Clinician beliefs and behaviors as modifiers of treatment response in multicenter clinical trials

Jeffrey Dale Lightfoot, Purdue University

Abstract

Overview. This project examined placebo response arising from clinician beliefs and behaviors through the development (Phase One) and validation (Phase Two) of a novel measure: the Assessment of Clinician Relationships and Expectations scale (ACRE). The central hypothesis was that clinician beliefs and behaviors, as measured by the ACRE, would have a significant relationship with patient outcome. Methods. In Phase One, theory-based clinician items and expert-opinion-based site items were developed and evaluated in a series of steps including item sorts, expert panel reviews, a qualitative pre-pilot evaluation, and a pilot investigation. The pilot evaluation tested the a priori factor structure, evaluated item characteristics and scale consistency, and derived an empirically-based factor structure. Phase Two examined the correlation between ACRE scores and outcomes for clinician-patient pairs that were created using three approaches. The first assigned a patient's outcome to a single clinician. The second assigned a patient's outcome to all clinicians contacted. The third assigned a patient's outcome based on the number of contacts. The ability of ACRE scores to discriminate patient outcome also was evaluated at the clinician, site, and item levels. Results. The a priori factor structure was not supported, but an empirically derived structure that contained four factors named Clinical Beliefs and Behaviors, Research Beliefs and Behaviors, Clinical Patient Beliefs, and Research Patient Beliefs was produced. The derived scale contained 50 items, had acceptable reliability (.85), and discriminated between sites on scale and factor totals. In Phase Two several hypotheses were tested, but generally not supported. The association between ACRE scores and patient outcome was in the expected direction, but not significant. The scale total was significantly related to Hamilton depression improvement scores, but only for those patients served by clinicians in high versus low ACRE total score groups. Conclusion. Phase One failed to create a fully theory-based scale. Phase Two failed to find a significant relationship between ACRE scores and outcome. Although this study did not clarify the role of the clinician in placebo response generation, it is still believed that a theory-based approach is the best way to gain an understanding of what drives such responses.

Degree

Ph.D.

Advisors

McGrew, Purdue University.

Subject Area

Clinical psychology

COinS