Identification of true drug response to antidepressants. Use of pattern analysis

Abstract

The purpose of this study was to develop a method for differentiating specific ("true") and nonspecific antidepressant drug response for the individual patient. Patterns of clinical response, based on weekly global ratings of clinical status, were generated for each of 185 patients participating in six-week placebo-controlled drug trials. We hypothesized and found that substantially more patients receiving active than placebo medication displayed treatment response patterns characterized both by two-week or greater delay in onset of initial improvement and nonfluctuating persistence of improvement once achieved. Identification of a distinctive pattern of clinical response to an active drug has both research and clinical applications. Pattern analysis may contribute to understanding the nature of drug mechanisms of action, may clarify some ambiguous treatment study outcomes, and in the individual case, may facilitate clinical management.