Heterogeneous treatment effects represent a major issue for medicine as they undermine reliable inference and clinical decision-making. To overcome the issue, the current vision of precision and personalized medicine acknowledges the need to control individual variability in response to treatment. In this paper, we argue that gene-treatment-environment interactions (G × T × E) undermine inferences about individual treatment effects from the results of both genomics-based methodologies—such as genome-wide association studies (GWAS) and genome-wide interaction studies (GWIS)—and randomized controlled trials (RCTs). Then, we argue that N-of-1 trials can be a solution to overcome difficulties in handling individual variability in treatment response. Although this type of trial has been suggested as a promising strategy to assess individual treatment effects, it nonetheless has limitations that limit its use in everyday clinical practice. We analyze the existing variability within the designs of N-of-1 trials in terms of a continuum where each design prioritizes epistemic and pragmatic considerations. We then support wider use of the designs located at the pragmatic end of the explanatory-pragmatic continuum.

异质治疗效果是医学的一个主要问题，因为它们破坏了可靠的推理和临床决策。为了克服这个问题，当前精准和个性化医疗的愿景承认需要控制个体对治疗反应的变异性。在本文中，我们认为基因-治疗-环境相互作用（G × T × E）破坏了基于基因组学方法（例如全基因组关联研究（GWAS）和全基因组研究）的结果对个体治疗效果的推断。相互作用研究（GWIS）和随机对照试验（RCT）。然后，我们认为 N-of-1 试验可以成为克服处理治疗反应个体差异的困难的解决方案。尽管此类试验被认为是评估个体治疗效果的一种有前景的策略，但它仍然存在局限性，限制了其在日常临床实践中的使用。我们根据连续体分析 N-of-1 试验设计中现有的变异性，其中每个设计都优先考虑认知和实用考虑。然后，我们支持更广泛地使用位于解释-实用连续体实用端的设计。