Revising and assessing axis II, Part I: developing a clinically and empirically valid assessment method.

OBJECTIVE

Personality pathology is difficult to measure. Current instruments have problems with validity and rely on a direct-question format that may be inappropriate for the assessment of personality. In addition, they are designed specifically to address current DSM-IV categories and criteria, which limits their utility in making meaningful revisions of those criteria. These problems suggest the need for consideration of alternative approaches to assessing and revising axis II.

METHOD

This article describes the development and validation of an assessment tool designed to allow clinicians to provide detailed, clinically rich personality descriptions in a systematic and quantifiable form (the Shedler-Westen Assessment Procedure, or SWAP-200). A total of 797 randomly selected psychiatrists and psychologists used the SWAP-200 to describe either an actual patient or a hypothetical, prototypical patient with one of 14 personality disorders (one of the 10 DSM-IV axis II disorders or one of four disorders included in the appendix or in DSM-III-R) or a healthy, high-functioning patient.

RESULTS

The data yielded aggregated descriptions of actual patients in each diagnostic category (N = 530) as well as aggregated descriptions of hypothetical, prototypical patients (N = 267). SWAP-200 descriptions of patients with personality disorders showed high convergent and discriminant validity on a variety of criteria. The diagnostic procedure lends itself to both categorical and dimensional personality disorder diagnoses. Descriptions of individual patients resemble MMPI profiles, based on the degree of match between the patient's profile and a criterion group, except that they are based on clinician observation rather than self-report.

CONCLUSIONS

The SWAP-200 represents an approach to the measurement and classification of personality disorders that has potential for refining axis II categories and criteria empirically in ways that are both psychometrically and clinically sound.