Research Article Received 21 February 2014, Accepted 11 March 2015 Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/sim.6494 Placebo non-response measure in sequential parallel comparison design studies Denis Rybin, a * † Gheorghe Doros, a,b Michael J. Pencina c and Maurizio Fava d The Sequential Parallel Comparison Design (SPCD) is one of the novel approaches addressing placebo response. The analysis of SPCD data typically classiies subjects as ‘placebo responders’ or ‘placebo non-responders’. Most current methods employed for analysis of SPCD data utilize only a part of the data collected during the trial. A repeated measures model was proposed for analysis of continuous outcomes that permitted the inclusion of information from all subjects into the treatment effect estimation. We describe here a new approach using a weighted repeated measures model that further improves the utilization of data collected during the trial, allowing the incorporation of information that is relevant to the placebo response, and dealing with the problem of possible misclassiication of subjects. Our simulations show that when compared to the unweighted repeated measures model method, our approach performs as well or, under certain conditions, better, in preserving the type I error, achieving adequate power and minimizing the mean squared error. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: SPCD; repeated measures; placebo effect 1. Introduction Accumulating empirical evidence suggests that in some cases, the gold standard of research, the classical double-blinded randomized clinical trial, may be suboptimal in the presence of robust placebo responses. [1–3] The placebo response may lead to dificulties in estimation of the true effect size and ultimately may prevent effective compounds from entering the market or existing drugs from obtaining new applications. As researchers struggle with this placebo response issue, new trial designs are becoming available, and new statistical approaches are needed to analyze the data. The Sequential Parallel Comparison Design or SPCD [1,4–8] is one of the novel approaches addressing placebo response. There are several variations of the SPCD, but in general, it is a design of two stages of equal duration comparing active drug and placebo. At the beginning of Stage I, subjects are randomized to receive either drug or placebo, typically with more subjects assigned to placebo, for example, a 2:1 or 3:1 ratio. At the end of the Stage I, those who were assigned to placebo are classiied as ‘placebo responders’ or ‘placebo non-responders’, based on their outcome. Then, ‘placebo non-responders’ are re- randomized to receive active drug or placebo during Stage II. Typically, those assigned to active treatment at Stage I continue on the active treatment at Stage II, and ‘placebo responders’ continue on placebo or are re-randomized to active drug or placebo. The overall treatment effect is a weighted average of the Stage I estimate in the full sample and of the Stage II estimate in the ‘placebo non-responders’. Despite its novelty, there is already a large body of statistical literature that deals with inference based on SPCD data. a Department of Biostatistics, Boston University, 801 Massachusetts Avenue, Boston, MA 02118, U.S.A. b Harvard Clinical Research Institute (HCRI), 930 Commonwealth Avenue, 3rd loor, Boston, MA 02215, U.S.A. c Duke Clinical Research Institute (DCRI), Biostatistics and Bioinformatics, Duke University, 2400 Pratt Street, Durham, NC 27705, U.S.A. d Harvard Medical School, Massachusetts General Hospital, 55 Fruit Street, Bulinch 351, Boston, MA 02114, U.S.A. * Correspondence to: Denis Rybin, Department of Biostatistics, Boston University, 801 Massachusetts Avenue, Boston, MA 02118, U.S.A. † E-mail: drybine@bu.edu Copyright © 2015 John Wiley & Sons, Ltd. Statist. Med. 2015