We set out to examine the concurrent validity of a mini-interview for personality disorder Structured Assessment of personality - Abbreviated Scale . In order to do this, we conducted a series of secondary analyses of data from a randomized controlled trial of personality disorder psychoeducation for substance use disorders. The subjects for this paper were patients randomized to the experimental condition in the study (n = 36), and 18 training cases. The 36 experimental cases were taken from consecutive admissions to the Centralized intake units where the randomized study took place. The 18 cases consisted of 10 cases who were admitted shortly before the trial began, plus 8 volunteers who were assessed for other purposes as part of their treatment for substance use disorders, who agreed to give consent for the use of their data for research purposes.
Subjects were approached by their caseworkers, and informed that they had the option to be assessed for personality disorders and other psychopathology as part of an ongoing study. Those who agreed were referred to a research technician, who explained the rationale of the study.
Subjects gave consent to participate in the study no earlier than 24 hours after being informed of the purpose of the study by the research technician. The subjects were told that the data collected for the study would be used for research purposes, and at the same time be used for their treatment. After assessment, subjects were first given an individual feedback about the results of the assessment, and then, if they expressed their interest in this, this feedback was passed on to their caseworker. Subjects were also informed that without their consent, no information from the interviews would be passed on to any third party.
The interviews were conducted on two different days. On the first day, the SAPAS, and all of the non-personality related instruments were administered, and on the second interview day, the AUDADIS, the PRISM and the NPI-16 were administered (see below for instrument descriptions).
The sample was 85% male, and the mean age was 32.5 (range: 19 to 54). Among the respondents, 74% had used alcohol in the last 30 days before interview, 66% had used cannabis, 9% amphetamine, 30% cocaine, 11% heroin or other opiates, and 6% benzodiazepines. All patients in the sample scored 3 or more on the Severity of Dependence scale, indicating substance dependence. The mean score for the Kessler 6 was 10.8 (range: 0 to 21), indicating elevated scores on depression and anxiety. The mean score for the ADD scale was 19.7 (range 2 to 34), where scores above 23 indicate likely ADD. The mean for the hyperactivity scale was 17.0 (range: 7 to 34) where scores above 23 indicate likely hyperactivity disorder.
The Structured Assessment of Personality Abbreviated Scale (SAPAS) is an eight-item screening interview for personality disorder . Each item is worded as a question to be answered with yes or no (e.g., item 1: "In general, do you have difficulty making and keeping friends?"). When the response is given that indicates pathology (i.e., yes to item 1), the interviewer must follow up by asking if that is true in general. With eight yes/no questions followed by up to eight follow-up questions, the SAPAS will normally be completed in less than a minute. As the SAPAS is a set of indicators covering multiple areas, it is not designed to be unidimensional. Rather, the SAPAS is designed to cover different areas of personality.
The Kessler 6 (K6) is a 6-item interview. Each question starts with the expression "How often in the past month did you feel ..." and offers specific symptoms such as "restless or fidgety," "nervous," and "so depressed that nothing could cheer you up?" The 5 possible responses range from "none of the time" to "all of the time" and are scored from 1 to 5; the items are summed to obtain a total score. The K6 has been found to correlate substantially with both the Comprehensive International Diagnostic Interview-Short Form and the World Health Organization Disability Assessment Schedule . Internal consistency alpha for the present sample was 0.79.
The Adult ADHD Self-Report Scale (ASRS)  is an 18-item self-report scale used to screen for adult symptoms of attention-deficit hyperactivity disorder. The ASRS has been found to possess excellent validity , including convergence with predicted genetic risk markers for ADHD . For the present sample, the internal consistency alpha was 0.84 for attention deficit symptoms and 0.74 for hyperactivity symptoms.
The Psychiatric Research Interview for Substance and Mental Disorders (PRISM) interview is a semi-structured interview assessing a range of disorders that are commonly co-morbid with substance use disorders. Each item is assessed by asking a question, and when positive answers are given, follow-up questions are asked to assess the severity and persistence of a symptom. Each item is scored as 1 (absent), 2 (sub-threshold) or 3 (clinically significant). The PRISM has demonstrated validity for antisocial and borderline personality disorder (e.g ). In the present study, we assessed the interrater reliability of 27 taped interviews. For individual criteria for borderline personality disorder, the average interrater agreement assessed as intraclass correlation ranged from 0.66 to 0.98. For the number of criteria satisfied, the intraclass correlation was 0.93. For individual antisocial personality disorder criteria, the intraclass correlation for individual items ranged from 0.63 to 0.98, and the intraclass correlation for number of criteria satisfied was 0.98.
The Alcohol Use Disorder and Associated Disabilities interview Schedule-IV (AUDADIS-IV) is a fully structured interview covering a range of disorders, including personality disorders. At the time of this study, only the proportions of the interview that covered avoidant, dependent, obsessive-compulsive, paranoid, schizoid, histrionic and antisocial personality disorder were published. The AUDADIS has demonstrated validity and reliability , although it has rarely been used in clinical samples. The AUDADIS asks one or more questions for each criterion for Axis II, and given an affirmative answer, the interviewer must ask "Did this ever trouble you or cause problems at work or school, or with your family or other people?" Since these questions must be answered yes or no, there is no clinical judgement with regard to the interview, and interrater reliability of the recordings was not assessed. Internal consistency alpha based on tetrachoric correlations for the present sample ranged from 0.69 to 0.94 for the AUDADIS scales.
The Narcissistic Personality Inventory-16 (NPI-16)  is an abbreviated version of the Narcissistic Personality Inventory. The NPI uses a forced-choice format with a narcissistic and a non-narcissistic response for each item (e.g., "I am apt to show off if I get the chance" and "I try not to be a show off"). The 16-item version was developed to capture the different aspects of narcissism measured by the original NPI, and has excellent convergent validity with the original version, and good predictive validity . Internal consistency alpha based on tetrachoric correlations for the NPI in this sample was 0.83.
For the patients (n = 54) who completed both the SAPAS and the full personality disorder assessment, we estimated the agreement between the SAPAS with the cut-off of 3 or more based on the original article . We also correlated the Spearman rank-order correlations between the SAPAS and number of personality disorder criteria by cluster (excluding schizotypal and narcissistic personality disorder), and for each individual criterion. We report correlations of 0.1-0.3 as low, 0.3-0.5 as moderate, and correlations >0.5 as large, following Cohen .
We conducted a series of linear regressions to assess the association between the SAPAS and number of personality disorder criteria controlling for various confounders, one for each cluster, and one for total number of personality disorder criteria. In each regression, the dependent variable was symptom count for personality disorders (by cluster, or in total). The covariates were age, gender, and severity of anxiety or depression symptomatology as measured by the Kessler 6 interview, and severity of attention deficit problems and hyperactivity as measured by the ADHD Self-Report Scale.
In terms of number of criteria, we used the sums of the PRISM borderline and antisocial and AUDADIS histrionic personality for cluster B pathology, the sum of avoidant, dependent and obsessive-compulsive personality disorder criteria for cluster C pathology, and the sum of paranoid and schizoid personality disorder criteria for cluster A pathology.
The NPI is not a diagnostic instrument per se, and while it has been shown to predict important indicators of narcissistic pathology, we did not include the NPI as a part of cluster B pathology in the analyses.
Danish IBRs do not evaluate studies that do not involve invasive procedures or the manipulation of pharmacotherapy or diet. Dr. Peter Ege, senior consultant of social medicine in the City of Copenhagen, did an informal review of the ethical implications of the study.