摘要：Adaptations to the gold standard randomised controlled trial (RCT) have been introduced to decrease trial costs and avoid high sample sizes. To facilitate development of precision medicine algorithms that aim to optimise treatment allocation for individual patients, we propose a new RCT adaptation termed the nested-precision RCT (npRCT). The npRCT combines a traditional RCT (intervention A versus B) with a precision RCT (stratified versus randomised allocation to A or B). This combination allows online development of a precision algorithm, thus providing an integrated platform for algorithm development and its testing. Moreover, as both the traditional and the precision RCT include participants randomised to interventions of interest, data from these participants can be jointly analysed to determine the comparative effectiveness of intervention A versus B, thus increasing statistical power. We quantify savings of the npRCT compared to two independent RCTs by highlighting sample size requirements for different target effect sizes and by introducing an open-source power calculation app. We describe important practical considerations such as blinding issues and potential biases that need to be considered when designing an npRCT. We also highlight limitations and research contexts that are less suited for an npRCT. In conclusion, we introduce the npRCT as a novel precision medicine trial design strategy which may provide one opportunity to efficiently combine traditional and precision RCTs.
