The heterogeneity of treatment effect lies at the heart of precision medicine. Randomized controlled trials are gold-standard for treatment effect estimation but are typically underpowered for heterogeneous effects. While large observational studies have high predictive power but are often confounded due to lack of randomization of treatment. We show that the observational study, even subject to hidden confounding, may empower trials in estimating the heterogeneity of treatment effect using the notion of confounding function. The confounding function summarizes the impact of unmeasured confounders on the difference in the potential outcomes between the treated and untreated groups accounting for the observed covariates, which is unidentifiable based only on the observational study. Coupling the randomized trial and observational study, we show that the heterogeneity of treatment effect and confounding function are nonparametrically identifiable. We then derive the semiparametric efficient scores and the rate-doubly robust integrative estimators of the heterogeneity of treatment effect and confounding function under parametric structural models. We clarify the conditions under which the integrative estimator of the heterogeneity of treatment effect is strictly more efficient than the trial estimator. We illustrate the integrative estimators via simulation and an application.
翻译:治疗效果的异质性是精密医学的核心。随机控制的试验是治疗效果估计的金质标准,但通常对多种效应的能量不足。虽然大型观测研究具有很高的预测力,但由于缺乏随机治疗而往往被混为一谈。我们表明,即使存在隐蔽的混杂现象,这种观察研究可能增强试验能力,用混杂功能的概念来估计治疗效果的异质性。混杂功能总结了未经测量的混杂者对被治疗和未经处理的组群之间潜在结果差异的影响,这些组群对观察到的同异质体的计算仅以观察性研究为依据,无法确定这种差异。将随机试验和观察研究结合起来,我们表明治疗效果和混杂功能的异质性是无法分辨的。然后我们得出半参数有效分数和高超度的综合估计值综合测算器对治疗效果和对准结构模型下功能的混合作用的影响。我们澄清了综合估测测算结果在何种条件下,仅以观察性研究为基础。通过随机试验和观察研究研究,我们表明治疗结果的精度的精准性分析结果的精准性比模拟性能性分析结果的精准性。