Placebo-associated improvements have been previously documented in small series of Parkinson's disease (PD) patients. Using a strict definition of placebo-associated improvement, we examined rates and timing of placebo responses to identify patient- and study-based characteristics, predicting positive placebo response in several PD clinical trials. We collected individual patient data from the placebo groups of 11 medical and surgical treatment trials involving PD patients with differing PD severities and placebo-assignment likelihoods. We defined a positive placebo response as > or = 50% improvement in total Unified Parkinson's Disease Rating Scale motor (UPDRSm) score or a decrease by > or = 2 points on at least two UPDRSm items compared to baseline. We calculated positive placebo response rates at early (3-7 weeks), mid (8-18 weeks), and late (23-35 weeks) stages of follow-up. Odds ratios for patient- and study-based characteristics were obtained from a model fitted using generalized estimating equations. There were 858 patients on placebo who met inclusion criteria for analysis. Three studies involved patients without need of symptomatic treatment, two involved patients without motor fluctuations needing symptomatic treatment, and six (three medical and three surgical) involved patients with motor fluctuations. The overall placebo response rate was 16% (range: 0-55%). Patients with higher baseline UPDRSm scores and studies that focused on PD with motor fluctuations, surgical interventions, or those with a higher probability of placebo assignment showed increased odds of positive placebo response. Placebo responses were temporally distributed similarly during early, mid, and late phases of follow-up. Placebo-related improvements occur in most PD clinical trials and are similarly distributed across all 6 months of follow-up. Recognition of factors that impact placebo response rates should be incorporated into individual study designs for PD clinical trials.
2008 Movement Disorder Society