Reliability of diagnosing complicated grief: a preliminary investigation.

Forty experienced mental health providers were asked to diagnose 4 case histories, each involving an aspect of difficult grief embedded in a larger psychopathology context. Half used the full Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev,; DSM-III-R; American Psychiatric Association, 1987); half used a forced choice from among anxiety, mood, adjustment, and personality disorders. For each case, participants also assigned 1 of 4 complicated grief labels derived from the thanatology literature or the label uncomplicated bereavement taken from the DSM-III-R. Interrater agreement was generally low using the DSM-III-R, regardless of instructional format, despite high reliance on relatively few categories. In contrast, interrater agreement was uniformly high using thanatological classifications. Results are discussed in relation to the growing literature differentiating grief from existing DSM classifications and supporting the argument for a separate taxonomy for grief reactions.