BACKGROUND: Clinical outcome assessments may not fully capture patients' perspectives of treatment benefit or tolerability. Incorporating individual exit interviews might enhance the description of the patient experience of drug effects.
OBJECTIVE: The objective of this study was to evaluate the patient treatment experience in a clinical trial of treatment-resistant depression utilizing exit interview methodology.
METHODS: Individual patient interviews were conducted with subjects exiting two phase II clinical trials involving investigational agents for treatment-resistant depression. Interviews included standardized questions about patients' perceptions of health changes and interest in continued use of the investigational agent. Constant comparative analysis of blinded data was used to identify, code, and categorize the data followed by a subsequent analysis of unblinded data to evaluate any potential treatment differences.
RESULTS: Ninety subjects completed exit interviews across the two trials. Most subjects (90%, Trial 2001; 74%, Trial 2002) reported at least one health change. Most subjects rated these changes to be at least moderately important, with most being rated "very important" to "extremely important." After unblinding, participants receiving active therapy alone reported most of the positive health changes (80% of overall positive changes in Trial 2001, 89% in Trial 2002), whereas patients taking placebo alone reported the majority of negative health changes (57% in Trial 2002). Positive changes included not only anticipated changes in mood but also potential cognitive benefits such as mental alertness, improved sleep, and better concentration.
CONCLUSIONS: Standardized interview data provided direct patient insight into the treatment experience from the patient perspective. Data from these interviews assisted in phase III endpoint selection by providing data on relevant concepts in the target treatment-resistant depression population receiving a new treatment, thus enabling the selection of tools to capture noted treatment effects and, by eliminating irrelevant constructs or measures, thereby reducing data "noise."
