Objective: Personality disorders involve patterns of behavior, mood, social interaction and impulsiveness that cause distress to the person experiencing them as well as to other people in their lives. DSM-IV-TR Personality Disorder categorical classification includes ten different personality disorders defined as cluster A, paranoid, schizoid, schizotypal, cluster B, histrionic, narcissistic, borderline, antisocial, cluster C avoidant, dependent, and obsessive-compulsive. In clinical settings, diagnosing personality disorders is essential for planning treatment, especially in treatment non-adherence and resistant groups. There are no measurable advantages for clinicianbased interview techniques as being more reliable and valid than self-report questionnaires, and both of the techniques are used. We aimed to compare SCID-2 (Structured Clinical Interview for DSM-IV Personality Disorder) with self-report questionnaires PBQ (Personality Belief Questionnaire) and MMPI-PD (Minnesota Multiphasic Personality Inventory) to determine consistency of these diagnostic tools in this study.
Methods: The participating thirty-nine patients were hospitalized at Ankara Diskapı Yildirim Beyazit Training and Research Hospital and Etlik Expert Hospital of Psychiatry Clinic; they were older than 18 years, literate, and had been diagnosed with personality disorder based on SCID-2 interview. Psychotic disorders were excluded. Socio-demographic data form, MMPI-PD, and PBQ were used. For data evaluation, non-parametric tests were used for descriptive statistics and Spearman correlation analysis for the relationship between tools.
Results: Thirty-nine patients met the inclusion criteria; of these, 19 (48.7%) were female, 20 (51.3%) were male. Mean age was 38.1±11.08. SCID-II diagnoses were clustered A, B, C for statistical evaluation. We found cluster B (61.5%) and borderline personality disorder were the most frequent (13 patients- 33.3%). The results of correlation analysis were statistically significant between self-report questionnaires PBQ, MMPI-PD and SCID-II for the correlation between PBQ-histrionic subscale and cluster B (r=0.34) and correlation between PBQ-antisocial and cluster C (r=-0.34). Comparing SCID-II personality disorder clusters with MMPI- PD subscales, moderate significant correlation has been obtained between cluster A and MMPI-PD paranoia subscale (r=0.50). Statistically significant correlation was not determined between other subscales.
Conclusion: In this study we aimed to compare diagnostic compatibility between self-report questionnaires PBQ, MMPI-PD and semistructured interview SCID-II. In an epidemiological study in Turkey, the prevalence of cluster A was found to be 7.7%, cluster B 4.1%, and cluster C 6.5%; in an outpatient setting it is reported that using PBQ (Personality Diagnostic Questionnaire) found high frequency in cluster C and low in cluster B. Hayward at al. showed that the most common personality disorder was cluster B personality disorder in patients diagnosed with axis 1 disorders. Similarly we found cluster B higher; this might reflect the fact that the participants were inpatients and co-morbid with axis 1 disorders. In conclusion, correlation between self-report questionnaires and interview techniques has been found low and reached different results in diagnosing personality disorders. More comprehensive studies are required to develop diagnostic tools for personality disorders.