Vulnerability (Dependent PD). In order to identify a PD each section begins with two screening questions; positive responses to these screening questions leads to additional exploratory questions probing for PD symptoms, leading to scoring the characteristics in question. The intervieweer not only uses the information obtained from the PAS-Q, but also relevant historical/background information from the patient. The PAS-Q is scored according to four levels of severity ranging from 0 to 3: 0 = no severity; 1 = personality difficulty; 2 = simple PD; and, 3 = diffuse or complex PD.
The present study focuses on the validity of the PAS-Q. The purpose of examining the PAS-Q was derived from the following considerations: 1) the PAS-Q is based on the universally accepted ICD-10 categories (as opposed to the majority of the available PD screeners, which are predominantly based on the DSM classification system); 2) the PAS-Q does not focus on the prediction of any PD (as the majority of PD screening instruments do), but provides the opportunity to obtain more specific prognoses of distinct PDs; and, 3) the PAS-Q response scales are not limited to a simple dichotomy (i.e., absence or presence of PD symptoms) but instead allow for increased nuances corresponding with level of severity. The researchers chose the Structured Clinical Interview for DSM-IV – II (SCID-II) to serve as the basis of comparison as it is internationally the most widely use and best known measure to assess for PDs (the SCID-I examines Axis I Disorders, while the SCID-II examines Axis II disorders–which includes PDs).
Materials and Methods
For this study, the researchers randomly recruited 207 participants from a large community mental health center in the city of Tilburg, the Netherlands. However, 12 participants dropped out during the study. Of the 195 participants that completed the study, 112 were female (57.4 %) and 83 were male (42.6 %). The mean age of the participants was 32.7 years. The researchers utilized both the PAS-Q and the SCID-II in order to evaluate the participants. The PAS-Q was completed first; subsequently the SCID-II was completed 1-2 weeks later. The PAS-Q was then completed a second time 2-3 weeks later. The same clinician evaluated all the participants in order to eliminate extraneous variables regarding evaluator differences. The test-retest reliability of each item on the PAS-Q, in addition to the overall score, was estimated using Pearson correlation coefficients. The dimensionality of the PAS-Q was assessed using factor analysis. The effect of changes in the cut-off score of the PAS-Q for the purpose of predicting SCID-II diagnoses were assessed using receiver operating characteristic (ROC) analysis.
Results
Although the study began with 207 participants, 12 dropped out, resulting in 195 participants that completed the study. Based on the SCID-II, a total of 97 of the 195 (50 %) participants received a PD diagnosis. In the group of participants with PD, the mean number of PDs was 1.8. The test-retest coefficient for the total score yielded a high score of 0.92. The section of Aloofness & Eccentricity had the lowest stability; the sections of Aggression & Callousness, Borderline, and Childishness & Lability had the highest stability over time. Overall internal consistency, as reflected by Cronbach alpha coefficient, for the total PAS-Q scale was 0.35. Internal consistency coefficients were low, ranging from 0.16 (Borderline) to 0.47 (Conscientiousness & Rigidity). These scores are suggestive that a high degree of heterogeneity exists between the different sections. The scores of the factor analysis were as follows: 0.43 (regarding the positive connections between Aggression and Impulsiveness & Borderline), 0.50 (regarding Resourcefulness & Vulnerability and Anxiousness & Shyness), and 0.40 (regarding Aloofness & Eccentricity and Suspiciousness and Sensitivity). The ROC analysis was used to determine the effect of the changing cut-off score on the PAS-Q in predicting a SCID-II PD diagnosis. The ROC scores, as demonstrated graphically by a curve (the ROC curve), had an area-under-the-curve of 83 % (with a 95 % confidence interval). This is stating that the cut-off score correctly identified 81 % of the participant pool as correctly having a PD.
Discussion:
In 81 % of the cases the PAS-Q was able to correctly identify the presence of a PD. The researchers state that its low overall consistency should not be interpreted that the PAS-Q is a test that performs poorly. The researchers suggest that latent variables between the sets of items may be implicated in the low homogeneity of the sections. Overall, the researchers were pleased with the outcome of the PAS-Q, believing that it can be a useful tool to identify PDs in adult psychiatry. They suggest that patients that receive a score of 2 (or higher) should be interviewed detailed structured, or semi-structured, interview for PDs.
A perceived limitation of the applicability of the study (regarding use in the United States) is that the PAS-Q only assessed for 8 of the recognized 10 personality disorders from the DSM classification system. Although this is not considered a limitation of the study itself, since an objective of the study was to assess the validity of an instrument grounded in the ICD-10 classification system (and it accounts for the eight primary PDs recognized by the ICD-10). The authors also did not disclose the success rate of the comparative method, the SCID-II. The data regarding which of the participants had a PD was already obtained as all of the participants were preexisting members of the community mental health agency. Therefore the success rate of 81 % from the PAS-Q was held against the prerecorded diagnoses of the patients from the mental health clinic. The article did not mention the success rate of the SCID-II (unless it was to be assumed that the SCID-II had a success rate of 100 % since that was, presumably, the method in which the mental health clinic obtained their diagnoses in the first place). Finally, while the fact that a single interviewer conducted all the interviews is considered a strength of the study, it can also simultaneously be considered a weakness due to time constraints. The clinician conducted all the interviews was forced to conduct a high number of interviews in a relatively low amount of time, therefore some interviews may have been rushed, in addition to the fact that the participants’ background information was not reviewed for any of the cases.
Germans, S., Van Heck, G., Hodiamont, P. (2011). Quick Personality Assessment Schedule (PAS-Q): Validation of a brief screening test for personality disorders in a population of psychiatric outpatients.
Australian and New Zealand Journal of Psychiatry, 45, 9, p 756-762