The prediction of short- and long-term improvement in depressive patients: Ethological methods of observing behavior versus clinical ratings

A considerable percentage of depressed patients do not respond to antidepressant treatment. Early indicators of prognosis clearly are needed. The aims of this study are to examine (1) whether the interpersonal behavior of patient and psychiatrist, as assessed by means of direct ethological observation of behavior during a clinical interview, might predict improvement, and (2) whether measures of psychomotor activation, as assessed by global clinical judgment, might predict improvement. An analytical procedure was designed that allowed investigators to determine the organization of the observed interpersonal behavior. Speech-pause behavior was taken as the basic structure of the interaction, and the relationship of looking and of movements of hands and head to this structure was determined. Baseline interviews conducted a few days after admission to the psychiatric hospital were studied on two groups of patients. In these groups, clinical improvement was predicted a posteriori over periods of 2 and 10 weeks of treatment, respectively. Evidence is presented that improvement of depressed patients over both periods can be predicted on the basis of ethological methods of directly observing and analyzing behavior. In contrast, global clinical judgment of psychomotor activation, the method frequently used clinically, had no predictive potency