Because Parkinson's disease (PD) has well-delineated clinical signs and is the frequent focus of controlled clinical trials, it is well suited for the analysis of positive placebo response rates. Multiple factors may influence these rates, including type of outcome measure, stated expectation of treatment (e.g., symptomatic versus restorative), patient/disease characteristics, intervention modality, enrolling site, and probability of assignment to placebo condition. Each of these factors may exert an effect independently or synergistically. In previous studies, we have applied a rigorous definition of placebo response in PD using a well validated outcome measure with both objective (trained rater) and subjective (participants' perception) outcomes, in two types of study designs: a symptomatic treatment design and a restorative treatment design without expectation of symptomatic relief. Consistent rates of placebo response were found for both outcome types and both study designs. Potential effects of sample/disease characteristics, intervention modality, enrolling site, and probability of assignment to placebo condition have not been examined.
The aim of this proposal is to investigate placebo response in a series of well-defined, multi-center, clinical trials varying across three clinical domains: severity of PD; medication or surgical intervention; and varying probability of assignment to placebo condition. Logistic regression, GEE methods for binary longitudinal data, and mixed models will be used to identify important subject and study characteristics related to placebo response and to quantify the additional variation associated with site. Defining these determinant influences will help enhance placebo responses in clinical practice and control them in clinical trials.
| Goetz, Christopher G; Wuu, Joanne; McDermott, Michael P et al. (2008) Placebo response in Parkinson's disease: comparisons among 11 trials covering medical and surgical interventions. Mov Disord 23:690-9 |
