Published online by Cambridge University Press: 13 June 2014
Objective: It is now widely recognised that personality disorders play a significant role in psychiatric practice. This study was set up to examine the diagnostic accuracy of the assessment of personality disorder in normal practice by a multidisciplinary community mental health team when compared with a formal assessment using a structured interview.
Method: Forty-four patients originally assessed by team members of the Early Intervention Service (EIS), were re-examined using the Personality Assessment Schedule (PAS ICD-10 and original version) at a mean period of 16.3 months after first assessment. Specifically we tested the extent to which three personality disorder (PD) clusters (flamboyant, odd/eccentric, anxious/fearful) could be differentiated from either no PD at all or from the remaining PD clusters.
Results: As expected, both flamboyant and odd/eccentric PD clusters were much more easily differentiated from no PD at all (overall accuracy of 87 % and 85 % respectively) than from other PD clusters (corresponding accuracy of 65 % and 50%) . In distinct contrast, the anxious/fearful cluster was much less accurately differentiated from no PD (accuracy of only 52%) than from other PD clusters (accuracy of 85%).
Conclusion: The diagnosis of personality disorders in routine clinical practice is fraught with difficulties, especially when mental status symptoms closely resemble personality traits. The present results suggest that greater knowledge, training and experience can facilitate diagnostic accuracy.