Since the completion of the human genome project, there has been increasing evidence of how pharmacogenomic variation can affect drug dosing, drug efficacy and safety.1 However, the implementation of pharmacogenomics into clinical practice has been slow—many papers have highlighted the reasons for the slow uptake, and this has led to a greater emphasis on implementation projects.2 Of note is the PREPARE study, an open-label, multicentre, controlled, cluster-randomized crossover implementation...