Diagnosing personality disorder in depressed patients. A comparison of patient and informant interviews

Personality disorder (PD) diagnosis were made in 66 depressed patients based on independent interviews of the patient and a close informant. The patients and informants markedly differed in their descriptions of the patients' normal personality. The kappa coefficient of agreement for the diagnosis of any PD was .13, and all kappa values for the individual PD diagnoses were below .35. Informants reported more pathologic conditions than the patients, such that PDs were diagnosed in 57.6% (38/66) of the patients, based on the informant interview, and in 36.4% (24/66), based on the patient interview. We also examined dimensional scores. In general, we found only modest correlations between the patient and informant dimensional scores and that the ratings based on the informant interviews were higher. These results varied by specific PD diagnoses. Consensus ratings, which were based on both sources of information, were sometimes more strongly associated with patient information and sometimes with informant information, and this, too, varied among the different PDs.