Is dimensional scoring important only for subthreshold levels of severity in personality disorders other than borderline?

Research assessing the utility of dimensional and categorical models of personality disorders (PDs) overwhelmingly supports the use of continuous over categorical models. Using borderline PD as an example, recent studies from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project suggested that continuous (criteria count) scoring of PDs is most informative for "subthreshold" levels of pathology, but is less important once a patient meets the diagnostic threshold. Using PD criteria count, the current study compared 7 indices of psychosocial morbidity for patients above and below diagnostic threshold for 3 additional PDs: paranoid, avoidant, and obsessive-compulsive. Results showed that for all tested PDs, only number of current Axis I disorders was more correlated with PD criteria in the sub-threshold group as compared to those who met criteria for the disorder. Results for the remaining 6 indices of psychosocial morbidity varied by PD tested.