Evidence needed to compare definitions of personality pathology: commentary on "a postmortem and future look at the personality disorders in DSM-5".

Comments on the original article by Widiger (see record 2013-45025-016). As Widiger notes, it is important to compare competing systems for the measurement of personality disorders (PD) using empirical data. The current author comments that classification systems are not right or wrong; they are simply more or less useful. What kind of evidence is needed for these comparisons? Clearly the relevant data will cover a wide range of topics, from basic descriptive information to behavior genetics and treatment outcome data. The current author makes three recommendations regarding studies of this sort. First, evidence regarding the presence of personality pathology should come from all sources, not simply self-report instruments. Second, investigators should study community samples as well as samples composed exclusively of patients who are receiving treatment. The third point is that investigators who compare diagnostic models should consider longitudinal evidence collected across all phases of the life span.