Optimal Designs in the N of 1 Trials
*Yin Li, University of Alberta 

Keywords: N of 1 trial, Optiaml design

The N of 1 trial is a randomised experiment in a single patient which is designed to study the response of an individual to the treatments and assist to make personal best clinical decision. To compare two treatment, A and B, a popular design in N of 1 trials is a single sequence which consists of AB pairs and BA pairs. In this study, we consider models with carryover effects. The efficiencies of designs are compared and the optimal N of 1 trials are presented under the models. Different covariance structures of the error terms in the models are also considered.